Animal experiments were performed in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health, which conforms to the provisions of the Declaration of Helsinki. The protocol was approved by the Ethics Committee of Fudan University, Shanghai, China. All surgerical procedures were performed under sodium pentobarbital anesthesia, and all efforts were made to minimize animal suffering.

Balb/c mice (Slac Laboratory Animal Co., Ltd., Shanghai, China) were deeply anesthetized before shaving the dorsal hair and cleaning the exposed skin with 70% ethanol. Subsequently, 5 mm full thickness punch wounds were created through the skin and the panniculus carnosus on the dorsal surface (one wound each on the left and right side) of each mouse. For further experiments, including immunohistochemical anlaysis, real-time PCR and western blot analysis, the healed tissues were removed en bloc with a 7 mm punch to include a margin of unwounded skin at the indicated time intervals; the unwounded back skin tissues removed at day 0 were used as controls [in the following results, the controls are presented as 'normal', EGF(−) day 0 or EGF(+) day 0].

A total of 60 Balb/c mice (8–9 weeks old) were randomly divided into 10 groups as follows: i) EGF(−) (wound model, but no EGF treatment) 1 day group; ii) EGF(−) 3 day group; iii) EGF(−) 7 day group; iv) EGF(−) 10 day group; v) EGF(−) 14 day group; vi) EGF(+) (wound model with EGF treatment) 1 day group; vii) EGF(+) 3 day group; viii) EGF(+) 7 day group; ix) EGF(+) 10 day group; and x) EGF(+) 14 day group. In the EGF(+) group, the wounds were treated with EGF (recombinant human epidermal growth factor derivative for external use, liquid (I), 2,000 IU/ml, S20010038; Shenzhen Watsin Genetech Co., Ltd. Shenzhen, China) once a day (spray once on each wound). In each group, 6 animals were included. At the end of the experiment, mice were sacrificed by cervical dislocation under anesthesia.

For immunohistochemical staining, fresh skin tissues were embedded by optimal cutting temperature (OCT) compound. Frozen sections of skin (10-µm-thick) were stained with rabbit polyclonal Kindlin-1 antibody (22215-1-AP; Proteintech Group Inc., Chicago, IL, USA) at a dilution of 1:100 overnight (4°C) and then incubated with Dako REAL™ EnVision™/HRP, Rabbit/Mouse (ENV) reagent (K5007; DAKO, Glostrup, Denmark). For tissue immunofluorescence staining (25,26), the frozen tissues were stained with F4/80 (a marker of macrophages), Keratin 6 (K6, a marker of keratinocytes), FSP1 (a marker of fibroblasts) and Kindlin-1 antibodies. For cell immunofluorescence staining, the cells were first fixed with 4% paraformaldehyde for 10 min at room temperature, rinsed with PBS and incubated in blocking solution for 60 min at room temperature. For Kindlin-1, activated integrin β1 [(integrin β1 (ac)] and phospho-FAK (p-FAK) labeling, the cells were incubated with the primary antibody for 2 h at room temperature, rinsed with PBS and incubated with the relevant secondary antibody for 1 h at room temperature; For actin staining, the cells were incubated with Phalloidin-iFluor 594 Conjugate (cat. no. 23122; 1:1,000; AAT Bioquest, Sunnyvale, CA, USA) followed by DAPI staining (Sigma, St. Louis, MO, USA). The following antibodies were used: rabbit polyclonal Kindlin-1 antibody (cat. no. 22215-1-AP; 1:50; Proteintech Group Inc.), mouse monoclonal activated integrin β1 antibody (cat. no. MAB2079Z; 1:300, Millipore, Bedford, MA, USA), mouse monoclonal Kindlin-1 antibody (cat. no. SAB4200465; 1:250; Sigma), rabbit monoclonal phospho-FAK (Tyr397) antibody (cat. no. EP2160Y; 1:400; Abcam, Cambridge, UK), DyLight 405 affinipure goat anti-rabbit IgG (H+L) (cat. no. A23120; 1:800; Abbkine, Los Angeles, CA, USA), DyLight 405 affinipure goat anti-mouse IgG (H+L) (cat. no. A23110; 1:800; Abbkine), Alexa Fluor 594 affinipure goat anti-mouse IgG (H+L) (cat. no. 115-585-003; 1:800; Jackson ImmunoResearch, West Grove, PA, USA), Alexa Fluor 594 affinipure goat anti-rabbit IgG (H+L) (cat. no. 111-585-003; 1:800; Jackson ImmunoResearch). Images were captured using a Nikon Eclipse 660 microscope (Nikon, Tokyo, Japan) or a confocal laser scanning microscope (Leica Microsystems, Wetzlar, Germany). A total of 6 samples from each group was selected. For each sample, at least 4 fields were captured.

The immortalized keratinocyte cell line, HaCaT, (KCB 200442YJ, Kunming Cell Bank of Chinese Academy of Science, Kunming, China), were kept in Dulbecco's modified Eagle's medium (DMEM), containing 10% fetal bovine serum and 1% penicillin/streptomycin. To inhibit the expression of Kindlin-1 in the HaCaT cells, shRNA targeting Kindlin-1 [EGFP-Kindlin-1 shRNA (hu)] lentiviral particles (Genechem, Shanghai, China), a pool of viral particles containing 3 target-specific constructs that encode 19 nt shRNA designed to knock down Kindlin-1 expression in the HaCaT cells, were transduced into the cells according to the manufacturer's instructions. Control lentiviral particles (Genechem) containing a scrambled shRNA sequence, were used as a negative control. The knockdown efficiency was confirmed by western blot analysis.

Real-time PCR was performed using the SYBR Primer Ex Taq™ II kit (Takara, Dalian, China). Data were obtained using the StepOne plus system (Applied Bio-systems, Foster City, CA, USA). The relative change in Kindlin-1 mRNA expression was determined by the fold change. The sequences of the primers used were as follows: Mouse Kindlin-1 sense primer (5′→3′), GGAGAGCAGCAGACAGAGAT; antisense primer (5′→3′), TGCTGAGGGGTGAAGAGAAG. Mouse GAPDH sense primer (5′→3′), GAGCGAGACCCCACTAACAT; antisense primer (5′→3′): TCTCCATGGTGGTGAAGACA.

Western blot analysis was carried out as previously described (25). Briefly, tissues and cell lysates were collected, and the protein concentrations were determined using a BCA™ Protein Assay kit (Pierce, Chemical Co., Rockford, IL, USA). Proteins were separated by 10% SDS-PAGE and electro-transferred onto PVDF membranes (Millipore). The membranes were first incubated with the indicated primary antibodies overnight (4°C), followed by HRP-conjugated secondary antibodies for 2 h (room temperature). Blots were visualized using an enhanced chemiluminescence detection kit (Tiangen Biotech Co., Ltd. Beijing, China). The following antibodies were used: mouse monoclonal Kindlin-1 antibody (cat. no. SAB4200465; 1:1,000, Sigma), rabbit monoclonal integrin β1 antibody (cat. no. NB110-57123; 1:1,000, Novus Biologicals, Littleton, CO, USA), mouse monoclonal activated integrin β1 antibody (cat. no. MAB2079Z; 1:1,000, Millipore), rabbit polyclonal FAK antibody (cat. no. 3285; 1:1,000, Cell Signaling Technology, Inc., Danvers, MA, USA), rabbit monoclonal Phospho-FAK (Tyr397) antibody (cat. no. 8556; 1:1,000, Cell Signaling Technology, Inc.), rabbit monoclonal GAPDH antibody (cat. no. 3683, 1:1,000; Cell Signaling Technology, Inc.). All experiments were repeated at least 4 times.

Fibronectin (FN, PHE0023; Gibco, Frederick, MD, USA), was used to induce cell adhesion. Briefly, dishes were coated with FN (10 µg/ml, 37°C) overnight. Detached serum-starved control HaCaT cells and Kindlin-1 shRNA-transfected HaCaT cells were stimulated with or without EGF (10 ng/ml) for 15 min and were then subjected to FN-coated dishes. Cells were harvested at 30 min, 1 h and 3 h, respectively, and FAK phosphorylation was detected by western blot analysis.

Migration assay was performed using Transwell cell culture systems (3422; Corning, NY, USA), as previously described (26). Briefly, the HaCaT cells were added into the upper chamber. In order to verify the role of FAK in HaCaT cell migration, normal cells were first treated with or without FAK inhibitor, PF573228 (S2013; Selleck, Houston, TX, USA) (1 µM, 30 min) and then treated with or without recombinant human EGF (236-EG; R&D Systems, Minneapolis, MN,USA) (10 ng/ml) for 24 h. In order to investigate the role of Kindlin-1 in HaCaT cell migration, normal, Kindlin-1 shRNA and control shRNA cells were treated with or without recombinant human EGF (10 ng/ml) for 24 h. The infiltrating cells were stained with crystal violet (Beyotime, Shanghai, China), and cell numbers were counted from 5 wells. All experiments were repeated 3 times.

Cell proliferation was determined using the Cell Counting Kit-8 (CCK-8; Dojindo, Kunamoto, Japan) following the instruction manual. In brief, the cells were treated with or without EGF (10 ng/ml) for 24 h and then incubated with CCK-8 solution for 1.5 h. The absorbance was determined at 450 nm using a Universal Microplate Reader (Bio-Tek Instruments Inc., Winooski, VT, USA). All assays were performed in triplicate, and the experiment was repeated 5 times.

The results are presented as the means ± SEM. Statistical analysis was performed using One-way ANOVA. Statistical significance was determined at the level of P<0.05.

In our initial set of experiments, we examined Kindlin-1 expression in skin tissues. Animals that underwent incisions were randomly divided into 2 main groups (with or without EGF treatment) and raised for different periods of time (Fig. 1A1). As shown in Fig. 1A2, the topical application of EGF improved the healing rate (72.73% vs. 90.54% on the 14th day). Kindlin-1 mRNA expression was assessed by real-time PCR. As shown in Fig. 1B, the mRNA expression of Kindlin-1 reached the highest level at 1 day post-injury. The administration of EGF resulted in the more rapid and prolonged increase of Kindlin-1 mRNA expression. Similar alterations were observed by western blot analysis (Fig. 1C1 and C2). The dynamic expression of Kindlin-1 suggested that it is involved in the process of skin wound healing and is very likely to contribute to EGF signaling.

The distribution of Kindlin-1 was detected by immunostaining. As shown in Fig. 2A, in normal skin, Kindlin-1 was mainly located in the epidermis with low intensity. During the natural healing process [EGF(−) group, Fig. 2B–D], strong Kindlin-1 staining was found 3 days after wounding, mainly distributed in the epidermis and sporadically in the new dermal fibroblasts (Fig. 2B). On the 7th day, moderate Kindlin-1-positive staining was observed in the regenerated epidermis, which decreased with the distance from the wound center (Fig. 2C). On the 14th day, only weak Kindlin-1 staining was observed in the wound epidermis (Fig. 2D). A more intense kindlin-1-positive signal was observed in the EGF(+) group (Fig. 2E–G); the most intense staining was found in the epidermal cells post-injury. Double labeling immunofluorescence (Fig. 3) further indicated that most of Kindlin-1 was located in the keratinocytes (Fig. 3A4), and to a lesser extent in fibroblasts (Fig. 3B4) and macrophages (Fig. 3C4). The specific accumulation of Kindlin-1 in keratinocytes encouraged us to further investigate the possible role of Kindlin-1 in re-epithelialization.

As mentioned above, keratinocytes are major cell components in re-epithelialization, EGF accelerates keratinocyte migration, induces rapid proliferation and exerts its effects in a paracrine manner (7,8). Therefore, in the following in vitro experiments, we used EGF and HaCaT cells to mimic the in vivo conditions of re-epithelization.

As shown in Fig. 4A1, in response to EGF stimuli, upregulated Kindlin-1 expression was detected 3 h after and sustained until the end of the observation (24 h, 2.47-fold vs. 0 h, Fig. 4A2). These data confirmed our hypothesis that Kindlin-1 may contribute to EGF-induced re-epithelialization through keratinocytes.

It has been documented that there is a highly complex regulation of cell signaling in re-epithelialization (2). To further explore the signal transdution of EGF-Kindlin-1 in HaCaT cells, we employed Kindlin-1 shRNA to search for downstream regulators. The interference and knockdown efficiency was confirmed by fluorescence microscopy (Fig. 4B) and western blot analysis (Fig. 4C). The Kindlin-1 knockdown population (Kindlin-1 shRNA) and the scrambled-RNAi population (Control shRNA) of HaCaT cells were used in the following assays.

Integrin β1 is one the best candidates of the downstream regulators either for its potential binding sites of Kindlin-1 or for its critical role in promoting re-epithelialization (27,28). In our study, EGF triggered integrin β1 expression and its activation was confirmed in Fig. 5A1–A4. Although Kindlin-1 knockdown yielded no obvious changes in integrin β1 protein expression (Fig. 5B1 and B3), reduced integrin β1 activation [integrin β1 (ac)] was observed in the EGF 0 h group, approximately 36% (Fig. 5B2, lane 3 vs. lane 2); a greater reduction was observed in the EGF 24 h group, an approximately 58% decrease when compared to the control shRNA group (Fig. 5B2, lane 6 vs. lane 5). Such findings were confirmed by immunofluorescence staining. Enhanced Kindlin-1 and integrin β1 (ac) staining was observed in the cells in the control shRNA group in response to EGF stimuli (Fig. 6B1 vs. A1 and B2 vs. A2), particularly at the cell peripheral sites. Intensive violet fluorescence in the inlays (Fig. 6B2 vs. A2) indicated the increased co-localization of Kindlin-1 and integrin β1 (ac) in the EGF-treated control shRNA-transfected cells. However in the cells transfected with Kindlin-1 shRNA, integrin β1 (ac) staining was less than that of the control shRNA-transfected cells, whether EGF was present in culture or not. These results implied that EGF evoked both the expression and binding of Kindlin-1 and integrin β1 (ac), while Kindlin-1 shRNA suppressed this effect. Together with the regulatory role of Kindlin-1 in integrin β1 trafficking (29), our data indicated that Kindlin-1 was very likely to be a mediator of EGF-induced integrin β1 signaling.

FAK is a widely expressed non-receptor tyrosine kinase which is responsible for multiple cell functions by mediating integrin signaling transduction (30–32). In our study, of note, although no obvious difference was observed in FAK expression between the groups, EGF-induced FAK activation was confirmed in the control shRNA-transfected HaCaT cells by both western blot analysis and immunofluorescence staining (Figs 7A1 and 8B2 vs. A2). In contrast, both adhesion-dependent FAK autophosphorylation (Figs 7A1 and 8C2 vs. A2,) and EGF-stimulated FAK phosphorylation were partly abolished in the Kindlin-1 knockdown group (Figs 7A1 and 8D2 vs. B2). Fold changes in FAK phosphorylation following exposure to EGF were quantified as shown in Fig. 7A2. Furthermore, exposure to EGF increased the co-localization of Kindlin-1 and p-FAK in the control HaCaT cells (inlay of Fig. 8B2 vs. A2). Combined with our findings shown in Fig. 6, the above-mentioned data proved that the Kindlin-1-integrin β1-FAK complex formed at cell peripheral sites and suggested the existence of an EGF-Kindlin-1-integrin β1-FAK signaling pathway in HaCaT cells.

Actin is another candidate of downstream regulators of EGF-Kindlin-1 either for its indirect interaction to Kindlin-1 through migfilin, or its direct binding to the intracellular part of integrin β1 (18,33). Through the adjustment of its organization, the actin cytoskeleton contributes to integrin-mediated outside-in and inside-out signal transduction. Our results demonstrated that, following culture with EGF, more obvious cell actin was concentrated at cell extensions and/or cell borders (Fig. 9B) when compared with EGF(−) deprived control cultures (Fig. 9A). However, in the Kindlin-1 knockdown group, the actin network was disorganized and disrupted (Fig. 9D vs. C), suggesting that EGF-induced actin reorganization may be partly dependent on Kindlin-1.

Although previous studies have proven that there is no catalytic activity, Kindlin-1 may regulate cell functions by integrating and transmitting cell signals (16,17,34). In this study, in the presence of EGF, the potential role of Kindlin-1 in actin re-organization and its increased binding to integrin β1, as well as FAK, strongly suggested that Kindlin-1 may influence cell behavior in wound healing by directly or indirectly binding to signaling proteins and mediating protein interactions.

To verify the above hypothesis, we then focused on examining the influence of Kindlin-1 on integrin β1 and FAK-related cell functions. During re-epithelization, keratinocytes migrate to the wound center and proliferate to fill the skin defect (35,36). Indeed, Kindlin-1 shRNA-mediated abnormal integrin regulation caused cell behavior defects: Transwell assay revealed that in the absence of EGF, PF-573228 suppressed EGF-induced HaCaT cell migration (Fig. 10A1, lane 3 vs. lane 2); likewise, Kindlin-1 shRNA attenuated HaCaT cell migration compared to the control shRNA group (Fig. 10A1, lane 5 vs. lane 4). More marked differences were observed between the Kindlin-1 shRNA- and control shRNA-transfected cells following stimulation with EGF (Fig. 10A1, lane 10 vs. lane 9). The fold induction of cell migration following exposure to EGF was quantified as shown in Fig. 10A2 [EGF(+)/EGF(−)], Kindlin-1 shRNA restrained cell migration compared to the control shRNA group (approximately 35% reduction, lane 5 vs. lane 4), indicating Kindlin-1 shRNA partly abolished EGF induced migration.

Similar changes were observed for cell proliferation. Under normal conditions, slightly impaired cell proliferation was observed in the Kindlin-1 shRNA-transfected cells (Fig. 10B1, lane 3 vs. lane 2), while in the presence of EGF, a more rapid difference was observed between the groups (Fig. 10B1, lane 6 vs. lane 5). The fold induction of cell viability following exposure to EGF was quantified as shown in Fig. 10B2 [EGF(+)/EGF(−)]; Kindlin-1 shRNA repressed cell proliferation compared to the control shRNA group (approximately 45% reduction, lane 3 vs. lane 2).

Taken together, the inhibition of Kindlin-1 expression in HaCaT cells partially weakened EGF-induced integrin β1-FAK activation and the recruitment of cytoskeletal proteins, which finally led to impaired cell functions, including migration and proliferation. These results confirm the role of Kindlin-1 in transmitting EGF activation signals in skin wound healing.