1. Introduction
Keloid and hypertrophic scars are defined as abnormal healing of injured or irritated skin during wound healing caused by a pathologically uncontrollable proliferation of fibroblasts in the dermis layer, which eventually results in overabundant accumulation of extracellular matrix (ECM) components [
1,
2,
3,
4]. Keloid scars form enlarged and extended scar tissues beyond the original wound margins and often recur after surgical excision, whereas hypertrophic scars do not extend beyond the wound boundary [
5,
6]. Although the biochemical and biophysical properties of keloid and hypertrophic scars are distinguishable, the pathophysiology of keloid formation and classification remain incompletely identified [
7]. Furthermore, many treatment methods for keloids have been introduced, including intraregional steroid injection, pressure garments, cryosurgery, laser therapy, and radiotherapy. Nevertheless, effective treatment of keloid scars with good clinical outcomes and high efficacy has been a daunting challenge in recent decades [
8,
9].
Transforming growth factor (TGF)-β1 signaling is a major pathway involved in keloid pathogenesis, with TGF-β1 representing an essential fibrotic cytokine involved in skin fibrosis and ECM remodeling [
10,
11]. Among various skin ECMs, collagens are the most abundant fibrillar proteins, and their proper reorganization and remodeling are crucial for regulating various resident cells within native skin tissues during wound-healing processes [
12]. Collagens are synthesized by activated myofibroblasts, with their synthesis and degradation orchestrated by various enzymes, including several catabolic matrix metalloproteinases (MMPs) for degradation of collagens as well as anabolic lysyl oxidase (LOX) and four lysyl oxidase-like (LOXL) family members that covalently form cross-links between collagens [
13]. Thus, emerging evidence reveals that the LOX and LOXL family are involved in various diseases related to pathogenic tissue fibrosis, including idiopathic pulmonary fibrosis (IPF), renal fibrosis, cardiac fibrosis, hepatic fibrosis and systemic sclerosis [
14,
15,
16,
17,
18]. However, their potential implications in keloid skin disorders are not fully understood. The epidermal growth factor (EGF) signaling pathway enhances fibroblast proliferation and the migration of vascular endothelial cells, as well as modulating the TGF-β1 signaling pathway [
19,
20]. A previous study reported that EGF treatment of human dermal fibroblasts downregulates ECM production, including expression of type I procollagen protein, and upregulates MMP-1 expression [
21].
Numerous reports have demonstrated the potential application of traction force microscopy (TFM) for discovering the important roles of cell–matrix/cell–cell interactions at cell–matrix interfaces in regulating various cellular behaviors, including cell adhesion, proliferation, differentiation, disease progression and tissue formation [
22,
23,
24]. There have been limited studies demonstrating the important aspects of pathophysiological imbalance between proteolytic enzyme-mediated ECM degradation and enzyme-associated ECM synthesis via collagen-cross-linking in keloid skin disorder. Therefore, in this study, we investigated whether exogenous EGF could ameliorate the fibrotic phenotypes of keloid and hypertrophic scar-derived dermal fibroblasts through modulation of ECM remodeling. Additionally, given the versatile capacity of TFM analysis to assess the cell motility of dermal fibroblasts, we employed fibronectin-conjugated polyacrylamide (PAA) hydrogels, with a normal skin-like Young’s modulus of 10.6 kPa, to measure cell–matrix traction stress in the presence of EGF in order to evaluate EGF-mediated changes in dermal fibroblast migration at cell–matrix interfaces.
3. Discussion
Although there have been numerous attempts to determine the molecular mechanism underlying skin fibrosis and myofibroblast activation in keloid skin disorders, little is known about keloid pathophysiology [
1,
3,
4,
6]. In this study, we characterized and compared the gene and protein expression profiles of scar tissue-derived fibroblasts isolated from NHK scar tissues. The initial findings provided strong evidence of distinct differences between normal versus hypertrophic and keloid scar tissues in terms of fibrotic ECM deposition at both the tissue and cellular levels. Such differences may be the consequence of aberrant ECM remodeling driven by imbalances between collagen-degrading catabolic enzymes and their collagen-cross-linking anabolic counterparts.
Despite similarities between hypertrophic and keloid scar-derived dermal fibroblasts, which exhibited an elevated expression of skin fibrosis-/myofibroblast activation-associated markers (
Figure 2) relative to normal scar-derived dermal fibroblasts, keloid dermal fibroblasts showed a higher degree of gene expression of collagen-cross-linking anabolic enzymes (
Figure 3), in agreement with previous studies [
38]. Similar to these results, previous studies also reported the overexpression of LOX and LOXL-2 in various pathological conditions characterized by fibrotic phenotypes, including IPF, renal fibrosis, cardiac fibrosis, skin aging and systemic sclerosis. Moreover, inhibition of LOX and LOXL-2 expression reduced fibrosis in animal models [
14,
15,
16,
17,
18]. These findings indicated that LOX and LOX-like family members can serve as potential therapeutic targets in skin fibrosis and keloid scar tissue formation.
Additionally, it has been extensively speculated that MMPs, as major ECM proteases, can play a pivotal role in collagen degradation and ECM remodeling during wound healing as well as scar tissue formation [
39]. Among various MMPs overexpressed in keloid scars, MMP-1, an interstitial collagenase primarily secreted by keratinocytes, plays an essential role in disrupting the collagen from its triple-helix structure and loosening cell-matrix adhesions within a wound matrix, thus promoting the re-epithelialization process [
13]. MMP-2 and MMP-9 are gelatinase proteins that degrade gelatin and remove the abnormal or unfolded collagen that previously has been cleaved by collagenase [
40]. More importantly, while these MMP-1/2/3 can promote migration of numerous cell types within skin tissue, such as fibroblasts, keratinocytes, smooth muscle cells, endothelial cells and macrophages, they are particularly elevated in keloid scar-derived fibroblasts, triggering abnormal ECM degradation which then leads to excessive ECM deposition [
41,
42]. MMP-3 is a stromelysin protein that cleaves not only collagen but also non-collagenous molecules such as proteoglycan, laminin and fibronectin, thus influencing wound contraction [
43].
We then investigated whether EGF treatment could reduce the fibrotic phenotypes of hypertrophic and keloid dermal fibroblasts via creating alterations in ECM remodeling. Previously, numerous studies demonstrated the important role of the TGF-β1 signaling pathway in keloid pathophysiology [
10,
11]. More specifically, it was reported that TGF-β1 promotes myofibroblast activation by inducing the overexpression of α-SMA, MMP-2, and MMP-9, while inhibiting MMP-1 expression, which results in excessive collagen accumulation [
12,
44]. On the other hand, the EGF signaling pathway has been shown to improve cell motility, dermal fibroblast proliferation and wound healing [
19,
21]. Interestingly, our results indicated that exogenous EGF alleviated the fibrotic phenotype only in keloid dermal fibroblasts, evident by a significantly decreased gene expression of FSP-1, α-SMA and vimentin, as well as decreased vimentin protein expression (
Figure 5). This could be achieved through modulation of LOX, LOXLs and MMP-1 expression (
Figure 6). The findings were also consistent with previous studies demonstrating the potential antifibrotic effect of exogenous EGF on various in vivo fibrosis models, including liver, heart and skeletal muscle models, where the supplemented EGF upregulated MMP-1 expression and subsequently alleviated the fibrotic phenotypes [
21,
45,
46,
47]. Although dermal fibroblasts respond to exogenous EGF, the degree of their responsiveness differs significantly based on their cellular phenotypes, such as aging and pathological conditions [
36,
37,
48]. Therefore, these findings suggest that the keloid dermal fibroblasts in our study exhibited reduced fibrotic phenotypes owing to decreased expression of LOX and LOXLs as well as significantly increased expression of MMP-1 with its higher sensitivity to EGF.
In addition to EGF-mediated ECM remodeling, we assessed the contribution of EGF to the cell–matrix interaction of dermal fibroblasts. Although emerging evidence has revealed the critical roles of focal adhesion-associated cell–matrix interactions in cell adhesion and migration across various cell types [
23,
49,
50], relatively less is known about the role of EGF in regulating hypertrophy and keloid dermal fibroblasts via the cell–matrix traction force. Here, TFM analyses indicated a strong correlation between cell–matrix traction stress and cell motility among various dermal fibroblasts. Additionally, culture of dermal fibroblasts on the normal skin-matching stiffness of the cell–adhesive matrix revealed a proportional degree of cell motility to the cell–matrix traction stress exerted by adhered dermal fibroblasts, with keloid dermal fibroblasts showing the highest cell–matrix traction stress (
Figure 4 and
Figure 7). However, treatment of keloid dermal fibroblasts with EGF decreased their cell–matrix traction stress. Similarly, recent advancements in the field of cell mechanics have enabled the observation that keloid dermal fibroblasts exhibit a greater magnitude of force generation than normal dermal fibroblasts through focal adhesion complexes during cell migration, as determined by atomic force microscopy measurements [
51]. Therefore, these TFM results suggest a correlation between the migration capacity of keloid dermal fibroblasts and its contribution to fibrotic phenotypes and myofibroblast activation. Moreover, the current findings highlight the potentially important role of cell–matrix interactions of keloid dermal fibroblasts within their microenvironment during in vivo wound healing and excess ECM production. Although characteristic differences were noted between NHK dermal fibroblasts, the principal mechanism underlying the development of keloid skin disorders remains unidentified. Nevertheless, our observations strengthen the current understanding of keloid pathophysiology and establish a basis for discovering effective therapeutic strategies to prevent and treat scar tissue formation.
In summary, this study characterized dermal fibroblasts within hypertrophic and keloid scar tissues in relation to ECM remodeling. Furthermore, EGF ameliorated the fibrotic phenotypes of keloid dermal fibroblasts by modulating collagen degradation and synthesis. In addition, the results of the cell–matrix traction force analysis indicated that this analytical method can effectively distinguish various cellular functions of dermal fibroblasts isolated from NHK scars, thus providing a reliable platform for the evaluation of subtle differences in cell–matrix interactions associated with the progression of skin-disease pathology.
4. Materials and Methods
4.1. Patients and Sample Collection
After obtaining written informed consent from all patients according to a protocol approved by the Institutional Review Board of Soonchunhyang University Bucheon Hospital (SCHBC_IRB_2017-08-010), skin tissues, including NHK scar tissues, from the central dermal layer of 20 patients (
Table 1) were obtained for fibroblast isolation and immunohistochemistry. For normal skin tissues, samples were obtained from tissue excision during breast reconstruction using a
Latissimus dorsi musculocutaneous flap. All experiments involving human subjects were conformed to the Declaration of Helsinki.
4.2. Isolation of Dermal Fibroblasts and In Vitro Cell Culture
Before enzymatic digestion of tissues, tissue fragments were placed in 15-mL conical tubes and washed with 10 mL of sterile phosphate-buffered saline (PBS) containing 1% penicillin-streptomycin (P/S; 10,000 U/mL of penicillin and 10,000 g/mL of streptomycin; Gibco-BRL, Gaithersburg, MD, USA) six times under vigorous agitation, with the tubes changed between each wash. Tissue fragments were then transferred to Petri dishes and chopped into small pieces (~1 mm3 in size) under sterile conditions. Tissue pieces were digested with serum-free high-glucose Dulbecco’s modified Eagle medium (DMEM; Gibco-BRL) containing 1% P/S and 1 mg/mL collagenase type I for 2 h at 37 °C in a shaking incubator. To stop the enzymatic digestion, an equal volume of DMEM containing 10% fetal bovine serum (FBS; Gibco-BRL) was added, and the cell suspension was passed through a 100-μm cell strainer (BD Falcon; BD Biosciences, Franklin Lakes, NJ, USA) to remove undigested tissue and cell aggregates. The filtered cells were washed twice with serum-free DMEM and seeded into 12-well plates at a seeding density of 5 × 103 cells/cm2. The cells were incubated in a humidified incubator with a 5% CO2 atmosphere at 37 °C in growth medium containing DMEM supplemented with 10% FBS, 1% L-glutamine (200 mM; Gibco-BRL), and 1% P/S.
4.3. H&E and Masson’s Trichrome Staining
Tissue specimens, which were previously fixed in 4% paraformaldehyde (PFA) for 24 h, were dehydrated using an ethanol series and embedded in paraffin. The paraffin-embedded tissues were sectioned into 10-µm-thick sections, which were stained with H&E and Masson’s trichrome for histopathologic evaluation. Tissue samples were first stained for nuclei with hematoxylin (Mayer’s modified; Abcam, Cambridge, UK) for 3 min and rinsed several times with an excess amount of water, followed by dipping several times in 1% (w/v) acid ethanol for destaining. Specimens were then stained with eosin (Sigma-Aldrich, St. Louis, MO, USA) for 20 s, dehydrated in a graded ethanol series, and treated with xylene prior to mounting. For Masson’s trichrome staining, the specimens were stained with iron hematoxylin and Biebrich Scarlet-Acid Fuchsin solution according to manufacturer’s instructions (Polyscience, Niles, IL, USA). All specimens were analyzed using an inverted microscope (Eclipse Ti-U; Nikon, Tokyo, Japan) at the Soonchunhyang Biomedical Research Core-Facility of the Korea Basic Science Institute (KBSI).
4.4. qPCR
Total RNA was extracted using Trizol reagent (Invitrogen, Carlsbad, CA, USA), and reverse transcription was performed using ReverTra Ace qPCR RT master mix with gDNA Remover (Toyobo, Osaka, Japan) according to manufacturer’s instructions. Briefly, for cDNA synthesis, 1 µg of RNA was used as a template in a 10-µL reaction, and qPCR was performed using SYBR Green real-time PCR master mix (Toyobo) on a StepOnePlus real-time PCR system (Applied Biosystems, Foster City, CA, USA) at the Soonchunhyang Biomedical Research Core-Facility of KBSI. All experiments were performed with at least three to seven biological replicates for each group, and the expression levels of genes of interest were normalized against glyceraldehyde-3-phosphate dehydrogenase (GAPDH). The ΔCt values were determined as follows: Ct
target − Ct
GAPDH, and relative fold changes were calculated using the 2
−ΔΔCt method [
52]. Primer sequences used in this study are presented in
Table 2.
4.5. Immunocytochemistry
Samples were fixed with 4% PFA for 15 min at room temperature, blocked with 1% (w/v) bovine serum albumin (BSA) in PBS, and permeabilized with 0.1% (v/v) Triton X-100 in PBS for 1 h at room temperature. Samples were incubated with primary antibody (diluted in 1% (w/v) BSA in PBS), mouse anti-vimentin (1:200; Santa Cruz Biotechnology, Dallas, TX, USA), overnight at 4 °C. Samples were then washed three times with PBS and incubated with the following secondary antibodies: antirabbit Alexa 555 (1:200; diluted in 1% BSA in PBS), anti-mouse Alexa 555 (1:200; Thermo Fisher Scientific, Waltham, MA, USA), or Alexa Fluor 488 phalloidin (1:200; Thermo Fisher Scientific) for 2 h at room temperature. Nuclei were stained with Hoechst 33342 (2 mg/mL; Thermo Fisher Scientific) for 10 min at room temperature. Fluorescence images were acquired using a confocal microscope (LSM 710; Carl Zeiss, Oberkochen, Germany) at the Soonchunhyang Biomedical Research Core Facility of KBSI.
4.6. Image Analysis
For immunofluorescence assays, cells were labeled with vimentin, F-actin and Hoechst 33342 and quantified using MATLAB (MathWorks, Natick, MA, USA). Briefly, the minimum pixel values of the images were subtracted as background noise. Individual cell and nucleus boundaries were determined from F-actin and Hoechst 33342 images using the region of interest and Otsu’s thresholding module. Mean intensity values of the expression of proteins of interest were quantified within these boundaries. The graphic representation of image analyses is presented in
Supplementary Figures S3 and S7. Additionally, morphological changes in cell shape were investigated by calculating the cell-shape index as cell circularity (4π × area/perimeter
2), as previously described [
53,
54].
4.7. Cellular Motility
To analyze the cellular mobility of fibroblasts from all three groups, a wound-healing assay was performed [
55]. Cells were seeded onto six-well plates and allowed to grow until they became fully confluent, after which a scratched straight line was created using a 1-mL pipette tip to generate wound gaps, with cell debris gently washed away with PBS. During a 48 h incubation, the wound gaps were measured daily until the gaps were filled. The wound-coverage percentage was calculated [vacant area at day 1 (or day 2)/initially scratched area at day 0] and assessed using ImageJ software (National Institutes of Health, Bethesda, MD, USA).
4.8. Cell Proliferation Assay
To examine the cell cycle of dermal fibroblasts, cells were stained and analyzed using the Click-iT EdU Alexa Fluor 488 imaging kit (Thermo Fisher Scientific). Prior to the EdU assay, all three groups of dermal fibroblasts were cultured in serum-free media overnight for cell-growth synchronization followed by incubation in EdU-containing growth medium for 3 h. The EdU assay was performed according to manufacturer instructions to indicate cells in S phase, followed by flow cytometry (Canto FACS Canto II; BD Biosciences) and immunofluorescence analyses at the Soonchunhyang Biomedical Research Core-Facility of KBSI.
4.9. Preparation of PAA Gels
PAA hydrogels with a Young’s modulus of 10.6 kPa [
30,
35] were prepared on top of glass-bottomed dishes. Briefly, the final concentrations of acrylamide and bis-acrylamide were 10% (
w/
v) and 0.1% (
w/
v), respectively, and the synthesized gels on top of the glass were 40- to 50-μm thick. Human fibronectin (cat# 356008; Corning, Oneonta, NY, USA) was conjugated with PAA using the bifunctional cross-linker N-sulfosuccinimidyl-6-[4-azido-2-nitrophenylamino] hexanoate (Sulfo-SANPAH) to ensure cell adhesion to the gel. As markers for traction force analysis, a final concentration of 0.01% (
w/
v) red fluorescent (excitation/emission = 580/605 nm) polystyrene beads (5 µm in diameter; Invitrogen) were added to the PAA solution before polymerization. The Young’s modulus of the PAA gels was confirmed using a previously described method [
56].
4.10. Cellular Traction Force Analysis
The cellular traction force was calculated as previously reported [
22,
24]. Briefly, prior to cell-seeding onto PAA gels, NHK fibroblasts were labeled with CellTracker Green CMFDA fluorescent probes (1:500; cat# C2925; Thermo Fisher Scientific), and cells were grown on a fibronectin-conjugated (10 µg/mL) PAA hydrogel in growth medium overnight to allow attachment of seeded cells to the substrate. Images of the fluorescent beads (excitation/emission = 580/605 nm) and cells were acquired in fluorescence and differential interference contrast modes using a laser confocal microscope (LSM 710; Carl Zeiss). A water immersion 40× objective lens (C-Apochromat; NA = 1.2; Carl Zeiss) was used to achieve the appropriate balance between image brightness and field-of-view size. The image pixel size was 0.10 μm/pixel. Displacement of the beads was determined by the particle image velocimetry method coded in MATLAB (MathWorks) by comparing the null-force state (in the absence of cells) and forced state (in the presence of cells) in the images. The window size was 32 × 32 pixels in the X and Y directions, resulting in a spatial resolution of 16 pixels (or 1.6 μm). The lateral stress exerted by the cells was determined from substrate deformation and equilibrium equations for the elastic substrate, as described previously [
24]. The partial differential equations were solved with the finite element method (FEM) using commercially available software (Abaqus; Dassault Systèmes, Vélizy-Villacoublay, France). The cellular traction stress was determined from the stress tensor acquired from FEM analysis.
4.11. Statistical Analysis
All values are shown as the mean ± standard error of mean of at least three to seven biological replicates for each group, and statistical significance was assessed by one-way analysis of variance (ANOVA) with Tukey’s multiple comparison test using GraphPad Prism software (* p < 0.05; ** p < 0.01; *** p < 0.001 of six or seven biological replicates).
