journal dental - gingival

Role of Transforming Growth Factor-beta1 in Cyclosporine-Induced Epithelial-to-Mesenchymal Transition in GingivalEpitheliumMartin M. Fu,*† Yu-Tang Chin,*‡ Earl Fu,* Hsien-Chung Chiu,* Li-Yu Wang,* Cheng-Yang Chiang,*and Hsiao-Pei Tu*§

Background: It has been proposed that cyclosporin A(CsA) may induce epithelial-to-mesenchymal transition(EMT) in gingiva. The aims of the present study are to con-firm the notion that EMT occurs in human gingival epithelial(hGE) cells after CsA treatment and to investigate the roleof transforming growth factor beta1 (TGF-b1) on this CsA-induced EMT.

Methods: The effects of CsA, with and without TGF-b1 in-hibitor, on the morphologic changes of primary culture ofhGE cells were examined in vitro. The changes of proteinand messenger RNA (mRNA) expressions of two EMTmarkers (E-cadherin and alpha-smooth muscle actin) inthe hGE cells after CsA treatment with and without TGF-b1inhibitor were evaluated with immunocytochemistry andreal-time polymerase chain reaction.

Results: The epithelial cells became spindle-like, elon-gated, and disassociated from neighboring cells and losttheir original cobblestone monolayer pattern when CsA wasadded. However, the epithelial cells stayed in their originalcobblestone morphology with treatment of TGF-b1 inhibitoron top of the CsA treatment. When CsA was given, the pro-tein and mRNA expressions of E-cadherin and a-SMA weresignificantly altered, and these alterations were significantlyreversed with pretreatment of TGF-b1 inhibitor.

Conclusions: CsA could induce Type 2 EMT in gingiva bychanging the morphology of epithelial cells and altering theEMT markers/effectors. The CsA-induced gingival EMT isdependent or at least partially dependent on TGF-b1. J Peri-odontol 2015;86:120-128.

KEY WORDS

Alpha-smooth muscle actin; cadherins; cyclosporine;epithelial-mesenchymal transition; gingiva; smad2 protein;smad3 protein; transforming growth factor beta1.

Cyclosporin A (CsA) is a widelyused immunosuppressant in organtransplantations but can cause

a number of significant side effects,including gingival overgrowth.1 WhenCsA-induced gingival overgrowth oc-curs, the epithelium is thickened, reteridges are elongated, and connectivetissue becomes fibrotic with an increasein fibroblasts and an accumulation ofextracellular matrix.2 Several mecha-nisms of CsA-induced gingival over-growth have been proposed.3,4 However,the exact underlying mechanism is stillunder investigation.

Epithelial-to-mesenchymal transition(EMT) is a process in which epithelial cellslose their original phenotype with a con-comitant development of mesenchymalphenotype in embryonic development,5

fibrogenesis,6 wound healing,7 or cancermetastases.8,9 When EMT occurs in adulttissues, the phenotype of epithelium isaltered by disaggregating epithelial unitsand reshaping for movement. The epi-thelial cells in EMT lose their apical-to-basal polarity, adherens junctions, tightjunctions, desmosomes, epithelial markers(e.g., E-cadherin and zonula occludens1).10 The original close-adhered, cu-boidal epithelial cells would transforminto unanchored, elongated mesenchy-mal, spindle-shaped, myofibroblast- orfibroblast-like phenotypic cells.10,11 The

* Department of Periodontology, School of Dentistry, National Defense Medical Center andTri-Service General Hospital, Taipei, Taiwan, ROC.

† Department ofOralMedicine, Infection, and Immunity;HarvardSchool ofDentalMedicine;Boston, MA.

‡ Institute for Cancer Biology and Drug Discovery, Taipei Medical University, Taipei,Taiwan, ROC.

§ Department of Dental Hygiene, China Medical University, Taichung, Taiwan, ROC.

doi: 10.1902/jop.2014.130285

Volume 86 • Number 1

120

role of EMT leading to tissue fibrosis has been clearlynoted. Evidence has been shown that EMT is asso-ciated with fibrosis in kidney,11 lung,12 eye,13 peri-toneum,14 and probably liver.6

Recently, it has been demonstrated that E-cadherin,an EMTmarker, and another two potential/debatableEMT markers (fibroblast specific protein-1 and fi-bronectin extra type III domain A)15,16 were alteredin the overgrown gingiva of patients taking CsA.17

Disrupted basal laminar structure indicating a possi-ble migration of epithelial cells across the basalmembrane was also observed in those CsA-inducedgingival tissues.18 The possibility of EMT partici-pating in the development of CsA-induced gingivalovergrowth was consequently proposed. On the otherhand, recent studies have shown that transforminggrowth factor beta1 (TGF-b1), commonly believed tobe a key mediator in EMT, may not necessarily beinvolved in renal EMT. However, in gingiva, the role ofTGF-b1 in CsA-induced EMT has never been in-vestigated. The available evidence of CsA-inducedEMT in gingiva is also very limited in the current lit-erature. In this study, therefore, the notion of CsA-induced EMT in gingiva is further evaluated in vitro. Theaims of the present study are: 1) to examine the effectsof CsA on themorphologic change of human epithelialcells and 2) to explore the role of TGF-b1 in CsA-induced gingival EMT by evaluating the changesof two common EMT markers after CsA treatmentwith and without the suppression of TGF-b1.

MATERIALS AND METHODS

Primary Culture of Human Gingival EpithelialCellsHuman gingival epithelial (hGE) cells in the currentstudy were obtained from three individuals (one maleand two females, aged 36, 52, and 63 years old,respectively) in either a crown lengthening or distalwedge procedure of non-inflamed periodontal tissuesat the Dental Clinic of the Tri-Service General Hospital,Taipei, Taiwan. The study protocol was approved bythe Institutional Review Board of Tri-Service GeneralHospital, and written informed consent for participa-tion in the study was obtained from each participant.After immersed in culture mediumi with 2 mg/mLcleavage supplement¶ and 10% fetal bovine serum at4�C for 2 days, the epithelial layer was separated fromthe underlying connective tissue. The epithelial frag-ments were then digested in serum-free mediumcontaining 0.05% trypsin-EDTA# at 37�C in 5% CO2

for 5 minutes. The hGE cell cultures were grown in anepithelium culture medium** with calcium chlorideand 1% of growth supplement†† for five to sevenpassages. The cells were stimulated by varied con-centrations (0, 500, 800, and 1,000 ng/mL) of CsA‡‡

in alcohol (as a solvent) for 48 or 72 hours. On the

culture plate, the epithelial cell was examined by in-verted microscopy and then categorized as cobble-stone or spindle according to its cell morphology.A total of 100 cells were counted in each cultureplate. After cell collection, themessenger RNA (mRNA)expression of EMT-associated genes, includingE-cadherin (a gene associated with epithelium) andalpha-smooth muscle actin (a-SMA) (a gene asso-ciated with mesenchyme), were examined by reverse-transcription polymerase chain reaction (RT-PCR).

RT-PCR and Real-Time PCRThe total RNA of hGE cells was extracted with re-agent,§§ quantified by spectrophotometry at 260 nm,and reversely transcribed into total complementaryDNA (cDNA) using a PCR system.ii The PCR re-actions involved initial denaturation at 94�C for 2minutes and 30 seconds, followed by 30 or 35 cyclesat 94�C for 30 seconds, annealing at 58�C to 62�C for30 seconds, and extraction at 72�C for 60 seconds.The real-time PCR reactions were performed by usinga kit¶¶ in the cycler device## and involved initialdenaturation at 95�C for 5 minutes followed by 40cycles of denaturing at 95�C for 5 seconds andcombined annealing/extension at 60�C for 10 sec-onds as described in the instructions. The PCRprimers were as follows: human E-cadherin, forward59-TGAAGGTGACAGAGCCTCTGGA-39 and reverse59-TGGGTGAATTCGGGCTTGTT-39 (GenBank ac-cession no. NM_004360.3); human a-SMA, forward59-GCGTGGCTATTCCTTCGTTAC-39 and reverse 59-CATAGTGGTGCCCCCTGATAG-3 (GenBank acces-sion no. NM_001613.2); and human GAPDH, forward59-AGCCGCATCTTCTTTTGCGTC-39 and reverse59-TCATATTTGGCAGGTTTTTCT-39 (GenBank ac-cession no. NM_002046). Primers for real-time PCRwere from commercial assays,*** including humanE-cadherin (QT00080143, GenBank accessionno. NM_004360), human a-SMA (QT00088102,GenBank accession no. NM_001613), human TGF-b1(QT00000728, GenBank accession no. NM_000660),and human GAPDH (QT00079247, GenBank acces-sion no. NM_002046). The exponential phases of RT-PCRs were determined from 30 to 35 cycles to allowquantitative comparison between the cDNAs. Am-plified RT-PCR products were then analyzed on1% agarose gels, visualized with ethidium bromide

i Leibovitz L-15, Sigma-Aldrich, St Louis, MO.¶ Dispase II, Roche Diagnostics, Indianapolis, IN.# Gibco BRL, Gibco, Thermo Fisher Scientific, Waltham, MA.** EpiLife, Gibco, Thermo Fisher Scientific.†† Human Keratinocyte Growth Supplement, Gibco, Thermo Fisher

Scientific.‡‡ Sigma-Aldrich.§§ Trizol, Thermo Fisher Scientific.ii GeneAmp PCR System 9700, Applied Biosystems, Thermo Fisher

Scientific.¶¶ Rotor-Gene SYBR Green PCR Kit, Qiagen, Hilden, Germany.## Rotor-Gene Q, Qiagen.*** QuantiTect Primer Assay, Qiagen.

J Periodontol • January 2015 Fu, Chin, Fu, et al.

121

staining by a camera system,††† and scanned withan analysis system.‡‡‡ The densities relative toGAPDH bands were determined.

Western Blot AnalysisWestern blot analysis was performed to quantifythe protein expression levels of phospho-Smad2/3(pSmad2/3) in the total cell lysates of hGEs aftertreatment with CsA (1,000 ng/mL), TGF-b1 (5 mg/mL)and TGF-b inhibitor (10 mM).§§§ Protein sam-ples were resolved on a 10% sodium dodecyl sulfatepolyacrylamide gel. A 20-mg quantity of protein wasloaded in each well with 4· sample buffer, and theprotein samples were resolved by electrophoresis at100 V for 2 hours. The resolved proteins were trans-ferred from the polyacrylamide gel to membranesiii

with a transfer system.¶¶¶ The membranes wereblocked with a solution of 2% bovine serum albumin(BSA)### and 1% polysorbate 20**** in Tris-bufferedsaline. The membranes were incubated with primaryantibodies to pSmad2/3†††† and GAPDH‡‡‡‡ at 4�Covernight and washed, and the proteins were detectedwith horseradish peroxidase–conjugated secondaryantibodies and reagent.§§§§ Images of the Westernblots were visualized, recorded, and analyzed.iiii

Immunocytochemistry andConfocal Laser ScanningMicroscopyThe hGE cells were cultured ineight-well chamber slides forimmunocytochemistry. The cells,after being treated with CsA,TGF-b1, or TGF-b1 inhibitor, werefixed with methanol, blocked with5% BSA in phosphate-bufferedsaline (PBS) for 20minutes, andthen exposed to the primaryantibodies of a-SMA¶¶¶¶ andE-cadherin#### (1:200 in 5%BSA)overnight at 4�C. The cells werethen washed in PBS, exposed tothe secondary antibody (fluores-cein isothiocyanate–conjugatedgoat anti-rabbit immunoglobu-lin G, 1:200 in 5% BSA) for 30minutes, and counterstained with49,6-diamidino-2-phenylindole(DAPI).***** The expressionsof a-SMA and E-cadherin wereobserved and compared underdifferent treatments by confocallaser scanning microscopy.†††††

Statistical AnalysesAll of the experiments werecarried out independently at

least three times. Student t tests or analysis of vari-ance were used for evaluating the differences be-tween the test and control (0 ng/mL CsA) groups orother groups when specified. P < 0.05 was selected asthe significance level.

RESULTS

CsA at 1,000 ng/mL Was Not Cytotoxic to hGECells but Induced Change of Cellular MorphologyThe hGE cell survival rate was not significantly re-duced if £1,000 ng/mL CsA was given to the cells(Fig. 1A). However, the survival rate was significantly

Figure 1.Cellular morphologic changes of hGE cells receiving CsA treatment. (A) Cytotoxicity of CsA in the gingivalepithelial cells: the human gingival epithelial cells received CsA (0, 500, 1,000, 1,500, or 2,000 ng/mL) for72 hours by MTS assay. *P <0.05, compared with 0 ng/mL CsA. Cellular morphology in the cells thatreceived solvent (B) or 1,000 ng/mL CsA (C) for 72 hours. (Original magnification ·100.)D) Statisticalcomparisons of distributions of the cells with cobblestone and spindle shapes after the cells receivedtreatments with and without CsA. *P <0.05.

††† Transilluminator/SPOT, Diagnostic Instruments, Sterling Heights,MI.

‡‡‡ ONE-Dscan 1-D Gel Analysis Software, Scanalytics, Fairfax, VA.§§§ TGF-b RI Kinase Inhibitor, Merck, Darmstadt, Germany.iii Immobilon-PSQ Transfer PVDF membranes, Millipore, Billerica, MA.¶¶¶ Hoefer, Holliston, MA.### Sigma-Aldrich.**** Tween 20, Sigma Aldrich.†††† Ser465/467 and Ser423/425, Cell Signaling Technology, Beverly,

MA.‡‡‡‡ GeneTex, San Antonio, TX.§§§§ Immobilon Western HRP Substrate Luminol Reagent, Millipore.iiii ChemiDoc XRS+ imaging system with Image Lab, Bio-Rad Labora-

tories, Hercules, CA.¶¶¶¶ Abcam, Cambridge, UK.#### BD Biosciences, Franklin Lakes, NJ.***** Sigma-Aldrich.††††† LSM 780, Carl Zeiss MicroImaging, Thornwood, NY.

Role of TGF-b1 in Cyclosporine-Induced Gingival EMT Volume 86 • Number 1

122

reduced when ‡1,500 ng/mL CsA was given to thecells. The hGE cells that received solvent (controlcells) presented typical cobblestone morphology(Fig. 1B), whereas the cells became spindle-like,elongated, and disassociated from neighboring cellsand lost their original cobblestone monolayer patternwhen CsA was added (Figs. 1C and 1D).

CsA-Induced Changes in EMT-Related mRNAExpression in hGE CellsThe mRNA expressions of E-cadherin were signifi-cantly reduced (P <0.05), in a dose dependent man-ner, after 1,000 ng/mL CsA treatment for 48 hours(Figs. 2A and 2B) and 72 hours (Figs. 2C and 2D) orafter 800 ng/mL CsA treatment for 72 hours. Thelevels of a-SMA and TGF-b1 were significantly in-creased (P <0.05), in a dose dependent manner aswell, after 1,000 ng/mL CsA treatment for 48 and72 hours or after 800 ng/mL CsA treatments for72 hours.

Effects of TGF-b1 Inhibition on mRNA Expressionof EMT Markers in hGE Cells Receiving CsAE-cadherin and a-SMA expression in hGE cells wasexamined with real-time PCR after a pretreatment ofTGF-b1 inhibitor (10 mg/mL) overnight and treat-ment of CsA (1,000 ng/mL) for 72 hours. The pre-treatment of TGF-b1 inhibitor alone did not alter the

expression of E-cadherin (Fig.3A). However, when 1,000 ng/mLCsA was given, expression ofE-cadherin was significantlyreduced, and this reduction wassignificantly preventedwhen therewas a pretreatment of 10 mMTGF-b1 inhibitor.

On the other hand, treatingwith 1 or 5 ng/mL TGF-b1 alonedid not significantly affect theexpression of E-cadherin, al-though a trend of reduction ofE-cadherin could be observed.Similar to the result of E-cadherin,the pretreatment of TGF-b1 in-hibitor alone did not alter theexpression of a-SMA (Fig. 3B),but when 1,000 ng/mL CsA wasgiven, the expression of a-SMAwas significantly increased,and this increase did not occurwhen there was a pretreatmentwith TGF-b1 inhibitor. In con-trast, the treatment with 1 or5 ng/mL TGF-b1 alone signif-icantly increased the expres-sion of a-SMA.

Effects of TGF-b1 Inhibition on pSmad2/3 in hGECells Receiving CsAThe expressions of Smad2 and Smad3 phosphory-lation in hGE cells were examined by Western blot-ting after pretreatment of TGF-b1 inhibitor (10 mg/mL)overnight and treatment of CsA (1,000 ng/mL) for24 hours. The pretreatment of TGF-b1 inhibitor alonedid not alter the expression of E-cadherin (Fig. 4).However, when 1,000 ng/mL CsA was given, theexpressions of pSmad2 and pSmad3 were signifi-cantly increased, and these increases did not occurwhen there was a pretreatment of TGF-b1 inhibitor.In contrast, the treatment with 5 ng/mL TGF-b1 alonesignificantly increased the expression of pSmad2and pSmad3.

CsA In Vitro Induced a Change of CellularMorphology in hGE Cells That Was Preventedby TGF-b1 InhibitorThe morphology of gingival cells that received thesolvent (control cells) presented a typical cobble-stone epithelial morphology (Fig. 5A). The mor-phology of cells that received CsA became elongatedand disassociated from the adjacent epithelial cells(Fig. 5B). However, the cells stayed with the originalcobblestone morphology when there was a treatmentof TGF-b1 inhibitor on top of the CsA treatment(Fig. 5C). Cells that had TGF-b1 inhibitor alone

Figure 2.In vitro mRNA expressions of E-cadherin,a-SMA, and TGF-b1 in hGE cells receiving CsA treatment for 48(A andB) or 72 (C andD) hours. The expressionwas semiquantitatively comparedwith relative densitiesto GAPDH. *Significant difference from control cells at P <0.05 for E-cadherin and a-SMA by post hocanalysis after one-way analysis of variance.


123

maintained typical epithelial morphology (Fig. 5D).The cells that received TGF-b1 stayed rounded, butwere more three-dimensional and disassociated fromthe adjacent epithelial cells (Figs. 5E and 5F).

TGF-b1 Inhibitor Prevented the Change ofExpression of EMT Markers Induced byCsA in hGE CellsFigure 6 shows the effect of TGF-b1 inhibitor on theprotein expression of EMT markers in hGE cells that

received CsA treatment. Withimmunocytochemistry, therewas no E-cadherin expressionobserved in the cell membranewhen hGE cells received a treat-ment of CsA or TGF-b1 (Figs.6D, 6F, and 6G), but the ex-pression of E-cadherin could beobserved in the control cells (withsolvent or TGF-b1 inhibitoralone) (Figs. 6B and 6C) or in thecells that received both CsA andTGF-b1 (Fig. 6E). Opposite re-sults were demonstrated fora-SMA. Positively stained a-SMAwas easily observed in cytosolwhen hGE cells received a treat-ment of CsA or TGF-b1 (Figs. 6K,6M, and 6N), but not observedin the control cells (with solvent orTGF-b1 inhibitor alone) (Figs. 6Iand 6J) or in the cells that re-ceived both CsA and TGF-b1(Fig. 6L).

Strong expression of DAPIwas observed in the nuclei of allthe hGE cells (Figs. 6O through6U).

DISCUSSION

The aims of the present studyare to examine the notion thatCsA could induce EMT in gin-giva and to explore the role ofTGF-b1 in CsA-induced gingivalEMT by evaluating the changesof two EMT common markersand two TGF-b1 transcriptionalmodulators after CsA treatmentwith and without the suppres-sion of TGF-b1. The effects ofCsA on the morphologic changeof human epithelial cells werefirst examined to confirm thenotion that CsA could induceEMT in gingiva.17 An increased

percentage of epithelial cells transforming the cellmorphology of cobblestone shape to spindle-like andelongated pattern was observed after CsA treatment(Fig. 1). The expression of a-SMA, a common EMTmarker, and TGF-b1, an epithelial-to-mesenchymaleffector, was significantly increased in CsA treatmentgroups, whereas E-cadherin, another common EMTmarker, was significantly reduced after CsA treat-ments (Fig. 2). The assumption that TGF-b1 playsa role in CsA-induced EMT in gingiva17 was also

Figure 3.Effect of TGF-b1 inhibitor on the expression of two EMTmarkers, E-cadherin (A) anda-SMA (B), in hGEcells receiving CsA treatment. The mRNA expressions of E-cadherin and a-SMA in hGE cells wereexamined with real-time PCR after treatment of CsA (1,000 ng/mL) with or without TGF-b1 inhibitor(10 mg/mL). *P <0.05 compared with control cells, †P <0.05 compared with CsA treatment.


124

investigated and verified in thecurrent study. The altered ex-pressions of E-cadherin anda-SMA were significantly pre-vented after TGF-b1 inhibitortreatment with examinations byreal-time PCR (Fig. 3) and im-munocytochemistry (Fig. 6). Theinhibited protein expressions ofpSmad2 and pSmad3 were alsosignificantly prevented after TGF-b1 inhibitor treatment (Fig. 4).The effects of CsA with TGF-b1inhibitor on the morphologicchange of hGE cells were alsoexamined. The change in mor-phology seen with CsA treat-ment was absent when TGF-b1inhibitor was added to the CsAtreatment (Fig. 5).

CsA has been widely usedin organ transplantation. Apartfrom its multiple systemic con-sequences,19-24 CsA has beenknown to cause gingival over-growth. At the histologic level,the induced gingival overgrowthis characterized by epithelialhyperplasia, interstitial fibrosis,and focal inflammatory cell in-filtration.25,26 At the molecularlevel, CsA increases the expres-sion of TGF-b1, interleukin-6,and other mediators in gin-giva.3,4 However, there is still avery limited understanding of theoccurrence of CsA-induced gin-gival overgrowth.

EMT is a process in whichepithelial cells lose their originalphenotype with a concomitantdevelopment of mesenchymalphenotype. At the cellular level,four fundamental steps are in-volved in EMT.27 These include:1) loss of epithelial cell–celladherens junction componentE-cadherin and actin cytoskeletalrearrangement; 2) de novo syn-thesis of myofibroblast-specificmarker proteins, e.g., a-SMA;3) disruption of the tubularbasement membrane; and 4)enhanced cell migration andinvasiveness. According to theclassification proposed at the

Figure 4.Effect of TGF-b1 inhibitor on Smad2 and Smad3 phosphorylation in hGE cells receiving CsA treatment.The phosphorylation of Smad2 and Smad3 in hGE cells after the treatments of CsA, TGF-b, and/orTGF-b1 inhibitor was evaluated by Western blotting (A). The expressions were semiquantitativelycompared with their relative densities to GAPDH. The statistic data of Smad2 (B) and Smad3 (C) areexpressed asmean and SD.*,†,‡,§Subsetswith a significant difference atP<0.05 obtained after post hocanalysis using one-way analysis of variance.

Figure 5.Effect of TGF-b1 inhibitor on cell morphology in hGE cells receiving CsA treatment. The cell morphology ofgingival epithelia in control cells treatedwith solvent or CsA (1,000 ng/mL)with orwithout TGF-b1 inhibitor(10 mg/mL). A) Solvent. B) CsA.C) CsA plus TGF-b1 inhibitor.D) TGF-b1 inhibitor. E) 1ng/mL TGF-b1.F) 5 ng/mL TGF-b1. (Original magnification ·100.)


125

EMT International Association(TEMTIA) meetings, EMT can beclassified into three subtypes.28,29

Type 1, so-called developmentalEMT, occurs only during embryo-genesis when primitive epithelialcells give rise to the mobile mes-enchymal cells, mesoderm, andendoderm. Type 2, also known asfibrogenic EMT, is associated withinflammation, wound healing,tissue regeneration, and organfibrosis in mature or adult tissue.Type 2 EMT is induced in re-sponse to inflammation or traumaand normally ceases once thestimulus stops.30 Type 3, alsoknown as metastasis-associatedEMT, occurs in epithelial cancercells and allows them to convert toa mesenchymal phenotype tomove to the invasive front of thetumors.28 Because of the absenceof metastasis, the CsA-inducedEMT in gingival overgrowth shouldbelong to Type 2 EMT accordingto the above classification.

However, whether EMT existsin a specific tissue could be iden-tified by detecting the changes ofEMT markers. Among all of themarkers, E-cadherin is well ac-knowledged and plays an essen-tial role in maintaining thestructural integrity of epitheliumand polarization.31 The suppres-sion of E-cadherin expression isan early and important step thatprecedes all other major eventssuch as induction of a-SMAduring TGF-b1–induced EMT.27

In gingival epithelium, significantchanges of some specific EMTmarkers,15,32 including E-cadherin,b-catenin,33 HSP47,34 and in-tegrins,35 have been observed inCsA histologic tissues in the au-thors’ previous study and others.Alterations of a few other EMTmarkers,36-39 e.g., upregulation ofSnail-1 and downregulation ofzonula occludens 1, have alsobeen lately observed in primarycultured hGE cells after CsAtreatment in the authors’ labora-tory ( EF and HPT; unpublished

Figure 6.Effect of TGF-b1 inhibitor on the protein expression of EMTmarkers in hGE cells receiving CsA treatment.Expression of E-cadherin anda-SMA in gingival epithelial cells was examined with immunocytochemistryafter treatment of CsA (1,000 ng/mL), TGF-b1 (5 ng/mL), and/or TGF-b1 inhibitor (10 mg/mL) for72 hours. (Original magnification ·100.) A through G) Expression of E-cadherin. H through N)a-SMA under different condition: Control (A and H), Solvent (B and I), TGF-b1 inhibitor (C and J), CSA(Dand K), CSAwith TGF-b1 inhibitor (E and L), TGF-b1 1ng/ml (F andM)and TGF-b1 5ng/ml (G andN).DAPI panel shows the location of the nucleus; DAPI on without antibody (O), solvent (P), TGF-b1inhibitor (Q), CsA (R), CsA plus TGFb1 inhibitor (S), 1ng/mL TGF-b1 (T), and 5ng/mL TGF-b1(U). Merged panel shows the results of combination of E-cadherin, a-SMA and DAPI panel; Mergedon without antibody (V), solvent (W), TGF-b1 inhibitor (X), CsA (Y), CsA plus TGF-b1 inhibitor(Z), 1ng/mL TGF-b1 (AA), and 5ng/mL TGF-b1 (BB).


126

observations), providing extra evidence of the exis-tence of CsA-induced EMT in gingival overgrowth.

TGF-b1 has been identified to induce EMT in thekidney, breast, and other organ systems.5,40 In ad-dition, TGF-b1 is capable of inducing EMT in epi-thelial fibrosis of the kidney,11,27 lung,12 and liver.6

TGF-b1 can also be influenced by CsA. After CsAapplications, it has been found that the expressions ofTGF-b1 were elevated in human gingival fibro-blasts,3,41 human gingival crevicular fluid,42 anddental plaque–free gingiva.43

Recently, it was demonstrated that E-cadherin andtwo fibroblast markers or debatable EMT markers15,16

(fibroblast-specific protein 1 and fibronectin extra typeIII domain A) were altered in the overgrown gingivaeof patients taking CsA,17 providing the first evidencethat EMT may be present in CsA-induced gingivalovergrowth. The present results are consistent withthose findings and support the notion. It has also beendemonstrated that TGF-b1 may trigger EMT in vitro inhealthy hGE cells with significant alterations of someEMT markers.17 However, it has been shown that,in renal tissues, CsA-induced EMT could be eitherTGF-b1 dependent or independent.44,45 Conse-quently, the model of TGF-b1-induced EMT may notbe ideal to represent the EMT induced by CsA. In thepresent study, the authors hypothesize that, in gin-gival epithelium, the CsA-induced Type 2 EMT isdependent or at least partially dependent on TGF-b1.These results demonstrate that the changes of E-cadherin, b-catenin, pSmad2, and pSmad3 after CsAtreatment were significantly prevented by addingTGF-b1 inhibitor, indicating that the CsA-inducedgingival EMT is TGF-b1 dependent. However, furtherinvestigations are still needed to clarify the detailedmechanisms of CsA-mediated EMT in gingiva.

CONCLUSIONS

These results suggest that CsA could induce Type 2EMT in gingiva by changing the morphology of epi-thelial cells and altering the EMT markers/effectors.CsA-induced Type 2 EMT in gingiva might be pre-vented by the inhibition of TGF-b1, indicating that CsA-induced gingival EMT is dependent or at least partiallydependent on TGF-b1. With a better understanding ofthe role of CsA-mediated EMT in gingiva, the authorshope to shed additional light on themechanism of CsA-induced gingival enlargement, which could have im-portant implications for clinical therapy.

ACKNOWLEDGMENTS

This study was supported by grants from the NationalScience Council (NSC-100-2314-B-016-037) andthe Department of National Defense (MAB-101-54), Taiwan, ROC. The authors report no conflictsof interest related to this study.

REFERENCES1. Wysocki GP, Gretzinger HA, Laupacis A, Ulan RA,

Stiller CR. Fibrous hyperplasia of the gingiva: A sideeffect of cyclosporin A therapy. Oral Surg Oral MedOral Pathol 1983;55:274-278.

2. Hallmon WW, Rossmann JA. The role of drugs in thepathogenesis of gingival overgrowth. A collective re-view of current concepts. Periodontol 2000 1999;21:176-196.

3. Chae HJ, Ha MS, Yun DH, et al. Mechanism ofcyclosporine-induced overgrowth in gingiva. J DentRes 2006;85:515-519.

4. Bostrom A, Bharath H, Saulewicz A, Narayanan AS.Cyclosporin A affects signaling events differentially inhuman gingival fibroblasts. J Dent Res 2005;84:532-536.

5. Hay ED. An overview of epithelio-mesenchymal trans-formation. Acta Anat (Basel) 1995;154:8-20.

6. Iwano M, Plieth D, Danoff TM, Xue C, Okada H, NeilsonEG. Evidence that fibroblasts derive from epitheliumduring tissue fibrosis. J Clin Invest 2002;110:341-350.

7. Weber CE, Li NY, Wai PY, Kuo PC. Epithelial-mesenchymal transition, TGF-b, and osteopontin inwound healing and tissue remodeling after injury.J Burn Care Res 2012;33:311-318.

8. Xue C, Plieth D, Venkov C, Xu C, Neilson EG. Thegatekeeper effect of epithelial-mesenchymal transitionregulates the frequency of breast cancer metastasis.Cancer Res 2003;63:3386-3394.

9. Guarino M, Micheli P, Pallotti F, Giordano F. Patholog-ical relevance of epithelial and mesenchymal pheno-type plasticity. Pathol Res Pract 1999;195:379-389.

10. Thiery JP, Sleeman JP. Complex networks orchestrateepithelial-mesenchymal transitions. Nat Rev Mol CellBiol 2006;7:131-142.

11. ZeisbergM, BonnerG,Maeshima Y, et al. Renal fibrosis:Collagen composition and assembly regulates epithelial-mesenchymal transdifferentiation. Am J Pathol 2001;159:1313-1321.

12. Willis BC, Liebler JM, Luby-Phelps K, et al. Induction ofepithelial-mesenchymal transition in alveolar epithelialcells by transforming growth factor-beta1: Potentialrole in idiopathic pulmonary fibrosis. Am J Pathol2005;166:1321-1332.

13. Saika S, Kono-Saika S, Ohnishi Y, et al. Smad3signaling is required for epithelial-mesenchymal tran-sition of lens epithelium after injury. Am J Pathol 2004;164:651-663.

14. Margetts PJ, Bonniaud P, Liu L, et al. Transient over-expression of TGF-beta1 induces epithelial mesenchy-mal transition in the rodent peritoneum. J Am SocNephrol 2005;16:425-436.

15. Kriz W, Kaissling B, Le Hir M. Epithelial-mesenchymaltransition (EMT) in kidney fibrosis: Fact or fantasy?J Clin Invest 2011;121:468-474.

16. Galichon P, Hertig A. Epithelial to mesenchymal tran-sition as a biomarker in renal fibrosis: Are we ready forthe bedside? Fibrogenesis Tissue Repair 2011;4:11.

17. Sume SS, Kantarci A, Lee A, Hasturk H, Trackman PC.Epithelial to mesenchymal transition in gingival over-growth. Am J Pathol 2010;177:208-218.

18. Kantarci A, Nseir Z, Kim YS, Sume SS, Trackman PC.Loss of basement membrane integrity in human gingi-val overgrowth. J Dent Res 2011;90:887-893.

19. Chang SH, Lim CS, Low TS, Chong HT, Tan SY.Cyclosporine-associated encephalopathy: A case report


127

and literature review. Transplant Proc 2001;33:3700-3701.

20. Galan AI, Fernandez E, Moran D, Munoz ME, JimenezR. Cyclosporine A hepatotoxicity: Effect of prolongedtreatment with cyclosporine on biliary lipid secretion inthe rat. Clin Exp Pharmacol Physiol 1995;22:260-265.

21. Nalesnik MA, Jaffe R, Starzl TE, et al. The pathology ofposttransplant lymphoproliferative disorders occurringin the setting of cyclosporine A-prednisone immuno-suppression. Am J Pathol 1988;133:173-192.

22. Woolfson RG, Neild GH. Cyclosporin nephrotoxicityfollowing cardiac transplantation. Nephrol Dial Trans-plant 1997;12:2054-2056.

23. Textor SC, Canzanello VJ, Taler SJ, et al. Cyclosporine-induced hypertension after transplantation. Mayo ClinProc 1994;69:1182-1193.

24. Rodino MA, Shane E. Osteoporosis after organ trans-plantation. Am J Med 1998;104:459-469.

25. Deliliers GL, Santoro F, Polli N, Bruno E, Fumagalli L,Risciotti E. Light and electron microscopic study ofcyclosporin A-induced gingival hyperplasia. J Peri-odontol 1986;57:771-775.

26. Ayanoglou CM, Lesty C. Cyclosporin A-induced gingi-val overgrowth in the rat: A histological, ultrastructuraland histomorphometric evaluation. J Periodontal Res1999;34:7-15.

27. Yang J, Liu Y. Dissection of key events in tubularepithelial to myofibroblast transition and its implica-tions in renal interstitial fibrosis. Am J Pathol 2001;159:1465-1475.

28. Kalluri R, Weinberg RA. The basics of epithelial-mesenchymal transition. J Clin Invest 2009;119:1420-1428.

29. Zeisberg M, Neilson EG. Biomarkers for epithelial-mesenchymal transitions. J Clin Invest 2009;119:1429-1437.

30. Thiery JP, Acloque H, Huang RY, Nieto MA. Epithelial-mesenchymal transitions in development and disease.Cell 2009;139:871-890.

31. Bush KT, Tsukamoto T, Nigam SK. Selective degrada-tion of E-cadherin and dissolution of E-cadherin-catenin complexes in epithelial ischemia.Am J PhysiolRenal Physiol 2000;278:F847-F852.

32. Quaggin SE, Kapus A. Scar wars: Mapping the fate ofepithelial-mesenchymal-myofibroblast transition. Kid-ney Int 2011;80:41-50.

33. Tu HP, Chen YT, Shieh YS, et al. Cyclosporin-induceddownregulation of the expression of E-cadherin duringproliferation of edentulous gingival epithelium in rats.J Periodontol 2006;77:832-839.

34. Chang TY, Tsai CH, Chang YC. The upregulation ofheat shock protein 47 in human gingival fibroblasts

stimulated with cyclosporine A. J Periodontal Res2010;45:317-322.

35. Walsh P, Hakkinen L, Pernu H, Knuuttila M, Larjava H.Expression of fibronectin-binding integrins in gingivalepithelium in drug-induced gingival overgrowth. J Peri-odontal Res 2007;42:144-151.

36. Venkov CD, Link AJ, Jennings JL, et al. A proximalactivator of transcription in epithelial-mesenchymaltransition. J Clin Invest 2007;117:482-491.

37. Teng Y, Zeisberg M, Kalluri R. Transcriptional regula-tion of epithelial-mesenchymal transition. J Clin Invest2007;117:304-306.

38. Cho HJ, Baek KE, Saika S, Jeong MJ, Yoo J. Snail isrequired for transforming growth factor-beta-inducedepithelial-mesenchymal transition by activating PI3kinase/Akt signal pathway. Biochem Biophys ResCommun 2007;353:337-343.

39. Cano A, Perez-Moreno MA, Rodrigo I, et al. Thetranscription factor snail controls epithelial-mesenchymaltransitions by repressing E-cadherin expression. Nat CellBiol 2000;2:76-83.

40. Thiery JP. Epithelial-mesenchymal transitions in tu-mour progression. Nat Rev Cancer 2002;2:442-454.

41. Cotrim P, Martelli-Junior H, Graner E, Sauk JJ, ColettaRD. Cyclosporin A induces proliferation in humangingival fibroblasts via induction of transforminggrowth factor-beta1. J Periodontol 2003;74:1625-1633.

42. Buduneli N, Kutukcxuler N, Aksu G, Atilla G. Evaluationof transforming growth factor-beta 1 level in crevicularfluid of cyclosporin A-treated patients. J Periodontol2001;72:526-531.

43. Chen YT, Tu HP, Chin YT, et al. Upregulation oftransforming growth factor-beta1 and vascular endo-thelial growth factor gene and protein expression incyclosporin-induced overgrown edentulous gingiva inrats. J Periodontol 2005;76:2267-2275.

44. Pallet N, Thervet E, Anglicheau D. c-Jun-N-terminalkinase signaling is involved in cyclosporine-inducedepithelial phenotypic changes. J Transplant 2012;2012:348604.

45. Berzal S, Alique M, Ruiz-Ortega M, Egido J, Ortiz A,Ramos AM. GSK3, snail, and adhesion moleculeregulation by cyclosporine A in renal tubular cells.Toxicol Sci 2012;127:425-437.

Correspondence: Asst. Prof. Hsiao-Pei Tu, Department ofDental Hygiene, China Medical University, 91 Hsueh-ShihRoad, Taichung, 40402, Taiwan, ROC. Fax: +886-2-87927145; e-mail: [email protected].

Submitted April 29, 2014; accepted for publication May26, 2014.


128

mailto:[email protected]

journal dental - gingival

Education