Flexible electrical stimulation device with Chitosan-Vaseline® dressing accelerates wound healing in diabetes

2.1. Manufacture of the flexible ES chitosan dressing

A total of 1.5 g of high MW (MW ~ 1,000,000 Da) and 1.5 g of low MW (MW ~5000 Da) chitosan (Golden-Shell Pharmaceutical Co., Zhejiang, China) was dissolved in a 1.0% glacial acetic acid solution (v/v) to generate a mixed chitosan solution (3%). Absorbent gauze was soaked in the mixed chitosan solution for 2 h to generate the chitosan gauze; this material was dried at 60–65 °C for 24 h to permit evaporation of the excess water. Vaseline® and paraffin oil were mixed in advance (2:1, w/w) and sterilized in a steam autoclave at 121 °C for 20 min. The dried chitosan gauze was then soaked in the Vaseline®-paraffin oil mix for 5 min. The ES device was manufactured using a flexible printing method based on the roll-to-roll process in the flexible printed circuit board factory using bendable PI as the substrate. Functional elements are integrated on the central part of the ES device, which were surrounded by flexible electrodes. Finally, the flexible ES device and CVG were fixed to one another by externally bonded spun-laced mesh-type cloth.

2.2. FTIR analysis and SEM

Fourier-transform infrared (FTIR) spectra were obtained on powders using a Bruker Tensor II Fourier-transform infrared spectrometer (BRUKER OPTICS, Ettlingen, Germany) at room temperature and ambient humidity with a spectral resolution of 4 cm⁻¹ that were averaged over 32 scans. For scanning electron microscopy (SEM), samples were cut into small sizes, and each was coated with a thick layer of platinum. SEM images were taken using the Nova Nano SEM 450 scanning electron microscope (FEI, Brno, Czech Republic).

2.3. Animals

Specific pathogen-free male SD rats (4–6 weeks old, 100–120 g) were used in the study (Laboratory Animal Center of Zhejiang Academy of Medical Sciences, Hangzhou, China). All animal experiments were approved by the Zhejiang University Animal Care Committee. The experimental procedures and animal maintenance were performed in accordance with the guidelines for animal experiments of the Institutional Animal Care and Use Committee of Zhejiang University.

2.4. T2D rat model establishment

All animals were fed a normal dry diet for 1 week during the acclimation period and then switched to a high fat and high glucose diet (Beijing Botai Hongda Biotechnology Co., Beijing, China). One month later, overnight-fasted animals were intraperitoneally injected with streptozotocin (STZ, Sigma-Aldrich, MO, USA) once with a dose of 35 mg/k [50,51]. Tail vein blood glucose was measured after 72 h and fasting blood glucose (FBG)≥300 mg/dL was set as the evaluation criterion for successfully established T2D rat model [52,53]. Rats that did not satisfy the FBG criterion for two successive times after injection after the STZ injection were excluded from further study. Model rats were continuously fed with the high fat and high glucose diet in further studies.

2.5. Animal grouping

In this study, three independent animal experiments were performed. The first experiment was focused on the identification of optimal electrical parameters. A total of 36 rats with T2DM were randomly assigned into groups (6 rats/group) and evaluated as follows: group 1, high voltage monophasic pulsed current (HVMPC, 40V, 100 pulses per second (pps), 1% duty ratio); group 2, low voltage monophasic pulsed current (LVMPC, 4V, 100 pps, 1% duty ratio); group 3, direct current (DC, 4V); group 4, high voltage monophasic pulsed current plus direct current (HVMPC + DC, 40V, 100 pps, 1% duty ratio; 4 V); group 5, high voltage biphasic pulse current (HVBPC, 40V, 100 pps, 1% duty ratio); and group 6, low voltage biphasic pulse current (LVBPC, 4V, 100 pps, 1% duty ratio). Our preliminary experiments revealed that DC at 40V was too high as the rats were burned directly by the electrical current; as such, the high-voltage DC group was removed from subsequent evaluation. The pulsed current in the experimental groups was generated using an isolated pulse stimulator (A-M systems, USA); direct current was provided via a direct current power supply (Longwei, China). Two wounds (one on the left side and one on the right side of the body) were generated in each rat; the wounds on the right side were treated with ES and those on the left side were used as controls. A non-invasive approach designed to avoid electrical burns using a surface electrode was introduced for the second set of experiments. A total of 48 rats with T2DM were assigned randomly into one of the six groups described above; 30 rats (5 per group) were used to evaluate wound healing and another 18 rats (3 per group) were euthanized on day 7 after surgery for Western blot analysis. Conditions and methods were as described above, save for the use of a surface electrode in place of the insertion electrode. For the third experiment, 40 rats with T2DM (10 per group) were used to examine the impact of the flexible ES-chitosan dressing on wound healing. Three rats were euthanized on day 7 and day 15, and the wounds tissue include 5 mm unwounded skin were collected. Each of the two wounds generated in a single rat was treated identically. Groups for the third experiment include group 1, control; group 2, CVG alone; group 3, HVMPC alone; and group 4, flexible ES-chitosan dressing (HVMPC + CVG).

2.6. Wound healing study

All animals were anesthetized with isoflurane (2%) and the hair on their backs was removed using an electric shaver and depilatory cream (Weet, French). Two 2 × 2-cm² full-thickness skin sections (including the dartos), which were parallel to and1 cm away from the midline, were excised on both sides of the dorsal skin. Then, wounds in experimental group was exposed to ES for 1 h once every 2 days, until the wound healed. Photographs of the wounds were taken. Surface areas of the wounds were measured and the relative wound areas were calculated using the Image J software (National Institutes of Health, USA). The healing rate of the wound was calculated according to the following formula:

2.7. Histopathological examination

Samples were fixed overnight in 4% paraformaldehyde. Then, the tissues were dehydrated with graded series of ethanol and embedded in paraffin blocks. Tissue sections of 4 μm were stained with hematoxylin and eosin (H&E) to analyze re-epithelialization and scar tissue formation. The H&E slices were scanned at 10 × magnification with an Olympus VS120 Virtual Slide Microscope. The actual measurements of wound widths were calculated with the Image-Pro Premier 3D software (Media Cybernetics, MD, USA).

2.8. Immunohistochemistry

Briefly, deparaffinized sections were rehydrated using a graded ethanol series and then incubated with 3% H₂O₂ for 10 min. The slides were then washed thoroughly with PBS and blocked with 3% BSA at room temperature for 30 min. Sections were then incubated overnight at 4 °C with the primary antibodies. The primary antibodies used are listed in supporting information Table S1. The next day, the slides were incubated for 50 min at room temperature with HRP-conjugated goat anti-rabbit IgG secondary antibody (Servicebio, Wuhan, China). The staining was visualized after incubation with a DAB-H₂O₂ solution for 5 min and counterstaining with hematoxylin. Images were scanned at 10 × magnification with an Olympus VS120 Virtual Slide Microscope. The positive IOD/area in wound was calculated with the Image-Pro Premier 3D software.

2.9. Western blotting

Wound tissues were homogenized in the RIPA buffer mixed with a protease inhibitor cocktail (Thermo Scientific) by the KZ-II tissue lapping instrument (Servicebio, Wuhan, China). Human Umbilical Vein Endothelial Cells (HUVECs) suspended in the lysis buffer with the same components were vibrated for 30 s every 10 min. The homogenates were then centrifuged at 12,000×g for 20 min at 4 °C and the supernatants were subsequently collected. Protein concentrations were determined using the Bradford assay kit (Bio-Rad, Hercules, CA, USA) and then diluted to 4 μg/μL. A total of 40 μg of protein were separated by the 4–12% gradient SDS-PAGE and transferred onto polyvinylidene membranes (Millipore, Bedford, MA, USA) at 250 mA for 90 min. After blocking with 5% nonfat milk at room temperature for 1 h, the membranes were incubated overnight at 4 °C with the primary antibody. The primary antibodies used are listed in supporting information Table S1. The next day, the membranes were incubated with a 1:5000 dilution of an HRP-conjugated anti-mouse or anti-rabbit secondary antibody (Proteintech, CHI, USA) at room temperature for 90 min. The membranes were then washed thoroughly with PBS three times and the immunoreactive bands were detected with an EZ-EL Kit (Biological Industries, Israel).

2.10. Cell culture and electrical stimulation in vitro

In order to examine the molecular mechanisms underlying the impact of monophasic pulsed current ES in vitro, the primary HUVECs (<8 passages, ScienCell, CA, USA) were cultured in endothelial cell medium (ECM, ScienCell, CA, USA) with 5% fetal bovine serum, 1% endothelial cell growth supplement (ECGS, ScienCell, CA, USA), and 1% penicillin/streptomycin solution (P/S, ScienCell, CA, USA) at 37 °C in a humidified atmosphere of 5% CO₂. The medium was changed every 2 days, and cells were passaged when cells reached 80% confluence. To mimic a high-glucose and high-fat environment, HUVECs were incubated in a medium supplemented with 25 mM d-glucose (Sigma-Aldrich, MO, USA) and 0.2–0.4 mM BSA-conjugated palmitate (Sigma-Aldrich, MO, USA) for 24h. HUVECs cultured in ECM medium containing 5.5 mM d-glucose, 19.5 mM l-glucose (Sigma-Aldrich, MO, USA), and 0.4 mM BSA were used as controls. For electrical stimulation, HUVECs in the experimental group were exposed to electrical stimulation environment using a C-Dish (Ion-Optics Co., MA, USA) and an isolated pulse stimulator (A-M systems, USA). The stimulating carbon electrodes were placed in the wells and submerged in the media.

2.11. Cell proliferation and apoptosis assays

Cell proliferation was analyzed using a Cell-Light 5-ethynyl-2′-deoxyuridine (EdU) Apollo567 In Vitro Kit (RiboBio, Shanghai, China) in strict accordance with the manufacturer's instructions. Images were acquired at 20 × magnification with a fluorescence microscope (Leica, Solms, Germany). EdU-positive cells were counted in the selected cells and calculated as the mean value from multiple fields of view. Cell apoptosis assays were performed using an Annexin V-PE 7-AAD Apoptosis Detection Kit (BD Biosciences, CA, USA) and terminal deoxynucleotidyl transferase-mediated dUTP-biotin nick end labeling assay (TUNEL assay, Beyotime, Shanghai, China) according to the manufacturer's protocol. For Annexin V-PE 7-AAD Apoptosis assay, the cells were analyzed the percentage of apoptotic cells, that cells undergoing early and late apoptosis, among 10,000–20,000 counted cells using a flow cytometer (BD Biosciences, CA, USA). For TUNEL assay, images of apoptosis cells were obtained using a microscope (Leica, Solms, Germany).

2.12. Cell migration assay

Cell migration testing was conducted using wound healing and Transwell assays. In order to obtain scratches of uniform length, a culture insert (ibidi, Munich, Germany) composed of two reservoirs was used in the wound healing assay. A cell suspension of 80 μL (3 × 10⁴ cells/mL) was placed into each well and incubated in the CO2 incubator (37 °C, 5% CO2). After cell adherence, the insert was gently removed, creating a cell-free gap of 500 μm. Cells were thoroughly washed with PBS to remove the crossed cells. Every well was then filled with 2 mL of ECM medium containing 1% FBS and maintained in the incubator. At 0 and 24 h, images were taken using a fluorescence microscope (Leica, Solms, German). Transwell assay was performed according to the manufacturer's protocol. Briefly, 200 μL of the treated cells (2 × 10⁵ cells) cultivated in serum-free DMEM were seeded into the upper chamber of the Transwell (Corning, NY, USA) with 8-μm pores and 500 μL of ECM medium containing 5% FBS were added to the lower chamber as a chemoattractant. After incubation for 24 h, the membrane was fixed in 4% paraformaldehyde and stained with 0.1% Crystal Violet (Solarbio, Beijing, China). The fixed cells in randomly selected fields were acquired at 20 × magnification using a microscope (Leica, Solms, German).

2.13. Statistical evaluation

Data were expressed as mean ± standard error of the mean (SE) and statistically analyzed by the GraphPad Prism 6.0 Software (GraphPad Software, CA, USA). The statistical difference significance between the stimulation groups and the control group was analyzed using the Student's t-test. Multiple group comparisons were analyzed by one-way ANOVA. P-values <0.05 were considered statistically significant.

3.1. Optimization of electrical parameters and electrodes to promote wound healing in diabetic rats

As wound healing progresses, the resistance of wound surface (primarily the epidermis) is seen to undergo a substantial increase [54]. An insertion electrode would have the capacity to apply voltage directly to the tissue without generating a current through the epidermis. However, using an insertion electrode in this series of experiments generated electrical burns in nearly all groups of rats. This complication was most apparent in the group subjected to DC, followed by those treated with high voltage (Supplementary Figs. 1a–c). However, the data revealed that administration of HVMPC resulted in accelerated wound closure (Supplementary Figs. 1d–h). When compared with the results from all other groups, rats treated with HVMPC underwent more rapid healing than did the control wounds (Supplementary Figs. 1i–k). As such, we concluded that HVMPC was the most effective electrical parameter among those evaluated although, as performed, it was associated with significant adverse events.

Then, we designed a separate animal experiment by using surface electrode; this modality has been currently the widely accepted ES therapy. Compared to the insertion electrode, surface electrode is non-invasive and easier to use; selection of electrical parameters for use with surface electrodes is more in-line with current clinical needs. Interestingly, similar results were obtained; HVMPC was selected as the better electrical parameter that might be used to accelerate wound healing. Our findings also revealed that administration of HVMPC significantly accelerated wound healing beginning on day 7 after surgery (Fig. 1) and did not result in electrical burns or any related adverse events.

Scar formation is an inevitable outcome after the injury of full-thickness skin; this process is characterized by the re-organization of the extracellular matrix (ECM) and the absence of normal skin appendages, such as hair follicles. H&E staining revealed that the healed wounds treated by stimulation with HVMPC were smaller than those of the control groups (Supplementary Fig. 2a). It also indicated that HVMPC has the better outcome of wound healing. Furthermore, wounds treated with HVMPC stimulation expressed higher levels of vascular endothelial growth factor (VEGF) compared to controls; this factor is deemed a critical activator of angiogenesis during wound healing (Supplementary Figs. 2b and c). Moreover, HVMPC stimulation resulted in an increase in the number of blood vessels at the wound sites compared to the controls (Supplementary Fig. 2d). Taken together, these data indicate that HVMPC may have accelerated wound healing in rats with T2DM via its capacity to promote angiogenesis.

3.2. Cell culture in high glucose/high palmitate to mimic the pathological environment associated with T2DM

We generated an in vitro model mimicking the microenvironment of wounds in patients with T2DM that included primary HUVECs cultured with high glucose (25 mM d-glucose) and a range of palmitate concentrations (0.2, 0.3, and 0.4 mM). TUNEL staining assay and Annexin V/7-AAD dual-color flow cytometry revealed that high glucose alone (25 mM d-glucose) in short-term culture had no impact on apoptosis of HUVECs when compared to cells in the control group (5.5 mM d-glucose, 19.5 mM l-glucose); by contrast, cellular apoptosis was observed in HUVEC cultures maintained in high levels of d-glucose with increasing concentrations of palmitate (0.2, 0.3, and 0.4 mM; Supplementary Figs. 3a–d). Wound healing and transwell assays have revealed that high levels of d-glucose alone had no impact on the migration of HUVECs; by contrast, diminished levels of HUVEC migration were observed in cultures that included a combination of high d-glucose and palmitate (Supplementary Figs. 3e–h). Likewise, proliferation of HUVECs was inhibited in the presence of both high d-glucose and palmitate; high levels of d-glucose alone had no impact on the HUVEC proliferation in short-term culture (Supplementary Figs. 3i and j).

3.3. HVMPC promoted the proliferation and migration capacity of HUVECs via activation of Akt and ERK1/2 signaling pathways

We performed experiments aimed at identifying the molecular mechanisms associated with HVMPC and its role in promoting angiogenesis. We detected a higher fraction of phosphorylated Akt (p-Akt) and phosphorylated extracellular signal-regulated kinase 1/2 (p-ERK1/2) in the cell cultures subjected to electrical stimulation compared to the controls (Supplementary Figs. 4a and 4b); no differences between the electrical stimulation groups were detected. Moreover, the results revealed that there were more proliferating HUVECs in the groups treated with HVMPC than were detected in the controls (Supplementary Figs. 4c and 4d). Moreover, as shown in Supplementary Figs. 4e and 4f, more cells were observed in the HVMPC group, which indicated the HVMPC group had faster cell migration.

In an effort to elucidate the mechanisms underlying ES using HVMPC and its role in promoting the proliferation and migration of HUVECs, the specific inhibitors of PI3K (LY294002) and ERK (U0126) were used to inhibit the activity of PI3K/Akt and ERK1/2 (Fig. 2a and b). HUVECs treated with either LY294002 or U0126 resulted in significant suppression of cell migration and proliferation (Fig. 2c–e). Furthermore, compared to the control group (i.e., no HVMPC, no inhibitors), more migrating cells was detected in the HVMPC group, and fewer were detected in any of the inhibitor groups (Fig. 2f). Moreover, EdU cell proliferation assay revealed that HVMPC promoted HUVECs proliferation (as determined by the number of EdU-positive cells); the impact of HVMPC was blunted in cultures treated with either LY294002 or U0126 (Fig. 2g). As such, we concluded that HVMPC may serve to promote proliferation and migration of HUVECs via activation of PI3K/Akt and ERK 1/2 signaling pathways.

3.4. Manufacture of the flexible-ES chitosan dressing

Based on aforementioned optimized results, we manufactured a flexible ES- chitosan dressing that consisted of a flexible ES device to be used together with CVG. CVG was generated by infiltration of gauze with both chitosan and Vaseline®. The result of FTIR spectroscopy of the chitosan-containing samples revealed a peak at 1555 cm⁻¹; this was attributed to C C stretching. Two additional peaks (between 2800 and 3000 cm⁻¹) were detected in the samples that contained Vaseline®; these peaks were attributed to C–H stretching (Fig. 3a). The spectrum of CVG included both characteristic peaks; this finding indicates that both chitosan and Vaseline® were components of the final CVG formulation. SEM images of the surface of the gauze samples have been taken (Fig. 3b). These images revealed that chitosan adheres to the gauze fibers in a net shape; by contrast, fibers of the Vaseline®-embedded gauze are smooth and waxy. As anticipated, the gauze fibers of CVG are rougher than those of the Vaseline® gauze due to infiltration with chitosan.

The flexible ES device has included the same functional elements that are used to generate the rigid circuit board. The system diagram of the pulse generating circuit is presented in Supplementary Fig. 5a. The pulse modulation module is a microcontroller unit (MCU; STC12C2052AD) programmed in C to produce a pulse signal at P1.2 with a frequency of 100Hz, a duty cycle of 1%, and an amplitude of 5 V. The boost circuit includes a DC-to-DC converter control circuit chip (MC34063AP). The chip can be specifically designed to be incorporated in either step-down, step-up, or voltage-inverting applications (the voltage-inverting model can change the phase of voltage from positive to negative and vice versa). In this study, we built a step-up converter to generate a 30–40V DC output. The pin P1.2 of MCU and the output of MC34063AP are connected with the gate and drain of N-channel Metal Oxide Silicon Field Effect Transistor (MOSFET; FDN86501LZCN-D), respectively, in order to generate a pulse output with adjustable amplitude.

The schematic diagram is shown in Fig. 4a. The DC output of the step-up converter is controlled by R2 and R1 as per the following formula:

According to the requirements of the study, the circuit should be capable of generating a pulse signal around 30–40V. Therefore, R1 and R2 were set at 10 kΩ and 390 kΩ, respectively, in order generate a 50V output. L2 and L7 were designed as a low-pass filters to reduce the ripple. In order to maintain flexible control of the output voltage, a variable resistor was set between the drain of the N-MOSFET and “boostout” pin. Because the maximum working current of the N-MOSFET at room temperature is 2.6A, and the maximum voltage required for this research application was 40V, the resistance value of the variable resistor should be no less than 15.3 Ω. The load will be connected parallel to the MOSFET. Therefore, when the output of the MCU is 0 V (gate-source voltage [VGS] < gate-source threshold voltage [VGS (th)]), the MOSFET will be closed, resulting in voltage applied on the load. By contrast, when the output of the MCU is 5 V (VGS > VGS [th]), the MOSFET will be opened, resulting in a short-circuited load.

The outputs of the MCU, the boost module, and the entire system were measured using a digital storage oscilloscope (SIGLENT SDS 2204X, Supplementary Figs. 5c–e). The actual output voltage of the boost module was restricted to 40 V due to the inductance of L2. By adjusting the variable resistance between the “boostout” pin and the “allout” pin, the voltage applied to the load can be adjusted from 0 to 40V to meet the application conditions.

After testing the performance of the rigid circuit board of the ES device (Fig. 4b and Supplementary Figs. 5b) and a flexible ES device was manufactured that included the same functional elements. This was generated with a flexible printing method using PI film specially designed with a serpentine trace structure. The flexibility is enabled by the thin film geometry and the serpentine trace give the stretchable property (Fig. 4c). The serpentine trace structure includes two free-standing columns between each electrode and the central control element. The hierarchical degree of buckling associated with the serpentine trace structure ensures low levels of strain on the wires; this facilitates a reliable and stretchable connection [55] and permits the device to be applied to wounds that vary in size. Functional elements are integrated into the central part of the ES device and surrounded by flexible electrodes, including a DC-DC boost module and a pulse modulation module. Given the nature of the programmable element, the duty cycle and stimulus frequency could be adjusted in situ and in real time. Finally, the flexible ES device and CVG were fixed by externally bonded spun-laced mesh-type cloth(Fig. 4d).

3.5. Flexible-ES chitosan dressing promotes wound healing via upregulation of VEGF and activation of PI3K/Akt and ERK1/2 signaling pathways

Our findings revealed that the flexible ES-chitosan dressing accelerated wound healing as early as day 3 after surgery (Fig. 5a). Statistical analysis of the healing times and healing rates revealed that the flexible ES-chitosan dressing promoted accelerated wound healing compared to all other conditions evaluated (Fig. 5b and c). On day 7 after surgery, the results of Western blotting of wound tissue revealed that rats in groups that underwent treatment with HVMPC or CVG expressed higher levels of VEGF than did those in the control group; likewise, higher levels of p-Akt and p-ERK1/2 were detected in wound tissues from the rats treated with HVMPC (Fig. 5d and e). Moreover, immunohistochemical staining revealed increased expression of VEGF in all three experimental groups and higher levels of p-Akt and p-ERK1/2 in the two groups that underwent treatment with HVMPC (Supplementary Figs. 6a and 6b). These data suggest that the flexible ES-chitosan dressing promotes the expression of VEGF and activation of the PI3K/Akt and ERK1/2 pathways.

3.6. Flexible ES chitosan dressing accelerated wound epithelialization and inhibited scar formation

As shown in Fig. 6a, there was more epidermal tissue generated in the wounds of experimental groups compared to the controls. H&E staining revealed that the experimental group generated well-formed epidermal tissues compared to those observed among the controls at 15 days after surgery; at this time point, the epithelium of the group treated with flexible ES-chitosan dressings and HVMPC was nearly completely closed (Fig. 6b). Wound epithelialization may be accelerated under conditions in which the cells are presented with an appropriate microenvironment like what had happened in rats treated with CVG. In addition, after healing, the scars visible among rats treated with flexible ES-chitosan dressings were visibly smaller than those of the controls (Fig. 6c). Treatment with either CVG alone or with the flexible ES-chitosan dressing resulted in smaller scars; these results suggest that administration of CVG may suppress scar formation (Fig. 6d and e). Furthermore, results from H&E staining revealed that the widths of the healed wounds were also smaller among the rats treated with CVG compared to the control groups (Fig. 6f). Although only a few of the findings reached statistical significance, results from Western blotting have demonstrated a trend suggesting that either CVG or HVMPC may suppress phosphorylation of Smad 2 (p-Smad 2) and Smad 3 (p-Smad 3, Supplementary Fig. 7a). Similar trends were also found in immunohistochemical staining assays (Supplementary Figs. 7b–d).