Improved in vitro wound healing in response to a superoxidised solution

Wound healing is a complex multifaceted process, comprising of four phases (haemostasis, inflammation, proliferation and remodelling) that involve host immune responses, inflammatory mediators and cellular migration.¹ Often compounding these processes are bacterial colonisation and formation of biofilms within the wound that not only protect invasive bacteria from elimination but can considerably impair the healing process.^2,3

Wound cleaning is a critical step in effective wound management and care, which can prevent chronic infection. Debridement and irrigation, the two main cleaning steps, are vital for the removal of dead tissue and residual debris, helping to prevent biofilm development and accelerate wound healing. Debridement methods can be selective (only targets unhealthy tissue) or non-selective (removes both healthy and unhealthy tissue),⁴ and include mechanical, biological and enzymatic approaches.⁵ Hypochlorous acid (HOCl) has been used in debridement solutions, which can improve wound healing rates by stimulating cellular migration, activating immunomodulatory transcription factors (such as transforming growth factor-β and fibroblast growth factor-2), reduce inflammation and contribute to increased oxygenation, thereby accelerating wound healing.^6,7 Furthermore, HOCl has antibacterial and antibiofilm properties, with in vitro efficacy demonstrated against Staphylococcus aureus, Pseudomonas aeruginosa, Escherichia coli and Candida albicans.^8,9,10

Superoxidised solutions are a non-alcoholic wound wash, debridement agent and cleaning solution for use on both acute and hard-to-heal (chronic) wounds. The product is indicated to clean, reduce microbial load, remove biofilm and support an improved wound healing process. Testing has shown that superoxidised solutions are non-cytotoxic, effective against bacteria (both Gram-negative and Gram-positive), fungi and spores, and can reduce biofilm by up to 99.9% in line with International Organization for Standardization (ISO) 0993, ASTM International (ASTM) 2315-03 and ASTM E2799-12 standards.¹¹

The use of reconstituted human epidermis (RHE) skin models for in vitro screening studies has been increasing, and can provide valuable information on the safety and efficacy of skin contact compounds. The EpiDerm system (MatTek Corporation, US) is a stratified full-thickness skin model comprising both dermal (fibroblast) and epidermal (keratinocyte) cells that are metabolically active, highly differentiated and organised into layers analogous to those found in vivo.¹²

In this study, we used EpiDermFT models, prewounded by the removal of a 3mm diameter section of the epidermal layer only, to track wound healing rate in response to the application of a superoxidised solution, through visualisation of both dermal fibroblasts and migrating keratinocytes using fluorescent immunohistochemistry. Select proinflammatory cytokine levels were also monitored at the same timepoints.

Method

Wound care test solution

A commercially available superoxidised solution (G21502, Hydrocyn aqua, Bactiguard South East Asia Sdn. Bhd., Malaysia), containing 0.003% HOCl as the active substance, was used in this study.

In vitro wounded skin tissue model and culture conditions

Prewounded, reconstructed full-thickness human skin models (EpiDermFT, MatTek Corporation, US) were chosen as they exhibit stratified epidermal components and a fully developed basement membrane, which resembles in vivo skin in both morphology and barrier function. Each tissue model was prewounded by the removal of a 3mm diameter section of the epidermal layer only. The skin tissues were transferred to six-well plates, with each well containing 2.5ml of maintenance medium (EFT-400-MM; MatTek, US). All plates were then incubated at 37°C; 5% carbon dioxide (CO₂) for 24 hours before commencing the wound healing assay.

In vitro wound healing assay

Triplicate prewounded tissues were treated with 10µl of either superoxidised solution or Dulbecco's phosphate buffered saline (DPBS), gently pipetted onto the wounded area. The plates were then incubated at 37°C; 5% CO₂. Models were transferred to new six-well plates containing 2.5ml fresh maintenance medium every 24 hours.

Models were subsequently treated with 10µl of superoxidised solution or DPBS every 48 hours. At each harvest timepoint (days 0, 1, 2, 5 and 7), three models for each treatment were removed for fixing and immunostaining. For measurement of proinflammatory markers, maintenance medium was collected every 24 hours following transfer of models to new six-well plates.

Immunostaining of wounded skin tissue models to assess wound closure

At each harvest time point, three wounded skin tissue models per treatment were removed and fixed with 250µl of 10% formalin for 30 minutes at room temperature. The models were then washed three times with 250µl of DPBS + 0.01% Triton X-100. Non-specific antibody binding was blocked by treating with 250µl of DPBS + 10% normal goat serum and 1% bovine serum albumin (BSA) for one hour at room temperature with shaking at 150rpm.

Primary antibodies to the target proteins cytokeratin-14 (CK14), a marker of keratinocytes (anti-CK14 antibody [mouse]; ab7800, Abcam, UK) and vimentin, a marker of dermal fibroblasts (anti-vimentin antibody [rabbit]; (ab92547, Abcam, UK) were simultaneously combined and diluted 1:250 in the same DPBS + 1% BSA. Diluted primary antibodies were then added to the models as a 250µl aliquot and incubated for two hours at room temperature with shaking at 150rpm. Following incubation, the antibody solution was removed and 250µl of DPBS + 0.1% normal goat serum was added to the models and incubated for 10 minutes at room temperature, with shaking at 150rpm. This was repeated twice for a total of three washes.

Secondary antibodies (goat anti-mouse Alexa Fluor 488 [ab150113, Abcam, UK] and Goat anti-rabbit Alexa Fluor 568 [ab175471, Abcam, UK]) were simultaneously diluted 1:400 in DPBS + 1% BSA. Subsequently, 250µl aliquots of diluted secondary antibodies were added to the models and incubated for two hours at room temperature, with shaking at 150rpm, and protected from light. Following secondary antibody incubation, the models were washed three times (two minutes per wash) with DPBS + 0.1% normal goat serum, with shaking at 150rpm, and protected from light. From each skin model, 8mm diameter biopsies were taken and then sampled for confocal microscopy.

Fluorescent confocal images were taken with a 10× objective lens using a Fluoview Fv10i confocal microscope (Olympus LifeSciences, UK). CK14 and vimentin proteins fluorescently labelled with Alexa Fluor 488 (green) and Alexa Fluor 568 (orange) (Abcam, UK) were detected using a 559-laser line and a 488-laser line. Images were combined and wound diameter was measured both vertically and horizontally using ImageJ. JS software v1.54f (https://imagej.net/ij/index.html) and expressed as arbitrary units (Figs 1 and 2).

Measurement of proinflammatory markers released by wounded skin tissue models for up to two days.

The release of proinflammatory markers by the wounded skin tissue models into the maintenance media was measured for up to two days using marker-specific enzyme linked immunosorbent assays (ELISAs). For detection of interleukin (IL)-1β (IL-1β) and tumour necrosis factor alpha (TNF-ɑ), neat maintenance media from each timpoint were analysed using commercial Human IL-1β and Human TNF-ɑ ELISA kits (Invitrogen, UK), according to manufacturer's instructions. For detection of IL-6, neat samples were first diluted 1:1000 using standard diluent (supplied by the manufacturer), and analysed using the commercial Human IL-6 ELISA kit (Invitrogen, UK) according to manufacturer's instructions. Plates were read on a Varioskan Lux microplate reader (Thermo Fisher Scientific, UK) at optical density (OD)450nm for protein detection, with each well normalised against background fluorescence at OD570nm. Obtained absorbance data was compared to a standard curve generated for each ELISA run to convert data into protein concentration (ngml or pgml).

Statistical analysis

Average vertical and horizontal wound diameters were presented as the mean±standard deviation (SD). A Student's two-tailed paired t-test was applied to assess statistical differences between the superoxidised solution samples and DPBS samples. Average ELISA protein detection data is presented as mean±SD for each proinflammatory marker per day. A one-way analysis of variance (ANOVA), with the Tukey–Kramer post hoc test, was used to assess statistical differences between the superoxidised solution samples and DPBS samples for each proinflammatory marker and timepoint. Data was considered statistically significant when p<0.05.

Results

Assessment of wound healing in response to superoxidised solution over seven days.

The baseline wound measurement was ascertained at day 0 prior to the addition of either superoxidised solution or DPBS (Table 1; Figs 3 and 4). At day 1, models treated with superoxidised solution had a significantly lower (p<0.05) vertical and horizontal diameter when compared with the day 0 measurement (Table 1; Figs 3 and 5).

At day 2, models treated with either superoxidised solution or DPBS had a significantly lower (p<0.05) vertical and horizontal diameter when compared with the day 0 measurement (Table 1; Figs 3 and 6). At day 2, both the horizontal and vertical measurements for superoxidised solution-treated samples were significantly lower (p<0.05) when compared to the DPBS-treated samples

At days 5 and 7, models treated with superoxidised solution demonstrated complete closure with no exposed dermal fibroblasts visible (data not shown). No specific immunostaining was observed with the DPBS-treated models; therefore, wound diameter could not be measured for days 5 and 7.

Measurement of proinflammatory markers released by wounded skin tissue models over seven days

Interleukin-1β

Tissue models treated with superoxidised solution demonstrated an average detectable IL-1β concentration of 33.60±7.20pgml on day 0, which reduced to 13.00±5.06pgml by day 2 (Fig 7a). Models treated with DPBS demonstrated an average detectable IL-1β concentration of 25.95±2.01 on day 0, which reduced to 23.98±10.00 by day 2. No statistical significance between treatment groups was identified at any timepoint.

Interleukin-6

Tissue models treated with superoxidised solution demonstrated an average detectable IL-6 concentration of 3.16±3.16ngml on day 0, which increased to 8.70±3.54ngml by day 2 (Fig 7b). Models treated with DPBS demonstrated an average detectable IL-6 concentration of 1.32±0.91 on day 0, which reduced to 0.93±0.73 by day 2. No statistical significance between treatment groups was identified at any timepoint.

Tumour necrosis factor-ɑ

Tissue models treated with superoxidised solution demonstrated an average detectable TNF-ɑ concentration of 2.74±0.07pgml on day 0, which reduced to 0.83±0.61pgml by day 2 (Fig 7c). Models treated with DPBS demonstrated an average detectable TNF-ɑ concentration of 2.70±0.92 on day 0, which reduced to 1.05±0.54 by day 2. No statistical significance between treatment groups was identified at any timepoint.

Discussion

Wound healing is a fundamental process that re-establishes tissue integrity and skin barrier function. A wound healing model was created by removing the top keratinocyte skin layer of a full thickness in vitro human skin model and exposing the dermal fibroblast layer. This model was used to evaluate the extent of wound closure over a period of seven days via immunostaining, in response to a treatment protocol using either superoxidised solution or DPBS. Each model was stained with primary antibodies against CK14, a marker of keratinocytes (green stain), and vimentin, a marker of dermal fibroblasts (orange stain).

Fibroblasts are known to be a primary modulator in the production of extracellular matrix,¹³ acting to produce new tissue in the wound, and keratinocytes are known to induce a supporting inflammatory reaction that play a major role in closing the injured area by aiding in the migration of additional keratinocytes and cell types.¹⁴ Successful and timely wound healing, including the avoidance of hard-to-heal wounds, requires a balance of both dermal fibroblasts and keratinocytes.¹⁵ A lack or imbalance of either cell type can lead to prolonged inflammatory responses and suboptimal production of ECM, leading to prolonged wound healing.¹⁶

Measurement of both vertical and horizontal wound diameters (as shown by immunostaining of dermal fibroblasts with anti-vimentin) indicated no significant difference between models treated with either superoxidised solution or DPBS, at days 0 or 1 (24 hours after first treatment). However, by day 2 (48 hours after first treatment), the average horizontal and vertical wound diameters of models treated with superoxidised solution were significantly reduced in comparison to DPBS-treated models, indicating an increased wound re-epithelialisation response for superoxidised solution-treated samples.¹⁷

At days 5 and 7, no specific immunostaining could be observed with the DPBS-treated models. These models showed no specific cytokeratin or vimentin binding, indicating that the DPBS tissue models died in culture after day 2 and no further wound closure occurred. If the models remained viable, specific antibody binding to keratinocytes and/or fibroblasts would have been observable. In contrast, models treated with superoxidised solution demonstrated a great abundance of specific CK14 immunostaining (data not shown) demonstrating complete wound coverage with fresh keratinocytes. Data presented in the current study demonstrates that immunostaining of dermal fibroblasts provides a suitable method for tracking keratinocyte presence for active healing. The presence of keratinocytes has been previously shown to drive polarised migration of basal and suprabasal cells, inducing differentiation and proliferation of keratinoctyes,¹⁸ indicating that treatment with superoxidised solution helps to promote active healing within an acute wound.

Previous work has demonstrated that IL-1β, IL-6 or TNF-ɑ presence is modulated during the typical wound healing process. IL-6 stimulates the migration of fibroblasts to the wound location, whereas IL-1β and TNF-ɑ are produced by M1 macrophages to induce keratinocyte growth factor.¹⁹ In the current study, no significant differences (p>0.05) were observed in the release of IL-1β, IL-6 or TNF-ɑ when comparing application of superoxidised solution-treated skin models to the DPBS-treated skin models over the first two days. Data indicates that treatment with superoxidised solution did not induce an increase in proinflammatory responses within the skin model, despite a significantly lower wound diameter by day 2. Additional components, such as tissue growth factors, vascular endothelial growth factors and monocyte chemoattractant proteins have also been shown to increase during wound healing,¹⁹ and the impact of superoxidised solution on such components could be further investigated.

Limitations

The current study demonstrates the early impact of a superoxidised solution on wound closure and the presence of certain proteins included in the wound healing process. Extending sampling to include additional timepoints would provide further information on cell migration and wound closure beyond the first two days, as examined here. Increasing replicate counts may also improve the discriminatory power of the data for statistical comparisons.

Conclusion

The application of superoxidised solution resulted in significantly improved wound closure over the first two days after treatment in comparison to treatment with DPBS only. By days 5 and 7, complete wound coverage with fresh keratinocytes could be observed following superoxidised solution treatment with the skin model. Furthermore, application of superoxidised solution did not induce significant proinflammatory effects over the first two days after treatment, despite the significantly reduced wound diameter. Overall, data confirmed that the superoxidised solution significantly increased wound closure without causing an increase in the inflammatory response during the first two days of application.

Reflective questions

• What are the advantages of using a superoxidised solution for the treatment of acute versus hard-to-heal wounds?
• How might the use of debridement solutions induce the expression of alternative cytokines?
• How would the findings from the in vitro model alter when applied to an in vivo model?