| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
From the Department of Oral Biological and Medical Sciences,* Laboratory of Periodontal Biology, Faculty of Dentistry, University of British Columbia, Vancouver, Canada; Research Pathology Department, Biogen, Inc. Cambridge, Massachusetts; the Department of Dermatology and Helsinki University Central Hospital, University of Helsinki, Helsinki, Finland; Genomic Pharmacology, Millenium Pharmaceuticals, Cambridge, Massachusetts; Viikki Transgenic Unit,|| University of Helsinki, Helsinki, Finland, the Department of Surgery,** Turku University Central Hospital, Turku, Finland; and the Department of Biology, University of Jyväskylä, Jyväskylä, Finland
 |
Abstract |
|---|
 Top Abstract Materials and MethodsResultsDiscussionReferences |
|---|
vß6 is an epithelial cell-specific receptor that is not normally expressed by resting epithelium but its expression is induced during wound healing. The function of vß6-integrin in wound repair is not clear. In the present study, we showed that ß6-integrin expression was strongly up-regulated in the epidermis in human chronic wounds but not in different forms of skin fibrosis. To test whether increased ß6-integrin expression plays a role in abnormal wound healing we developed four homozygous transgenic mouse lines that constitutively expressed human ß6-integrin in the epithelium. The mice developed normally and did not show any histological abnormalities in the skin. The rate of experimental skin wound closure was unaltered and the wounds healed without significant scar formation. However, during breeding program 16.1 to 27.0% of transgenic mice developed spontaneous, progressing fibrotic chronic ulcers. None of the wild-type animals developed these lesions. The chronic lesions had areas with severe fibrosis and numerous activated macrophages and fibroblasts expressing transforming growth factor (TGF)-ß. The level of TGF-ß1 was significantly increased in the lesions as compared with normal skin. The findings suggest that increased vß6-integrin in keratinocytes plays an active part in abnormal wound healing possibly through a mechanism involving increased activation of TGF-ß.
Cell adhesion receptors of the integrin family play a major role in the regulation of keratinocyte function and in wound healing. Integrins are heterodimeric adhesion molecules that are composed of an - and ß-subunit. Integrins play a role in initiating intracellular signaling cascades in response to extracellular ligands, participate in growth factor activation, and collaborate with growth factor receptor signaling.6
On wounding, integrin expression by keratinocytes is up-regulated and they also start to express novel integrins, including vß6.6-8
The function of vß6-integrin in wound healing is not known. In cell culture, vß6-integrin facilitates cell adhesion and migration on fibronectin, vitronectin, and tenascin that are components of the early wound provisional matrix.9-13
Early expression of vß6-integrin during keratinocyte migration into wound suggests that vß6-integrin may regulate this process also in vivo.8,13-15
However, ß6-integrin-deficient mice did not show any change in wound closure rate.16,17
Although vß6-integrin is expressed by migrating keratinocytes in early wounds, the maximal expression coincides with basement membrane and granulation tissue formation when the migrating edges of the wound epithelium have joined.8,13,18
The temporal localization of vß6-integrin in epithelium suggests that vß6-integrin may also be involved in epithelial cell differentiation, regeneration of basement membrane, regulation of inflammatory reaction, and formation of granulation tissue. Interestingly, integrin vß6 can localize transforming growth factor (TGF)-ß1 to cell surface by binding to the latency-associated peptide leading to TGF-ß activation presumably by a nearby cell.19
TGF-ß1 plays an important role in wound repair by regulating re-epithelialization, suppressing inflammation, and promoting connective tissue regeneration and scar formation.20
In the absence of vß6-integrin there seems to be a deficiency of TGF-ß1 activation by the epithelium leading to exaggerated inflammation in response to injury or infection. Transgenic mice with targeted deletion of ß6-integrin developed excess lung inflammation that was reversed by restoring ß6 expression,11
and they were compromised in their ability to mount an inflammatory response to experimental infection in the gut.21
Activated TGF-ßs may also regulate other cellular functions critical for wound healing, including integrin expression, cell migration, proliferation, differentiation and, matrix synthesis and degradation, in an autocrine or paracrine manner.22-24
Chronic, nonhealing wounds are characterized by deficient re-epithelialization and excess formation of granulation tissue. This is associated with ongoing inflammation that maintains the chronic state and prevents healing.25
On the other hand, hypertrophic scars and skin fibrosis result from excess healing response and associate with increased TGF-ß1 expression and activation. This leads to enhanced extracellular matrix production by fibroblasts and excess matrix accumulation.26,27
We hypothesized that prolonged expression of vß6-integrin expressed by keratinocytes may lead to abnormal healing. To this end, we studied the expression of vß6-integrin in human aberrant wound healing (chronic wounds and hypertrophic scars) and skin fibrosis. The findings showed that in human chronic skin wounds, but not in hypertrophic scars or in skin fibrosis, expression of vß6-integrin was up-regulated in the epithelium. The vß6-integrin-expressing keratinocytes were in close proximity with activated fibroblasts in the underlying connective tissue. To study the function of vß6-integrin in wound healing, we created transgenic mice harboring human ß6-integrin gene under the control of cytokeratin 14 promoter to target constitutive expression of vß6-integrin in the epidermal basal cells. The homozygous mice showed normal wound repair speed, and wound healing did not result in scar formation. However, during breeding of the colony, the transgenic animals developed (unlike wild-type mice) spontaneous chronic wounds that were surrounded by progressing fibrosis. These lesions were characterized by numerous activated fibroblasts and macrophages and abundant expression of TGF-ß in connective tissue underneath the vß6-integrin-expressing epithelial cells. In chronic lesions from transgenic animals, the level of TGF-ß1 was significantly increased. These findings suggest that epithelial integrin vß6 plays a role in abnormal wound healing possibly through its ability to activate TGF-ß and thereby control cell recruitment and matrix deposition.
|
Materials and Methods |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
A 2.6-kb cDNA containing the entire coding region for human ß6-integrin was subcloned into pCDNAIneo plasmid.28 The construct was linearized with PvuII so that a 505-bp anti-sense probe corresponding to nucleotides 2139 to 2644 was obtained using sp6 polymerase whereas linearization with SspI and transcription with T7 gave a 721-bp sense probe corresponding to nucleotides 211 to 932. The specificity of the probe was determined by sequencing.28 The production and specificity of the anti-sense human stromelysin-2 probe used as positive control in the experiments has been described.29 The cDNAs were transcribed in vitro using a commercial kit (Promega Corp., Madison, WI) and labeled with 35S-UTP, as previously described.30
Samples of fibrotic and chronic nonhealing wounds and normal skin wounds were obtained from patients of the Department of Dermatology, University of Helsinki, Helsinki, Finland. The procedures were approved by the corresponding ethical committee. The tissue biopsies were fixed in 10% formalin and processed for paraffin embedding. In situ hybridization was performed as described in detail elsewhere.31 After deparaffinization and rehydration, 5-µm sections cut from paraffin blocks were pretreated with proteinase K and washed in 0.1 mol/L of triethanolamine buffer containing 0.25% acetic anhydride. The sections were covered with 25 to 70 µl of hybridization buffer containing 4 x 104 cpm/L of 35S-labeled anti-sense or sense RNA probe and incubated at 50 to 55°C in a humidified chamber for 18 hours. After hybridization, the slides were washed under stringent conditions, including treatment with RNase A to remove unhybridized probe. After 30 to 50 days of autoradiography, the photographic emulsion was developed, and the slides were stained with hematoxylin and eosin (H&E). All samples were processed in at least two experiments. The samples were examined under dark-field illumination and the bright-field view of a microscope. Samples previously positive for stromelysin-2 (chronic human wounds) were used as positive controls. No signal was detected with the sense ß6-integrin probe.
Transgenic Mouse Construction
The entire (2.6 kb) coding region of human ß6-integrin cDNA28 was inserted into pUC19 plasmid downstream of the cytokeratin 14 promoter to generate pUC19K14-hß6 plasmid. The 5.0-kb K14-hß6 expression cassette was then removed from the plasmid by digestion with XhoI and SalI. The expression cassette was injected into fertilized oocytes from FVB/N F0 mice.32 Founders harboring the transgene in their genome were crossbred back to establish lines of wild-type, homozygous, and heterozygous animals. Animals were kept in an established 12 hour:12 hour light:dark cycle in the animal facility.
Southern Blotting
We used Southern blotting to screen for the presence of the transgene in F1 or later generation animals and to determine the copy number of the transgene. Genomic DNA from mouse tail snips was isolated and digested with BamHI, which cuts twice to yield a 2.6-kb band. Digested DNA (4 µg) was electrophoresed on 1% agarose gel, blotted onto nylon, and probed with a human ß6-integrin probe consisting of a 2.6-kb NotI fragment of the human ß6-integrin insert. Copy numbers were determined as previously described.32
Polymerase Chain Reaction
Newborn mice were euthanized with inhalation of CO2, and their skins were collected and homogenized on liquid nitrogen. Total RNA was isolated using Trizol (Life Technologies, Inc., Gaithersburg, MD) followed by digestion with DNase I and purification with the RNeasy kit (Qiagen, Valencia, CA). A reverse transcriptase-polymerase chain reaction kit (Life Technologies, Inc.) was used to amplify 0.25 µg of total RNA to cDNA using two primers specific for human ß6-integrin to yield a 245-bp band: 5'-TCTGGAGTTGGCGAAAGG-3' (5' primer) and 5'-TCCACCGGGTAGTCCTCA-3' (3' primer). Amplified DNA fragments were analyzed by electrophoresis through 0.8% agarose gel. As a positive control, total RNA isolated from human keratinocytes treated with TGF-ß1 was used.12 Negative control was RNA from the skin from homozygous animals that was not reverse-transcribed.
Excisional Wound Healing in Transgenic Mice
All animal studies were conducted in compliance with Canadian Council on Animal Care guidelines and approved by the University of British Columbia Animal Care Committee. Mice were anesthetized with inhalation of halothane (MTC Pharmaceuticals, Cambridge, Ontario, Canada) and the dorsal skin shaved and then depilated (Nair Cantor Horner, Inc., Mississauga, Ontario, Canada). Four full-thickness, 4-mm bunch biopsy wounds were created on the dorsal surface of age- and sex-matched adult (more than 3 months old) wild-type and homozygous mice (ß6F1, ß6F2, ß6F3, and ß6F4). The normal skin was collected for histology and immunohistochemistry. At days 1, 3, 4, 6, 10, 12, 14, 28, 30, 44, 58, and 210 after wounding six to eight animals in each group were euthanized by CO2 inhalation, and the wounds including 5 mm of surrounding unwounded skin, as well as samples of normal tail and ear skin were collected.
At the time of wounding and at days 3, 4, 8, and 12 after wounding, standardized images of 8 to 16 wounds in wild-type and ß6F1-transgenic animals were recorded using a digital camera for analysis of wound closure rate. The surface areas of the wounds were quantitated by using NIH Image software.
Wound Breaking Strength
The wound strength was measured by using a previously described method with slight modifications.33 Briefly, for each time point, a 20-mm long full-thickness incisional wound was made to the standardized area of the mid-back in 8 to 14, 4-month-old, sex-matched wild-type and hß6F1 mice, and the wounds were closed using two surgical clips. After 4, 7, and 14 days the mice were sacrificed and the entire dorsal pelt was excised and cut into a standardized transverse strip perpendicular to the original wound. Unwounded skin collected as above served as control. Perpendicular breaking strength of the wounded or unwounded skin was then measured using an Instron 4301 tensometer with a 500-N maximum load cell (Instron Corp., Canton, MA) and analyzed using Series IX (version 8.12) software (Instron).
Histochemistry and Immunohistochemistry
TGF-ß1, -2, and -3 were localized with a monoclonal antibody recognizing the active forms of TGF-ß1, -ß2, and -ß3 (R&D Systems, Inc., Minneapolis, MN). Myofibroblasts were localized with a monoclonal antibody against -smooth muscle actin (1A4; Sigma Chemical Co., St. Louis, MO). To localize macrophages, sections were stained with a monoclonal antibody F4/80 (Serotec Inc., Raleigh, NC). Activated fibroblasts in human paraffin sections were stained with a rat antibody to N-terminus of the type I procollagen (PC-1) molecule (mAb 1912; Chemicon, Temecula, CA). In frozen mouse sections, human ß6-integrin was localized with a monoclonal rabbit antibody (B1) against human vß6-integrin.11
This antibody recognizes both the human and mouse ß6-integrin subunit. Cytokeratin 14 was stained with a monoclonal anti-CK14 antibody (LL002, Serotec, Inc.). To localize v integrin subunits in mouse tissues, a polyclonal anti-v antibody was used (Ab 1930, Chemicon).
Immediately after collection, fresh tissue samples were mounted in Histoprep (Fisher Scientific, New Jersey, NJ) and snap-frozen in liquid nitrogen. Frozen sections (6 µm) were cut and fixed with -20°C acetone for 5 minutes and stored at -70°C until used for staining. A set of samples were fixed for 2 hours with 10% formalin and processed for paraffin embedding. After deparaffinization and rehydration, 5-µm sections cut from paraffin blocks were used for staining.
Immunostaining was performed using the avidin-biotin-peroxidase complex or immunofluorescence techniques.13
The sections were incubated in 0.3% hydrogen peroxide in methanol for 30 minutes and washed in phosphate-buffered saline (PBS) containing 0.1% bovine serum albumin (BSA; Sigma Chemical Co.). Sections were rinsed and incubated with normal blocking serum (Vectastain; Vector Laboratories, Burlingame, CA) for 60 minutes at room temperature and then incubated with the primary antibody in PBS/BSA in a humid chamber at 4°C for 16 hours. To stain macrophages, slides were pretreated with Proteinase K (Sigma) 20 µg/ml in 50 mmol/L of PBS (pH 7.4) for 5 minutes at room temperature before staining. After washing with PBS/BSA, sections were incubated with an appropriate biotinylated secondary antibody for 60 minutes and then reacted with ABC avidin/peroxidase reagent (Vectastain Elite kit, Vector Laboratories). Chromogen was developed using diaminobenzidine (Chemicon) or VIP (Vector Laboratories) as chromogenic substrate and hematoxylin as counterstain. Reactions were monitored until suitable color development was achieved and then washed with distilled water. The sections were mounted using Permount (Fisher Scientific, Pittsburgh, PA). After immunostaining with the anti-v integrin antibody, sections were incubated with a fluorescent secondary antibody (Alexa 546; Molecular Probes Inc., Eugene, OR) and mounted with cyanoacrylate glue (Krazy Glue). Control sections were stained as above but the incubation with the primary antibody was omitted.
A set of paraffin sections of mouse skin samples were stained with H&E, phosphotungstic acid hematoxylin, Gomori’s methenamine silver, periodic acid-Schiff, or acid fast stains using standard methods.
Analyses of TGF-ß
Viscose cellulose sponges (Cellomeda Oy, Turku, Finland) were used as an inductive matrix for connective tissue formation.34,35
Two pieces of sponge (0.5 x 0.5 cm) were implanted into the dorsal skin of each of the control animals and ß6F1- to ß6F4-transgenic mice (four age- and sex-matched adult animals in each group). Sponges were removed after 21 or 36 days and subjected to analysis for TGF-ß. Normal skin from wild-type transgenic animals was used as a control. The amount of biologically active TGF-ß in granulation tissue sponges was studied as described previously using the plasminogen activator inhibitor-I/luciferase assay.36
For this purpose, cellulose sponges were homogenized on liquid nitrogen. TGF-ß was extracted by eluting the tissue homogenates (250 mg of tissue/ml) with serum-free Dulbecco’s modified Eagle’s medium containing glutamine (2 mmol/L), pyrogen-free BSA (0.2%), aprotinin (1 µg/ml), and antibiotics at 37°C for 20 hours. The tissue debris was then removed by centrifugation. For TGF-ß assays, two samples from each eluate were used. One sample was analyzed as such representing the proportion of active TGF-ß in tissue. The other sample representing the total amount (active plus latent) of TGF-ß present in the tissue was heated at 80°C for 5 minutes to activate the latent form of TGF-ß. The granulation tissue eluate samples were added in duplicate to mink lung epithelial cell culture that were stably transfected with TGF-ß-inducible PAI-1 promoter fused to luciferase reporter gene37
and incubated overnight in the elution medium at 37°C. Cells were washed and lysed in Promega passive lysis buffer (Promega, Madison, WI). The lysates were diluted with PBS containing BSA (1 mg/ml) and transferred to a 96-well luminometer plate. An equal volume of Bright-Glo luciferase reagent (Promega) was added to the samples, and the luciferase activity was determined using a microtiter plate luminometer (Polarstar/Fluostar Optima; BMG Labtechnologies Inc., Durham, NC). The assay system is specific for TGF-ß1, -2, and -3 and the detection limit is 
5 to 10 pg/ml.36
To compare the levels of TGF-ß1 in skin, normal skin samples were collected from six wild-type animals, from eight ß6-integrin-overexpressing mice (line ß6F1, n = 4; line ß6F2, n = 4) and from chronic lesions from six ß6-integrin-overexpressing mice (line ß6F1, n = 3; line ß6F2, n = 3). The samples were immediately frozen in liquid nitrogen and stored at -80°C until used. For the analyses, the samples were homogenized by grinding over liquid nitrogen. The homogenized tissue was weighted, and the soluble TGF-ß was eluted by adding a fixed weight-per-volume ratio of Dulbecco’s modified Eagle’s medium containing 1 µg/ml aprotinin, 0.1% pyrogen-free BSA, and antibiotics (50 µg/ml streptomycin sulfate, 100 U/ml penicillin) and incubating at 37°C for overnight with occasional vortexing. Eluents were cleared of tissue debris by centrifugation. TGF-ß1 in the clear supernatants was activated by acidification and quantitated with Quantikine TGF-ß1 immunoassay using human TGF-ß1 as a standard (R&D Systems).
Statistical Analysis
The wound breaking strength of wild-type and transgenic mice was compared using Student’s t-test. One way analysis of variance was used to compare wound surface areas, TGF-ß levels, and prevalence of chronic lesions from different groups of animals. The significance of difference between samples from ß6-transgenic mice and wild-type mice was analyzed using Bonferroni’s post test. P values <0.05 were considered significant.
|
Results |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
To analyze whether expression of ß6-integrin associates with fibrosis or chronic wounds, we analyzed expression of ß6-integrin in human samples by in situ hybridization. In keloids, hyperthrophic scars, scleroderma, and a case of dermatofibroma, no ß6-integrin expression was demonstrated in the epithelium overlying the lesion (Table 1)
. However, in 85.7% of samples from nonhealing, chronic wounds, ß6-integrin was expressed by the epithelial cells (Table 1)
. These lesions were histologically characterized by a moderately inflamed open ulcer surrounded with hypercellular, fibrotic connective tissue. In both areas, the epithelium was hyperplastic. Most abundant expression of ß6-integrin mRNA localized to the basal keratinocytes at the edge of the open wound and overlying fibrotic, hypercellular connective tissue distant from the open wound (Figure 1)
. In these lesions, ß6-expressing keratinocytes were localized next to fibroblasts that showed strong intracellular immunoreactivity for type I procollagen (Figure 1)
. In acute dermal wounds, ß6-integrin was first expressed 3 days after wounding and remained up-regulated for at least 9 days (Table 1)
. The most abundant expression localized to the basal epithelial cells of the migrating epithelium in early wounds and to the basal and few suprabasal layers in the joined wound epithelium (data not shown).
|
|
To study whether increased expression of ß6-integrin leads to abnormal wound healing we developed transgenic mouse lines that constitutively express human ß6-integrin (hß6) in the basal keratinocytes. To this end, fertilized mouse oocytes were microinjected with the K14-hß6 construct. Screening of progeny by Southern blotting revealed that six of them carried the hß6 gene. The six founder mice were mated with normal mice to determine stability of the transgene. Four founder mice produced progeny that were heterozygous for the hß6-integrin. The transgene was inherited autosomally. Mating of heterozygous animals produced homozygous animals that were further bred to produce four mouse lines (ß6F1 to ß6F4) that were homozygous for hß6-integrin. The K14-hß6 transgene in each mouse line was screened by Southern blotting to confirm the homozygous inheritance and determine the gene copy numbers (Figure 2A)
. Expression of hß6-integrin in newborn mouse skin was also confirmed by reverse transcriptase-polymerase chain reaction (Figure 2B)
. The expression of hß6 was further studied using immunostaining and in situ hybridization (Figure 3)
. Expression of hß6 gene and positive immunoreactivity for hß6-integrin was found in the back skin of all ß6F1 to ß6F4 animals. The relative expression level was highest in the ß6F1 line in which the copy number of the hß6 gene was also the highest (Figure 2)
. The expression of cytokeratin 14 and hß6-integrin co-localized in the epidermis and at the outer root sheath cells of hair follicles (Figure 3)
. In the epidermis, hß6 immunostaining localized to the cell membranes of basal and one to two suprabasal cell layers. Immunostaining of v integrin co-localized with hß6-integrin at the basal epithelial cells (Figure 3)
. In addition to various locations of skin, basal cells of tongue epithelium were positive for hß6 expression (data not shown). Control animals did not express ß6-integrin in the epidermis but some expression in the cells of the outer root sheath of hair follicles was noted (Figure 3)
. The expression of the transgene did not have any apparent effect on the development, weight, behavior, or fertility of the animals. Histologically, the skin in newborn and adult transgenic mice did not show any alterations as compared with wild-type animals (Figure 3)
.
|
|
To study whether constitutive expression of ß6-integrin in skin affects wound healing, we created excisional wounds to the back skin of the transgenic (ß6F1 to ß6F4) and normal animals. Wounds in all animals were completely closed at day 12 after wounding and no difference was found in the relative wound areas at different time points (Figure 4A)
or in the wound breaking strength between wild-type and human ß6-integrin-transgenic mice (Figure 4B)
. In histological examination, no differences in the re-epithelialization speed, inflammatory reaction, or reorganization of connective tissue was observed (Figure 5)
. Both wild-type and transgenic animals developed collagen-rich wound connective tissue that was histologically evident by 30 days after wounding (Figure 5)
. Macroscopically, the skin in both transgenic and wild-type animals healed without hypertrophic scar formation. The results were consistent among different human ß6-integrin-transgenic mouse lines.
|
|
. There was also a weak positive immunoreaction at the cell membranes of suprabasal cells (Figure 6A)
. Seven days after wounding when the migratory epithelial fronts had just joined to cover the wound, basal epithelial cells still expressed ß6-integrin (Figure 6C)
. At day 14, the expression of ß6-integrin was strongly down-regulated and no expression was seen at the 28- or 44-day-old wound epithelium of the wild-type mice (Figure 6
; E, G, and I). In contrast, expression of ß6-integrin was strongly increased at all layers of the early migrating epithelium in the transgenic animals 3 days after wounding (Figure 6B)
. Unlike in the control animals, basal and some suprabasal epithelial cells expressed ß6-integrin still strongly from day 14 until day 28 after wounding (Figure 6, F and H)
. After 44 days the wound epithelium was morphologically identical with unwounded epithelium and expressed abundantly ß6-integrin similar to unwounded epithelium of the transgenic animals (Figures 6J and 3A
, respectively).
|
Development of Chronic Lesions in Transgenic Mice
During breeding of the colony of transgenic ß6 mice, they started to develop spontaneous chronic lesions (Figure 7)
. The lesion usually started to develop at 10 to 20 weeks of age and was initially characterized macroscopically by hair loss in the lower back area (Figure 7A)
. Throughout time (usually within 22 to 72 weeks) this affected area of skin slowly developed into a fibrotic lesion that also had areas with open, superficial chronic ulcerations. The lesion was progressive and no regression was observed. The mice did not usually develop multiple lesions but the single lesions could become relatively large in size (Figure 7A)
. The fibrotic area of skin was very thick and progressed often to the abdominal area where it caused constriction (Figure 7A)
. The prevalence of the chronic lesions was followed for up to 2 years. The wild-type mice did not develop chronic lesions (Figure 7B)
. Three wild-type mice showed very mild cases of chronic lesions involving some hair loss and some thickening of the skin, but these lesions did not progress to form constricting fibrosis. In the transgenic animals, 16.1% (ß6F3) to 27.0% (ß6F1) of the animals developed chronic lesions. The prevalence was the highest in males of ß6F1 mice (50.0%, Figure 7B
). The difference in the prevalence of chronic lesions was significantly higher (P < 0.0001, Student’s t-test) in human ß6-integrin-transgenic mice (38 of 173 hß6F1, hß6F2, hß6F3, and hß6F4 mice had lesions) as compared with wild-type mice (none of 64 mice had lesions). When the human ß6-integrin-transgenic mouse lines were compared individually against the wild-type mice, hß6F1 (P < 0.01), hß6F2 (P < 0.05), and hß6F4 (P < 0.01) mouse lines showed significantly increased prevalence of chronic lesions (analysis of variance). The difference in the prevalence of the lesions between hß6F3 mouse line and wild-type mice was not statistically significant (P > 0.05, analysis of variance).
|
or focal hyperkeratosis and granulation tissue indicative of recent and resolving ulceration. Stainings of representative lesions with acid fast did not reveal acid fast bacilli in the lesions (data not shown). Likewise, Gomori’s methenamine silver and periodic acid-Schiff stains were negative for microorganisms and fungi in the dermis (data not shown). The ulcerations were surrounded by hyperplastic epithelium and strongly thickened connective tissue (Figures 8 and 9)
. The epithelium covering these fibrotic areas and at the edge of open wounds expressed vß6-integrin in the basal and suprabasal epithelial cells (Figure 9)
. Strong localization of ß6-integrin mRNA was also demonstrated at the basal epithelial cells. In connective tissue surrounding the ulcerations, all lesions showed a vast expansion of the papillary and reticular dermis with a mixed infiltrate of activated fibroblasts and abundant macrophages (Figure 9)
. Macrophage infiltration was evident in some cases as deep as the subcutaneous fat. The macrophages in the dermis were diffusely scattered between fibroblasts and were not immediately obvious by H&E staining. The connective tissue next to the ulcerations showed abundant collagen accumulation (Figures 8 and 9)
and contained fibroblasts that showed positive immunoreactivity for -smooth muscle actin (Figure 9)
. Immunostaining with an antibody against TGF-ß1, -ß2, and -ß3 isoforms localized in the cytoplasms of connective tissue cells in the fibrotic areas with heavy collagen accumulation. We then analyzed levels of total TGF-ß1 in the healthy skin in wild-type and ß6F1- and ß6F2-transgenic animals as compared with the chronic scars from the same ß6-integrin-overexpressing mouse lines by using the TGF-ß1 immunoassay. The chronic lesions from ß6-integrin-overexpressing mice contained significantly higher levels of TGF-ß1 than healthy skin from transgenic and wild-type animals (Figure 10)
. No difference in the level of TGF-ß1 was found between healthy skin from wild-type and transgenic animals (Figure 10)
.
|
|
|
|
Discussion |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
vß6-integrin associated with chronic human skin wounds but not with keloids, hypertrophic scars, scleroderma, or a single case of dermatofibroma. This suggested to us that expression of vß6-integrin was either induced secondarily because of chronic state of the wound or vß6-integrin could be actually a contributing factor for the development of these lesions. The chronic wounds were characterized by open ulcers that were surrounded by hypercellular, fibrotic connective tissue. In these lesions, expression of ß6-integrin was localized at the basal epithelial cells. Underneath the ß6-integrin-expressing keratinocytes, strong immunoreactivity for type I procollagen localized at the cytoplasms of fibroblasts suggesting that the cells were actively secreting collagen. Integrin vß6 can bind latent TGF-ß to the cell membrane and induce its activation.19
TGF-ß can potently stimulate recruitment and growth of connective tissue cells and induce collagen deposition by fibroblasts.2
Therefore, it is possible that vß6-integrin-expressing keratinocytes actively modulate connective tissue cell functions in chronic wounds through activation of TGF-ß and therefore actively contribute to the development of the lesions.
To test whether vß6-integrin plays an active role in abnormal wound healing, we developed transgenic mouse lines that constitutively expressed human ß6-integrin under the control of cytokeratin 14 promoter. The localization of human ß6-integrin expression in the basal and some suprabasal epidermal cell layers and in outer root sheath cells of the hair follicles was consistent with previous reports of cytokeratin 14 promoter expression.38
The hß6 integrin was localized at the cell membranes of keratinocytes and co-localized with endogenous v integrin. The formation of a heterodimer from v- and ß6-integrin subunits is the prerequisite for the assembly of a functional, integral cell membrane vß6-integrin receptor.39
The transgenic animals did not show any developmental abnormalities and histologically the skin in newborn and adult animals was unaltered as compared with wild-type mice. Additionally, the speed of normal wound healing was unchanged as compared with wild-type mice. In human cutaneous and mucosal wounds, vß6-integrin is up-regulated during early keratinocyte migration and remains up-regulated at least until day 14 when re-epithelialization is complete.8,13-15
As in human wounds, ß6-integrin expression was induced by basal epithelial cells of the migrating epithelium at day 3 but strongly down-regulated after 14 days of healing. Transgenic mice showed, however, increased expression of ß6-integrin in all cell layers of migrating epithelium at day 3 and the expression remained high throughout the healing period. Thus, although the endogenous ß6-integrin was down-regulated in wound epithelium within 14 days, the transgene expression remained high at the wound epithelium during later phases of wound healing. It seems obvious, therefore, that the different expression levels of vß6-integrin do not modulate normal wound healing because ß6 -/-16
or ß6-overexpressing (our study) animals do not show abnormal wound healing response. It is possible that vß6-integrin plays a role only when wound healing is compromised by infection, immunosuppression, or other conditions as suggested by our data from chronic human wounds.
Interestingly, when the mouse colonies were bred, 16.1 to 27.0% of the transgenic mice developed spontaneous chronic wounds. Throughout the time these lesions developed into progressing, fibrotic scars with areas of nonhealing open ulcerations. In an advanced stage, the fibrosis caused constriction of the scar tissue that often progressed to the abdominal region at the time when the experiment was terminated. We noted that when transgenic or control mice were caged together, they occasionally injured each other to the skin in the lower back during territorial fighting. Therefore, wounding as a result of territorial fighting may have contributed to the development of the chronic wounds in the ß6-integrin-overexpressing mice. All of the chronic lesions appeared to be associated with ulcerative trauma and, therefore, it is possible that superficial infection contributed to the lesions. However, no microorganisms or fungi were present in the histological sections from the lesions.
In mice, bleomycin treatment leads to lung fibrosis because of increased activation of TGF-ß.19
The vß6-integrin plays an active role in this process because in mice lacking ß6-integrin, bleomycin failed to induce lung fibrosis.19
In culture, the ß6-integrin -/- cells could not activate latent TGF-ß1,19
and the ß6-integrin-deficient mice did not show the increased expression of TGF-ß1-regulated genes as was detected in the wild-type animals.40
Therefore, it is possible that constitutive expression of vß6-integrin by keratinocytes provides a mechanism that promotes TGF-ß1 activation also in the skin that in turn could lead to altered wound healing response. Increased expression of TGF-ß1 has been shown to associate with reduced rate of wound re-epithelialization and increased collagen deposition.40
In our study, closure speed of experimental wounds was unaltered in the transgenic animals, and the wounds did not heal by excessive scar formation. Additionally, the levels of active and latent TGF-ß1, -2, and -3 in subcutaneously implanted cellulose sponges from 21 and 36 days after wounding were similar in the ß6-integrin-overexpressing and wild-type mice. Therefore, activation of TGF-ß by vß6-integrin may not play as important role in normal skin wound healing as in bleomycin-induced lung fibrosis. However, in a compromised healing situation, as in the case of development of chronic lesions in this study, activation of TGF-ß by vß6-integrin may play a role. The chronic lesions in this study were composed abundantly of activated macrophages and fibroblasts. Many of the cells underneath the vß6-expressing epithelium showed positive intracellular staining with an antibody recognizing TGF-ß1, -2, and -3. Additionally, the level of TGF-ß1 in the lesions of transgenic animals was significantly higher than in healthy skin. There were also many myofibroblasts present in a collagen-rich extracellular matrix. It is possible that in chronic lesions, keratinocytes constitutively expressing vß6-integrin could activate TGF-ß leading to induction of fibroblast differentiation into myofibroblasts. These activated fibroblasts could then produce more TGF-ß that would stimulate collagen deposition. Additionally, one possibility is that in the chronic lesions, increased TGF-ß at the dermoepidermal junction causes a sufficient increase in MCP1 secretion by dermal fibroblasts41
leading to situation in which induced macrophage influx fails to reach a normal equilibrium. However, we cannot rule out the possibility that increased TGF-ß in the lesions is a result of chronic inflammation and wound environment and not directly linked to increased activation of TGF-ß by vß6-integrin-expressing keratinocytes.
Because the transgenic mice, unlike wild-type mice, developed chronic progressing scars and ulcerations, we suggest that it is possible that the expression of vß6-integrin by the epithelium contributes to the development of the chronic lesions. The idea that epithelial cells could play a significant role in the development of lesions in the skin has been proposed previously. For example, suprabasal integrin expression resulted in development of psoriasis-like changes in transgenic mice,4
and excess production of latent TGF-ß1 by keratinocytes enhances collagen production by connective tissue cells.42
Based on the previous findings with ß6-integrin knockout animals16
and on our study, we hypothesize that epidermal vß6-integrin is not critical for normal wound repair but it may play a role when wound healing is compromised for example in conditions involving chronic inflammation or immunosuppression.
|
Acknowledgements |
|---|
|
Footnotes |
|---|
Supported by the Canadian Institutes for Health Research (grant MOP-12589).
The present address of M. L. is the A. I. Virtanen Institute, University of Kuopio, Kuopio, Finland; the present address for U.S-K. is the Department of Dermatology, Karolinska Institute, Soder Hospital, Stockholm, Sweden; and the present address for H.R. is Department of Biosciences and Institute of Biotechnology, Neuroscience Center, University of Helsinki, Finland.
Accepted for publication September 29, 2003.
|
References |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
9 integrin subunit and vß6 integrin during wound healing in human oral mucosa. J Histochem Cytochem 2000, 48:985-998This article has been cited by other articles:
 |
|
 |
|
  T. Tabata, H. Kawakatsu, E. Maidji, T. Sakai, K. Sakai, J. Fang-Hoover, M. Aiba, D. Sheppard, and L. Pereira Induction of an Epithelial Integrin {alpha}v{beta}6 in Human Cytomegalovirus-Infected Endothelial Cells Leads to Activation of Transforming Growth Factor-{beta}1 and Increased Collagen Production Am. J. Pathol., April 1, 2008; 172(4): 1127 - 1140. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. J. Santulli, W. A. Kinney, S. Ghosh, B. L. DeCorte, L. Liu, R. W. A. Tuman, Z. Zhou, N. Huebert, S. E. Bursell, A. C. Clermont, et al. Studies with an Orally Bioavailable {alpha}V Integrin Antagonist in Animal Models of Ocular Vasculopathy: Retinal Neovascularization in Mice and Retinal Vascular Permeability in Diabetic Rats J. Pharmacol. Exp. Ther., March 1, 2008; 324(3): 894 - 901. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. S. Horan, S. Wood, V. Ona, D. J. Li, M. E. Lukashev, P. H. Weinreb, K. J. Simon, K. Hahm, N. E. Allaire, N. J. Rinaldi, et al. Partial Inhibition of Integrin {alpha}v 6 Prevents Pulmonary Fibrosis without Exacerbating Inflammation Am. J. Respir. Crit. Care Med., January 1, 2008; 177(1): 56 - 65. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. S. Agren and M. Werthen The Extracellular Matrix in Wound Healing: A Closer Look at Therapeutics for Chronic Wounds International Journal of Lower Extremity Wounds, June 1, 2007; 6(2): 82 - 97. [Abstract] [PDF]
|
|
|
|
 |
|
 D. DiCara, C. Rapisarda, J. L. Sutcliffe, S. M. Violette, P. H. Weinreb, I. R. Hart, M. J. Howard, and J. F. Marshall Structure-Function Analysis of Arg-Gly-Asp Helix Motifs in {alpha}vbeta6 Integrin Ligands J. Biol. Chem., March 30, 2007; 282(13): 9657 - 9665. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Hakkinen and A. Csiszar Hereditary Gingival Fibromatosis: Characteristics and Novel Putative Pathogenic Mechanisms J. Dent. Res., January 1, 2007; 86(1): 25 - 34. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Hahm, M. E. Lukashev, Y. Luo, W. J. Yang, B. M. Dolinski, P. H. Weinreb, K. J. Simon, L. Chun Wang, D. R. Leone, R. R. Lobb, et al. {alpha}v{beta}6 Integrin Regulates Renal Fibrosis and Inflammation in Alport Mouse Am. J. Pathol., January 1, 2007; 170(1): 110 - 125. [Abstract] [Full Text] [PDF]
|
|
|
|
