Diabetic wound tissue was obtained from debridement or amputation of diabetic foot patients over a course of more than three months. Non-diabetic wound tissue in the control group was obtained from debridement of acute trauma injury or plastic surgery. Written informed consent was obtained from participating patients. The procedure was approved by the ethics committee of Shanghai Jiao Tong University School of Medicine (SJTUSM) (Number:2016-105-T54).

HaCaT and 293T cells were obtained from Fu Heng Biology (Shanghai, China). Cells were cultured in defined keratinocyte-serum free medium (K-SFM, 10744–019, Gibco, Waltham, MA, USA) supplemented with high glucose (30 mM) or normal glucose (9 mM) for 5 d at 37°C in 5% CO₂. HaCaT differentiation was induced by increasing the concentration of CaCl₂ to 2.8 mM. Cells were pre-treated with 20 μM 10058-F4 (S7153, Selleck, Houston, TX, USA) or 25 μM KYA1797K (S8327, Selleck) or 1 μM S100A6 recombinant protein(10939-HNAE, Sino Biological Inc. Beijing, China) for 24 h. Overexpression of c-Myc (c-Myc ^OE), knock-down c-Myc (shMyc,Target Seq : GAGGCGAACACACAACGTCTT) and knock-down S100A6 (shS100A6,Target Seq: TGCAAGGCTGATGGAAGACTT) were introduced into HaCaT by transfection with lentivirus synthesised at Shanghai GeneChem Company (Shanghai, China). Stable cell clones were obtained through selection on 2 μg/ml puromycin. Cells were harvested when cell confluence reached to about 80%. Cells were collected for protein or mRNA extraction or fixed with 10% polyformaldehyde for immunofluorescence analysis.

Thirty rats were obtained and housed at the Animal Science Center of SJTUSM. The procedure was approved by the Institutional Animal Care and Use Committee of SJTUSM. Rats were randomly divided into three groups: the control group (Ctrl, n =10), diabetic group (DM, n =10) and diabetic rats treated with an c-Myc inhibitor group (DM+10058-F4, n =10). A high-fat diet combined with streptozotocin injections was used to establish a diabetic rat model as previously described (15). Briefly, rats in the control group were fed a normal chow diet, while rats in the diabetic group were fed a high-fat diet containing 60% (kcal) fat, 20% (kcal) carbohydrates, and 20% (kcal) protein (Research diets, D12492, Research Diets Inc., New Brunswick, NJ, USA) for 8 weeks. Subsequently, the rats in the diabetic group were fasted for 16 h and then intraperitoneally injected with 10 mg/kg streptozotocin (ALX-380-010, Enzo, Farmingdale, NY, USA) dissolved in 0.1 M citrate buffer for 4 consecutive days. Rats were then allowed to develop diabetes for 4 weeks and resumed a high-fat diet. Rats in the control group received intraperitoneal injections of saline. Random blood glucose level >16.6 mM was considered as an indicator of diabetes.

Rats were anaesthetized with a single intraperitoneal injection of 3% sodium pentobarbital (60mg/kg) and the hair was removed from the backs of the rats. Four full-thickness skin wounds on the mid-back were created with a 9-mm punch biopsy. The wounds of rats in the control and DM groups were treated with saline, while the c-Myc inhibitor group was topically administered 30 mg/kg 10058-F4 for 11 consecutive days. Wound and an additional 5 mm of the surrounding normal skin tissues were collected on day 11 post-injury. Half of the tissues were fixed in 4% paraformaldehyde solution and the other half were frozen in liquid nitrogen. Paraffin sections of the wound tissues were stained with haematoxylin and eosin (H&E).

Protein was extracted from HaCaT cells and measured with the BCA assay. Proteins (20 μg) were separated by Biofuraw™ Precast Bis-Tris Gel (180-8008H, Tanon, Shanghai, China) and then transferred to nitrocellulose membranes. After blocking with 5% skimmed milk for 1 h, the membranes were then incubated overnight at 4°C with the following primary antibodies at the dilution of 1:1000: c-Myc (10828-1-AP, Proteintech, Rosemont, IL, USA), transglutaminase 1 (TGM1,12912-3-AP, Proteintech), loricrin (LOR, 55439-1-AP, Proteintech), keratin 1(K1,16848-1-AP, Proteintech), β-catenin (51067-2-AP, Proteintech), H3 (A2348, ABclonal, Woburn, MA, USA), S100A6 (10245-1-AP, Proteintech) and β-actin (66009-1-Ig, Proteintech). The next day, the membranes were incubated with the HRP-conjugated secondary antibody (1:5000) for 1 h at room temperature. Finally, western blot bands were visualised using electrochemical luminescence. The band density analysis was performed using Image‐Pro Plus.

Total RNA was extracted using Trizol and cDNA was synthesised using the HiScript^® III RT SuperMix kit (R323-01, Vazyme, Nanjing, China). Quantitative PCR (qPCR) was performed in triplicate using the ChamQ Universal SYBR qPCR Master Mix (Q711-02, Vazyme). The primers were synthesised by Shanghai Sangon Biotech (Shanghai, China) and shown in Table 1 . The relative gene expression was calculated using the 2^−ΔΔCT method.

Cells were fixed in 4% paraformaldehyde for 30 min at room temperature. Paraffin sections were deparaffinized and then antigen repaired with sodium citrate. Slides were incubated with 3% hydrogen peroxide for 20 min to eliminate endogenous peroxidases. Slides were washed with phosphate buffered saline (PBS) three times and permeabilized with 0.1% Triton X-100 for 10 min. Slides were then incubated with 10% BSA in PBS for 30 min and incubated with primary antibodies against TGM1, LOR, K1, or S100A6 at a dilution of 1:100 at 4°C overnight. The next day, cells were incubated with the donkey anti-rabbit IgG secondary antibody Alexa Fluor 568 (A10042, Invitrogen, Carlsbad, CA, USA) at the dilution of 1:500 for 1 h at room temperature and with DAPI for 5 min.

The pre-processing of the slides was performed as described in the previous Immunofluorescence section. Slides were then incubated with primary antibodies against c-Myc at the dilution of 1:100 at 4°C overnight and with the biotinylated secondary antibody on the next day for 1h. Finally, slides were incubated with 3,3′-diaminobenzidine tetrahydrochloride and the reaction time was controlled under the microscope. Brown particles were considered positive.

HaCaT cells (1×10⁷) were used for ChIP according to a standard protocol using the Enzymatic Chromatin IP Kit (9003, CST). DNA immunoprecipitated with an anti-c-Myc antibody (9402S, CST), IgG or H3 was examined by qPCR using a pair of primers that encompassed the potential c-Myc binding site in the S100A6 promoter: Forward GCCTTCACTCCCCCGTAAA; Reverse CCTCAGTGCCCCAAATTCCA. The ChIP enrichment efficiency was calculated using the following formula: percent input = 2% × 2^{(CT 2%Input Sample–CT IP Sample)}. Finally, the qPCR products were separated with 1.5% agarose gel electrophoresis and imaged under ultraviolet light.

pMyc-TA-luc vector was obtained from Beyotime Biotechnology (D2198, Shanghai, China). A 2.0-kbp promoter region of S100A6 (S100A6-WT) or with mutation in the potential c-Myc binding site of (S100A6-Mut) was cloned into a pGL3-Basic luciferase reporter vector (Umibio Co. Ltd, Shanghai, China). The c-Myc overexpression plasmid and its control plasmid were obtained from the Shanghai GeneChem Company. S100A6 promoter activity was normalised by co-transfection with a Renilla luciferase reporter (Umibio, Shanghai, China). Luciferase activities for firefly and Renilla were measured in 293T cells with a dual-luciferase reporter kit (MA0518, Meilunbio, Dailan, China) 36 h after transfection.

All values are expressed as means ± SEM. Data was analysed using SPSS 24 (IBM Corp., Armonk, NY, USA). Statistical comparisons were performed using the Student’s t-test or one-way analysis of variance (ANOVA) followed by the Tukey’s post-test. P values < 0.05 were considered statistically significant.

There are few reports on the effect of high glucose on the differentiation of HaCaT. Thus, we determined the effect of high glucose on the differentiation of HaCaT cells and keratinocytes at diabetic wound margin. The expressions of differentiation makers such as TGM1, LOR, and K1 in the high glucose (HG) group were significantly lower than those in the normal glucose (NG) group ( Figure 1A ). Similarly, the expression levels of TGM1, LOR and K1 in full-thickness skin defect wound margin tissues from diabetic rats and patients were significantly lower than those in the control group ( Figures 1B–G ). Altogether, these results indicated that high glucose or diabetes impaired HaCaT or keratinocytes differentiation.

c-Myc is a transcription factor encompassing a helix-loop-helix structure that forms heterodimers with Max. c-Myc regulates a variety of fundamental cell biological processes including proliferation, apoptosis, differentiation, and metabolism (16). Firstly, the expression of c-Myc in HaCaT cells cultured in high glucose and keratinocytes at diabetic wound margin was observed. The expression of c-Myc in the HG condition was higher than that in the NG condition ( Figure 2A ). Similarly, c-Myc was upregulated in the wound margin of diabetic patients and rats ( Figure 2B ). pMyc-TA-luc has multiple c-Myc binding sites (E-box DNA binding element) inserted into the multiple cloning sites of pGL6-TA plasmid, which can be used to detect c-Myc transcriptional activity (17). The transcriptional activity of c-Myc in the HG condition was higher than that in the NG condition ( Figure 2C ). To assess whether c-Myc affects HaCaT differentiation, we constructed HaCaT cells with stable c-Myc overexpression ( Figures 2D, E ) or knock-down ( Figures 2D, F ). The mRNA ( Figure 2D ) and protein ( Figures 2G, I ) levels of differentiation-related genes (TGM1, LOR and K1) in HaCaT overexpressing c-Myc were significantly decreased, while those in knock-down c-Myc were significantly increased. Furthermore, knockdown c-Myc ( Figure 2H ) could reverse the decrease of TGM1, LOR and K1 protein levels caused by high glucose ( Figure 2I ). These results indicated that overexpression of c-Myc caused differentiation dysfunction, while knocking down c-Myc promoted differentiation.

c-Myc is the target gene of the WNT/β-catenin pathway (18). Therefore, we speculated that high glucose induced increased c-Myc expression through the WNT/β-catenin pathway. The total protein expression of β-catenin in the HG condition was higher than that in the NG condition ( Figure 3A ). Similarly, the expressions of β-catenin and c-Myc in the nucleus in the HG condition were remarkably up-regulated than those in the NG condition ( Figure 3B ). In addition, immunofluorescence showed up-regulated expression of β-catenin in the nucleus ( Figure 3C ). The hallmark of WNT pathway activation is the nuclear translocation of β-catenin (19). The higher expression of β-catenin in the nucleus in the HG condition could be a consequence of increased total level of β-catenin. Furthermore, the expression changes of c-Myc and differentiation indicators (TGM1, LOR and K1) were observed after inhibiting the WNT pathway of HaCaT cells. KYA1797K is a potent and highly selective Wnt/β-catenin inhibitor (20). KYA1797K abolished the increased expressions of β-catenin and c-Myc caused by high glucose ( Figure 3D ). Similarly, KYA1797K prevented the decrease in TMG1, LOR and K1 expressions caused by high glucose ( Figure 3E ). These results showed that high glucose increased the expression of c-Myc and inhibited differentiation in HaCaT cells by activating the WNT/β-catenin pathway.

We speculated whether the inhibition of differentiation by c-Myc was related to its function as a transcriptional regulator. 10058-F4 specifically inhibits the c-Myc-Max interaction and prevents transcriptional activation of c-Myc target genes (21). Overexpression of c-Myc in HaCaT cells resulted in the decrease of TGM1 and K1 protein levels, which was prevented by the addition of the c-Myc inhibitor 10058-F4. The expression of LOR was significantly decreased in HaCaT overexpressing c-Myc. Although the LOR expression of the overexpressing c-Myc cells treated with 10058-F4 showed an upward trend, there was no statistical difference ( Figure 4A ). Similarly, 10058-F4 prevented the decrease of TGM1, LOR and K1 caused by high glucose ( Figure 4B ). Topical application of 10058-F4 to the wounds of diabetic rats (DM+10058-F4) also prevented the decrease of TGM1 ( Figure 4C ), LOR ( Figure 4D ) and K1 ( Figure 4E ) in keratinocytes at the wound margin as shown in the DM group. Taken together, these results indicated inhibition of c-Myc transcriptional activity alleviated differentiation dysfunction caused by overexpression of c-Myc or high glucose.

Since the differentiation process is related to c-Myc transcriptional regulation, it is critical to identify the specific target gene of c-Myc that is involved in regulating differentiation. Firstly, we found in a ChIP-seq database (Cistrome Data Browser: 8122) that S100A6 was most likely to be the target gene of c-Myc (12). qPCR, Western blotting, ChIP and dual luciferase reporter assay were designed to verify that S100A6 was the target gene of c-Myc. We found that both the mRNA and protein levels of S100A6 were significantly increased in HaCaT overexpressing c-Myc, and significantly reduced in HaCaT with knocked-down c-Myc ( Figures 5A, B ). Overexpression of c-Myc or high glucose caused a significant increase in the S100A6 protein level compared with the control and could be abolished by 10058-F4 treatment ( Figures 5C, D ). We predicted c-Myc binding sites in the S100A6 promoter region with Jaspar (http://jaspar.genereg.net/). A 2.0-kbp promoter region of S100A6 (S100A6-WT) or with mutation in the potential c-Myc binding site of (S100A6-Mut) was cloned into a pGL3-Basic luciferase reporter vector ( Figure 5E ). ChIP-qPCR products were visualized by agarose gel electrophoresis and the result demonstrated that c-Myc directly bound to the S100A6 promotor ( Figure 5F ). Higher amounts of S100A6 promoter DNA in the HG condition were pulled down by the anti-c-Myc antibody than those in the NG condition ( Figure 5G ). Dual-luciferase assay shown increased fluorescence ratios in the S100A6-WT+c-Myc^OE group or S100A6-WT+HG group and decreased fluorescence ratios in the S100A6-Mut+c-Myc^OE group or S100A6-Mut+HG group, respectively ( Figures 5H, I ). These results indicated that S100A6 was directly transcriptionally regulated by c-Myc and high glucose promoted S100A6 transcription.

Finally, we examined the expression of S100A6 and its effect on differentiation. The expression of S100A6 was increased in HaCaT cells cultured with high glucose ( Figure 6A ) and in keratinocytes at the wound margin from diabetic rats ( Figure 6B ) or humans ( Figure 6C ). Morphologically, undifferentiated HaCaT cells exhibited a more spindle-like shape with a more loosely connected phenotype while differentiated HaCaT cells shown a more cuboidal appearance with tightly close packing cell–cell tight junctions ( Figure 6D ). The mRNA ( Figure 6E ) and protein ( Figures 6F, G ) expressions of c-Myc and S100A6 were decreased in differentiated HaCaT cells compared with those of undifferentiated HaCaT cells. HaCaT cells pre-treated with S100A6 recombinant protein showed decreased protein levels of TGM1, LOR and K1 ( Figure 6H ). The mRNA and protein levels of TGM1, LOR and K1 in knock-down S100A6 cells were higher than those in the control group ( Figures 6I, J ). High glucose or overexpression of c-Myc caused the decrease of TGM1, LOR and K1 protein levels, which was prevented by knocking down S100A6 ( Figures 6K, L ). These results suggested that S100A6 was highly expressed in HaCaT cells cultured with high glucose and diabetic wound margin, thereby inhibiting differentiation.