suppression of oxidative stress Inhibitory effect of kaempferol on skin ﬁ brosis in systemic sclerosis bythe Journal of Dermatological Science

(1)

Original Article

Inhibitory effect of kaempferol on skin ﬁ brosis in systemic sclerosis by the suppression of oxidative stress

Akiko Sekiguchi

^a,1

, Sei-ichiro Motegi

^a,

*

^,1

, Chisako Fujiwara

^a

, Sahori Yamazaki

^a

, Yuta Inoue

^a

, Akihiko Uchiyama

^a

, Ryoko Akai

^b

, Takao Iwawaki

^b

, Osamu Ishikawa

^a

aDepartmentofDermatology,GunmaUniversityGraduateSchoolofMedicine,Maebashi,Japan

bDivisionofCellMedicine,DepartmentofLifeScience,MedicalResearchInstitute,KanazawaMedicalUniversity,Ishikawa,Japan

ARTICLE INFO

Articlehistory:

Received13May2019

Receivedinrevisedform30July2019 Accepted14August2019

Keywords:

Fibrosis Kaempferol Oxidativestress Systemicsclerosis

ABSTRACT

Background:Thereisgrowingevidencethatvasculopathy-inducedhypoxiaandoxidativestressenhance theprocessofﬁbrosisinsystemicsclerosis(SSc).Kaempferolisanaturalﬂavonoidwidelyfoundin variousvegetablesandfruits,andhasbeenreportedtohaveexcellentantioxidantactivity.

Objective:Objectivewastoelucidatetheeffectofkaempferolonskinﬁbrosisandthemechanismofthe inhibitoryregulationofﬁbrosisbykaempferol.

Methods:Weassessedtheeffectofintraperitoneallyadministeredkaempferolonbleomycin-induced dermalﬁbrosisinmice.Theeffectofkaempferolonoxidativestressinbleomycin-treatedmiceandSSc ﬁbroblastswasassessedinvivoandinvitro.

Results: We identifiedthat kaempferol injection significantlyinhibited bleomycin-induced dermal fibrosisinmice.ThenumberofαSMA⁺myofibroblasts,CD3⁺T-cells,andCD68⁺macrophagesinlesional skinwassignificantlydecreasedbykaempferolinjections.Kaempferoladministrationalsosignificantly suppressedthebleomycin-inducedoxidativestresssignalinOKD48mice.Additionally,mRNAlevelsof oxidativestress-associatedfactors,suchasHO-1andNOX2,aswellasinflammatoryandpro-fibrotic cytokines,includingIL-6,TGF-βandTNFαinscleroticskinweresignificantlydecreasedbykaempferol.

Kaempferolalsoreducedbleomycin-inducedTUNEL⁺apoptoticcellsinthelesionalskinofbleomycin- treatedmice.Furthermore,theoxidant-inducedintracellularaccumulationofreactiveoxygenspecies (ROS) in SSc ﬁbroblasts was inhibited by kaempferol treatment. In addition, the oxidant-induced apoptosisofSScﬁbroblastswasdecreasedbykaempferolinvitro.

Conclusion: Kaempferol might improve bleomycin-induced fibrosis by reducing oxidative stress, inflammation,andoxidativecellulardamage.Administrationofkaempferolmightbeanalternative treatmentforskinfibrosisinSSc.

1.Introduction

Systemic sclerosis (SSc) is a connective tissue disorder characterized by the development of ﬁbrosis in the skin and internal organs, as well as microvascular dysfunction [1,2].

Vasculopathyoccursatanearly stepofSSc.Almostallpatients present Raynaud's phenomenon as an initial symptom, and frequently show other manifestations of vascular diseases, includingischemicdigitalulcers,pulmonaryarterialhypertension, andrenalarterialinvolvementassociatedwithmalignanthyper- tension and renal failure. Although the pathogenesis of SSc remains unknown, it has been elucidated that vasculopathy- inducedhypoxiaandoxidativestressplayanimportantroleinthe prognosis of SSc [3–8]. Vasculopathy in patients with SSc is characterized by the activation of endothelial cells (ECs) and altered vascular tone. These pathological changes involve the inductionofpro-inﬂammatorycytokinesandabnormalangiogen- esis/vasculogenesis regulators, as well as the loss of redox regulation,leadingtooxidativestressandhypoxia[3].Thereare manypreviousreportswhichdemonstratedthattheserumlevels of oxidative stress-related factors, such as nitric oxide, Abbreviations:ATP,adenosinetriphosphate;ECs,endothelialcells;HO-1,Heme

Oxygenase 1; IL-6, interleukin-6; Nox, nicotinamide adenine dinucleotide phosphate(NADPH)oxidase;Nrf2,NF-E2-relatedfactor2;ROS,reactiveoxygen species;SSc,systemicsclerosis;TGF-β,transforminggrowthfactor-β;TNF-α,tumor necrosisfactor-α;Trx2,thioredoxin2;TUNEL,terminaldeoxynucleotidetransferase dUTPnickend-labeling.

*Corresponding authorat: Department of Dermatology, Gunma University GraduateSchoolofMedicine,3-39-22,Showa,Maebashi,Gunma,371-8511,Japan.

E-mailaddress:[email protected](S.-i.Motegi).

1Theseauthorscontributedequallytothiswork.

https://doi.org/10.1016/j.jdermsci.2019.08.004

ContentslistsavailableatScienceDirect

Journal of Dermatological Science

j o u r n a lh o m e p ag e :w w w . j d s j o u r n a l . c o m

(2)

malondialdehyde, asymmetric dimethylarginine, hydroperoxide, and homocysteine, increased in patients with SSc [4,5]. Anti- oxidativebiomarkers,suchassuperoxidedismutaseandvitaminC, werelowerintheserumofpatientswithSSccomparedtohealthy controls[4].Inaddition,micetreatedwithagentsthatinducethe releaseoffree radicals,suchashydroxylradicalsand peroxyni- trites,exhibitedcutaneousandlungfibrosis[6]. Furthermore,it hasbeenreportedthatedaravone,afreeradicalscavenger,hasa significant inhibitory effect on fibrosis both in the bleomycin- inducedSSc mouse modeland in tight skinmice [7]. Selective nicotinamideadeninedinucleotidephosphate(NADPH)oxidases4 (NOX4)inhibitorexertedahighlyeffectiveanti-fibroticeffectin various animal models of tissue fibrosis [8]. These pieces of evidence indicate that vasculopathy-induced hypoxia and oxidativestressmightpromotefibrosisinSSc.

Flavonoids are polyphenol compounds mainly contained in plants and play an important role in our health. Kaempferol (3,4⁰,5,7-tetrahydroxyflavone)is a natural flavonoidabundantly found in tea, broccoli, apples, strawberries, and beans [9]. A numberofstudieshaveshownthatkaempferolhasawiderangeof beneficial pharmacological activities, including excellent anti- oxidation[10–13],anti-inflammation[14],andanti-tumorgrowth [15].However,theeffectofkaempferolontheskinfibrosisinmice model in vivoand themechanisms of theinhibitory effect for fibrosisby kaempferol areunclear. Therefore,in this study, we examinedtheeffectofkaempferolonskinfibrosisinbleomycin- inducedfibrosismicemodelinvivoandthemechanismsof the inhibitory effects of oxidative stress and inflammation by kaempferolinvivoandinvitro.

2.Materialsandmethods

2.1.Animals

All experiments were approved by the Gunma University AnimalCareandExperimentationCommittee(#17-047,#15-053), and carried out in accordance with the approved guidelines.

C57BL/6 micewere purchased fromthe SLC (Shizuoka, Japan).

Eight-to12-week-old micewereusedfor allexperiments.The micewere maintainedin the Instituteof Experimental Animal Research of Gunma University under speciﬁc pathogen-free conditionsandwerehandledinaccordancewiththeanimalcare guidelinesofGunmaUniversity.

2.2.Bleomycin-inducedskinﬁbrosismodel

Dermalﬁbrosiswasinducedin8-week-oldC57BL/6orOKD48 (Keap1-dependent oxidative stress detector, NO-48) mice with injectionsofbleomycin.Injectionsof300

m

^l^of^bleomycin^(Nippon

Kayaku)ataconcentrationof1mg/mlweregiven5timesperweek for2weeksaspreviouslydescribed[16,17].Toexaminetheeffect ofkaempferol,micereceivedintraperitoneallykaempferol(40mg/

kg/day)dissolvedin100

m

^lôf^DMSOôr^DMSOâlone⁵^times^per

weekfor2weeks.

2.3.Quantitativeassessmentofcollagencontent

TotalsolublecollagenintheskinwasquantiﬁedusingaSircol collagen assay (Biocolor, County Antrim, UK) according to the manufacturer’sprotocoland the previouslydescribed protocols [17,18].

2.4.Westernblotting

Western blot analyses were performed according to the previouslydescribedprotocols[16,17].Afterwashingwithice-cold

PBS,theskinsamplesweredisruptedinaRIPAbuffer.Lysateswere centrifuged,andtheresultingsupernatantsweresubjectedtoSDS- PAGE,followedbyimmunoblotanalysisusinganti-connectivetissue growthfactor(CTGF)Ab(Santacruz),anti-hemeoxygenase1(HO-1) Ab(Abcam),anti-activecaspase-3Ab(Abcam)andanti-GAPDHAb (Santa Cruz). HRP-conjugated secondary antibodies (Jackson ImmunoResearch) wereusedwith ECLprime(GEhealthcare)to imageimmunoblots.

2.5.Histologicalexamination

Sections(4

m

^m^thick)ôf^mouse^skinⁱⁿ^paraffinwerestained withhematoxylinandeosin(H&E)orMasson’strichrome.Skin fibrosiswasquantifiedbymeasuringthethicknessofthedermis, whichwasdefinedas thedistance fromtheepidermal-dermal junction to the dermal-subcutaneous junction, at 6 randomly selected microscopicfields. For immunohistochemicalstaining, tissuesectionsofmouseskinweretreatedforantigenretrieval with a pressure cooker. After blocking, the sections were incubated with anti-αSMA antibody (Sigma-Aldrich, St. Louis, MO),anti-CD3antibody(Abcam,Cambridge,UK)andanti-CD68 antibody (Bio-Rad, Hercules, CA). After washing, the sections wereincubatedwithahorseradishperoxidase-labeledpolymer- conjugatedsecondaryantibody(EnVision⁺;Dako). Finally,color was developed with 3,3⁰-diaminobenzidine tetrahydrochloride.

Numbers of αSMA⁺, CD3⁺ and CD68⁺ cells in the dermis were determinedbycountinginsixrandommicroscopicﬁeldsin4–6 micepergroup.

2.6.RNAisolationandquantitativereversetranscription-polymerase chainreaction

To analyze the mRNA expression levels in the bleomycin- inducedﬁbroticskinbyusingRT-PCR,wholeskinsamplesfromthe bleomycininjectedsitewereused.TotalRNAwasisolatedusing RNeasy Mini Kits (Qiagen, Valencia, CA) and was subjected to reversetranscriptionwithaSuperScriptIIIFirst-StrandSynthesis System for RT-PCR (Invitrogen). Quantitative RT-PCR was per- formedwiththeSYBRsystem(AppliedBiosystems,FosterCity,CA) byusingABI7300real-timePCRinstrumentation.SYBRprobesand primers werepurchased fromSigmaand TakaraBio Inc. (Otsu, Japan).Normalizationandfoldchangeswerecalculatedusingthe comparativeCtmethod.

2.7.Detectionofluminescentsignals

Detection of luminescent signals in OKD48 mice was performedasdescribedpreviously[19–21].Miceweresacriﬁced andtheskinwassurgicallyremovedandimmersedin0.3mg/mL VivoGlo^TM Luciferin, in vivo Grade (Promega, Tokyo, Japan) dissolved withPBS. Assoonas possible,thecollectedskinwas placedintheimagingchamberoftheinvivoimagingsystem(IVIS;

Perkin Elmer, Waltham, MA). Data were collected with high- sensitivity/10-min exposure and analyzed using Living Image software(Xenogen).

2.8.Apoptosisassay

The presence of apoptotic cells in the skin sections were assessedusingaterminaldeoxynucleotidetransferasedUTPnick end-labeling(TUNEL)stainingkit(RocheDiagnostics,Indianapolis, IN)asdescribedpreviously[20–23].Images(sixﬁeldspersection) were taken and visualized with a FV10i-DOC confocal laser scanning microscope(Olympus). The number of apoptoticcells

(3)

was determined by counting TUNEL and DAPI double-positive nucleiintheﬁeld,aspreviouslyreported[20–23].

2.9.ROSdetectionassayinvitro

Human dermalﬁbroblasts wereobtainedbyskinbiopsiesof affecteddorsalforearmareasfrom3diffusecutaneoustypeofSSc patients,accordingtotheclassiﬁcationbyLeRoyetal.,[24]and age,raceandgendermatchedhealthyvolunteers.Thestudywas approved by the local research ethics committee of Gunma University.ThisstudywasconductedaccordingtotheDeclaration ofHelsinkiprinciples.Cellswerestimulatedwith1mMH2O2with orwithoutkaempferolfor2h,andthenthereactiveoxygenspecies (ROS) levels were measured with the DCFDA Cellular ROS Detection Assay Kit (Abcam) according to the manufacturer’s protocolandaspreviouslyreported[20,21,25].

2.10.Apoptosisandnecrosisanalysiswithﬂowcytometry

Flow cytometric analysis of apoptosis was performed as described previously [20,21,25]. SSc fibroblasts wereincubated incontrolmediumorkaempferolwithorwithoutH2O2(0.5mM) for12h.Cellsweretreatedwithfluoresceinisothiocyanate(FITC)- conjugatedAnnexinV(BDBioscience)and7-amino-actinomycinD (7-AAD),andanalyzedwithaFACSCaliburflowcytometer(Becton Dickinson,SanJose,CA).CellsthatstainedpositiveforAnnexinV andnegativefor7-AADwereconsideredtobeapoptoticcellsand staineddoublepositiveforAnnexinVand7-AADwereconsidered tobenecroticcells.

2.11.Statisticalanalysis

Pvalues werecalculated usingone-wayanalysisofvariance, followedbyBonferroni’s　posttest.Errorbarsrepresentstandard errorsofthemean,and thenumbersofexperiments(n)areas indicated.

3.Results

3.1.Kaempferolsuppressedthebleomycin-induceddermalﬁbrosisin mice

First,toassesstheeffectofkaempferolonskinfibrosisinvivo, wecomparedthebleomycin-inducedskinfibrosistreatedwithor withoutkaempferol. Bleomycin-enhanceddermalthicknesswas significantly suppressed by kaempferol injections (Fig. 1A, B).

Dermalthicknessinmicewithoutbleomycintreatmentwasnot changedbykaempferolinjections(datanotshown).Weconfirmed that the amount of collagen in the skin revealed by Masson’s trichromestainingwasenhancedbybleomycin,andthisenhance- mentwasinhibitedbykaempferolinjections(Fig.1C).Inaddition, theincrease in the amountof collagenin the skininduced by bleomycintreatmentwas alsosignificantlyinhibitedbykaemp- feroltreatment(Fig.1D).Connectivetissuegrowthfactor(CTGF)is the member of matricellular proteins associated with fibrosis.

CTGFisoverexpressedinSScfibroblastsandplaysanimportant roleintheproductionandmaintenanceoffibroticlesionsinSSc [26].Wefoundthatbleomycin-inducedCTGFexpressionwasalso inhibitedbykaempferoltreatment(Fig.1E).ThenumberofαSMA⁺ myofibroblasts in lesional skin in bleomycin-treated mice was increased, and these cells were significantly decreased by kaempferol injections (Fig. 1F). These results suggested that kaempferol might have the potential to suppress skin fibrosis inducedbybleomycininjectioninvivo.

3.2.Kaempferolinjectionreducedoxidativestressinthebleomycin- inducedﬁbroticskininvivo

Sincetheexcellentanti-oxidantactivityofkaempferoliswell known[10–13], we next examined theeffectof kaempferol on oxidativestressintheskinusingthebleomycin-inducedfibrosis mousemodel.OKD48micehaveatransgeneencodingamodified Nrf2,whichisanessentialtranscriptionfactorfortheexpressionof anti-oxidative stress genes[19,27]. By using this mouse strain, oxidative stress in vivo could be detected with luminescence signals [19–21]. After injectionof bleomycin for 5 days, strong luminescence signal was detected in the bleomycin-induced fibroticskin,andthissignalwasmarkedlyreducedbykaempferol injection (Fig. 2A, B). Additionally, mRNA expression levels of oxidativestress-associated factors, includingHO-1, NADPH oxi- dase2(NOX2),andthioredoxin2(Trx2)inthebleomycin-induced fibrotic skin were examined using real-time PCR. HO-1 is an importantenzymewithantioxidanteffect[28].Biologicalfunction ofNOXenzymesisthegenerationofreactiveoxygenspecies(ROS) [29]. Trx2 is a multi-functional protein which is essential for protectionagainstoxidant-inducedapoptosis[30].Wefoundthat mRNAexpressionsofHO-1andNOX2weresignificantlyenhanced bybleomycin treatment, andbleomycin-inducedmRNA expressions of HO-1 and NOX2 were significantly inhibited by the administrationofkaempferol(Fig.2C).Bleomycin-inducedprotein expressionofHO-1wasalsosignificantlyinhibitedbykaempferol treatment(Fig.2D).However,Trx2mRNAlevelsintheskintreated withbleomycinwerenotchangedbykaempferoltreatment.These resultssuggestedthatbleomycin-inducedoxidativestressinthe skinmightbeinhibitedbykaempferolinjection.

3.3.Kaempferolsuppressedthenumberofinflammatorycellsandthe inflammatoryandpro-fibroticcytokinesintheskinofbleomycin- treatedmiceinvivo

Ithasbeenreportedthatbleomycin-treatedmicealsomimic the inflammatory component of patients with SSc [31]. Since kaempferolhasbeenreportedtobringbenefitsofattenuationof inflammation[14,32],wenextinvestigatedtheeffectofkaemp- ferolonbleomycin-inducedinflammatorycells andcytokinesin mice.Afterinjectionofbleomycinfor5days,thenumberofCD3⁺T cellsandCD68⁺macrophagesinthedermisofbleomycin-treated micewassignificantlyhigherthanthoseinthenon-treatedmice, and kaempferol treatment significantly reduced thenumber of thesecells(Fig.3A).Additionally,themRNAexpressionlevelsof inflammatory and pro-fibrotic cytokines, including interleukin (IL)-6, tumor necrosis factor alpha (TNF-α), and transforming growth factor beta (TGF-β), in the lesional skin of bleomycin- treatedmicewereenhanced,andtheselevelsweresignificantly reducedbykaempferoltreatment(Fig.3B).Theseresultssuggested thatkaempferolmightsuppresstheinfiltrationofinflammatory cells and the inflammatory and pro-fibrotic cytokines in the lesionalskinofbleomycin-treatedmiceinvivo.

3.4.Kaempferolreducedbleomycin-inducedapoptosisinvivoand inhibitedtheoxidant-inducedintracellularaccumulationofROSand apoptosisinSScﬁbroblastsinvitro

Wenextexaminedtheeffectofkaempferolonthenumberof apoptoticcellsinthelesionalskinofbleomycin-treatedmice.After injectionofbleomycinfor5days,theincreaseinthenumberof TUNEL⁺ apoptotic cells in the bleomycin-injected site was signiﬁcantlydecreasedbykaempferolinjection(Fig.4A).Wealso examined caspase-3 activity, which is the apoptosis-related marker.Bleomycin-inducedcaspase-3activityinthelesionalskin was signiﬁcantly inhibited by kaempferol treatment in vivo

(4)

Fig.1.Kaempferolsuppressedthebleomycin-induceddermalfibrosisinmice.(A,C)RepresentativeimagesofH&Estaining(A)orMasson-Trichromestaining(C)oftheskin inmicetreatedwithsubcutaneousinjectionsofcontrolPBSorbleomycin,andtreatedwithintraperitonealinjectionofkaempferolorcontrolDMSO.Scalebar=50m^m.^(B) Quantificationofdermalthicknessofthelesionalskininmice.Valuesweredeterminedinthreerandommicroscopicfieldsinn=5–6micepergroups.(D)Theamountof solublecollagenamountintheskintissuetreatedwithcontrolPBSorbleomycin,andtreatedwithintraperitonealinjectionofkaempferolorcontrolDMSO.n=3miceper

(5)

(Fig. 4B). It has been reported that ROS is produced from SSc fibroblasts by oxidative stress, and ROS can cause several abnormalities,suchasECsdamageorenhancedplateletactivation, leadingtoupregulationoftheexpressionofadhesionmoleculesor secretionofinflammatoryorfibrogeniccytokines,includingPDGF and TGF-β [33]. Therefore, we next investigated the effect of kaempferol onH2O2-induced intracellular ROSaccumulation in SScfibroblasts.H2O2-inducedintracellularROSaccumulationwas suppressedbykaempferoltreatment(Fig.4C).Next,theeffectof kaempferol on H2O2-induced apoptosis in SSc fibroblasts was examined usingflowcytometry. The number of apoptotic cells

(Annexin V⁺ and 7-AAD) was increased by H2O2 treatment (Fig. 4D). On the other hand, administration of kaempferol significantly reduced these cells in a dose-dependent manner (Fig.4D).Spadonietal.reportedthattheexpressionlevelsofNOX2 increased in SSc skin fibroblasts compared to normal skin fibroblasts[34].Therefore,wenextexaminedthemRNAexpres- sionsofNOX2inthenormalandSScfibroblaststreatedwithor withoutkaempferol.Similartopreviousresults,weidentifiedthat theexpressionofNOX2wasincreasedinSScfibroblastscompared tonormalskinfibroblasts(Fig.4E).Inaddition,wefoundthatthe increasedexpressionofNOX2inSScfibroblastswassignificantly Fig.2.Kaempferolinjectionreducedoxidativestressinthebleomycin-inducedfibroticskininvivo.(A)RepresentativeimageofluminescencesignalsintheskinofOKD48 micetreatedwithsubcutaneousinjectionsofcontrolPBSorbleomycin,andtreatedwithintraperitonealinjectionofkaempferolorcontrolDMSOfor5days.Thecolorscale barshowsthephotoncounts(photon(p)/sec/cm²/sr).(B)QuantificationofluminescencesignalsinbleomycinorPBS-injectedskinsiteinOKD48mice.n=3–5ineachgroup.

(C)mRNAlevelsofoxidativestress-associatedfactors,HO-1,NOX2andTrx2intheskintreatedwithsubcutaneousinjectionsofbleomycinofcontrolPBS,andtreatedwith intraperitonealinjectionofkaempferolorcontrolDMSOfor5days.n=4–8miceineachgroup.(D)ProteinlevelsofHO-1intheskintissuetreatedwithcontrolPBSor bleomycin,andtreatedwithintraperitonealinjectionofkaempferolorcontrolDMSObyimmunoblotting.n=5micepergroups.QuantiﬁcationofrelativelevelsofHO-1were accomplishedviadensitometryusingImageJ.mRNAandproteinlevelsincontrolmicewereassignedvaluesof1.ValuesrepresentmeanSEM.**P<0.01,*P<0.05.

groups.(E)ProteinlevelsofCTGFintheskintissuetreatedwithcontrolPBSorbleomycin,andtreatedwithintraperitonealinjectionofkaempferolorcontrolDMSOby immunoblotting.n=3micepergroups.QuantificationofrelativelevelsofCTGFwereaccomplishedviadensitometryusingImageJ.Theexpressionlevelincontrolmicewas assignedavalueof1.(F)ThenumberofαSMA⁺myofibroblastsinthedermiswasdeterminedbycountingcells.RepresentativeimagesofαSMAstainingintheskin.Values weredeterminedinsixrandommicroscopicfieldsinn=4–6micepergroups.*P<0.05,**P<0.01.Scalebar=50m^m.

(6)

inhibited by kaempferol treatment (Fig. 4E). These results suggestedthat kaempferol might reduce apoptoticcells invivo as well as ROS production and oxidative stress-induced cell apoptosisinvitro.

4.Discussion

Asoxidativestressimpactsallaspectsofthepathophysiologyof SSc,itconstitutesaninterestingtherapeutictarget,andthereis increasingevidencethattheefﬁcacyofanti-oxidativetherapyis

very encouraging [7,8,35,36]. Yoshizaki et al. reported that edaravone, a free radical scavenger,hasa signiﬁcantinhibitory effectonﬁbrosisinSScmousemodel[7].Piera-Velazquezetal.

reportedthatselectiveNOX4inhibitorexertedahighlyeffective anti-ﬁbroticeffectinvariousanimalmodelsoftissueﬁbrosis[8].

Toyamaetal.recentlyreportedthatdimethylfumarate(DMF),an FDA-approved anti-oxidative and anti-inflammatory agent,was showntohavea beneficialeffectforskinfibrosisinbleomycin- induced fibrosis mice model [35]. Polyphenols are natural antioxidants present in many plant foods, and among them, Fig.3.Kaempferolsuppressedtheamountofinflammatorycellsandtheinflammatoryandpro-fibroticcytokinesintheskinofbleomycin-treatedmiceinvivo.(A)The numberofinfiltratingCD3⁺TcellsandCD68⁺macrophagesinthedermiswasdeterminedbycountingcells.RepresentativeimagesofCD3andCD68stainingintheskin.

Valuesweredeterminedinsixrandommicroscopicfieldsinn=4–6micepergroups.Scalebar=50mm.(B)mRNAlevelsofinflammatoryandpro-fibroticcytokines,IL-6,TNF- αandTGF-βintheskintreatedwithsubcutaneousinjectionsofbleomycinofcontrolPBS,andtreatedwithintraperitonealinjectionofkaempferolorcontrolDMSOfor5days.

n=4–8miceineachgroup.ThenumberofcellsormRNAlevelsincontrolmicewereassignedvaluesof1.ValuesrepresentmeanSEM.*P<0.05,**P<0.01.

(7)

epigallocatechin-3-gallate(EGCG)presentingreenteaextractsisa scavengeroffreeradicalsandiseffectiveindecreasingoxidative stressinSSc[36].

Kaempferolisanaturalflavonoid,andtheanti-inflammatory effect of kaempferol has been demonstrated in several inflammatorydiseasesinanimalmodels.Forexample,kaempferol markedly reduced the symptoms of lipopolysaccharide (LPS)-inducedmastitisinmiceandinhibitedtheexpressionsof

inﬂammatory cytokines, such as TNF-α, IL-6, and IL-1β [37].

Regarding theanti-oxidative effectof kaempferol,kaempferol can actas a potentscavengerof freeradicals andsuperoxide radicals,resultinginthepreventionandtreatmentofoxidative stress[10–13].Ithasbeenreportedthatkaempferol inhibited cytokine- or glutamate-induced ROS production in human umbilicalvein endothelialcells orneuronalcells,respectively [11,12].Additionally,thediethylnitrosamine-inducedreduction Fig.4.Kaempferolreducedbleomycin-inducedapoptosisinvivoandinhibitedtheoxidant-inducedintracellularaccumulationofROSandapoptosisandnecrosisinSSc ﬁbroblastsinvitro.(A)After5daysinjectionofbleomycin,thenumberofapoptoticcellsinbleomycin-injectedskinwasdeterminedbycountingbothTUNELandDAPIpositive cells.Valuesweredeterminedin8–9randommicroscopicﬁeldsinn=3micepergroups.Thenumberofapoptoticcellsincontrolmicewasassignedavalueof1.Scalebar

=20m^m.^(B)Âfter⁵^daysînjectionôf^bleomycin,^theâmountôfâctive^caspase-3ⁱⁿ^the^control^PBSôrbleomycin-injectedskintreatedwithorwithoutkaempferolwas examinedbyimmunoblotting.n=3micepergroups.Quantificationofrelativelevelsofactivecaspase-3wereaccomplishedviadensitometryusingImageJ.(C)Quantification ofH2O2-inducedintracellularROSproductioninSScfibroblaststreatedwithorwithoutkaempferol.ROSformationincellswithouttreatmentwasassignedavalueof1.n=6 ineachgroup.(**P<0.01,*P<0.05relativetocellstreatedwithH2O2andwithoutkaempferol)(D)Quantitationofthetotalamountofapoptoticcells(AnnexinV⁺and7- AAD)inSScfibroblastswithorwithoutH2O2and/orkaempferol.(E)mRNAlevelofNOX2expressioninnormalandSScfibroblaststreatedwithorwithoutkaempferol.n=3.

ValuesrepresentmeanSEM.**P<0.01,*P<0.05.

(8)

ofmRNAexpressionofantioxidantenzymes,such ascatalase, glutathione peroxidase,and glutathione-S-transferase, can be restored by kaempferol [13]. Based on these findings, we hypothesized that kaempferol might suppress the fibrosis of skin in bleomycin-induced scleroderma mice model. In the present study, for the first time, we have shown that the intraperitonealinjectionofkaempferolsignificantlysuppressed bleomycin-inducedskinfibrosis,aswellasbleomycin-induced oxidativestress,andinflammationinvivo.

InSSc,thefibroticprocessisthoughttobeinitiatedbyvascular injurywithischemia-reperfusionphenomenawhichisaccompa- niedbyoveractivationofNOXinECsandasubsequentreleaseof ROS [38]. Fibroblasts from patients with SSc constitutively produce a greater amountof ROS, suchas H2O2,through NOX activation, and ROS acts as signaling molecules and increases geneexpressionsofcollagentypeIandαSMA[39,40].Consistent withthesefindings,weidentifiedthattheamountofcollagenand the number of αSMA⁺ myofibroblasts in the lesional skin in bleomycin-treated mice were significantly reduced by kaemp- ferolinjection.

We haveshown for thefirst time that oxidativestress was increased at skin fibrotic lesion induced by bleomycin using a visualization model. Furthermore, we confirmedthat oxidative stresswasreducedbykaempferolinjection.NOX2appearstobe the most widely distributed among NOX isoforms, such as neutrophils, macrophages, ECs, fibroblasts, and hematopoietic cells. NOX2 is crucial for the differentiation of monocytes to macrophages and for the polarization of M2 profibrotic type macrophages [41]. In addition, it hasbeenreportedthat NOX2 contributestoROSgenerationinSScfibroblasts,playingacritical roleincellactivation,includingexpressionsofcollagentypeIand αSMA [34]. In our results, mRNA expressions of NOX2 were significantlyenhanced bybleomycin treatment,and bleomycin- inducedmRNAexpressionsofNOX2weresignificantlyinhibitedby theadministrationofkaempferol,suggestingthattheregulationof NOX2bykaempferolmightbeassociatedwiththeregulationofthe fibroticresponses.However,Trx2mRNAlevelsintheskintreated withbleomycinwerenotchangedbykaempferoltreatment.We

alsofoundthatbleomycinandkaempferoltreatmentsigniﬁcantly increasedtheexpressionofTrx2intheskincomparedtocontrol.

However,itisdifﬁculttointerpretthisphenomenon,andfurther investigationiswarranted.

IncreasedproductionofROSisresponsiblefortheactivationof lymphocytes, macrophages, and fibroblasts, and triggers the productionofpro-inflammatorycytokinesfromthesecells,such asIL-1β,TNF-α,andIL-6 [42].Additionally,ithasbeenreported thatTGF-βincreasedoxidativestressbyreducingtheantioxidant synthesisandpromotingROSproduction[43].Weidentifiedthat kaempferoltreatment significantlyreduced thenumberofCD3⁺ TcellsandCD68⁺macrophages,aswellasmRNAexpressionlevels ofinflammatoryandpro-fibroticcytokines,includingIL-6,TNF-α, andTGF-βinthelesionalskin,suggestingthatkaempferolmight suppresstheproductionofROS,resultinginthesuppressionofthe infiltrationofinflammatorycellsandtheinflammatoryandpro- fibrotic cytokine production in the lesional skin of bleomycin- treatedmiceinvivo.

IncreasedROSproductionintheskincauses oxidativestress, whichcanleadtoapoptosis.Ithasbeenreportedthatoxidative stress-inducedapoptoticcells wereincreasedin thebleomycin- inducedfibroticskin,andtransplantationof mesenchymalstem cellsoverexpressingTrx-1significantlyinhibitedoxidativestress and dermalfibrosisinducedbybleomycin [44]. Inaddition,SSc fibroblasts wereextremelysensitivetooxidativestress-induced apoptosis[39].Similartothesepreviousresults,wedemonstrated that kaempferolinjectionreducedapoptoticcells inbleomycin- treatedskininmiceinvivoaswellasROSproductionandoxidative stress-inducedcellapoptosisinSScfibroblastsinvitro.

On the basis of our results, we propose a model of the mechanism by which kaempferol suppresses skin fibrosis in bleomycin-inducedSScmousemodel(Fig.5).Inthelesionalskin of bleomycin-treated mice, vascular injury-induced hypoxia might increase ROS production, leading to the activation of fibroblasts,Tcells,and macrophages,and subsequentlyincrease collagenproductionfromfibroblasts.Kaempferolinjectionmight significantlyreducetheamountofROS,thenumberofmyofibroblasts, T-cells, macrophages, and inflammatoryand pro-fibrotic

Fig.5. Modelofthemechanismbywhichkaempferolsuppressesskinfibrosisinthebleomycin-inducedSScmousemodel.Inthelesionalskinofbleomycin-treatedmice, vascularinjury-inducedhypoxiamightincreaseROSproduction,leadingtotheactivationoffibroblasts,Tcellsandmacrophages,andsubsequentlyincreasecollagen productionfromfibroblasts.KaempferolinjectionmightsignificantlyreducetheamountofROS,thenumberofmyofibroblasts,T-cellsandmacrophages,andinflammatory andpro-fibroticcytokines,includingIL-6,TGF-βandTNFαinlesionalskin.Thesebeneficialeffectofkaempferolmightimprovebleomycin-inducedskinfibrosis.Additionally, kaempferolalsoinhibitsextracellularATP-inducedIL-6/collagentypeIproductioninvivobytheinhibitionofATP:P2Y2receptorsignaling.

(9)

cytokines,includingIL-6, TGF-β, and TNFα in thelesional skin.

Thesebeneﬁcialeffectsofkaempferolmightimprovebleomycin- inducedskinﬁbrosis.

In addition, kaempferol has a function of purinergic P2Y2

receptorantagonist[45,46].P2Y2receptorisoneofthereceptorsof extracellularadenosinetriphosphate(ATP)[47].ATP,theubiqui- toussourceofenergyincells,canbereleasedfromcellsbyvarious stimulations,suchasmechanicalstress,tissueinjury,inflammation, and hypoxia [47–49]. Recent evidence suggests that extracellular ATP can serve as damage-associated molecular patterns(DAMPs), whichinitiatesaninflammatoryresponseby autocrine/paracrinesignaling[49].ThereleasedATPisrecognized byplasmamembrane-localizedpurinergicreceptors,suchasP2X andP2Y[47].P2XreceptorsareATP-gatedionchannels,andP2Y receptor is a G protein-coupled receptor [47]. We recently identifiedthat hypoxia enhanced ATP release,and extracellular ATP-inducedphosphorylationofp38viaP2Y2receptorenhanced IL-6and collagentype Iproduction in SScfibroblasts [50]. Our previousfindings suggested that kaempferol alsomight inhibit extracellularATP-inducedIL-6/collagentypeIproductioninvivo bytheinhibitionofATP:P2Y2receptorsignaling.

In conclusion, kaempferol inhibited collagen production via through reducing oxidative stress, inﬂammation, and oxidative cellular damage in bleomycin-induced skin ﬁbrosis model. In addition,kaempferolmightinhibitATP-inducedIL-6production, leadingtoadecreaseincollagenproductioninthismodel.This studysuggeststhatkaempferolcouldbeanalternativetreatment for skin sclerosis in patients with SSc. Because it has been consideredthatoxidativestressinjury mightbeassociatedwith thevasculopathy[3,4,20,21],ourresultssuggestthatkaempferol alsomaybeeffectiveforvasculopathyinSScpatients.However, furtherexaminationisrequired.

Fundingsources

Nofundingsourcesforthiswork DeclarationofCompetingInterest

Theauthorsdeclarenoconﬂictsofinterest.

AppendixA.Supplementarydata

Supplementarymaterialrelatedtothisarticlecanbefound,inthe onlineversion,atdoi:https://doi.org/10.1016/j.jdermsci.2019.08.004.

References

[1]C.P.Denton,D.Khanna,Systemicsclerosis,Lancet390(2017)1685–1699.

[2]O.Ishikawa,H.Ishikawa,Macrophageinﬁltrationintheskinofpatientswith systemicsclerosis,J.Rheumatol.19(1992)1202–1206.

[3]D.Abraham,O.Distler,Howdoesendothelialcellinjurystart?Theroleof endothelininsystemicsclerosis,ArthritisRes.Ther.9(2007)S2.

[4]J.Y.Luo,X.Liu,M.Jiang,H.P.Zhao,J.J.Zhao,Oxidativestressmarkersinbloodin systemicsclerosis:ameta-analysis,Mod.Rheumatol.27(2017)306–314.

[5]S.Motegi,S.Toki, K.Yamada, A.Uchiyama,O. Ishikawa,Elevatedplasma homocysteinelevelispossiblyassociatedwithskinsclerosisinaseriesof Japanesepatientswithsystemicsclerosis,J.Dermatol.41(2014)986–991.

[6]A.Servettaz,C.Goulvestre,N.Kavian,C.Nicco,P.Guilpain,C.Chéreau,etal., SelectiveoxidationofDNAtopoisomerase1inducessystemicsclerosisinthe mouse,J.Immunol.182(2009)5855–5864.

[7]A.Yoshizaki,K.Yanaba,A.Ogawa,Y.Iwata,F.Ogawa,M.Takenaka,etal.,The speciﬁcfreeradicalscavengeredaravonesuppressesﬁbrosisinthebleomycin- inducedandtightskinmousemodelsofsystemicsclerosis,ArthritisRheum.

63(2011)3086–3097.

[8]S. Piera-Velazquez,S.A. Jimenez,Role of cellular senescence and NOX4- mediatedoxidativestressinsystemicsclerosispathogenesis,Curr.Rheumatol.

Rep.17(2015)473.

[9]S.M.Somerset,L.Johannot,DietaryﬂavonoidsourcesinAustralianadults, Nutr.Cancer60(2008)442–449.

[10]M.Erben-Russ,W.Bors,M.Saran,Reactionsoflinoleicacidperoxylradicals withphenolicantioxidants:apulseradiolysisstudy,Int.J.Radiat.Relat.Stud.

Phys.Chem.Med.52(1987)393–412.

[11]I.Crespo,M.V.García-Mediavilla,B.Gutiérrez,S.Sánchez-Campos,M.J.Tuñón, J.González-Gallego,Acomparisonoftheeffectsofkaempferolandquercetin oncytokine-inducedpro-inﬂammatorystatusofculturedhumanendothelial cells,Br.J.Nutr.100(2008)968–976.

[12]E.J.Yang,G.S.Kim,M.Jun,K.S.Song,Kaempferolattenuatestheglutamate- inducedoxidativestressinmouse-derivedhippocampalneuronalHT22cells, FoodFunct.5(2014)1395–1402.

[13]I. Crespo, M.V. García-Mediavilla, M. Almar, P. González, M.J. Tuñón, S.

Sánchez-Campos,etal.,Differentialeffectsofdietaryﬂavonoidsonreactive oxygenandnitrogenspeciesgenerationandchangesinantioxidantenzyme expressioninducedbyproinﬂammatorycytokinesinChangLivercells,Food Chem.Toxicol.46(2008)1555–1569.

[14]P.Rajendran,T.Rengarajan,N.Nandakumar,R.Palaniswami,Y.Nishigaki,I.

Nishigaki,Kaempferol,apotentialcytostaticandcureforinﬂammatory disorders,Eur.J.Med.Chem.86(2014)103–112.

[15]Q.Dang,W.Song,D.Xu,Y.Ma,F.Li,J.Zeng,etal.,Kaempferolsuppresses bladdercancertumorgrowthbyinhibitingcellproliferationandinducing apoptosis,Mol.Carcinog.54(2015)831–840.

[16]Y.Yokoyama,A.Sekiguchi,C.Fujiwara,A.Uchiyama,A.Uehara,S.Ogino,etal., InhibitoryregulationofskinﬁbrosisinsystemicsclerosisbyApelin/APJ signaling,ArthritisRheumatol.70(2018)1661–1672.

[17]C.Fujiwara,A.Uehara,A.Sekiguchi,A.Uchiyama,S.Yamazaki,S.Ogino,etal., SuppressiveregulationbyMFG-E8oflatenttransforminggrowthfactorβ- Inducedfibrosisviabindingtoαvintegrin:significanceinthepathogenesisof fibrosisinsystemicsclerosis,ArthritisRheumatol.71(2019)302–314.

[18]N.M.Thoua,E.C.Derrett-Smith,K.Khan,A.Dooley,X.Shi-Wen,C.P.Denton, Gutﬁbrosiswithalteredcoloniccontractilityinamousemodelof scleroderma,Rheumatology(Oxford)51(2012)1989–1998.

[19]D.Oikawa,R.Akai,M.Tokuda,T.Iwawaki,Atransgenicmousemodelfor monitoringoxidativestress,Sci.Rep.2(2012)229.

[20]S.Motegi,A.Sekiguchi,A.Uchiyama,A.Uehara,C.Fujiwara,S.Yamazaki,etal., Protectiveeffectofmesenchymalstemcellsonthepressureulcerformationby theregulationofoxidativeandendoplasmicreticulumstress,Sci.Rep.7(2017) 17186.

[21]A.Sekiguchi,S.Motegi,A.Uchiyama,A.Uehara,C.Fujiwara,S.Yamazaki,etal., BotulinumtoxinBsuppressesthepressureulcerformationincutaneous ischemia-reperfusioninjurymousemodel:possibleregulationofoxidative andendoplasmicreticulumstress,J.Dermatol.Sci.90(2018)144–153.

[22]A.Uchiyama,K.Yamada,B.Perera,S.Ogino,Y.Yokoyama,Y.Takeuchi,etal., Topicalbetamethasonebutyratepropionateexacerbatespressureulcersafter cutaneousischemia-reperfusioninjury,Exp.Dermatol.25(2016)678–683.

[23]A.Uchiyama,K.Yamada,B.Perera,S.Ogino,Y.Yokoyama,Y.Takeuchi,etal., ProtectiveeffectofMFG-E8aftercutaneousischemia-reperfusioninjury,J.

Invest.Dermatol.135(2015)1157–1165.

[24]E.C.LeRoy,C.Black,R.Fleischmajer,S.Jablonska,T.Krieg,T.A.MedsgerJr,etal., Scleroderma(systemicsclerosis):classiﬁcation,subsetsandpathogenesis,J.

Rheumatol.15(1988)202–205.

[25]S.Motegi,A.Uchiyama,K.Yamada,S.Ogino,Y.Yokoyama,B.Perera,etal., Increasedsusceptibilitytooxidativestress-andultravioletA-induced apoptosisinﬁbroblastsinatypicalprogeroidsyndrome/atypicalWerner syndromewithLMNAmutation,Exp.Dermatol.25(Suppl.3)(2016)20–27.

[26]C.P.Denton,D.J.Abraham,Transforminggrowthfactor-betaandconnective tissuegrowthfactor:keycytokinesinsclerodermapathogenesis,Curr.Opin.

Rheumatol.13(2001)505–511.

[27]K.Itoh,T.Chiba,S.Takahashi,T.Ishii,K.Igarashi,Y.Katoh,etal.,AnNrf2/small MafheterodimermediatestheinductionofphaseIIdetoxifyingenzymegenes throughantioxidantresponseelements,Biochem.Biophys.Res.Commun.236 (1997)313–322.

[28]J.Alam,D.Stewart,C.Touchard,S.Boinapally,A.M.Choi,J.L.Cook,Nrf2,a Cap’n’Collartranscriptionfactor,regulatesinductionofthehemeoxygenase-1 gene,J.Biol.Chem.274(1999)26071–26078.

[29]K.Bedard,K.H.Krause,TheNOXfamilyofROS-generatingNADPHoxidases:

physiologyandpathophysiology,Physiol.Rev.87(2007)245–313.

[30]T.Tanaka,F.Hosoi,Y.Yamaguchi-Iwai,H.Nakamura,H.Masutani,S.Ueda, etal.,Thioredoxin-2(TRX-2)isanessentialgeneregulatingmitochondria- dependentapoptosis,EMBOJ.21(2002)1695–1703.

[31]T. Yamamoto, Thebleomycin-induced sclerodermamodel:what have welearned forsclerodermapathogenesis?Arch.Dermatol.Res.297(2006)333–344.

[32]L.Kong,C.Luo,X.Li,Y.Zhou,H.He,Theanti-inﬂammatoryeffectofkaempferol onearlyatherosclerosisinhighcholesterolfedrabbits,LipidsHealthDis.12 (2013)115.

[33]P.S.Tsou,N.M.Talia,A.J.Pinney,A.Kendzicky,S.Piera-Velazquez,S.A.Jimenez, etal.,Effectofoxidativestressonproteintyrosinephosphatase1Bin sclerodermadermalﬁbroblasts,ArthritisRheum.64(2012)1978–1989.

[34]T. Spadoni, S. Svegliati Baroni, D. Amico, L. Albani, G. Moroncini, E.V.

Avvedimento,etal.,Areactiveoxygenspecies-mediatedloopmaintains increasedexpressionofNADPHoxidases2and4inskinﬁbroblastsfrom patientswithsystemicsclerosis,ArthritisRheumatol.67(2015)1611–1622.

[35]T.Toyama,A.P.Looney,B.M.Baker,L.Stawski,P.Haines,R.Simms,etal., TherapeutictargetingofTAZandYAPbydimethylfumarateinsystemic sclerosisﬁbrosis,J.Invest.Dermatol.138(2018)78–88.

[36]A.Dooley, X.Shi-Wen, N.Aden,T. Tranah, N.Desai, C.P. Denton,et al., ModulationofcollagentypeI,ﬁbronectinanddermalﬁbroblastfunctionand