A rat model of type 1 diabetes was generated using a previously described method (11). A total of 12 male Sprague-Dawley rats (age, 8 weeks; weight, 150 to 200 g) were housed in cages (12 h light/12 h dark cycle; free access to water and food) at 25°C and 30% humidity, divided into 2 groups (n=6/group) and were injected intraperitoneally with streptozotocin (STZ; Sigma-Aldrich; Merck KGaA, Darmstadt, Germany; 50 mg/kg) which was dissolved in sodium citrate buffer (pH 4.5) (12). STZ-injected rats with blood glucose levels >16.7 mmol/l were considered to have developed DM. To generate the control group, the same volume of sodium citrate without STZ was injected into the male rats. To generate skin wounds, the rats were anesthetized with pentobarbital (Sigma-Aldrich; Merck KGaA; 45 mg/ml), and two circular wounds (~200 mm²) that were 1 cm deep were created on the lower back of each rat. Ethical approval was obtained from the ethics committee of Wenzhou Medical University (Wenzhou, China).

Following 8 weeks of DM onset, the wound area was evaluated every 4 days for 16 days. The skin wounds from 3 diabetic and 3 control rats were photographed with rulers, and the area of the unhealed regions was calculated using these images together with TINA 2.0 software (DesignSoft, Inc., Budapest, Hungary; http://www.designsoftware.com/) (11).

Following anesthesia with pentobarbital (45 mg/ml; Sigma-Aldrich; Merck KGaA), the dorsal area of diabetic rats was totally depilated with Na2S (8.0%, w/v; Sigma-Aldrich; Merck KGaA) and two full-thickness circular wounds (~300 mm² each) were created on the lower back of each rat using a pair of sharp scissors and a scalpel. Skin tissues (2 g) that were 1-cm deep were collected from the diabetic and control rats for protein extraction 2 weeks following the establishment of DM. The following reagents were used for protein extraction: PBS, 9.5 M urea, 0.1% (w/v) DTT, 2% (w/v) CHAPS and 0.8% (w/v) pharmalyte (Sigma-Aldrich; Merck KGaA). The extracts were dissolved in immobilized pH gradient (IPG) rehydration buffer [8 M urea, 2% (w/v) 3-[(3-cholamidopropyl) dimethylammonio]-1-propanesulfonate hydrate, 2% (w/v) IPG buffer, 0.04 M dithiothreitol, 1X nuclease solution and 0.1% (w/v) of bromophenol blue] prior to isoelectric focusing electrophoresis. The following reagents were used for 2-D gel analysis: 1% DTT, 50 mM Tris-HCl, pH 8.8, 6 M urea, 30% (v/v) glycerol, 2% (w/v) SDS, 0.1% (w/v) bromophenol blue, equilibration buffer (2.5% (w/v) iodoacetamide (Sigma-Aldrich; Merck KGaA). The MS/MS spectra were processed using Proteome Discoverer software (version 1.3; Thermo Fisher Scientific, Inc., Waltham, MA, USA) and the database search was performed using the Mascot search engine (Matrix Science Mascot version 2.3; http://www.matrixscience.com/) against a concatenated forward-decoy approach (13).

Protein spots were visualized using ImageMaster 2D Platinum version 6.0 (GE Healthcare Bio-Sciences, Pittsburgh, PA, USA). A total of 77 protein spots of interested were excised and digested using a 1% protease cocktail blue (Sigma-Aldrich; Merck KGaA), which were further analysed using MS. In-gel digestion was performed in 3 steps according to the procedure described by Russell et al (14). The following reagents were used for analysis: 50% (v/v) acetonitrile (ACN), 10 mM DTT containing 100 mM ammonium bicarbonate and 1% trypsin solution (Sigma-Aldrich; Merck KGaA).

The proteins extracted from the gel sections were digested using 1% trypsin solution (Sigma-Aldrich; Merck KGaA) and 4 µl of samples were submitted to online nanoflow liquid chromatography using the easy-nano LC system (Proxeon Biosystems; Thermo Fisher Scientific, Inc.) with 10 cm capillary columns of an internal diameter of 75-µm, filled with 3-µm Reprosil-Pur C18-A2 resin (Dr. Maisch GmbH, Ammerbuch-Entringen, Germany) (14). The gradient consisted of 10–30% (v/v) ACN in 0.1% (v/v) formic acid at a flow rate of 200 nl/min for 45 min, 30–100% (v/v) ACN in 0.1% (v/v) formic acid at a flow rate of 200 nl/min for 1 min and 100% CAN in 0.1% formic acid at a flow rate of 200 nl/min for 10 min. The elution was electrosprayed through a Proxeon nanoelectrospray ion source by (electrospray ionization) ESI-MS/MS analysis on a Thermo Fisher LTQ Velos Pro (Thermo Fisher Scientific, Inc.) using full ion scan mode over the m/z range 200–1800. Collision-induced dissociation (CID) was performed in the linear ion trap using a 4.0-Th isolation width and 35% normalized collision energy with helium as the collision gas. Five independent MS/MS scans were performed on each ion using dynamic exclusion. The precursor ion that was selected for CID was dynamically excluded from further MS/MS analysis for 30 sec. Further corresponding protein identification was performed using the Swiss-Prot protein sequence database (version 54.5; http://web.expasy.org/docs/swiss-prot_guideline.html). LC-MS/MS and bioinformatics analyses were performed by LC Bio, Inc. (Hangzhou, China).

Skin tissues (100 mg) were ground using liquid nitrogen and an ice-cold lysis solution (7 M urea, 2 M thiourea, 2% CHAPS, 40 mM Tris base, 40 mM dithiothreitol, and 1% protease inhibitor; Sigma-Aldrich; Merck KGaA) was added to procure whole cell extracts. Total proteins (20 µg) from skin samples with or without induced diabetes 2 weeks prior, were separated on a 12% SDS-PAGE gel and electrotransferred to Immobilon^®-P PVDF Transfer Membranes (EMD Millipore, Billerica, MA, USA). The membranes were blocked in 1X Tris-buffered saline containing 5% skim milk and 0.05% Tween-20 for 2 h. All of the reagents for SDS gel electrophoresis and membrane blocking were purchased the Sigma-Aldrich; Merck KGaA. To analyse protein levels in the samples the membranes were incubated with the following primary antibodies at room temperature (25°C) for 2 h: Anti-eNOS antibody (cat. no. ab76198; 1:2,000, Abcam, Cambridge, MA, USA), anti-cleaved (c)-caspase-3 antibody (cat. no. ab2302; 1:2,000, Abcam), anti-plasminogen activator inhibitor (PAI) −1 antibody (cat. no. ab125687; 1:2,000; Abcam), anti-cluster of differentiation (CD) 34 antibody (cat. no. ab81289; 1:2,000; Abcam), anti-collagen1 antibody (cat. no. ab34710; 1:2,000; Abcam), anti-annexin A2 antibody (cat. no. ab41803; 1:1,000; Abcam), anti-JNK antibody (cat. no. ab179461; 1:2,000; Abcam), anti-phosphorylated (p) -JNK antibody (cat. no. ab4821; 1:2,000; Abcam) and anti-GAPDH antibody (cat. no. ab8245; 1:2,000; Abcam). The membranes were then incubated with either the anti-mouse or anti-rabbit horseradish peroxidase-conjugated secondary antibody (cat. no. 7074; 1:2,000; Cell Signaling Technology, Inc., Danvers, MA, USA) at room temperature (25°C) for 2 h. Antigen-antibody complexes were visualized using an electrochemiluminescence kit (BioTrand, Crystal Lake, Illinois, USA). Protein levels were normalized against GAPDH and protein expression was analyzed using Image J2 version 2.0 software (13).

Human fibroblast cell culture was performed as previously described (5). Human foreskin tissues were collected from 3 patients diagnosed with redundant prepuce (all three patients were healthy without DM or other disease diagnoses; ages, 26, 30 and 32 years old) who were admitted in January 2014 to the Department of Dermatology at the First Affiliated Hospital of Wenzhou Medical University (Wenzhou, China). The present study was approved by the ethics committee of Wenzhou Medical University, and written informed consent was obtained from all patients involved. The fat was first removed from the tissue samples, when were subsequently cut into 3×2-mm strips and incubated overnight at 4°C in 0.05% Dispase I (Sigma-Aldrich; Merck KGaA). The epidermis was then removed and the dermis was placed into 25 cm² flasks pre-treated with foetal bovine serum (FBS; HyClone; GE Healthcare Life Sciences, Logan, UT, USA). The flasks were positioned horizontally for 1 h, and then vertically for 3 h in a culture chamber at 37°C with 5% CO2. The cells were cultured in Dulbecco's modified Eagle's medium (DMEM; HyClone; GE Healthcare Life Sciences) containing 5.5 mM glucose (Sigma-Aldrich; Merck KGaA), 10% FBS and 1% penicillin-streptomycin (Gibco; Thermo Fisher Scientific, Inc.); the medium was refreshed every 3 days. When cell confluence had reached 70 to 80% the cells were digested and passaged using 0.25% trypsin (Gibco; Thermo Fisher Scientific, Inc.). Cells were cultured for 3 days in 5.5 mM glucose medium, before they were transferred to media with or without 25 µM SP600125 (Cell Signaling Technology, Inc.) for up to 1 day prior to the migration assay, or for 1 h prior to western blot analysis.

For overexpression of annexin A2, the open reading frame (ORF) sequences of the annexin A2 gene (ANXA2; GenBank Accession no. NM_001002858.2; National Center for Biotechnology Information, https://www.ncbi.nlm.nih.gov/) was amplified for 28 cycles of 95°C for 30 sec, 60°C for 30 sec and 72°C for 90 sec using the following primers: Forward, 5′-ATCGTAGGATCCATGGGCCGCCAGCTAG-3′, and reverse, 5′-AGCTATTCTAGATCAGTCATCTCCACCACACAGGTA-3′. The amplified polymerase chain reaction (PCR) fragments were then cloned into the Bacillus amyloliquefaciens and Xanthomonas badrii restriction enzyme sites of the pcDNA3 expression vector (Invitrogen; Thermo Fisher Scientific, Inc.). Primer design and sequencing were performed by Sangon Biotech Co., Ltd. (Shanghai, China; http://www.sangon.com/). A total of 2 or 3 µg pcDNA3-ANXA2 plasmid was transformed into human foreskin primary fibroblasts from the 3 patients using a Lipofectamine 2000 kit (Invitrogen; Thermo Fisher Scientific, Inc.) according the manufacturer's instructions. Empty vector pcDNA3 using Lipofectamine 2000 transformed cells was used as control. The 2 and 3 µg of pcDNA3-ANXA2 plasmid transformed cells were termed OX1 and OX2, respectively. Following 2 days of transformation, annexin A2 levels in the fibroblast cells were examined by western blotting, and cell migration tests were performed.

Primary fibroblast cells were seeded onto 6-well plates at 80–90% confluency and incubated at 37°C overnight in DMEM containing 0.5% FBS and 5 µg/ml mitomycin-C (Sigma-Aldrich; Merck KGaA). Linear scratch wounds were subsequently created in the confluent fibroblast monolayer using a sterile 200 µl pipette tip (Ningbo MFLab Medical Instruments Co., Ltd., Ningbo, Zhejiang, China). The medium was immediately replaced with prewarmed (37°C) fresh DMEM containing 0.5 % FBS and 5 µg/ml mitomycin-C. At 0, 12 and 24 h following wound establishment, images were captured using a light microscope (IX70; Olympus Corporation, Tokyo, Japan) equipped with a CCD camera (CoolSNAP HQ; Nippon Roper, Toyko, Japan), which was controlled by MetaMorph 7.1 software (Molecular Devices, LLC, Sunnyvale, CA, USA). To quantify cell migration, 10 cells on the border of the wound area were randomly selected from each well and the migration distance was measured using ImageJ version 14.8 software (https://imagej.nih.gov/ij/) at the indicated time points.

Statistical analyses were performed using GraphPad Prism software (version 5.0; GraphPad Software, Inc., La Jolla, CA, USA). Comparisons between 2 groups were performed using a Student's t-test, while significant differences among 3 groups were analysed using one-way analysis of variance, followed by Bonferroni's multiple comparison test. P<0.05 was considered to indicate a statistically significant difference.

Diabetes-induced skin wound healing was analyzed in a rat model of diabetes. STZ, an inducer of type 1 diabetes was used to induce diabetes in rats, and wound healing was assessed by monitoring and comparing the size of ~200 mm² circular wounds created on the waist of each rat at 8 weeks following the onset of diabetes (8). Wound repair was then analysed over the course of 16 days, and the results indicated that the rate of wound healing was significantly reduced in diabetic rats when compared with non-diabetic normal rats (Fig. 1).

In order to gain an improved understanding of the molecular mechanisms underlying DM-mediated impairment of skin wound healing, comparative proteomic assays were preformed to identify differentially expressed proteins in the skin of DM rats and normal rats using 2-D gel electrophoresis. Total skin protein extracts were first subjected to 2-D gel analysis followed by identification by LC/MS. This led to the detection of proteins altered by DM. Proteins extracted from the normal rat skin were used as a control (Fig. 2). Optimization of the 2-D gels and image processing identified 36 and 41 protein spots (representing differentially expressed proteins) that were induced and repressed by DM, respectively (Fig. 2). The experiments were repeated 3 times and only the reproducible alterations were taken into account. To identify differentially expressed proteins, the spots were excised from the 2-D gels and identified by LC/MS and trypsin digestion. The results presented in Table I demonstrate that 72 proteins were identified among the 77 spots observed in the 2-D gel (Fig. 2). The major group of proteins identified, either overexpressed or repressed, were observed to be involved in regulation of the cytoskeleton (tubulin and actin formation), carbohydrate metabolism (glycolysis and the tricarboxylic acid cycle) and protein folding. In the present study, the calcium binding proteins, calmodulin and calreticulin, were induced in DM (Table I). Formation of the ECM is an important step in the wound healing process (3). DM reduced the levels of collagen α1 and dermatopontin, two ECM-associated proteins. In addition to the major group of proteins, RNA metabolism, adenosine triphosphate (ATP) synthesis and cell migration-associated protein levels were altered (Table I).

To verify the results identified by proteomic assay, western blot analysis was performed using specific antibodies to detect select marker proteins in the pathways identified. The results revealed that the expression levels of CD34, collagen-1 and endothelial NO synthase were decreased (NO synthesis), whereas PAI-1 (inflammation), c-caspase3 (apoptosis) and annexin A2 were increased in DM rat skin when compared with normal rat skin (Fig. 3). These results were similar to those identified by the proteomic analysis.

Annexin A2 was induced in DM rat skin (Fig. 3), and the transcript of its orthologue gene in humans was previously demonstrated to be induced by high-glucose treatment of foreskin fibroblast cells (9). Therefore, the function of annexin A2 in fibroblast cells was examined further, by forced overexpression of annexin A2. Human ANXA2 (GenBank Accession no. NM_001002858.2) ORF sequences were amplified by PCR, and the PCR products were then cloned into the pcDNA3 expression vector. Following sequencing, 2 and 3 µg pcDNA3-ANXA2 plasmid were transformed into human foreskin primary fibroblast cells using the Lipofectamine 2000 system to generate two independent transformants, termed OX1 and OX2, respectively. Compared with the control cells (untransformed), OX1 and OX2 cells exhibited a significantly increased cell migration rate following wound generation (Fig. 4A and B). To verify the expression levels of annexin A2, western blot analysis was performed. The results demonstrated that the level of annexin A2 was increased in OX1 and OX2 when compared with the controls (Fig. 4C). In addition, annexin A2 overexpression maintained an increased cell migration rate when compared with the control cells when they were cultured in high-glucose containing medium (data not shown). However, the level of p-JNK, a key regulator of cell migration, was not altered by the overexpression of annexin A2 (Fig. 4C).

As annexin A2 overexpression was not observed to alter p-JNK protein expression levels, SP600125, an inhibitor of JNK was used to analyze its effect on annexin A2 expression. SP600125 significantly inhibited fibroblast cell migration (Fig. 5A and B), which is consistent with results presented in a previous study (11). In addition, SP600125 suppressed p-JNK expression while maintaining normal total JNK expression levels (Fig. 5C). SP600125 treatment marginally repressed annexin A2 protein expression levels (Fig. 5C).

Due to the observed inhibitory effects of SP600125 on cell migration and annexin A2 expression levels, its effects were further examined in OX1 fibroblast cells overexpressing ANXA2. SP600125+OX1 was associated with a significant increase in cell migration when compared with SP600125-treated controls at 12 and 24 h. In addition, SP600125 treatment did not alter the expression levels of annexin A2 in OX1 transformed cells (Fig. 6A and B). Western blotting results demonstrated that SP600125 markedly suppressed p-JNK levels in OX1 cells (Fig. 6C). Collectively, these results indicate that annexin A2 may function downstream of JNK to promote fibroblast cell migration.