If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Department of Pathology, University of Colorado Denver, Aurora, ColoradoDepartment of Dermatology, Yueyang Hospital of Integrated Traditional Chinese and Western Medicine, Shanghai University of Traditional Chinese Medicine, Shanghai, China
Address reprint requests to Xiao-Jing Wang, M.D., Ph.D., Department of Pathology, Bldg. RC1-N, Rm P18-5128, Mail Stop 8104, University of Colorado Denver, Aurora, CO 80045-0508
The expression of Smad7, a tumor growth factor-β (TGFβ) antagonist, is increased during cutaneous wound healing. To assess this significance, we temporally induced Smad7 transgene expression in wounded skin in gene-switch-Smad7 transgenic (Smad7 tg) mice. Smad7 induction in epidermal keratinocytes caused an increase in keratinocyte proliferation with reduced Smad2 activation, indicating that Smad7 abrogated TGFβ-mediated growth inhibition. Additionally, wounded skin from Smad7 tg mice exhibited accelerated re-epithelialization, with increased activation of extracellular signal-regulated kinase (Erk), and an in vitro migration assay revealed that Erk activation contributed to Smad7-mediated keratinocyte migration. Notably, epidermis-specific Smad7 transgene expression also has a profound effect on the wound stroma, resulting in reduced inflammation, angiogenesis, and production of type I collagen. Reduced Smad2 activation was observed in wounded stroma from Smad7 transgenic (Smad7 tg) mice, possibly owing to fewer infiltrated TGFβ-producing leukocytes compared to those in wounds from control mice. Because Smad7 is not secreted, these effects could reflect functional changes in Smad7 tg keratinocytes. Supporting this notion, the activation of NF-κB, a nonsecreted protein complex that transcriptionally activates inflammatory cytokines, was reduced in wounded epidermis from Smad7 tg mice compared to that in wounded wild-type epidermis. In sum, epidermal Smad7 overexpression accelerated wound healing through its direct effects on keratinocyte proliferation and migration, and through indirect effects on wound stroma.
Cutaneous wound healing progresses through three overlapping phases: inflammation, tissue formation, and tissue remodeling.
These are dynamic processes that involve interactions among the epidermis, leukocytes, extracellular matrix, and dermal fibroblasts. In response to skin injury, blood clots, infiltrated inflammatory cells, and other cell types in the wound release multiple cytokines and chemokines. These cytokines initiate fibroblast proliferation and synthesis of extracellular matrix that fill the wound deficit and lead to wound closure. Meanwhile, keratinocytes at the wound edge begin to proliferate and migrate to cover the wound surface. Underneath the re-epithelialized epidermis, new stroma, called granulation tissue, begins to fill the wound space, which contains provisional extracellular matrix, inflammatory cells, fibroblasts, and blood vessels. Once the wound area is filled with the granulation tissue and covered by newly re-epithelialized epidermis, the process of wound closure is completed. Later on, the wound gradually returns to normal strength and texture through tissue remodeling.
Among the many molecules known to influence wound healing, transforming growth factor-β (TGFβ) has the broadest spectrum of action, affecting all cell types that are involved in all stages of wound healing.
When a ligand binds to TGFβ type I and type II receptors (TGFβRI and TGFβRII), TGFβRI phosphorylates Smad2 and Smad3. Phosphorylated Smad2 and Smad3 bind a co-Smad, Smad4, forming heteromeric Smad complexes that translocate into the nucleus to regulate transcription of TGFβ target genes.
In contrast, the introduction of exogenous Smad3 to wound sites to enhance TGFβ signaling also accelerated wound healing in a rabbit dermal ulcer model.
Skin wounds in Smad4-deficient mice exhibit dramatically increased inflammation and angiogenesis, causing a delay in wound closure and excessive scar formation.
Transient adenoviral gene transfer of Smad7, an antagonist of TGFβ signaling, in corneal epithelium and stroma resulted in accelerated corneal wound healing with reduced inflammation.
Further, Smad7 gene transfer to the lens epithelium and stroma prevented injury-induced epithelial–mesenchymal transition of lens epithelial cells and suggests a potential role of Smad7 in prevention of capsular fibrosis.
These studies suggest that the effects of TGFβ signaling components, such as Smad7, on wound healing are complex and highly context specific. Given the completely opposite effects of Smad3 and Smad4 on skin wound healing, and the opposite effects of Smad7 on corneal wound healing and balloon injury, it is difficult to predict the effects of Smad7 overexpression that blocks signaling from all signaling Smads, on cutaneous wound healing. Additionally, the effect of Smad7 may not always be explained by its role in TGFβ signaling. For instance, Smad7 also interacts with components of the Wnt/β-catenin
In the present study, we sought to assess the effect of epidermal-specific Smad7 overexpression on cutaneous wound healing. Our study reveals the positive role of Smad7 overexpression in cutaneous wound healing by affecting multiple biological processes.
Materials and Methods
Wound Healing Experiments
The gene-switch-Smad7 transgenic mice, which allow keratinocyte-specific Smad7 transgene expression by topical application of the inducer RU486, were generated by cross-breeding the K14.GLp65 activator line and the tata.Smad7 target line to generate bigenic mice containing both transgenes as we previously described.
The wound healing experiments were performed using a protocol approved by the Institutional Animal Care and Use Committee at University of Colorado Denver. Fifty-four 8-week-old bigenic mice were enrolled in the present study. Forty-five 8-week-old monogenic (K14.GLp65 or tata.Smad7) or wild-type mice were used as controls. Skin wounding was performed as previously reported.
Briefly, the back skin of the mice was shaved and treated with RU486 (10 μg/mouse) to induce Smad7 transgene expression in bigenic mice. One hour after RU486 treatment, the mice were anesthetized and their backs were wiped with 70% ethanol. Four full-thickness excisional wounds were generated on the dorsal skin of mice using a 6-mm-diameter dermal punch. The dorsal skin (including wound area) was treated with 10 μg of RU486 daily until day 14 after wounding. Control mice, which included wild-type and monogenic littermates, received the same treatment (wounding and RU486 applications). The wounded skin, including a 2-mm margin, was excised for analysis. Wounded skin from 8 to 10 mice/genotype was harvested at each time point. Wound closure was evaluated by macroscopic and microscopic measurements of the wound diameter through its midline. Wound remodeling was evaluated by wound stromal area, which was defined as the region beneath the migrating epidermis and above the subcutaneous fat tissue, and measured using MetaMorph software (Molecular Devices, Sunnyvale, CA).
In Situ Hybridization
In situ hybridization was performed using digoxigenin-11-dUTP–labeled probes as we have previously described.
Histology, Immunohistochemistry, and Immunofluorescence
For general histological analyses, tissue samples were fixed in 10% neutral-buffered formalin, embedded in paraffin, sectioned from the midline of wounds, and stained with hematoxylin and eosin (H&E). Immunofluorescence or immunohistochemistry was performed on optimal cutting temperature compound–embedded frozen sections or paraffin-embedded sections as previously described.
The primary antibodies used in the present study were guinea pig anti-K14 (RDI-Fitzgerald, Acton, MA), rat anti-CD31, -CD45, -CD4, and BM8 (anti-F4/80) (BD Bioscience, San Jose, CA), rabbit anti-NFκBp50, and proliferating cell nuclear antigen (PCNA) (Santa Cruz Biotechnology, Santa Cruz, CA), rabbit anti-pSmad2 and pErk (Cell Signaling Technology, Danvers, MA), and rabbit anti-Smad7.
For immunohistochemistry, the sections were incubated with secondary biotinylated antibodies (4 μg/mL) for 30 minutes at room temperature, and the Vectastain ABC kit (Vector Laboratories, Burlingame, CA) was used for detection. Hematoxylin QS (Vector Laboratories) was used as the counterstain. For immunofluorescence, Alexa Fluor 594 (red) or 488 (green) conjugated secondary antibodies (Invitrogen, Carlsbad, CA) were used and incubated for 30 minutes at room temperature. All slides are mounted in Fluoromount G (Southernbiotech, Birmingham, AL) with coverslips before microscopic examination. To quantify pSamd2, PCNA, and NFκBp50 nuclear staining, cells with positive staining in four to five images from different wound sections were counted using MetaMorph software and expressed as the number of positive cells in the migrating epidermis (in mm) or per mm2 stromal area. CD31 stained vessels were quantified using ImageJ software (NIH, Bethesda, MD) and results presented as percentage of stroma covered by vessels.
Picrosirius Red Staining and Imaging
Picrosirius red staining was performed as previously described.
Briefly, paraffin sections were de-waxed and rehydrated. The sections were stained in picrosirius red solution (0.1% sirius red F3B in saturated picric acid solution; Sigma, St. Louis, MO) for 1 hour at room temperature. After two quick washes in 0.1 N acetic acid, the slides were dehydrated and mounted with polymount-xylene (Polyscience Inc, Warrington, PA). Images were taken under a microscope with and without polarized light.
Western Blot Analysis
Tissue and cultured cells were homogenized in Complete Lysis-M buffer containing protease inhibitor together with phosphatase inhibitors (Roche Applied Science, Indianapolis, IN). Extracts were cleared by centrifugation and protein concentrations were determined by Bio-Rad protein assay kit (Bio-Rad Laboratories, Hercules, CA). Sixty micrograms of total protein from lysates was loaded into each lane of a 10% SDS gel for electrophoresis. After transfer to a nitrocellulose membrane, the membrane was blocked and incubated with primary antibodies at 4°C overnight. Primary antibodies used include rabbit anti-Smad7,
rabbit anti-VEGF (Santa Cruz Biotechnology, Santa Cruz, CA), rabbit anti-pErk or total Erk and rabbit anti-GAPDH (Cell Signaling Technology, Danvers, MA). IRDye conjugated donkey anti-rabbit IgG (Rockland Immunochemicals Inc., Gilbertsville, PA) was used as secondary antibody and incubated with membrane at room temperature for 1 hour. Gray-scale images and quantification were obtained with Odyssey v.1.2 software (LI-COR Biosciences, Lincoln, NE).
RNA Extraction and Analyses
Total RNA was isolated from wound samples using RNAzol B (Tel-Test, Inc., Friendswood, TX) as previously described.
Results from four samples of each genotype are shown.
Primary Keratinocyte Culture and in Vitro Keratinocyte Migration Assays
Primary keratinocytes including gene-switch-Smad7 (bigenic) and control cells (wild-type or monogenic) were isolated from neonatal mouse skins with different genotypes, as previously described.
Expression of a dominant-negative type II transforming growth factor beta (TGF-beta) receptor in the epidermis of transgenic mice blocks TGF-beta-mediated growth inhibition.
The cells were cultured in Progenitor Cell Targeted Epidermal Keratinocyte medium (CELLnTEC USA, La Vista, NE) to maintain keratinocyte proliferation. For keratinocyte migration assays, cultured keratinocytes at nearly full confluency were treated with 10−7 mol/L RU486 for 12 hours to induce Smad7 expression
and treated with mitomycin C (Sigma) to block cell proliferation. A 1.2-mm scratch wound was made using a pipette tip. Cell migration was evaluated in the presence of RU486 and photographed every 24 hours after wounding. To study whether blocking Erk activation affects Smad7-mediated acceleration of cell migration, cell migration was further evaluated in the presence of the Erk inhibitor U0126 (10 μmol/L; EMD Chemicals Inc, Gibbstown, NJ) with or without Smad7 induction. Wound closure was calculated by the area occupied with migrating cells at different time points determined by ImageJ software.
Statistical Analysis
Statistical differences were analyzed using the Student's t-test. The data were presented by mean ± SD.
Results
Smad7 Is Increased during Cutaneous Wound Healing in Wild-Type Mice
We examined Smad7 expression during normal cutaneous wound healing. In comparison with unwounded mouse skin, the level of Smad7 mRNA was significantly increased in day 3 wounded skin after a 6-mm full-thickness skin wound, and reached its peak at approximately day 7 after wounding, ie, ∼eightfold higher than Smad7 in unwounded skin as determined by RT-qPCR (Figure 1A). Thereafter, Smad7 levels gradually returned to normal by day 18 after wounding, at which point, wound closure was completed as determined by macroscopic and microscopic observation. The peak level of Smad7 expression during wound healing was detected by Western blot analysis (Figure 1B). To localize the source of increased Smad7 expression during cutaneous wound healing, we compared Smad7 expression in wounded tissues with unwounded skin in wild-type mice. As shown in Figure 1, C and D, Smad7 protein was barely detected in normal skin, but was significantly increased in both migrating epidermal cells and proliferative fibroblasts in wound stroma with cytoplasmic and/or nuclear staining.
Figure 1Smad7 expression is increased during cutaneous wound healing in wild-type mice. A: RT-qPCR analysis of Smad7 mRNA levels in normal cutaneous wound tissues at different time points after wounding. *P < 0.05. B: Western blot of Smad7 protein in unwounded skin (day 0), day 7, and day 18 wounded tissues after wounding. C: Locations of increased endogenous Smad7 in wounded skin (lower panel) compared with nonwounded skin (upper panel), as determined by immunofluorescence staining of Smad7 (green). Antibody against K14 (red) was used to counterstain. D: Enlarged area of the red box in C showing Smad7 staining in migrating epidermal keratinocytes with a higher magnification. Scale bar = 80 μm (B and C). Dotted lines indicate the borderline of epidermis and dermis. White arrow in C and D indicates the migration direction of re-epithelialization.
Four excisional wounds with a 6-mm diameter were created on the dorsal skin of 8-week-old bigenic and control mice. Smad7 transgene expression in keratinocytes was induced in bigenic skin by daily topical application of RU486, beginning 1 hour before wounding. Wild-type littermates treated with the same dose of RU486 were used as controls. Smad7 transgene expression was four- to sixfold higher than endogenous Smad7 levels in control skin, as determined by RT-qPCR (not shown). Grossly, Smad7 tg skin exhibited faster wound closure at all time points in comparison with control mice (Figure 2A). Histologically, epidermal migrating tongues in Smad7 tg skin wounds were more advanced than in control skin wounds at days 3 and 5 after wounding, resulting in a significant decrease in the wound width compared with control skins (Figure 2, B and C). By day 7, when 50% to 75% of the wounded area in control wounds was closed, Smad7 tg wounds were completely closed (Figure 2D). By day 11, when wounds in control skins were completely closed, Smad7 tg wounds had already grown new hair follicles (Figure 2D). The wound stromal area, defined as the area below the migrating epidermal tongues and above the subcutaneous fat, was significantly reduced in Smad7 tg wounds from day 7 to 11 after wounding in comparison with control wounds (Figure 2, D and E), indicating accelerated wound healing and remodeling in Smad7 tg mice.
Figure 2Smad7 overexpression accelerated wound healing. A: Representative macroscopic appearance of wound closure in Smad7 transgenic and control mice at different time points after wounding. B: H&E staining of wound sections from Smad7 tg and control wounds at day 3 and day 5 after wounding illustrating accelerated re-epithelialization in Smad7 wounded skin. Dotted lines indicate re-epithelialized epidermis. C: Quantification of wound width on day 3 and day 5 wounds from Smad7 tg and control mice (five skin samples from each group). *P < 0.05. D: H&E staining of skin sections from Smad7 and control wounds at day 7 and day 11 after wounding, illustrating reduced wound stroma in Smad7 wounded skin. Dotted lines indicate re-epithelialized epidermis. E: Quantification of wound stromal area in day 7 and day 11 wounds from Smad7 tg and control mice (5 skin samples from each group). Scale bar = 100 μm (B and D). *P < 0.05.
Enhanced Keratinocyte Proliferation and Reduced Smad Activation in the Epidermis of Smad7 tg Wounds
Previously, we demonstrated that Smad7 overexpression in keratinocytes caused epidermal hyperproliferation as a result of its blockade of TGFβ signaling.
Because wounded skin generally exhibits increased epidermal proliferation, we sought to determine whether the observed accelerated wound closure of Smad7 wounds could be due, at least in part, to an additional increase in epidermal proliferation in wounds. Indeed, Smad7 wound epidermis exhibited a significant increase in the number of PCNA-positive cells in basal and suprabasal layers of the regenerated epidermis in comparison with control wounds from after wounding day 5 and day 7 (Figure 3, A and B). To determine whether increased epidermal proliferation in Smad7 wounds is due to blockade of TGFβ signaling, we examined phospho-Smad2 (pSmad2), a surrogate marker for activation of Smad-dependent TGFβ signaling. Beginning on day 3 after wounding, increased pSmad2 nuclear staining in control wounds was observed in the epidermis at the wound edge in comparison with nonwounded skin, which correlated with increased levels of endogenous TGFβ1 (data not shown). Increased pSmad2 was more obvious on days 5 and 9 in control wounds (Figure 3, C and D). In contrast, the number of pSmad2-positive cells in the epidermis of Smad7 wounds was significantly reduced compared to controls (Figure 3, C and D). Additionally, pSmad2 staining in stromal cells was also reduced in Smad7 tg wounds compared to control wounds (Figure 3, C and E), which could be a result of the accelerated healing and a reduction in inflammatory cells (see below) that secrete TGFβ1.
Figure 3Enhanced keratinocyte proliferation and reduced Smad activation in the epidermis of Smad7 tg wounds. A: PCNA staining of day 5 and day 7 wounds from Smad7 tg and control mice shows an increased number of PCNA-positive cells (brown) in the regenerated epidermis of Smad7 wounded skin. Dotted lines indicate re-epithelialized epidermis. The bar represents 40 μm for all images. B: Quantification of PCNA positive cells in the regenerated epidermis of Smad7 and control wounds. *P < 0.01. The data represent mean ± SD of five separate wounds analyzed. C: Immunostaining of pSmad2 in day 5 and day 9 Smad7 and control wound tissues. Note a reduction of pSmad2 nuclear staining (brown) in the epidermis and stroma of Smad7 tg skin at both time points. Dotted lines indicate re-epithelialized epidermis. Scale bar = 40 μm. D: Quantification of pSmad2-positive cells in the regenerated epidermis of Smad7 and control wounds. *P < 0.01. The data represent mean ± SD from analysis of four separate wounds. E: Quantification of pSmad2-positive cells in wound stroma of Smad7 and control wounds. *P < 0.01. The data represent mean ± SD from analysis of four separate wounds.
Erk Activation Contributes to Accelerated Keratinocyte Migration in Smad7 tg Skin Wounds
To determine whether the accelerated wound closure observed in Smad7 tg wounds was the result of increased epidermal proliferation alone or if Smad7 also directly promotes keratinocyte migration, we performed an in vitro migration assay using normal and Smad7 tg keratinocytes. We first compared endogenous Smad7 expression in migrating and nonmigrating keratinocytes. Wild-type keratinocytes were cultured in chamber slides to confluency and subjected to a scratch wound. Immunofluorescence staining with a Smad7 antibody revealed that the number of Smad7-positive cells and the intensity of staining increased among keratinocytes migrating into the wound area compared to keratinocytes away from the wound margin (Figure 4A). This experiment suggests that migrating keratinocytes express more Smad7 than nonmigrating keratinocytes. To determine whether Smad7 transgene expression blocks only Smad signaling, not Smad-independent TGFβ signaling involved in cell migration, we examined Erk, which is a major TGFβ target activated by phosphorylation (pErk) through Smad-independent TGFβ signaling,
p38 and ERK1/2 coordinate cellular migration and proliferation in epithelial wound healing: evidence of cross-talk activation between MAP kinase cascades.
Interestingly, pErk was not abrogated by Smad7 overexpression; instead, increased nuclear staining of pErk was observed in Smad7 tg wounds on day 5 to day 9 (Figure 4, B and C) in comparison with control wounds. We then performed in vitro migration assays using normal and Smad7 tg keratinocytes to determine whether increased Erk activation in Smad7 wounds promotes keratinocyte migration. Control and bigenic keratinocytes were isolated from the epidermis of littermates, cultured to confluency, and then treated with 10−7 mol/L RU486 for 12 hours to induce Smad7 overexpression in bigenic keratinocytes. The cells were then treated with mitomycin C for 2 hours to stop cell proliferation, and a scratch wound was introduced. By 24 hours after wounding, when control keratinocytes began to migrate into the wound area, more than 50% of the wound area was covered by migrating cells in the Smad7 tg keratinocyte culture (Figure 4, D and E). By 48 hours, Smad7 tg keratinocytes had almost completely covered the wound area, whereas control keratinocytes only covered ∼30% of the wound area. To evaluate whether Erk activation contributes to Smad7-mediated acceleration of keratinocyte migration, we treated Smad7 tg and control keratinocytes with the Erk inhibitor U0126 (10 μmol/L) for 8 hours in scratch assays. Western blot analysis showed Smad7 induction increased pErk1/2 fourfold without affecting the amount of total Erk, and U0126 treatment reduced pErk to a nearly undetectable level in both control and Smad7 tg cells (Figure 4F). Consistent with previous reports, this dose of U0126 treatment inhibited normal keratinocyte migration in both humans and mice
; in vitro migration assay shows that control keratinocytes treated with U0126 completely stopped migration (Figure 4, D and E), indicating that Erk activation is a major contributor of keratinocytes migration. Further, this dose of U0126 was also sufficient to block migration of Smad7 tg keratinocytes (Figure 4, D and E), which had a higher pErk level (Figure 4, B, C, and F).
Figure 4Smad7 overexpression promotes keratinocyte migration contributed by Erk activation. A: Immunofluorescence staining of wild-type keratinocytes with an antibody against Smad7 (green) shows more Smad7-positive cells in migrating keratinocytes than nonmigrating keratinocytes. K14 (red) was used as a counterstain. Smad7-positive cells appeared yellow when Smad7 and K14 staining overlapped. Dotted lines indicate the wound margin. Black arrows indicate the direction of cellular migration. B: Representative immunofluorescence staining of pErk (green) in day 5 wound epidermis from Smad7 tg and control mice. Antibody against K14 was used for counterstain (red). C: Quantification of pErk-positive cells in the migrating epidermis of Smad7 and control wounds. *P < 0.01. The data represent mean ± SD from analysis of four separate wounds. D: Cell migration in Smad7 and control keratinocytes at 0, 24, and 48 hours after in vitro wounding with and without the Erk inhibitor U0126 treatment. Dotted lines indicate the wound margin. E: Quantification of migrating cells into wounded area. Migration is expressed by percentage of the wounded area covered by migrating keratinocytes compared with 0 hour wounded area. Three representative images were analyzed at each time point. *P < 0.05. F: Western blot analysis of pErk and total Erk expression in Smad7 and control cells with and without U0126 treatment for 8 hours. Scale bar = 40 μm (A and B).
Inflammation, Angiogenesis, and Collagen Production Are Reduced in Wounded Skin from Smad7 tg Mice
The wound healing process involves stromal changes orchestrated with re-epithelialization. Although inflammation, angiogenesis and collagen synthesis are part of the normal healing process, excessive stromal responses delay healing.
To determine whether this Smad7 function affects wound stroma, we compared inflammation, angiogenesis, and collagen synthesis between Smad7 tg and wild-type wounds. Immunofluorescence staining with antibodies against CD45 (pan-leukocyte marker), CD4 (T cell marker), and BM8 (monocyte/macrophage marker) revealed reduced total leukocytes on day 3 of Smad7 tg wounds, although a substantial number of BM8-positive macrophages, which play an important role in wound healing,
Topical vascular endothelial growth factor accelerates diabetic wound healing through increased angiogenesis and by mobilizing and recruiting bone marrow-derived cells.
were still present (Figure 5A). Similar changes lasted till day 9 after wounding; thereafter, leukocyte infiltration was significantly reduced to normal level in either control or Smad7 tg skin (not shown). Because Smad7 is not secreted, we assessed whether this anti-inflammatory effect is mainly the result of functional changes in keratinocytes. We examined NF-κB, a nonsecreted protein complex controlling transcriptional regulation of inflammatory cytokines. Consistent with previous findings that NF-κB is activated during wound healing,
wild-type wounds exhibited increased nuclear NF-κB p50 subunit in most epithelial and stromal cells (Figure 5, B and C). In contrast, the number of nuclear NF-κB p50 subunit–positive cells was significantly reduced, primarily in migrating keratinocytes of Smad7 wounds (Figure 5, B and C).
Figure 5Reduced inflammation and NF-κB p50 in wounded skin from Smad7 tg mice. A: Immunofluorescence staining of inflammatory cells (green) with antibodies against CD45 (pan-leukocyte marker), CD4 (T cell marker), and BM8 (monocyte/macrophage marker) on sections of day 3 wound edge from Smad7 tg and control mice. K14 was used as a counterstain (red) to highlight the epidermis. Dotted lines indicate the wound margin. B: Immunofluorescence staining of NF-κB p50 (green) in day 5 and day 9 wounds from Smad7 tg and control mice. Antibody against K14 was used for counterstain (red). The bar represents 40 μm for all panels. C: Quantification of NF-κB p50-positive cells in the migrating epidermis of Smad7 and control wounds. *P < 0.01. The data represents mean values ± SD from analysis of four separate wounds. Scale bar = 40 μm (A and B).
We examined angiogenesis using CD31 staining. On days 1 and 3 after wounding, Smad7 and control wounds showed similar degrees of angiogenesis in the wound bed (Figure 6, A and B) and at the wound edge (not shown). Starting on day 5 after wounding, Smad7 wounds exhibited a considerable reduction in angiogenesis within the granulation tissue compared to control wounds (Figure 6, A and B). A further reduction in angiogenesis was observed during later stages of wound healing in Smad7 mice, particularly underneath the migrating epidermis (Figure 6, A and B). Because angiogenesis could be an important step to initiate healing, and angiogenesis factor VEGF-A shows therapeutic effects on wound healing through promoting angiogenesis and epidermal proliferation,
Topical vascular endothelial growth factor accelerates diabetic wound healing through increased angiogenesis and by mobilizing and recruiting bone marrow-derived cells.
we examined whether Smad7 transgene expression affects VEGF-A production. On day 3 after wounding, VEGF-A expression was elevated at similar levels in wild-type and Smad7 wounds, but was significantly less in Smad7 wounds at late-stage wound healing, beginning on day 9 after wounding (Figure 6C). These data suggest that Smad7 overexpression does not affect early-stage angiogenesis, and reduced angiogenesis at later time points is a consequence of faster re-epithelialization during wound healing.
Figure 6Evaluation of angiogenesis and VEGF-A expression in Smad7 tg and control wounds. A: Immunofluorescence staining with CD31 antibody (green) in sections of day 3 wound bed, and day 5 and day 9 wound tissues from Smad7 and control mice. K14 was used as a counterstain (red) for the epidermis. Scale bar = 40 μm. B: Quantification of CD31 staining in the stromal area of Smad7 tg and control wounds. *P < 0.05. The data represent mean ± SD from analysis of four separate wounds. C: Western blot analysis of VEGF-A expression in day 3, day 9, and day 18 wound tissues showing reduced VEGF-A protein in Smad7 wounds at day 9 and day 18 time points.
Type-I collagen produced by activated fibroblasts is increased during wound healing as part of stromal repair and remodeling, and is a major component of granulation tissue and is reduced after healing is completed. Since fibroblast activation and collagen production can be affected by keratinocyte-secreted factors and/or inflammation,
we examined whether keratinocyte-specific Smad7 induction affects type-I collagen production. In situ hybridization shows fewer COL1A1 mRNA-positive cells in the wound area of Smad7 tg skin compared to control skin wounds at days 5, 7, and 11 after wounding (Figure 7A). To determine whether reduced COL1A transcripts affect collagen organization thus altering the quality of wound stromal remodeling, we performed picrosirius red staining. On day 5, the collagen fibers in Smad7 tg wounds were more abundant compared to control wounds but were comparable with day 7 control wounds (Figure 7B). This result suggests that reduced COL1A mRNA (Figure 7A) is a consequence of faster healing. In control wounds, collagen fibers were thicker on day 7 than day 11 (Figure 7B). The thickness of collagen fibers in Smad7 wounds on day 7 and day 11 were similar to those in wild-type day 11 wounds, but overall, the collagen was reduced in Smad7 wounds at these two time points compared to controls (Figure 7B). These data suggest that the kinetics of new collagen production, collagen fiber formation, and remodeling occurs more quickly in Smad7 tg wounds than wild-type wounds.
Figure 7Detection of type-IA1 collagen (COL1A1) mRNA and collagen production in wound sections from Smad7 tg and control mice. A: In situ hybridization for COL1A1 mRNA at different time points showing reduced COL1A1 synthesis in Smad7 wounds. Dotted lines highlight the boundary between newly re-epithelialized epidermis and the stroma. B: Picrosirius red staining for collagen (bright images or images with polarized light) in wild-type and Smad7 wounds at different time points showing early collagen formation with fast resolution in Smad7 wound stroma in parallel with accelerated wound healing. Scale bar = 80 μm (A and B).
In the present study, we demonstrated that Smad7 overexpression in keratinocytes affected multiple biological processes during cutaneous wound healing to coordinately promote healing. Further analyses revealed that the mechanisms underlying Smad7 functions in cutaneous wound healing involve blocking the negative effects of TGFβ signaling and NF-κB signaling but increasing Erk activation during wound healing.
Smad7 Wounded Skin Demonstrates Reduced Smad Signaling Which Abrogated the Negative Effects of TGFβ1 on Re-Epithelialization
Our previous studies have shown that in nonwounded skin, Smad7 overexpression in keratinocytes causes epidermal hyperproliferation as a result of blocking TGFβ-induced growth inhibition.
We have also previously shown that during wound healing, TGFβ1 is elevated and that TGFβ1 overexpression in wounded skin inhibits keratinocyte proliferation at the wound edge.
Our current study shows that Smad7 transgene induction, at levels four- to sixfold higher than endogenous Smad7, was sufficient to override the growth inhibitory effect of increased endogenous TGFβ1 in wounds, which is evident by reduced Smad activation in wounded skin keratinocytes. This finding is consistent with the report that mice lacking Smad3 showed accelerated re-epithelialization during wound healing.
In addition to inducing keratinocyte hyperproliferation, Smad7 overexpression appeared to have a direct effect on keratinocyte migration. First, Smad7 expression was increased in migrating cells in comparison with nonmigrating cells. Second, Smad7 tg keratinocytes migrated faster than control keratinocytes even when proliferation was attenuated by mitomycin C treatment. Previously, activation of Erk and NF-κB pathways have been shown to positively regulate keratinocyte proliferation and migration.
p38 and ERK1/2 coordinate cellular migration and proliferation in epithelial wound healing: evidence of cross-talk activation between MAP kinase cascades.
Targeted deletion of the integrin beta4 signaling domain suppresses laminin-5-dependent nuclear entry of mitogen-activated protein kinases and NF-kappaB, causing defects in epidermal growth and migration.
In our current study, Erk expression was increased, whereas that of NF-κB was decreased in Smad7 tg keratinocytes. We show that increased Erk activation in Smad7 tg keratinocytes contributed to accelerated keratinocyte migration. Thus, increased Erk activation along with blocking TGFβ-induced growth arrest appeared sufficient to overcome any negative effect, if any, from reduced NF-κB, on keratinocyte migration and proliferation. TGFβ signaling activates Erk through a Smad-independent mechanism,
and Smad7 transgene expression does not appear to block this process. Additional mechanisms by which Erk was further activated in Smad7 tg keratinocytes are yet to be determined. Nevertheless, Smad7 overexpression appears to selectively block the negative effects while maintaining the positive effects of TGFβ signaling on wound healing. With respect to NF-κB down-regulation found in Smad7 keratinocytes, this change is linked to reduced inflammation that could create a microenvironment to facilitate healing (see below).
Stromal Effects of Epithelial-Specific Smad7 Overexpression Contribute to Accelerated Wound Tissue Remodeling
In addition to accelerated re-epithelialization, epithelial-specific Smad7 overexpression has a profound effect on wound stroma (ie, reducing inflammation, angiogenesis, and collagen production). Reduced inflammation and angiogenesis have also been reported in corneal wounds with exogenous Smad7 expression.
However, in contrast to that study, in which Smad7 was introduced to all cell types in the wound, Smad7 transgene expression in our current study was restricted to keratinocytes. Because NF-κB signaling, the central mediator of inflammation, was reduced primarily in the epidermis of wounds from Smad7 tg mice (evidenced by reduced p50), this change could contribute significantly to reduced inflammatory cytokine production in keratinocytes. Further, blocking TGFβ signaling (evidenced by reduced pSmad2) could also contribute to reduced inflammation in Smad7 tg wounds. Studies have shown that although TGFβ1 has an anti-inflammatory effect in internal organs as evidenced by autoimmune responses in TGFβ1 knockout mice,
8-methoxypsoralen plus ultraviolet A therapy acts via inhibition of the IL-23/Th17 axis and induction of Foxp3+ regulatory T cells involving CTLA4 signaling in a psoriasis-like skin disorder.
Consistently, keratinocyte-specific Smad7 overexpression blocked T-cell infiltration more significantly than macrophage infiltration (Figure 5), which could enable the remaining macrophages to promote wound repair.
Topical vascular endothelial growth factor accelerates diabetic wound healing through increased angiogenesis and by mobilizing and recruiting bone marrow-derived cells.
Additionally, although Smad7 transgene is not expressed in the stroma, reduced pSmad2 staining was observed in the stroma of Smad7 tg wounds. This could be a result of reduced leukocyte infiltration or accelerated healing in Smad7 tg wounds, which consequently have fewer inflammatory cells and activated fibroblasts. Either one of these possibilities or a combination of the two would, in turn, produce less TGFβ1 in the stroma, resulting in reduced Smad activation in the stroma. As a result, reduced inflammation in Smad7 tg wounds could significantly contribute to accelerated wound healing. Supporting this, TGFβ1 transgene expression in keratinocytes results in delayed wound healing, mainly through excessive inflammation.
Reduced inflammation in Smad7 tg wounds could be responsible, at least in part, for reduced angiogenesis and collagen production, as both occurred after the reduction in inflammation was first observed. Many inflammatory cytokines and chemokines are released by leukocytes, which are angiogenic and fibrogenic,
could be reduced in Smad7 wounds as a result of reduced leukocyte infiltration. As a result, early-stage angiogenesis and collagen production required for wound repair were not perturbed, whereas prolonged angiogenesis and collagen production were prevented in Smad7 tg wounds. These changes could possibly accelerate wound stromal remodeling and prevent excessive scarring due to unresolved inflammation or collagen overproduction.
In summary, we found that temporal induction of Smad7 overexpression in wounded epidermis improves multiple aspects of cutaneous wound healing. Our current study also highlights the importance of future studies to evaluate whether temporal Smad7 induction in skin wounds can be used as a therapeutic strategy for impaired wound healing related to defects in keratinocyte migration, excessive inflammation and scarring.
Acknowledgments
We thank Pamela Garl for proofreading the manuscript.
References
Martin P.
Wound healing: aiming for perfect skin regeneration.
Expression of a dominant-negative type II transforming growth factor beta (TGF-beta) receptor in the epidermis of transgenic mice blocks TGF-beta-mediated growth inhibition.
p38 and ERK1/2 coordinate cellular migration and proliferation in epithelial wound healing: evidence of cross-talk activation between MAP kinase cascades.
Topical vascular endothelial growth factor accelerates diabetic wound healing through increased angiogenesis and by mobilizing and recruiting bone marrow-derived cells.
Targeted deletion of the integrin beta4 signaling domain suppresses laminin-5-dependent nuclear entry of mitogen-activated protein kinases and NF-kappaB, causing defects in epidermal growth and migration.
8-methoxypsoralen plus ultraviolet A therapy acts via inhibition of the IL-23/Th17 axis and induction of Foxp3+ regulatory T cells involving CTLA4 signaling in a psoriasis-like skin disorder.
Supported by National Institutes of Health grants GM70966, CA79998, and CA87849 to X.-J.W. G.H. is a recipient of the Dermatology Foundation Research Fellowship award.