{"paper_id":"dfbddfe8-48ad-44c4-b0fc-f962ad88f082","body_text":"Int. J. Biol. Sci. 2021, Vol. 17 \n \n \nhttp://www.ijbs.com \n2449 \nInternational Journal of Biological Sciences \n2021; 17(10): 2449-2460. doi: 10.7150/ijbs.60167 \nResearch Paper \nCombination of Ferulic Acid, Ligustrazine and \nTetrahydropalmatine attenuates Epithelial-mesenchymal \nTransformation via Wnt/β-catenin Pathway in \nEndometriosis \nChengling Zhang1,2,3,4*, Ying Zhang1,2,3,4*, Haiying Pan5*, Yi Tan1,2,3,4, Qinghua Wei1,2,3,4, Xueshan Dai1,2,3,4, \nJiahui Wei1,2,3,4, Yi Chen1,2,3,4 \n1. College of Pharmaceutical Sciences & Chinese Medicine, Southwest University, Chongqing, China. \n2. Chongqing Key Laboratory of New Drug Screening from Traditional Chinese Medicine, Chongqing, China. \n3. Pharmacology of Chinese Materia Medica – the Key Discipline Constructed by the state Administration of Traditional Chinese Medicine, Chongqing, China. \n4. National Demonstration Center for Experimental Pharmacy Education (Southwest University), Chongqing, China. \n5. Sichuan Jinxin Women & Children Hospital, Chengdu 610066, China. \n*These authors contributed equally to this work. \n Corresponding author : Yi Chen, College of Pharmaceutical Sciences & Chinese Medicine, Southwest University, No.2 Tiansheng Road Beibei District, \nChongqing 400715, China. Tel: +862368251225; Fax: +862368251225. E-mail: rachelcy@swu.edu.cn. \n© The author(s). This is an open access article distributed under the terms of the Creative Commons Attribution License (https://creativecommo ns.org/licenses/by/4.0/). \nSee http://ivyspring.com/terms for full terms and conditions. \nReceived: 2021.03.06; Accepted: 2021.05.22; Published: 2021.06.11 \nAbstract \nPreviously the potential therapeutic action of ferulic acid, ligustrazine and tetrahydropalmatine (FLT) are \ndiscovered with unclear mechanism in  rat autograft  endometriosis. However, the effect of FLT on \nendometrial cells a nd allograft endometriosis is still unclear. This study is designed to  elucidate the \ninfluence of FLT on epithelial-mesenchymal transformation in allograft endometriosis and endometrium \ncells. In vivo, fluorescent xenogeneic endometriosis model was established. In vitro, invasion and metastasis \nwere analyzed after treating FLT. E pithelial-mesenchymal transformation and Wnt/ β-catenin pathway \nwere inspected in vitro  and in vivo. Activator or inhibitor of Wnt/ β-catenin signaling was performed  to \ninspect mechanism of epithelial-mesenchymal transformation. In vivo, FLT not only decreased fluorescent \nintensity and volume of ectopic lesion, but also ameliorated pathological morphology. E2 and PROG levels \nin serum were reduced by FLT. In endometrial cells, FLT significantly  inhibited the invasion and \nmetastasis. Meantime, epithelial-mesenchymal transformation was reversed, accompanied by suppression \nof Wnt/β-catenin pathway. In-depth study, activation of Wnt/ β-catenin pathway lead to promotion of \nepithelial-mesenchymal transformation, which was reversed by FLT. FLT prevented fluorescent allograft \nendometriosis and endometrium cells, which was related to suppress epithelial -mesenchymal \ntransformation through inactivating Wnt/β-catenin pathway. The findings disclose molecular mechanism \nof epithelial-mesenchymal transformation in endometriosis by FLT, and contribute to further application. \nKey words: Ferulic acid; Ligustrazine; Tetrahydropalmatine ; Allograft endometriosis; Epithelial-mesenchymal \ntransformation; Wnt/β-catenin pathway \nIntroduction \nEpithelial-mesenchymal transformation is the \ncell process from epithelial to mesenchymal \nphenotype, then acquisition of invasion and \nmetastasis. In epithelial-mesenchymal transformation, \nE-cadherin attenuates as one of the epithelial markers. \nAs mesenchymal markers, N- cadherin, Vimentin, \nSnail, ZEB1, Twist, and Slug expand for mesenchymal \nphenotype [1]. As a common gynecologic disease, \nendometriosis (EMS) is characterized with dynamic \nendometrium developing in extrauterine s ites. \nRecently, epithelial-mesenchymal transformation  is \nregarded as an important pathological factor in EMS  \n[2]. Wnt/β -catenin signaling is also abnormally \nstimulated in EMS, including promoting EMS lesions, \n \nIvyspring  \nInternational Publisher \n\nInt. J. Biol. Sci. 2021, Vol. 17 \n \n \nhttp://www.ijbs.com \n2450 \nand fibrosis [3]. Notably, influence of Wnt /β-catenin \nsignaling on epithelial- mesenchymal transformation \nremains unexplored in EMS. \nIn Chinese medicine, EMS belongs to blood \nstasis syndromes, especially based on the blood stasis \nand obstruction of uterus  [4]. As the famous \ngynecological prescription on EMS, Foshou san has the \neffects of invigorating blood and fusing stasis  [5]. \nFerulic acid and ligustrazine are the active ingredients \nof Foshou san . Both of them join with \ntetrahydropalmatine to form FLT, a new Chinese \nherbal monomer recipe. FLT has displ ayed the \nconceivable reaction on autograft EMS [6]. However, \nthe effect of FLT is still unclear in endometrium cells \nand allogenic EMS. \nIn the study, t herapeutic effect of FLT was \nobserved in vivo and in vitro. Using endometrial cells, \nscratch test and transwell assays were performed after \ntreatment with FLT. Regulations of  epithelial- \nmesenchymal transformation and Wnt /β-catenin \nsignaling were detected by FLT in endometrium cells \nand allogenic EMS. \nMaterials and methods \nAnimals and chemicals \nROSAmT/mG mice were kindly provided by Pro. \nLin Chen of Army Medical University from Jackson \nLab. 18-22 g C3H mice were supplied for cross \nbreeding (Vital River Laboratory Animal Technology, \nAuthorization SCXK [Jing] 2016-0006, Beijing, China). \n18-22 g female nude mice were used for allograft EMS \n(Silaike Jingda Animal, Authorization SCXK [Xiang] \n2016-0002, Hunan, China). The operation of mice was \napproved by Care and Use of Laboratory Animals of \nSouthwest University (Authorization 201702). \nAnesthesia and other necess ary methods were \nprovided to reduce suffering. \nFLT were composed of 99.8% ferulic acid, 99.3% \nligustrazine, and 98.1% tetrahydropalmatine (Zelang \nMedical Technology, Nanjing, China ), with the ratio \nof 1:0.5:0.3. FLT were dissolved at 0.5% CMC -Na or \nDMSO fo r mice administration or endometrial cell \ntreatment separately. The positive control selected \ngestrinone (Zizhu Pharmaceutical Co., Beijing, \nChina). LiCl (Sigma-Aldrich, USA) or XAV-939 (MCE, \nUSA) was provided as activator or inhibitor of \nWnt/β-catenin pathway. \nFluorescent allograft EMS model and \ntreatment \nAllograft operation was slight ly modified from \nthe previous studies  [7]. Fluorescence b ilateral \nuteruses were collected from estrus  female \nROSAmT/mG mice. Then, 4 mm 2 fluorescent uteruses \nwere transferred to subcutaneous abdomens in nude \nmice. After operation, nude mice were \nintramuscularly administrated with 2 mg·Kg −1 \nestradiol in 5 days’ interval.  28 days later, ectopic \nlesions were captured by Fusion -FX7 imaging system \n(Vilber Lourmat, France). Fluoresc ent intensity more \nthan 9.58E+6 were regarded as successful EMS model. \nAccording to fluorescent intensity, EMS mice were \nrandomly into 5 groups. Firstly, EMS group was \ntreated with CMC -Na. Secondly, there were 3  FLT \ngroups using 90, 180, and 360 mg·Kg −1 FLT. Finally, \nanother group was treated with 2 mg·Kg−1 gestrinone. \nThen, no operation mice were performed for control \nwith same treatment as EMS group. In the  end of 28 \ndays’ treatment, Evolution -Capt software (Version \n18.2, Vilber Lourmat, France) was proce ssed to \ninvestigate and analyze fluorescence intensity of \nectopic lesions. Meanwhile, cubage of xenogeneic \nEMS was estimated with vernier caliper. \nH&E staining, E2 and PROG detection by \nELISA \nEutopic and allograft endometrium were \ngathered from control and other 5 groups separately. \nThen tissues were marked with hematoxylin and \neosin after fixation with paraformaldehyde.  \nEndometrial morphology was observed in the \nmicroscope (DFC310 FX, Leica, Germany). \nAccording to ELISA kit instruction, prepared \nserums were added into the equilibrated plates with \ndilution reagents. Then the plates were annexed with \nHRP-conjugate reagents, and incubated for 30 min. \nAfter incubating with chromogenic reagent, the plates \nwere processed to stop reaction. Then absorbency was \ndetected at 450 nm in spectrophotometer reader (Bio \nTek, USA). \nCell culture \nPro. Xiao -hong Chang of Peking University \nprovided hEM15A cells, the human endometrial \nstromal cells from EMS patients  [8]. The endometrial \ncarcinoma HEC1 -B cells were supplied by Chinese \nCentre for Type Cultures Collections (Wuhan, China). \nhEM15A or HEC1 -B cells were cultured in DMEM/  \nF12 or MEM (Gibco, Grand Island, NY, USA) with \n10% FBS (Hyclone, Shanghai, China) separatel y. The \ntwo cells grew in a 37 °C incubator supply of 5% CO\n2. \nScratch wound assay \nIn 24-well microplate, 6 -9×104 endometrial cells \nwere cultured in each well. After nearly 80% cell \nconvergence, 1 ml pipette tip were utilized for \nwounded traces in plate. Treated with different doses \nof FLT, microplates were detected under 50 ×  \nmagnification in different time points. Image software \n\nInt. J. Biol. Sci. 2021, Vol. 17 \n \n \nhttp://www.ijbs.com \n2451 \n(6.0, Media Cybernetics, USA) was utilized for \ncalculation of scratch size. Migration rate = (average \nscratching size in 0 h -  average scratching size in 24 \nh)/average scratching size in 0 h × 100%. \nTranswell assay \n3-5×104 endometrial cells were cultured in each \nwell with serum-free FLT medium in transwell inserts \n(Corning, New York, USA). The lower chambers of \n24-well plate were distributed with 5% serum \nmedium. Then plate with transwell inserts were \nincubated at 37 °C  for 24 h. The bottom of transwell \ninserts were dyed with gentian violet for observation \nof infiltrating cells excluding surface cells. Under the \nmicroscope, 5 vision were randomly observed for  \ninfiltrating cell count. \nRNA isolation and RT-qPCR \nTRIzol reagent (Invitrogen, CA, USA) was  \nutilized RN A isolation  in vivo  and in vitro . Then \nPrimeScript™ RT Reagent Kit was  purchased from \nTakara (China) for cDNA synthesis. SYBR™ Mix of \nThermo Fisher (USA) was applied for gene expressing \nwith 2\n-∆∆CT calculation. Primer sequences of mRNA \nwere produced by Shenggong Biotechnology \n(Shanghai, China) (Table S1). β -actin or GAPDH was \nperformed as the control reference. \nWestern blotting test \nProteins of allografts and cells were processed to \nwestern blotting following previous method  [9]. \nSDS-PAGE were used to divide the protein samples. \nAfter separation, protein of tissues and cells were \ndelivered onto PVDF membrane. Commercial \nprimary antibodies reacted with the protein on \nmembranes at 4  °C overnight (Table S2). Then goat \nsecond antisubstance was provide by Multi S ciences \n(China) with 1:5,000 dilution. Imaging system of \nTanon were utilized for chemical luminescence \ndetection. Inner reference was selected with β-actin or \nβ-tubulin. \nStatistical method \nOne-way ANOVA approach was performed for \ndata analysis using SPSS 2 1.0 software. P values, less \nthan 0.05 threshold, was regarded as statistical \ndifference. \nResults \nFLT inhibited ectopic lesions, E2 and PROG \nlevels in fluorescent allograft EMS \nAfter allotransplantation for 28 days, 25 of 30 \nnude mice were testfied fluorescent allograft EMS \nwith 83.33% success rate. Before treatmeant, distinct \ndifference of fluorescent intensity were not found in \nEMS, FLT, and gestrinone groups.  After treating fo r \n28 days, the fluorescent intensity showed no diversity \ncompared to pretreatment in EMS group. Using FLT, \nremarkable decrease of fluorescent intensity were \ndetected in recipient nude mice (Fig. 1A, B). Then \nfluorescent intensity and volume of isolated ect opic \nissue were investigated through the second \nlaparotomy. Compared with EMS group, FLT \nreduced the fluorescent intensity and volume of \nisolated ectopic issue  (Fig. 1C-F). Meanwhile, \ngestrinone showed the similar effects. FLT shows the \nprohibition of allograft EMS growth. \nEctopic endometrium of EMS group expressed \nthe similar structure as eutopic endometrium in \ncontrol group. In ectopic endometrium, epithelial \nglands were surrounded by endometrial stromal \ntissue. Using 360 mg·Kg\n-1 FLT, pathological \nmorphology of ectopic lesion were ameliorated, \naccompany with less and smaller ectopic glands, \nfewer microvascular and inflammatory infiltration \n(Fig. 1G). \nEMS is regarded as a benign gynecological \ndisease with the characteristics of estrogen and \nprogesterone abn ormity[10]. In EMS, E 2 and PROG \nlevels in serum were remarkably enhanced more than \ncontrol group. Using FLT, E\n2 contents were \nremarkably diminished less than EMS group. \nHowever, PROG contents in FLT groups were \nsignificantly reduced vs EMS group. It was no tenable \nthat only 360 mg·Kg\n-1 FLT could decreased both E 2 \nand PROG levels, same as  gestrinone group (Fig. 1H, \nI). \nFLT prohibited epithelial-mesenchymal \ntransformation in vivo \nDuring epithelial -mesenchymal transformation, \nE-cadherin is considered as the epi thelial marker. \nMeanwhile, the mesenchyme biomarkers include \nN-cadherin, Twist, Slug, Snail, ZEB1, and Vimentin \n[1]. After treatment with FLT, epithelial-mesenchymal \ntransformation markers were detected in endometrial \nallotransplantation. In EMS group, gene expression of \nE-cadherin showed remarkably decreasing by \nqRT-PCR ( P<0.01). Different concentrations of FLT \nexpressed the different effect, which were lower in 90 \nmg·Kg\n-1 group, and higher in 180 or 360 mg·Kg -1 \ngroups (Fig . 2A). In EMS group, 6 mesenchymal \nmRNA levels, including N -cadherin, Twist, Slug, \nSnail, ZEB1, and Vimentin, obviously expanded vs \ncontrol group. FLT obviously downregulated gene \nexpression of N -cadherin, ZEB1, Twist and Slug, \nwhile mRNA levels of Vimentin and Snail showed the \nopposite changes (Fig . 2B, C). Furthermore, \nE-cadherin protein was remarkably reduced with \nraising N -cadherin, Vimentin, Snail in EMS group. \n\nInt. J. Biol. Sci. 2021, Vol. 17 \n \n \nhttp://www.ijbs.com \n2452 \nFLT obviously promoted E -cadherin protein \nexpression, accompanied by demotion of N -cadherin, \nSnail, Vimentin (Fig. 2D-F). The diverse results of \nVimentin and Snail genes and protein were indicated \nthat FLT might have the post -translational \nmodification on them. \n \n \nFigure 1. FLT inhibited fluorescent allograft EMS. (A-D) After treatment for 28 days, fluorescent intensity of ectopic lesions was detected in vivo imaging system. (E, F) \nVolume of isolated ectopic lesions were calculated by vernier caliper. (G) Using H&E staining, pathological changes were observed in control, EMS and 360 mg·Kg-1 FLT group. \n(H, I) E2 and PROG levels in serum were investigated by ELISA assay. # P< 0.05 to pretreatment, ## P< 0.01 to pretreatment, ∗ P< 0.05 to EMS, ∗∗ P< 0.01 to EMS. Columns, \nmean (n=5). Bars, SD. Upper scale bar=100 μm. Lower scale bar=100 μm. EMS, endometriosis; FLT, ferulic acid, ligustrazine and tetrahydropalmatine; GTN, gestrinone. \n\n\nInt. J. Biol. Sci. 2021, Vol. 17 \n \n \nhttp://www.ijbs.com \n2453 \n \nFigure 2. Downregulation of epithelial-mesenchymal transformation and Wnt/β-catenin pathway by FLT in vivo. (A-C) The mRNA levels of E-cadherin, N-cadherin, Vimentin, \nSnail, ZEB1, Twist, and Slug were detected by RT-qPCR. (D-F) Protein expression of E-cadherin, N-cadherin, Vimentin, and Snail were measured by western blot. (G, H) mRNA \nlevels of APC, GSK3β, β-catenin, c-Myc, and CyclinD1 were detected by RT-qPCR. (I-K) Protein expression of Wnt3a, GSK3β, p-GSK3 β, β-catenin, p-β-catenin, and c-Myc \nwere analyzed by western blot. # P< 0.05 to control, ## P< 0.01 to control, ∗ P< 0.05 to EMS, ∗∗ P< 0.01 to EMS. Columns, mean (n=3). Bars, SD. EMS, endometriosis; FLT, ferulic \nacid, ligustrazine and tetrahydropalmatine. \n\n\nInt. J. Biol. Sci. 2021, Vol. 17 \n \n \nhttp://www.ijbs.com \n2454 \nInhibition of Wnt/β-catenin signaling by FLT in \nvivo \nRecently, EMS formation has been efficiently \narrested by down -regulation of Wnt /β-catenin \npathway [11]. Compared with control, Wnt/β-catenin \nsignaling was mobilized in EMS, through suppression \nof APC and GSK3 β, enhancement of downstream \nβ-catenin, c -Myc and CyclinD1. Using FLT, the \nmRNA level of APC and GSK3β obviously expanded, \nwhile β-catenin, c-Myc and CyclinD1 attenuated (Fig. \n2G, H). In addition, the protein expressed similarly as \ngenes of Wnt /β-catenin pathway. In EMS, higher \nlevels of Wnt3a, p -GSK3β, β-catenin, and c-Myc were \nobserved. Moreover, lower levels of GSK3 β and \np-β-catenin were inspect ed vs control group. FLT \nregulated W nt/β-catenin pathway step by step. \nFirstly, FLT extremely attenuated Wnt3a. Then FLT \ninduced the total protein of GSK3 β, while decreasing \np-GSK3β. Thirdly, down-regulation of β-catenin were \nconcomitant with up -regulation of p -β-catenin. \nFinally, c-Myc was lessened by FLT (Fig. 2I-K). \nFLT restrain invasiveness and migration in \nvitro \nAfter treated with FLT for 24 h, hEM15A and \nHEC1-B cells were prevented to cicatrize, especially in \nhigh dose ( P<0.01) (Fig. 3A -D). Cell capac ity of \nmigration and invasiveness is investigated in \ntranswell test. Migrating cells were significantly \ndecreased across transwell insert membrane in FLT vs \ncontrol (Fig. 3E -H). All above data implied that cell \nmigration and invasion were resisted by FLT \nconcentration-dependently. \nSuppression of epithelial-mesenchymal \ntransformation by FLT in vitro \nEpithelial-mesenchymal transformation is \ncorrelated to invasion and metastasis [12]. In hEM15A \ncells, E-cadherin gene expression expanded after 24 h \nusage of 240  μg.mL-1 FLT. The mRNA levels of  \nN-cadherin and  Vimentin obviously attenuated in \nFLT groups compared to control group (Fig. 4A).  \nMeantime, i n HEC1 -B cells, FLT increased mRNA \nlevel of E -cadherin, and decreased mRNA levels of  \nSnail and Slug (Fig. 4B). Moreover, using FLT, \nE-cadherin protein expressed higher than control. \nSimultaneously, the protein  of mesenchyme \nbiomarkers was downregulated in hEM15A cells, \nsuch as N-cadherin, Slug, and Vimentin (Fig. 4C). At \nthe same time, the protein  of N-cadherin, Snail, and \nSlug decreased, with accumulation of E- cadherin in \nHEC1-B cells (Fig. 4D). The data suggested that FLT \nhad promotion on epithelial marker, and interruption \nof mesenchymal markers in endometrial cells. \nArrest of Wnt/β-catenin signaling in \nendometrium cells treated with FLT \nAfter using FLT for 24 h, there was no obvious \neffect of FLT on GSK3 β gene expression in hEM15A \ncells. mRNA level of β-catenin gene was remarkably \nreduced especially using 480 and 960 μg.mL -1 FLT. \nTreated with 480 μg.mL -1 FLT, the mRNA level of \nCyclinD1 were significantly down-regulated (Fig. 4E). \nOnly 313 μg.mL -1 FLT increased the expression of \nGSK3β gene in HEC1-B cells, while 1250 μg.mL -1 FLT \nreduced β-catenin mRNA level. The gene expression \nof CyclinD1 became less in  625 and 1250 μg.mL-1 FLT \ngroups than control group (Fig. 4F).  Protein level of \nGSK3β increased in FLT group compared with control \ngroup in hEM15A cells. Simultaneously, the protein \nlevel of p -GSK3β, β-catenin, and CyclinD1 decreased \n(Fig. 4G). In HEC1 -B c ells, the protein levels of \np-GSK3β, β-catenin, and CyclinD1 reduced in FLT \ngroups, accompanied with accumulation of GSK3 β \n(Fig. 4H). \nFLT reversed epithelial-mesenchymal \ntransformation through regulating Wnt/ \nβ-catenin pathway \nDuring canonical Wnt  signaling, β- catenin is \nsubsequently degraded after phosphorylated by \nGSK3β [13]. LiCl is a special inhibitor of GSK3β, \nwhich could promote β -catenin [9]. In endometrium \ncells, stimulation of Wnt /β-catenin signaling lead to \nepithelial-mesenchymal transfor mation induced by \nLiCl. Consequently, LiCl inevitably promoted cell \nmigration and invasion. FLT reversed LiCl -activated \nWnt/β-catenin pathway. Therefore, FLT restrained \nepithelial-mesenchymal transformation inducing by \nLiCl (Fig. 5A, F). At the same time, cell invasiveness \nand migration were in hibited by FLT in scratching \nand transwell assay (Fig. 5B-E, G-J). As a consequence, \nLiCl activated the Wnt/β-catenin pathway, promoted \nepithelial-mesenchymal transformation, invasiveness \nand migration. FLT had the antagonistic effect with \nLiCl. \nXAV-939, a tankyrase inhibitor, can increase \nphospho-β-catenin through inducing degradation \n[14]. Wnt /β-catenin signaling was obviously \nsuppressed by XAV -939 in hEM15A and HEC1 -B \ncells. Dysfunction of Wnt /β-catenin signaling caused \nprohibition of epithelial-mesenchymal transformation \nthrough increasing E -cadherin and decreasing \nVimentin in both cells, down- regulating Snail in \nhEM15A cells. Simultaneously, XAV -939 inevitably \nrestricted invasiveness and migration in endometrial \ncells. FLT enhanced XAV -939-induced inactivation of \nWnt/β-catenin pathway through raising GSK3β, \ndiminishing phospho- GSK3β (Ser9), β -catenin and \nCyclinD1. Additionally, FLT strengthened the effect \n\nInt. J. Biol. Sci. 2021, Vol. 17 \n \n \nhttp://www.ijbs.com \n2455 \nof XAV-939 on E-cadherin, N-cadherin and Snail (Fig. \n6A, F). Cell migration and invasion were prevented by \nFLT and XAV -939 (Fig. 6B -E, G-J). In brief, XAV -939 \nreversed epithelial -mesenchymal transformation \nthrough repression of Wnt /β-catenin signaling, \nresulting in inhibition of invasion and metastasis. FLT \nhad the synergistic effect with XAV-939. \n \n \nFigure 3. Suppression of invasion and metastasis using FLT in endometrial cells. (A-D) After treated with FLT, cell migration of hEM15A and HEC1 -B cells was observed in \nscratch wound assay. (E-H) Transwell assay were performed to investigate migration and invasion ability. ∗ P< 0.05 to control, ∗∗ P< 0.01 to control. Columns, mean (n=3). Bars, \nSD. Scale bar=200 μm. FLT, ferulic acid, ligustrazine and tetrahydropalmatine. \n\n\nInt. J. Biol. Sci. 2021, Vol. 17 \n \n \nhttp://www.ijbs.com \n2456 \n \nFigure 4. FLT downregulated epithelial-mesenchymal transformation and Wnt/β-catenin pathway in endometrial cells. (A, B) The mRNA levels of E -cadherin, N-cadherin, \nVimentin, Snail and Slug were detected by RT-qPCR. (C, D) Protein levels of E-cadherin, N-cadherin, Vimentin, Slug and Snail were measured by western blot. (E, F) The mRNA \nlevels of GSK3β, β-catenin and CyclinD1 were analyzed by RT-qPCR. (G, H) Protein levels of GSK3β, p-GSK3β, β-catenin, and CyclinD1 were observed with western blot. ∗ \nP< 0.05 to control, ∗∗ P< 0.01 to control. Columns, mean (n=3). Bars, SD. FLT, ferulic acid, ligustrazine and tetrahydropalmatine. \n\n\nInt. J. Biol. Sci. 2021, Vol. 17 \n \n \nhttp://www.ijbs.com \n2457 \n \nFigure 5. FLT reverses LiCl-induced activation of epithelial-mesenchymal transformation. (A, F) After using LiCl or FLT, protein levels of Wnt/β-catenin pathway and epithelial- \nmesenchymal transformation were observed by western blot. (B-E, G-J) Cell migration and invasion ability were measured by scratch wound and transwell assay in 24h. ∗ P< \n0.05 to control, ∗∗ P< 0.01 to control. Columns, mean (n=3). Bars, SD. FLT, ferulic acid, ligustrazine and tetrahydropalmatine. \n\n\nInt. J. Biol. Sci. 2021, Vol. 17 \n \n \nhttp://www.ijbs.com \n2458 \n \nFigure 6. FLT enhances XAV-939-mediated attenuation of epithelial-mesenchymal transformation. (A, F) After using XAV-939 or FLT, protein levels of Wnt/β-catenin pathway \nand epithelial-mesenchymal transformation were observed by western blot. (B-E, G-J) Treating with XAV-939 or FLT for 24h, cell migration and invasion ability were measured \nby scratch wound and transwell assay. ∗ P< 0.05 to control, ∗∗ P< 0.01 to control. Columns, mean (n=3). Bars, SD. FLT, ferulic acid, ligustrazine and tetrahydropalmatine. \n\n\nInt. J. Biol. Sci. 2021, Vol. 17 \n \n \nhttp://www.ijbs.com \n2459 \nDiscussion \nIn this experiment, FLT markedly inhibited \nfluorescent ectopic issue growth, E2 and PROG levels, \nameliorated pathological change in allograft EMS  \nmodel. Meanwhile, FLT prevented invasiveness and \nmigration in endometrial cells, which was correlated \nto repress epithelial -mesenchymal transformation via \nWnt/β-catenin pathway. \nNowadays, EMS has the high morbidity in \nreproductive age women with equivocal  \npathogenesis. Recognized phenomenon recovers that \nimplantation of active endometrium cells mainly \ndepends on invasion and metastasis in menstrual \nreflux hypothesis  [15, 16] . Epithelial -mesenchymal \ntransformation leads to enhancement of invasion and \nmetastasis. During this process, epithelial marker, for \nexample E -cadherin, decreases. Mesenchymal \nmarkers, for example N -cadherin, Vimentin, ZEB1, \nZEB2, Snail, Slug, and Twist increase  [12, 17] . \nRecently, Epithelial-mesenchymal transformation has \nbeen significantly elevated in EMS. Inhibition of \nepithelial-mesenchymal transformation reverses the \nprogress of EMS  [18-20]. Consistently, in our \nexperiment, interstitial biomarkers, including \nVimentin, N -cadherin, Snail, Slug, ZEB1 and Twist \nenhanced, when E -cadherin as epithelial marker \nsimultaneously attenuated in fluorescent allograft \nEMS mice. Interestingly, FLT restrained the growth of \nectopic lesions, at the same time suppressed invasion \nand metastasis in  vitro. In depth study, after treated \nwith FLT, epithelial -mesenchymal transformation \nwas prevented through ascendant E- cadherin, \ndescendant interstitial biomarkers, such as \nN-cadherin, Vimentin, Snail, Slug, Twist, and ZEB1 in \nvitro. These data indicated the underlying mechanism \nof FLT on invasiveness and migration may be related \nto inhibit epithelial-mesenchymal transformation. \nThe canonical Wnt signaling is prevalent in \ndevelopment, differentiation, and proliferation  [21]. \nRecently, Wnt/β -catenin signali ng is reported to \ncorrelate with epithelial-mesenchymal transformation \nin serval kinds of cancer. Wnt /β-catenin signaling \nactivation appears that accumulation of β -catenin is \ntransformed to nuclear. After binding with the \ndownstream transcription factors, β-catenin triggers \nthe target gene expression, such as Slug and Twist. \nMeanwhile, downregulation of E-cadherin causes less \ncombination with β-catenin in membrane. That leads \nto promotion of epithelial -mesenchymal \ntransformation [22-24]. Although Wnt /β-catenin \nsignaling dysfunction has been found in EMS, \nmechanism of Wnt /β-catenin signaling on  \nepithelial-mesenchymal transformation is still await \ninvestigation. To uncover the mechanism, \nendometrial cells were treated with activator and \ninhibitor of Wnt/β-catenin signaling. Activation of \nWnt/β-catenin signaling facilitated \nepithelial-mesenchymal transformation, invasion and \nmetastasis. By contrast, constraint of Wnt/β -catenin \npathway interrupted epithelial -mesenchymal \ntransformation, invasion and metastasis. \nFerulic acid, ligustrazine and tetrahydro -\npalmatine are the components of FLT, the novel \nChinese herbal monomer recipe. In previous research, \nferulic acid represses metastasis through withdraw \nepithelial-mesenchymal transformation in breast \ncancer [25]. Furt hermore, combining with other \ncomponents, ferulic acid induces Wnt/β -catenin \nsignaling, resulting in promotion of bone matrix in \nosteoblasts [26]. Beyond that, ligustrazine restrains \nepithelial-myofibroblast transformation by TGF -β/ \nSmads signaling in rena l tubular epithelial cells  [27]. \nBut tetrahydropalmatine on epithelial -mesenchymal \ntransformation or Wnt/β -catenin signaling has not \nbeen reported. In our results, as the combination of \nferulic acid, ligustrazine, and tetrahydropalmatine, \nFLT restricted Wn t/β-catenin signaling, causing \ndownregulation of epithelial- mesenchymal \ntransformation in allograft EMS and endometrial \ncells. FLT appeared the antagonistic effect with \nWnt/β-catenin pathway activator; on the contrary, \nFLT had the synergistic effect with  Wnt/β-catenin \npathway inhibitor.  It is worthwhile to explore other \nmechanism of FLT on EMS. \nConclusion \nFLT prevented invasiveness and migration in \nendometrium cells, and ectopic growth of allograft \nEMS. FLT suppressed epithelial -mesenchymal \ntransformation through inhibiting Wnt/β -catenin \nsignaling. These results reveal the pharmacologic \nmechanism for FLT further research. \nAbbreviations \nFLT: ferulic acid, ligustrazine and \ntetrahydropalmatine; EMS: endometriosis; GTN:  \ngestrinone; E2 : estrogen; PROG : progesterone; H&E: \nhematoxylin and eosin; CMC -Na: carboxymethyl \ncellulose sodium; DMSO:  dimethyl sulfoxide; FBS: \nfetal bovine serum; CO2:  carbon dioxide; RT -qPCR: \nReverse transcription -quantitative real -time PCR; \nGAPDH: glyceraldehyde-3-phosphate dehydro -\ngenase; SDS -PAGE: sodium dodecyl sulfate - \npolyacrylamide gel electrophoresis; HRP: horseradish \nperoxidase; GSK3β:  glycogen synthase kinase -3β; \nAPC: adenomatous polyposis coli; PVDF:  \npoplvinylidene fluoride. \n\nInt. J. Biol. Sci. 2021, Vol. 17 \n \n \nhttp://www.ijbs.com \n2460 \nSupplementary Material  \nSupplementary tables.  \nhttp://www.ijbs.com/v17p2449s1.pdf  \nAcknowledgements \nFunding statement \nThis work was supported by National Natural \nScience Foundation of China [grant numbers \n81773984, 81402441]; Chongqing Natural Science \nFoundation of China [grant numbers \nCSTC2020JCYJ-MSXM1639]; Traditional Chinese \nmedicine research project of Chongqing Health \nBureau [grant numbers 2020ZY023665]. \nCompeting Interests \nThe authors have declared that no competing \ninterest exists. \nReferences \n1. Dongre A, Weinberg RA. New insights into the mechanis ms of \nepithelial-mesenchymal transition and implications for cancer. Nat Rev Mol \nCell Biol. 2019; 20: 69-84. \n2. Konrad L, Dietze R, Riaz MA, Scheiner -Bobis G, Behnke J, Horne F, et al. \nEpithelial-Mesenchymal Transition in Endometriosis-When Does It Happen? J \nClin Med. 2020; 9: 1915. \n3. Shao X, Wei X. FOXP1 enhances fibrosis via activating Wnt/beta- catenin \nsignaling pathway in endometriosis. Am J Transl Res. 2018; 10: 3610-8. \n4. Liao F. Herbs of activating blood circulation to remove blood stasis. Clin \nHemorheol Microcirc. 2000; 23: 127-31. \n5. Bai H, Li PQ, Liu J, Liu XP. [Association analysis on traditional efficacy and \nmodern research of Foshou San]. Chinese Traditional Patent Medicine. 2014; \n36: 601-4. \n6. Wei JH, Zhao BX, Zhang CL, Shen BB, Zhang Y, Li CX, et al. Jiawei Foshou San \nInduces Apoptosis in Ectopic Endometrium Based on Systems Pharmacology, \nMolecular Docking, and Experimental Evidence. Evid-Based Compl Alt. 2019; \n2019: 2360367. \n7. Attaman JA, Stanic AK, Kim M, Lynch MP, Rueda BR, Styer AK. The \nanti-inflammatory impact of omega-3 polyunsaturated Fatty acids during the \nestablishment of endometriosis -like lesions. Am J Reprod Immunol. 2014; 72: \n392-402. \n8. Chen Y, Li H, Cheng HY, Rui -Qiong M, Ye X, Cui H, et al. Fibrinogen alpha \nchain is up -regulated and affects the pathogenesis of endometriosis. Reprod \nBiomed Online. 2019; 39: 893-904. \n9. Chen Y, Yu DK, Zhang CX, Shang BY, He HW, Chen JJ, et al. Lidamycin \nInhibits Tumor Initiating Cells of Hepatocellular Carcinoma Huh7 Through \nGSK3 beta/beta-Catenin Pathway. Mol Carcinogen. 2015; 54: 1-8. \n10. Barra F, Scala C, Biscaldi E, Vellone VG, Ceccaroni M, Terrone C, et al. Ureteral \nendometriosis: a systematic review of epidemiology, pathogenesis, diagnosis, \ntreatment, risk of malignant tran sformation and fertility. Hum Reprod \nUpdate. 2018; 24: 710-30. \n11. Zhu H, Cao XX, Liu J, Hua H. MicroRNA -488 inhibits endometrial glandular \nepithelial cell proliferation, migration, and invasion in endometriosis mice via \nWnt by inhibiting FZD7. J Cell Mol Med. 2019; 23: 2419-30. \n12. Pastushenko I, Blanpain C. EMT Transition States during Tumor Progression \nand Metastasis. Trends Cell Biol. 2019; 29: 212-26. \n13. Matsuzaki S, Botchorishvili R, Pouly JL, Canis M. Targeting the \nWnt/beta-catenin pathway in endome triosis: a potentially effective approach \nfor treatment and prevention. Mol Cell Ther. 2014; 2: 36. \n14. Lietman C, Wu B, Lechner S, Shinar A, Sehgal M, Rossomacha E, et al. \nInhibition of Wnt/beta-catenin signaling ameliorates osteoarthritis in a \nmurine model of experimental osteoarthritis. JCI Insight. 2018; 3(3):e96308. \n15. Mihailovici A, Rottenstreich M, Kovel S, Wassermann I, Smorgick N, Vaknin \nZ. Endometriosis-associated malignant transformation in abdominal surgical \nscar: A PRISMA-compliant systematic review. Medicine (Baltimore). 2017; 96: \ne9136. \n16. Chui MH, Wang TL, Shih IM. Endometriosis: benign, malignant, or something \nin between? Oncotarget. 2017; 8: 78263-4. \n17. Chen T, You Y, Jiang H, Wang ZZ. Epithelial -mesenchymal transition (EMT): \nA biologica l process in the development, stem cell differentiation, and \ntumorigenesis. J Cell Physiol. 2017; 232: 3261-72. \n18. Wu RF, Huang ZX, Ran J, Dai SJ, Lin DC, Ng TW, et al. Lipoxin A4 Suppresses \nEstrogen-Induced Epithelial -Mesenchymal Transition via ALXR -Dependent \nManner in Endometriosis. Reprod Sci. 2018; 25: 566-78. \n19. Chatterjee K, Jana S, DasMahapatra P, Swarnakar S. EGFR -mediated matrix \nmetalloproteinase-7 up -regulation promotes epithelial -mesenchymal \ntransition via ERK1 -AP1 axis during ovarian endometri osis progression. \nFASEB J. 2018; 32: 4560-72. \n20. Qi S, Yan L, Liu Z, Mu YL, Li M, Zhao X, et al. Melatonin inhibits \n17beta-estradiol-induced migration, invasion and epithelial -mesenchymal \ntransition in normal and endometriotic endometrial epithelial cells . Reprod \nBiol Endocrinol. 2018; 16: 62. \n21. Martin-Orozco E, Sanchez -Fernandez A, Ortiz-Parra I, Ayala- San Nicolas M. \nWNT Signaling in Tumors: The Way to Evade Drugs and Immunity. Front \nImmunol. 2019; 10: 2854. \n22. Pearlman RL, Montes de Oca MK, Pal HC, Afaq F. Potential therapeutic \ntargets of epithelial- mesenchymal transition in melanoma. Cancer Lett. 2017; \n391: 125-40. \n23. Wu C, Zhuang Y, Jiang S, Liu S, Zhou J, Wu J, et al. Interaction between \nWnt/beta-catenin pathway and microRNAs regulates epithelial-mesenchymal \ntransition in gastric cancer (Review). Int J Oncol. 2016; 48: 2236-46. \n24. Fang D, Kitamura H. Cancer stem cells and epithelial -mesenchymal transition \nin urothelial carcinoma: Possible pathways and potential therapeutic \napproaches. Int J Urol. 2018; 25: 7-17. \n25. Zhang X, Lin D, Jiang R, Li H, Wan J, Li H. Ferulic acid exerts antitumor \nactivity and inhibits metastasis in breast cancer cells by regulating epithelial to \nmesenchymal transition. Oncol Rep. 2016; 36: 271-8. \n26. Li M, Zhang ND, Wang Y, Han T, Jiang YP, Rahman K, et al. Coordinate \nregulatory osteogenesis effects of icariin, timosaponin B II and ferulic acid \nfrom traditional Chinese medicine formulas on UMR-106 osteoblastic cells and \nosteoblasts in neonatal rat calvaria cultures. J Ethnop harmacol. 2016; 185: \n120-31. \n27. Yuan XP, Liu LS, Fu Q, Wang CX. Effects of ligustrazine on ureteral \nobstruction-induced renal tubulointerstitial fibrosis. Phytother Res. 2012; 26: \n697-703.","source_license":"CC0","license_restricted":false}