Curcumin alleviates hypertrophic scarring by inhibiting fibroblast activation and regulating tissue inflammation

Hypertrophic scar (HS) that can lead to defects in appearance and function is often characterized by uncontrolled fibroblast proliferation and excessive inflammation. Curcumin has been shown to have anti‐inflammatory and anti‐oxidative effects and to play an anti‐fibrotic role by interfering transforming growth factor‐β1 (TGF‐β1)/Smads signaling pathways.


| INTRODUC TI ON
Normal wound healing and scar maturation are inseparable from timely and orderly occurrences of fibroblast proliferation, collagen secretion, and myofibroblast differentiation.However, persistent fibroblast activation leads to extracellular matrix (ECM) accumulation and severe pathological scars that affect patients' aesthetics and function. 1[4] Treatment strategies for pathological scars mainly include physical compression, drug therapy, laser therapy and surgery.
Intralesional steroid injections are widely applied in clinic but limited by complications such as skin atrophy, telangiectasia, and hyperpigmentation. 5Many herbal ingredients have been found to possess biological activity against hypertrophic scar (HS) with negligible side effects.Curcumin is a natural polyphenolic compound extracted mainly from the rhizomes of Curcuma longa (Zingiberaceae).Curcumin usually appears as an orange-yellow crystalline powder, and it is insoluble in water but can be dissolved in dimethylsulfoxide (DMSO).Traditionally used in food coloring, curcumin is getting more attention in medical treatment because of its anti-inflammatory, antioxidant and immunomodulatory effects. 6The dressings coated with curcumin nanoformulations have great therapeutic potential for burn wounds and infectious wounds. 7Over the past decade, many studies have demonstrated that curcumin can treat fibrotic diseases by regulating transforming growth factor-β1 (TGF-β1)/Smads signaling pathway.Hsu et al. found that curcumin reduced TGF-β1/Smad2 pathway and inhibited ECM secretion from keloid-derived fibroblasts. 8Du et al. discovered that curcumin increased the radiosensitivity of urethral scars by suppressing the expression of Smad4, a common-partner Smad helping stabilize the receptor-regulated Smads. 9Curcumin can regress oxidative stress-mediated liver fibrosis by inhibiting nuclear factor-κB (NF-κB) and TGF-β1/Smad3 pathways. 10wever, few studies have focused on the roles of curcumin in regulating the TGF-β1/Smad3 pathway of fibroblasts in HS.In this study, the changes of cell proliferation and migration and the expression of α-smooth muscle actin (α-SMA, a marker of myofibroblast differentiation) of human dermal fibroblasts (HDFs) treated with TGF-β1 and curcumin were investigated.In addition to TGF-β1 and Smad3, the expression levels of TGFβ-R1 and -R2, as well as Smad4, were also detected because of their involvements in Smad3 phosphorylation and TGF-β1 signal transduction. 11,12Unlike the contractile healing process in rodents, wound repair in rabbit ears is mainly dependent on tissue regeneration, which is similar to the situation in human. 13Therefore, a rabbit ear scar model was used to further explore the in vivo regulatory effects of curcumin on fibroblasts and inflammatory cells.Cells were incubated with 2 ng/mL TGF-β1 for 24 h to simulate the microenvironment of fibroblasts in HS. 14 Curcumin (MedChemExpress, NJ, USA) was simultaneously added to the culture medium of some of the cells to final concentrations of 5, 10 or 25 μmol/L.The preliminary effects of curcumin on TGF-β1-induced fibroblast abilities including proliferation and migration were evaluated (details below).Cells with no drug treatment served as negative control.Curcumin concentration of 25 μmol/L was considered appropriate and was used in the other subsequent experiments.

| Cell Counting Kit-8 (CCK-8) assay
In order to determine a rough concentration range of curcumin, cells seeded in a 96-well plate (3 × 10 3 cells/well) were treated with TGF-β1 and gradient concentrations of curcumin (1, 5, 10, 25, 50 and 100 μmol/L) for 24 h.Afterwards, cell viability was measured using a CCK-8 assay.Briefly, cells were incubated in 100 μL of complete medium per well containing 10% CCK-8 reagent (Dojindo Molecular Technologies, Tokyo, JPN) at 37°C for 1 h.The result was presented as the absorbance at 450 nm which was measured using a microplate reader (Molecular Devices, CA, USA) and normalized to the control group.

| 5-Ethynyl-2′-deoxyuridine (EdU) staining
According to a previous method, 15 EdU assay was conducted to detect the activity of DNA replication which reflected the change of proliferation ability of HDFs.Results were expressed as the percentage of EdU-positive cells.

| Transwell migration assay
For each well, a total of 2 × 10 4 cells in 100 μL DMEM were transfered into the upper compartment of a Transwell chamber (8μm pore size, Corning, NY, USA), using 600 μL complete medium with 10% FBS as a chemoattractant in the lower chamber.After 24 h of culture, cells fixed were stained with 0.1% crystal violet for 20 min.
Afterwards, the cells staying on the upper surface of the membrane insert were removed using a cotton swab, and the number of the cells migrating to the lower surface was counted.

| Immunofluorescence
The paraformaldehyde-fixed cells were washed with phosphate buffer saline and blocked with 10% normal goat serum for 30 min at room temperature.Then cells were incubated overnight at 4°C with primary antibodies against α-SMA (1:200, rabbit anti-human) and/or t-Smad3 (1:200, mouse anti-human), followed by incubation in the dark for 1 h with goat anti-rabbit-594 IgG and goat anti-mouse-488 IgG secondary antibodies (both 1:200, Solarbio Bioscience).After stained with Heochest 33 342 (Sigma, MO, USA), cells were observed under an inverted fluorescence microscope (Nikon, Tokyo, JPN).

| Rabbit ear scar model and curcumin treatment
A total of twenty healthy adult female New Zealand rabbits aged 6 months (3.0-3.5 kg) were anesthetized by 3% pentobarbital sodium (35 mg/kg).A longitudinal rectangular wound (5.0 cm × 1.0 cm) was made on the ventral surface of a rabbit ear.After 4 weeks, the wounds almost healed with similar scars, a few of which showed small residual areas.At this moment (denoted as day 0), curcumin pre-dissolved in DMSO was diluted with sterile saline to a concentration of 25 μmol/L.One of the two ear scars of a rabbit was randomly selected as the treatment group which accepted intralesional injection of 1 mL of curcumin working solution, while the other ear scar was treated with equivalent amount of DMSO/saline solution without curcumin as the control group.Before the samples were collected, supplementary injection of drug was conducted every week (Figure 3A).

| Histopathological analysis
By the end of 4 weeks of drug treatment, there were not ulcer, skin atrophy and other side-effects occurred, and all samples (n = 20) were included in analysis.Samples including the whole scar tissue, the attached cartilage and some surrounding normal skin were completely excised, formalin-fixed, paraffin-embedded and sectioned into 2.5 μm slices, which were then subjected to hematoxylin and eosin (H&E) staining, Masson's trichrome staining and immunohistochemical staining with primary antibodies (1:100 for CD45, p-Smad3 and t-Smad3; 1:200 for α-SMA, PCNA, TGF-β1, IL-1β, CD68 and CD163).
Scar elevation index (SEI) was expressed as the ratio of the thickness of central scar (H1) to that of adjacent normal dermis (H2) (Figure 3D).Collagen content was assessed by calculating the total area of collagen fibers using ImageJ 1.8.0 software (NIH, MD, USA).
For immunohistochemistry, the field of view of central region of the scar was selected, and the immunoreactivity was expressed as the ratio of the number of immunopositive cells to that of total cells and normalized by the control group (DMSO only).

| Statistical analysis
Results are expressed as mean ± SD (n = 3 for in vitro assays and n = 20 for in vivo assays).Data were analyzed by SPSS 22.0 software (IBM Corp., CA, USA) using one-way analysis of variance method followed by Tukey's HSD test.Statistically significant difference was considered at a p value of <0.05.

| Curcumin represses TGF-β1-induced proliferation, migration and α-SMA expression of HDFs
The result of CCK-8 assay showed that incubation of HDFs with TGF-β1 increased cell activity (p < 0.001), which was inhibited by curcumin treatment in a concentration-dependent manner.Specifically, the cell activity of groups treated with curcumin at 5 and 10 μmol/L was lower than that of TGF-β1 only group (p < 0.01 and 0.001, respectively), and fell back to a level similar to the untreated control group (p > 0.05).When curcumin concentration was 25 μmol/L or above, the cell activity was dramatically depressed compared with that of TGF-β1 only group (p < 0.001) (Figure 1A).Western blotting and immunofluorescence showed that incubation with TGF-β1 upregulated the expression of α-SMA (p > 0.05) and PCNA (p < 0.01) and increased the proportion of EdU-positive cells (p < 0.05).These promotion effects were gradually diminished with the increase of curcumin concentration (Figure 1B-F).Transwell migration assay showed that motility of fibroblasts was not altered by TGF-β1 and curcumin (5 and 10 μmol/L) (p > 0.05), but was significantly inhibited by curcumin at the concentration of 25 μmol/L (p < 0.01) (Figure 1G,H).

| Curcumin inhibits phosphorylation and nuclear translocation of Smad3 in vitro
As shown by western blotting, the protein levels of endogenous TGF-β1 (p < 0.01) and its receptors (TGFβ-R1 and TGFβ-R2) (p < 0.001) were remarkably elevated in HDFs after incubation with TGF-β1.

| Curcumin reduces scar elevation and collagen deposition in rabbit ear scar model
Gross observation demonstrated that curcumin accelerated the healing of residual wounds and markedly reduced the width and thickness of the final scars (Figure 3B).H&E staining showed that the scars of the untreated control group exhibited a great thickness, a dense structure and large numbers of fibroblasts and inflammatory cells.Masson's trichrome staining showed the robust collagen fibers but with poor organization.Significant neovascularization was also observed.After curcumin treatment, scar samples presented lower SEI values (p < 0.01) and less collagen content (p < 0.01) relative to the control group.In addition, the collagen arrangement was improved, the cell density and blood vessels were reduced, and a few skin appendages were visible (Figure 3C-F).

| Curcumin regulates fibroblast activation and inflammation in rabbit ear scar tissue
The result of immunohistochemistry showed that PCNA was mainly expressed in the nucleus, while α-SMA was dominantly expressed in the cytoplasm and interstitium.PCNA-and α-SMA-positive cells were both distributed in the superficial layer of the dermis.Curcumin significantly inhibited the immunoreactivities of PCNA (p < 0.05) and α-SMA (p < 0.001), indicating lower proliferative activity of fibroblasts and degree of myofibroblast differentiation (Figure 4A,B).Curcumin simultaneously inhibited TGF-β1/Smad3 pathway in scar tissue, as manifested in the decreased staining intensities of TGF-β1 (p < 0.001), p-Smad3 (p < 0.01) and t-Smad3 (p < 0.05) (Figure 4C,D).In addition, the amounts of CD45-and IL-1β-positive cells infiltrated into scar tissue of the curcumin group were much less than that of the control group (p < 0.01 for both), which suggested the tissue inflammation was relieved.Compared with the pan-macrophage marker CD68 (p > 0.05), the inhibition of CD163 by curcumin was more apparent (p < 0.01), indicating a lower M2 polarization of macrophages (Figure 4E,F).

| DISCUSS ION
TGF-β1/Smad3 signaling pathway participates in fibroblast activation and is an important pathogenic mechanism of HS. 12  the downstream molecule of TGF-β1, can mediate up-regulation of α-SMA transcription by binding with Smad3-Binding element 1. 16 Unlike constitutively phosphorylated Smad2, the phosphorylation of Smad3 is more likely to be inducted by TGFβ, facilitating the aggregation of α-SMA in focal adhesions and the stress fibers of cytoskeletal organization. 17Induction of TGF-β1 in vitro promotes cell proliferation and α-SMA expression of fibroblasts, leading to excessive fibroblast-to-myofibroblast differentiation. 14On the contrary, the TGFβ antagonist peptide reverses TGF-β1-mediated cell migration and contraction of fibroblast-populated collagen lattice. 18e in vitro experimental part of our study revealed that curcumin did not cancel TGF-β1-induced expression of intracellular endogenous TGF-β1 and receptors 1 and 2 (TGFβ-R1 and TGFβ-R2), but still independently inhibited the phosphorylation of downstream signaling molecule Smad3, which prevents the formation and nuclear translocation of transcriptionally active Smad3-Smad4 complex. 11,12terestingly, intralesional injection of curcumin significantly reduced the content of TGF-β1 in rabbit ear scar tissue, and the result inconsistent with the in vitro study might be caused by the changes in tissue components other than fibroblasts.Scar formation is a complex process that is affected not only by fibroblasts but also by local inflammatory infiltration.Curcumin has been proved to shorten the inflammatory period and accelerate the healing process by inducing leukocyte apoptosis and restoring the imbalance between pro-oxidant and anti-oxidant activities. 19In the present study, residual wounds on the scars treated with curcumin exhibited a faster healing rate, which was probably the result of attenuation of tissue inflammation.
Macrophages coordinate the transition of micro-environment from early to latter phases of wound repair, where M1 macrophages are pro-inflammatory phenotype while subsequent M2 macrophages are associated with inflammation resolution and scar remodeling. 20However, postponement and prolongation of M2-dominant phase increase susceptibility to abnormal scar pathogenesis. 21TGF-β1 secreted by M2 macrophages is able to act in a paracrine manner to mediate myofibroblast transdifferentiation of fibroblasts, 22 and in an autocrine manner to activate M2 macrophages themselves. 23 contrast, the reduction of macrophage infiltration by the usage of integrin-αL antibody can prevent severe scar formation secondary to burn wounds. 24Therefore, timely switching from M1 to M2 polarization and the disappearance of macrophages are conducive to wound healing.In this study, we found that treating rabbit ear scars with curcumin not only lowered the infiltration of inflammatory cells (CD45+) and inflammatory factor IL-1β, but also decreased the number of macrophages (CD68+), especially M2-type macrophages (CD163+), which could explain for the lower TGF-β1 content and less scarring observed.
In this study, 25 μmol/L was confirmed as an effective therapeutic concentration of curcumin in both in vitro and in vivo conditions.However, high concentrations of curcumin induce fibroblast apoptosis by increasing the generation of reactive oxygen species (ROS), while low concentrations improve heme oxygenase-1 activity and protect cells from oxidative injury. 25Our team has recently reported that the cultivation of fibroblasts with curcumin at 10 μmol/L can promote autophagy and defend cells against ROS attack after hydrogen peroxide exposure. 26More curcumin concentrations and fibroblasts from different phases of wound repair need to be studied to elaborate the effect of curcumin on the oxidant-antioxidant balance of fibroblasts, which determines whether it is applied for wound healing or scar treatment.In addition, high-dose oral curcumin is considered safe for clinical use, which also indicates its poor systemic bioavailability. 27Topical application of drugs is a popular way to treat skin diseases.By blocking phosphorylase kinase and the NF-κB pathway, smearing curcumin gel on burn wound can reduce inflammation, accelerate healing and prevent residual scarring. 28wever, for wounds that have completely healed with scar formation, topical drugs on skin surface are difficult to penetrate and deposit into the tissues underneath due to the epithelial barrier. 29In the present study, injection of curcumin into scar solid at the early stage of hyperplasia markedly alleviated the degrees of scar elevation and collagen deposition in the rabbit ear scars, accompanied with no local adverse reactions.Therefore, we speculate that early injection of curcumin is likely to improve the prognosis for patients with continuous hypertrophic scarring, and multiple injections may be safe.However, curcumin is still expected to be rigorously evaluated for the appropriate concentration, interval and treatment course in humans before it can be used in the clinic.This article has some limitations.Firstly, this is a preliminary study, and further study is needed to increase the sample size, optimize dosing and extend observation time.Secondly, the regulatory mechanism of curcumin on inflammation control and macrophage polarization needs to be further explored.Finally, the use of novel drug delivery systems, such as injectable sustained-release microspheres, 30 may help prolong the retention of curcumin and increase its efficacy and duration of action.

| CON CLUS ION
Our study demonstrated that 25 μmol/L curcumin reduced the activation of fibroblasts induced by TGF-β1 and those in HS by simultaneously inhibiting TGF-β1 expression and Smad3 phosphorylation.
In addition, curcumin alleviated inflammatory response and modulated macrophage phenotype in scar tissue.These results indicate that curcumin may possess the potential to treat hypertrophic scarring through multiple mechanisms (Figure 5).

E TH I C S S TATEM ENT
The animal experiment was approved by the Research Ethics
Although 25 μmol/L curcumin did not change the expression of TGF-β1, TGFβ-R1 and TGFβ-R2 (p > 0.05), it significantly attenuated TGF-β1-induced Smad3 phosphorylation in HDFs (p < 0.001).The expression of Smad4 was not regulated by TGF-β1 and curcumin (p > 0.05) (Figure2A,B).In addition, dual immunofluorescence showed that curcumin treatment blocked nuclear translocation of Smad3 in HDFs with concomitant decrease in the fluorescence of α-SMA (Figure2C).These results suggested that curcumin still repressed α-SMA expression and myofibroblast differentiation by suppressing Smad3 phosphorylation and nuclear translocation, even if the level of TGF-β1 was not declined.

F I G U R E 1
Curcumin inhibits the activity of TGF-β1-induced HDFs in a concentration-depended manner.(A) A CCK-8 assay was conducted to detect the effects of different concentrations of curcumin (1, 5, 10, 25, 50, 100 μmol/L) on the viability of HDFs induced by TGFβ-1.(B, C) The protein levels of α-SMA and PCNA in HDFs treated with curcumin (5, 10, 25 μmol/L) were detected by Western bloting.The intensities of immunoblot bands were normalized to GAPDH and expressed as the fold-change of the control group.(D, E) The result of EdU staining was presented as the percentage of EdU-positive cells whose nuclei were stained in red.Scale bar = 50 μm.(F) Immunofluorescence staining for α-SMA.Scale bar = 20 μm.(G, H) The fold-change of the number of migrating cells on the Transwell membrane relative to the control group.Scale bar = 20 μm.Data are reported as means ± SD for each group (n = 3).*p < 0.05, **p < 0.01, ***p < 0.001, versus the control group; #p < 0.05, ##p < 0.01, ###p < 0.001, versus TGF-β1-only treatment group.TGF-β1: transforming growth factor-β1; CUR: curcumin (the same below).

F I G U R E 3
Curcumin attenuates the elevation and collagen deposition of rabbit ear scars.(A) The rabbit ear scar model was established 4 weeks after wound formation and treated with equal amount of DMSO or curcumin every week for 4 weeks.(B) Gross view of scars was conducted at 0, 2 and 4 weeks post initial treatment.Scale bar = 5 mm.(C, D) The first row of panels on the left represent H&E staining.The blue dashed lines show the demarcation between dermis and cartilage.Scale bar = 200 μm.The dashed boxes are enlarged in the lower panels where epidermis, dermis, and skin appendage are marked.Scale bar = 50 μm.Scar elevation index of scar was calculated as the ratio of the thickness of central scar (H1) to that of normal skin (H2, not shown).(E, F) Masson's trichrome staining.Collagen fibers were stained in blue, muscle fibers were in red, and new blood vessels were marked by red arrows.Scale bar = 100 μm.Data are the means ± SD (n = 20).**p < 0.01, versus DMSO-only group.DMSO: dimethylsulfoxide (the same below).
Committee of Huzhou Institute forFood and Drug Control (approval number: SYXK (Zhejiang) 2018-0015), and was carried out according to the Guide for the Care and Use of Laboratory Animals in Huzhou Institute for Food and Drug Control.

F I G U R E 5
Mechanism diagram.Curcumin suppresses fibroblast activation including reduced proliferation, migration and α-SMA expression, which is associated to the regulation of TGF-β1/Smad3 pathway.In addition, inhibition of macrophage M2 polarization may contribute to the decline of expression of TGF-β1 in scars treated with curcumin.