Characterization of an ablative fractional CO2 laser‐induced wound‐healing model based on in vitro 3D reconstructed skin

This study describes the development and characterization of a novel in vitro wound‐healing model based on a full‐thickness reconstructed skin by exposing the tissue to fractional ablative laser treatment.


| INTRODUC TI ON
Over the past decade, the number of medical aesthetics procedures performed had drastically increased, among which laser resurfacing procedure is one of the most popular procedures due to its longlasting benefits and minimal invasiveness. 1 Laser resurfacing procedure treats a variety of skin conditions ranging from wrinkles, sunspots to acne scar, and effectively delivers significant resurfacing improvement over a relative short amount of time. 2 Among the laser procedures, fractional laser resurfacing is one of the latest developments and has garnered high popularity by its capability to control the width, depth, and density of thermal damage zone while leaving adjacent tissue completely intact. 3,4 This feature facilitates skin repair post-procedure and allows the patient to enjoy minimal downtime. Many clinical trials using fractional laser resurfacing demonstrated promising performance in the treatment of wrinkles, skin rejuvenation, acne scars, and burn scars. [5][6][7][8][9][10] However, despite the clinical advancement, the molecular mechanism of this procedure is yet to be fully elucidated. Moreover, it lacks appropriate preclinical models to aid the development of pre/post-procedure skin care products.
Animal models have the advantages of available skin samples and similar skin morphology and molecular response compared with human skin. 11 However, the use of animals for research applications is inconsistent with the principle of humane experimental techniques and is a recent point of emphasis for the FDA to pivot away from the use of animal studies. 12 Ex vivo skin models reflected the intrinsic wound-healing process to some degree but are hampered with limited access to skin biopsies and donor-to-donor variations. 13,14 In vitro reconstructed three-dimensional (3D) skin models, which exhibits similar morphology and functionality to normal human skin, are well-recognized as good in vitro alternatives. [15][16][17] Relatively convenient to access, standardized reconstruction process and and consistent biological responses made them promising tools to study skin wound healing.
A number of in vitro skin models have been utilized to study the wound-healing process different types of professional lasers.
Marquardt et al. investigated the efficacy of a calcium pantothenatecontaining emulsion on a reconstructed skin model wounded with fractional CO 2 laser treatment and demonstrated that the formula was able to enhance the kinetics of wound closure. 18 Vaughan et al.
studied the ability of keratinocytes epithelialization in a full-thickness wound model induced by erbium-YAG laser, a high-powered, and a low-powered excimer laser. 19 In their study, it was shown that the keratinocytes were able to migration from the cut wound induced by the low-powered excimer laser but not from the erbium-YAG or high-powered excimer laser due to thermal denaturing. Schimitt et al.
explored the morphological and molecular changes induced by erbium:YAG laser on a low nutrition 3D skin model. 20 Utilizing the same skin model, they further studied time-dependent molecular changes for epidermis recovery and dermis remodeling post-fractional CO2 laser treatment. 21 Amann et al. reported the kinetics of epidermal recovery at gene and protein expression level post-non-ablative fractional erbium glass laser treatment. 22 These models exhibited some epidermal wound-healing features post-laser treatment, from histological level and molecular expression profile. However, a comprehensive understanding of laser-induced wound healing which covers inflammation, epithelialization, and dermis rebuilding phases will aid to disclose the dynamic of skin wound healing and further contribute to the discovery of adjunct active material that complements the laser procedure.
In this study, we developed an in vitro wound-healing model with patterned wound areas based on a well-established reconstructed full-thickness skin model, by exposing to fractional ablative CO 2 laser with controlled energy. It allowed qualitative observation of wound area by H&E staining and OCT noninvasive scanning. We further developed OCT-based quantitative method to follow the wound volume in dermis. Through the analysis of critical biomarkers, the wound-healing process of the reconstructed skin including inflammation, re-epithelialization and regeneration, and dermal remodeling has been characterized.

| Skin cell cultures
Normal human keratinocytes and fibroblasts were isolated from foreskin of adult volunteers as previously described. 23 Briefly, epidermal keratinocytes were obtained and cultured with a feeder layer of irradiated 3 T3 fibroblasts as described by Reheinwald and Green and used at early passages. 24 Normal human fibroblasts (NHF) were isolated from the dermis by explant culture method and were cultured in DMEM (Gibco) supplemented with 10% fetal bovine serum (FBS, Hyclone, Germany). Media were changed every 2 days, and the cells were cultured at 37°C and 5% CO 2 .

| Full-thickness skin equivalent reconstruction
The full-thickness skin model was reconstructed as described previously. 16 In brief, dermal equivalent was obtained by culturing a lattice of 1.5 mg/ml native collagen type I containing 0.14 × 10 6 cells/ml normal human fibroblasts for 4 days at 37°C under 5% CO 2 .
After 4 days contraction, 0.15 × 10 6 normal human keratinocytes were seeded on top of each dermal lattice and cultured for 7 days in submerged conditions, and then raised to the air-liquid interface for a further 7 days to achieve a well-stratified and differentiated epidermis.

| Fractional ablative CO 2 laser irradiation
Laser irradiation was conducted using CO 2 RE® system (Syneron, Candela). The energy level was set to Deep mode (fractional coverage 5%, core energy 30 mJ, core energy fluence 170 J/cm 2 ). The 3D full-thickness skin models were positioned on agarose gel and topically treated with clinical laser adaptor for 0.5 seconds, left with a pattern size of 9 mm in diameter. After laser treatment (Day 0), skin models were transferred to 6-well plates with culture medium, main-

| Transepidermal water loss measurement
The reconstructed skin models with or without laser treatment

| Histology and immunofluorescent staining
The laser-treated center area (9 mm in diameter) of each skin model was cut with a 10 mm-diameter punch for histological analysis and immunofluorescence staining. Half of the punched samples were fixed in 10% formaldehyde and processed for paraffin embedding. The embedded samples were cut into 5 μm sections using a microtome (Leica, RM2255) and stained with hematoxylin and eosin. For immunofluorescent staining, skin samples were embedded in OCT for cryosection. Seven micrometers sectioned slices were dried for 1 h at room temperature and then fixed in cold acetone for 5 min. After washing with DPBS twice, the slices were blocked with 0.2% BSA in DPBS for another 10 min. Ki67 (M7240, Dako, mouse 1:100) and Filaggrin (MA5-13440, Thermo Fisher Scientific, mouse, 1:100) primary antibodies were diluted in DPBS containing 1% donkey serum and incubated at room temperature for 2 h in humid chamber, followed by washing with DPBS for three times. Then, the secondary antibodies (donkey anti-mouse Alexa Fluor 488, Invitrogen A21202, 1:300) were diluted in DPBS containing 1% donkey serum incubated for 1 h at room temperature and followed by DPBS wash for three times and then stained with DAPI (Invitrogen D1306, 1:10000), followed by DPBS washing. Nikon Eclipse Ti fluorescence microscope was used for imaging.

| OCT imaging and quantitative analysis
Optical measurements were performed daily since the day of laser treatment using optical coherence tomography (OCT, Thorlabs), with an axial resolution of 6 μm (930 nm center wavelength and 2.2 mm penetration depth). For each sample, the vertical 2D scanning dimension was 9.42 × 1.5 mm, with pixel of 1024 × 512. The 3D scanning dimension was 9.42 × 9.42 × 1.5 mm, with pixel of 1024 × 512 × 697.
Epidermal thickness measurement was conducted utilizing vertical 2D scanning images with Image J. One 2D OCT image for each sample was taken, and three replicates in each group were included.
Volume calculation of dermis cavities on Day 0 and cell inclusion from Day 1 to Day 7 was performed using an internally developed software. Principally, a clear structure margin was displayed on OCT scanning images when there is physical structure changes, including laser-caused dermis cavities and cell inclusions within dermis. Dermis amorphous shape within 90 original OCT images in total from Days 0, 2, and 4 were manually labeled, referred to morphology from H&E staining. After computational neural network deep learning and training, 512 original OCT images were labeled, simulated images were generated, and corresponding volume was calculated ( Figure S1).

| Protein assays
To measure inflammation factors and secreted cytokines, tissue culture medium was collected for multifactor ELISA assay. For ICAM-1, IGFBP-2, and DKK-1 analysis, Luminex Assay (R&D Systems) was applied in a panel following the manufacturer's instructions.
To measure the ECM proteins, 10 mm in diameter center area of the skin models was punched. The samples were emerged with Tissue Lysis Buffer (R&D Systems) and cryogenic grinded for three times and then further lysed on ice for 30 min. Cellular debris was removed by centrifugation at 10000 × g for 10 minutes at 4°C. Protein concentration was determined by BCA Protein Array Kit (Absin, abs9232-500) and normalized before ELISA assay. The following ELISA kits were used: Collagen type I (Human) ELISA kit (BioVison, E4617-100), Collagen type III (Human) ELISA kit (Novus, NBP2-75855), and alpha SMA (Human) ELISA Kit (Abcam, ab240678). Medium supernatant was also collected for ELISA assay of Procollagen type I (Procollagen type I C-peptide (PIP) EIA Kit, Takara, MK101).

| Data analysis
All the fluorescent images were processed using Image J software.
All the data were statistically analyzed by GraphPad Prism and presented as mean ± standard deviation unless otherwise specified. A one-tailed Welch t-test was used for data statistical analysis. * and ** denoted significant differences with p-Value less than 0.05 and 0.01 respectively.

| Optical and histological morphology of lasertreated full-thickness skin model
The deep mode of CO 2 RE® system created a wound zone of 9 mm in diameter, containing 56 minor wounds and patterned in three circles ( Figure 1A). The diameter of each individual micro wound was 200 μm with a depth of 500-700 μm. Fractional ablative CO 2 laser treatment caused microscopic vertical channels of damaged tissues in both epidermis and dermis as revealed by HE staining ( Figure 1B).
The cavity was due to the tissue vaporization from high temperature caused by laser. The surrounding pink areas were caused by the denature effect of central high energy. The tissue areas between two micro-wounds were kept intact and undamaged and thus left health interspace tissues for fast post-wound recovery.  IGFBP-2, a cytokine secreted by dermis fibroblast to participate in keratinocyte proliferation and differentiation, was reported to correlate with epidermis thickness. 27 Multifactor ELISA assay of IGFBP-2 from tissue culture medium on Day 1 displayed downregulated expression (~0.5-fold) of IGFBP-2 by laser treatment ( Figure 5D), well correlated with the decreased epidermis thickness after laser treatment measured by OCT ( Figure 2C). interleukin family (IFN-gamma, IL-1α, IL-1ra, and ILs), and inflammation triggered dermis degradation factor (MMP-9) were investigated within a Human XL Cytokine Array Kit. Surprisingly, our current study did not yield any significant changes with the inflammatory cytokines (> twofold) triggers by laser ( Figure 6). The reported result is normalized to the respective protein level of each cytokine for the untreated control tissue.

| Extracellular matrix remodeling
One of the key hallmarks of skin wound healing is the prospect of extracellular matrix regeneration in the new tissue. Whole skin tissue extraction from the wound zone was utilized to quantify content of type I and type III collagen in skin ( Figure 7A Figure 7E). DKK-1 is a key inhibitor of Wnt signaling pathway, and a reduced level of DKK-1 suggests that the laser-wounded tissue may be prone to potential fibrosis during wound-healing process.

| DISCUSS ION
In this study, an in vitro wound-healing model was developed based on a full-thickness skin model by using fractional ablative CO 2 with stable energy. The selected equipment CO 2 RE® system, Syneron around 700 μm in total, in order to ensure that the laser did not damage the bottom membrane beneath the dermis layer. This reconstructed skin model is able to follow the key wound-healing stages: inflammation, re-epithelialization, and dermis matrix remodeling. 25,26,28,29 Under clinical conditions, fractional CO 2 laser intervention on non-facial skin reported de- Inflammation is the early phase of wound healing, where the innate immune response is activated, and diverse cell types are involved: neutrophils, monocytes, macrophages, fibroblast, and lymphocytes, which gave rise to cytokines and inducing inflammation. 29 Clinical study on ablative CO 2 laser reported instant inflammation appeared on Day 1 after laser treatment, the expression of IL-1β and The lack of ECM remodeling observed in the in vitro model could be due to the early timing of sampling. Investigating the dermal remodeling process after procedure would need longer term culture of skin models. As expected, the expression of α-SMA increased on Day 2 in the laser-treated tissues, indicating the transition of fibroblast to myofibroblast during wound healing. 30,38 In our study, an OCT image-based noninvasion method to monitor the wound closure in dermis was developed. Results from our study exhibited discrete dermis cavities on Day 0 by laser, amorphous dermis cell inclusions post-Day 0, and gradually debris extrusion from stratum corneum during wound-healing process. The wound-healing curve generated based on calculated dermis cell inclusion volume, provided valuable information to investigate dermis recovery, and might be used to evaluate the efficacy of novel active on dermal wound closure.

| CON CLUS ION
In this study, an in vitro wound-healing model was developed with patterned wound areas based on a well-established reconstructed F I G U R E 7 Extracellular matrix remodeling markers expression. (A, B) Protein quantification of Collagen type I & III within tissue on Day 0 and Day 4 post-laser treatment by ELISA assay. (C) Protein quantification of Procollagen I secretion in tissue medium on Day 2 and Day 4 by ELISA assay. (D) Protein quantification assay of α-SMA within tissue on Day 2 post-laser treatment by ELISA assay. (E) Protein quantification of DKK-1 secretion in tissue culture medium on Day 1 by multifactor ELISA assay. Mean value and SD of three replicates from each group are presented. **p < 0.01 (one-tailed Welch's t-test).
full-thickness skin model, by exposing to fractional ablative CO 2 laser with controlled energy. The wound-healing process including inflammation, re-epithelialization, and dermal remodeling has been characterized. The expression of skin regeneration markers Ki67 and Filaggrin was triggered by laser treatment, and the migration ICAM-1, epidermal thickness, as well as the dermal fibrosis biomarkers DKK-1 and α-SMA was regulated upon laser treatment. This model together with the methods and biomarkers gave us a deeper understanding of skin wound-healing process post-laser procedure from both morphology and molecular level, provided molecular evidence for the skin regeneration triggered by laser treatment, and also highlighted the necessity of active intervention post-procedure to enhance and consolidate resurfacing performance.

AUTH O R CO NTR I B UTI O N S
C.HE. and, W.ZHANG.designed the research study, performed the research, analyzed the data, and wrote the paper. Y.TU. designed the research study. L.ZHONG. and, R.WANG. contributed essential reagents or tools. Y.TENG. and, I.LIAO. wrote the paper. C.DING. designed the research study and, wrote the paper.
All authors have read and approved the final manuscript.

ACK N OWLED G M ENTS
The authors wish to acknowledge Wei WANG BS for his great support in providing performing histological section and staining of the tissue samples.

CO N FLI C T O F I NTE R E S T
This work was funded by L'Oreal Research & Innovation group. The authors declare that they have no competing financial interests.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.

E TH I C A L A PPROVA L
The authors confirm that the ethical policies of the journal, as noted on the journal's author guidelines page, have been adhered to. Cells for skin reconstruction were obtained from surgical waste with written informed patient consent with ethics approval from the review board of the Skin Research Institute of Singapore.