ERβ e CLDN6 Inibem Migração e Invasão de Células de Câncer de Mama Via Autofagia

RESEARCH Open Access Estrogen receptor β inhibits breast cancer cells migration and invasion through CLDN6mediated autophagy Peiye Song Yanru Li Yuan Dong Yingying Liang Huinan Qu Da Qi Yan Lu Xiangshu Jin Yantong Guo Yiyang Jia Xinqi Wang Wenhong Xu and Chengshi Quan Abstract Background Estrogen receptor β ERβ has been reported to play an anticancer role in breast cancer but the regulatory mechanism by which ERβ exerts this effect is not clear Claudin6 CLDN6 a tight junction protein acts as a tumor suppressor gene in breast cancer Our previous studies have found that 17βestradiol E2 induces CLDN6 expression and inhibits MCF7 cell migration and invasion but the underlying molecular mechanisms are still unclear In this study we aimed to investigate the role of ERβ in this process and the regulatory mechanisms involved Methods Polymerase chain reaction PCR and western blot were used to characterize the effect of E2 on the expression of CLDN6 in breast cancer cells Chromatin immunoprecipitation ChIP assays were carried out to confirm the interaction between ERβ and CLDN6 Dual luciferase reporter assays were used to detect the regulatory role of ERβ on the promoter activity of CLDN6 Wound healing and Transwell assays were used to examine the migration and invasion of breast cancer cells Western blot immunofluorescence and transmission electron microscopy TEM were performed to detect autophagy Xenograft mouse models were used to explore the regulatory effect of the CLDN6 beclin1 axis on breast cancer metastasis Immunohistochemistry IHC was used to detect ERβCLDN6beclin1 expression in breast cancer patient samples Results Here E2 upregulated the expression of CLDN6 which was mediated by ERβ ERβ regulated CLDN6 expression at the transcriptional level ERβ inhibited the migration and invasion of breast cancer cells through CLDN6 Interestingly this effect was associated with CLDN6induced autophagy CLDN6 positively regulated the expression of beclin1 which is a key regulator of autophagy Beclin1 knockdown reversed CLDN6induced autophagy and the inhibitory effect of CLDN6 on breast cancer metastasis Moreover ERβ and CLDN6 were positively correlated and the expression of CLDN6 was positively correlated with beclin1 in breast cancer tissues Conclusion Overall this is the first study to demonstrate that the inhibitory effect of ERβ on the migration and invasion of breast cancer cells was mediated by CLDN6 which induced the beclin1dependent autophagic cascade Keywords Estrogen receptor β CLDN6 Autophagy Migration Invasion Breast cancer The Authors 2019 Open Access This article is distributed under the terms of the Creative Commons Attribution 40 International License httpcreativecommonsorglicensesby40 which permits unrestricted use distribution and reproduction in any medium provided you give appropriate credit to the original authors and the source provide a link to the Creative Commons license and indicate if changes were made The Creative Commons Public Domain Dedication waiver httpcreativecommonsorgpublicdomainzero10 applies to the data made available in this article unless otherwise stated Correspondence quancsjlueducn The Key Laboratory of Pathobiology Ministry of Education College of Basic Medical Sciences Jilin University 126 Xinmin Street Changchun 130021 Jilin China Song et al Journal of Experimental Clinical Cancer Research 2019 38354 httpsdoiorg101186s1304601913599 Background Estrogen plays an important role in hormonedependent breast cancer progression and metastasis The effects of estrogen are primarily mediated through the estrogen re ceptors ERs ERα and ERβ 1 The contribution of ERα to the normal development of the mammary gland and the tumorigenesis and progression of breast cancer is essential 2 ERα expression in normal breast epithelial cells is ap proximately 10 but increases to 5080 in breast cancer cells 3 Loss of ERα in breast cancer patients indicates poor prognosis and ERα has been the principal biomarker for endocrine therapy in breast cancer 4 However only 70 of ERαpositive breast cancers respond to tamoxifen ER antagonist treatment and 3040 of patients relapse during treatment and become resistant to endocrine ther apy 5 ERβ has the same structural domains as ERα but its function is not exactly the same as ERα The role of ERβ in breast cancer remains elusive and ERβ is currently not used in the diagnosis or treatment of breast cancer patients 6 Although a few studies claim that ERβ expression pro motes the invasion and metastasis of breast cancer and that high ERβ level is linked with poor prognosis 7 multiple studies have demonstrated that ERβ is an antioncogene in breast cancer In contrast to those of ERα clinical studies showed that the levels of ERβ were high in mammary epi thelial tissues and decreased during tumor progression 3 In triple negative breast cancer TNBC high expression of ERβ was significantly associated with good clinical outcome in patients treated with tamoxifen 8 In vitro studies showed that ERβ expression inhibited the cell proliferation and the migratory and invasive properties of breast cancer cells 9 10 Therefore ERβ may be a potential target for novel therapeutic avenues in breast cancer Nevertheless the molecular mechanisms underlying the inhibitory effects of ERβ on breast cancer remain unidentified and need to be explored Recent studies have suggested that ERβ could also trig ger autophagy 11 12 Autophagy plays a key role in the maintenance of cellular homeostasis 13 Dysregula tion of autophagy has been implicated in cancers A re cent study reported that ERβ activation could inhibit breast cancer cell proliferation by reducing the G2M phase as well as triggering autophagy 11 ERβinduced damage regulated autophagy modulator 2 DRAM2me diated autophagy has been associated with a reduction of cancer cell proliferation in Hodgkin lymphoma HL cells 12 However few studies have reported that the inhibitory role of ERβ on migration and invasion is dir ectly related to the modulation of autophagy in breast cancer Furthermore the regulatory mechanism of ERβ inducedautophagy is still unclear Intriguingly in this study we found that ERβ induced autophagy and inhib ited migration and invasion through claudin6 CLDN6 in breast cancer cells CLDN6 is a tight junction TJ protein that belongs to a family of transmembrane proteins with 4 transmem brane domains and 27 members of this family have been identified 14 15 As an important component of TJs CLDNs not only play important roles in the classic barrier and fence functions of TJs but are also involved in regulat ing cellular communication and signaling 16 CLDNs pos sess a carboxyterminal PDZbinding motif This domain allows CLDNs to interact with cytoplasmic scaffolding pro teins PDZ domaincontaining proteins such as zonula oc cludens ZO1 and afadin which are important for CLDNs to communicate with a multitude of signaling pro teins 17 In previous studies our group first cloned the CLDN6 gene from mammary epithelial cells of COP rats and identified CLDN6 as a breast cancer suppressor gene 18 19 We have reported that CLDN6 expression induces apoptosis and inhibits tumor growth migration and inva sion in breast cancer cells 2022 Moreover in our recent studies we found that CLDN6 overexpression not only strengthened the tight junctions in breast cancer cells but also induced a large number of autophagic vacuoles ob served under transmission electron microscopy TEM A series of subsequent experiments demonstrated that CLDN6 induced autophagy whereas the relationship be tween CLDN6induced autophagy and breast cancer re mains poorly investigated Our previous studies have shown that 17βestradiol E2 upregulates CLDN6 expression and hinders MCF7 cell mi gration and invasion but the molecular mechanisms are still unclear Interestingly in this study we found that the expression of CLDN6 was increased and migration and in vasion were hindered in both MCF7 ERαERβ GPR30 and MDAMB231 ERαERβGPR30 cells after E2 treatment Thus we supposed that this E2induced effect was not ERαdependent and we wanted to explore the role of ERβ in this process In view of the abovemen tioned reports we presumed that ERβ regulated CLDN6 expression and that ERβinduced autophagy affected migra tion and invasion in breast cancer cells In addition little is known about the role of CLDN6 in ERβinduced autoph agy Hence the current study aimed to explore the regula tory role of ERβ on CLDN6 expression in breast cancer cells and the mechanism related to its biological functions In this study we demonstrated a novel finding that ERβ in duced autophagy and inhibited migration and invasion via CLDN6 in breast cancer cells We also analyzed the mo lecular mechanisms underlying this effect in some details Methods Cell culture and reagents The human breast adenocarcinoma cell lines MCF7 MDA MB231 and SKBR3 were obtained from the Cell Bank of the Chinese Academy of Sciences Shanghai China and cul tured in DMEM GIBCO USA supplemented with 10 Song et al Journal of Experimental Clinical Cancer Research 2019 38354 Page 2 of 18 fetal bovine serum FBS Biological Industries Israel at 37 C with an atmosphere of 5 CO2 For estrogenfree cell culture cells were cultured in phenol redfree DMEM and serumfree HyClone USA for 12 h before specific treat ments The 17βestradiol E2 10006315 and PHTPP 16025 were purchased from Cayman Chemical Denver USA Diarylpropionitrile DPN H5915 was purchased from Sigma St Louis MO USA ICI 182780 HY13636 Chloroquine CQ HY17589 and 3methyladenine 3MA HY19312 were purchased from MCE New Jersey USA RTPCR and quantitative realtime PCR qRTPCR Total RNA was isolated from cells using TRIzol Invitrogen Carlsbad CA USA and converted to cDNA using First Strand cDNA Synthesis kit TransGen Biotech Beijing China Semiquantitative RTPCR and qRTPCR were car ried out as Liu et al 23 described previously The primers were synthesized by Sangon Shanghai China PCR primer sequences were as follows CLDN6 PCRforward TTCATCGGCAACAGCATC GT CLDN6 PCRreverse GGTTATAGAAGTCCCGGA TGA CLDN6 qRTPCRforward CCCTTATCTCCTTCGC AGTG CLDN6 qRTPCRreverse ATGCTGTTGCCGATGA AAG βactinforward CAGAGCCTCGCCTTTGCCGA TCC βactinreverse CCTTGCACATGCCGGAGCCGT Western blot analysis Protein extraction and western blot analysis were per formed as described by Yang et al 24 The following primary antibodies were used CLDN6 V118 11000 Bioworld Technology USA ERα ab66102 11000 ERβ ab288 11000 Abcam Cambridge UK LC3B D11 1 1000 beclin1 D40C5 11000 Atg5 D5F5U 11000 Atg16L1 D6D5 11000 Cell Signaling Technology Dan vers MA USA ZO1 217731AP 11000 UVRAG 195711AP 11000 βactin 600081Ig 15000 Pro teintech Group USA The signals were visualized using ECL reagent Millipore Billerica MA USA Wound healing assay The cells from each group were seeded onto 6well cul ture plates with corresponding treatment When cells had grown to full confluence wounds were created through the monolayers by using a sterile pipette tip The wounded areas were imaged at 0 h and 24 h by an inverted microscope Olympus Japan Image J software NIH USA was employed to analyze the wound widths from the images Transwell assay Transwell chambers Costar USA with matrigel were used to perform cell invasion assays An equal number of cells 1 105 cells were loaded into matrigel BD Bio sciences USA precoated chambers in 200 μl serumfree media NIH3 T3 conditioned medium as the chemo attractant was placed in the lower compartment of the chamber After 24 h incubation cells were fixed in 4 paraformaldehyde and stained with 01 crystal violet No invading cells were removed with cotton swabs Im ages were photographed in three randomly selected fields of view at 200 magnification NIH3 T3 condi tioned medium was prepared by culturing NIH3 T3 fi broblasts cells for 48 h in serumfree media 25 The supernatants were collected and stored at 80 C Immunofluorescence Cells were grown on coverslips and then incubated with or without DPN for 24 h and fixed in 4 formaldehyde The cells were blocked and incubated with primary anti bodies CLDN6 1200 ERβ 1200 and LC3B 1500 overnight at 4 C Cells were stained with Alexa Fluor 594488 goat antirabbit IgG Cell Signaling Technology Danvers MA USA for 1 h at room temperature Finally samples were counterstained with 4 6diamino2pheny lindole DAPI Sigma Images were taken with fluores cence microscopy Olympus Japan Transmission electron microscopy TEM Cells were exposed to DPN with or without 100 nM for 24 h The treated cells were fixed with 4 glutaraldehyde and postfixed in 1 OsO4 Samples were dehydrated through a graded series of ethanol solutions and embed ded in Eponate 12 epoxy resin Ultrathin sections were counterstained with uranyl acetate and lead citrate Ob servation and photography carried out by transmission electron microscopy FEI Tecnai Spirit USA Transfection Three candidate target sequences of ERβ were cloned into U6MCSUbiquitinCherryIRESpuromycin vectors Gene Chem Shanghai China Scrambled RNAs were used as a negative control NC for nonsequencespecific effects Infor mation on CLDN6 overexpression and CLDN6 short hairpin RNA shRNA plasmids has been described previously 26 The ERβ overexpression plasmid was generated by cloning ERβ cDNA into the pLentiCMVGFPPuro vector PPL008074a PPL Genebio Technology Nanjing China Beclin1 shRNA was cloned into the pPLKGFP Puro vec tor PPL000433b PPL Genebio Technology Nanjing China All lentiviral plasmids were cotransfected into 293 T cells together with packaging plasmids pMD2G and psPAX2 The supernatants containing lentiviral particles were collected 48 h after transfection and used to transduce Song et al Journal of Experimental Clinical Cancer Research 2019 38354 Page 3 of 18 cells according to the manufacturers protocol of Lipofecta mine 2000 Invitrogen USA The target for human ERβ shRNA sequences were as follows shRNA15 TGCTTTGGTTTGGGTGATT 3 shRNA2 5 TTCTCCTTTAGTGGTCCAT 3 shRNA3 5 GTAAACAGAGAGACACTGAAA 3 The target for human beclin1 shRNA was 5 GTGG ACACGAGTTTCAAGATT 3 Nuclearcytosol fractionation assay 2 106 MDAMB231 cells were seeded in 10cm culture plates and treated with DMSO control or DPN 100 nM for 24 h respectively The nuclear and cytosolic fractions were isolated with a NuclearCytosol Fractionation Kit Beyotime Biotechnology Shanghai China according to the manufacturers instructions The subcellular protein extracts were then analyzed by western blot Chromatin immunoprecipitation ChIP assay MDAMB231 cells were grown in 10cm culture plates and cultured with serumfree medium for 12 h and then treated with 100 nM DPN for 24 h Following treatment ChIP assay was performed as described previously 23 with minor modifications The precleared chromatin was incubated overnight with ERβ ab288 Abcam Cam bridge UK or Sp1 D4C3 Cell Signaling Technology Danvers MA USA and immunoprecipitated with protein AG magnetic beads A normal rabbit IgG Millipore Bil lerica MA USA was used as negative control Amplified DNA fragments were visualized on a 2 agarose gel PCR primer sequences were as follows ChIPCLDN61 forward GCTTAAGTGGTGAAGC GGAG ChIPCLDN61 reverse CAGACGTCCAGACTCACC CA ChIPCLDN62 forward TGTGCGTGTTGGAGAG ACG ChIPCLDN62 reverse CGAAGGACCCTATCACCT CG Dual luciferase reporter assay MDAMB231 cells were seeded in 6well culture plates and transfected with pGL3CLDN6 plasmid which containing CLDN6 promoter fragment 2000 250 bp and renilla lu ciferase reporter plasmid pRLTK After transfection for 48 h cells were treated with DPN Luciferase activities were measured using the dualluciferase reporter assay system Promega San Luis Obispo CA USA according to the manufacturers protocol Firefly luciferase activity was nor malized to renilla luciferase activity Reporter plasmids were purchased from GeneChem Co Ltd Shanghai China Coimmunoprecipitation coIP assay After DPN treatment cells were harvested and lysed with NP 40 lysis buffer Protein lysates were incubated with the anti UVRAGantiZO1 antibody or normal rabbit IgG and ro tated overnight at 4 C to form the immunecomplex Reac tion mixture was incubated with protein AG plusagarose beads Santa Cruz Biotechnology Santa Cruz CA USA for 3 h at 4 C Agarose beads were washed five times with cold washing buffer and heated with 40 μl 1 SDS buffer to 100 C for 5 min The samples were analyzed by Western blot Immunohistochemistry IHC Human breast cancer tissue specimens HBreD070CS02 were purchased from Shanghai Outdo Biotech CO The tissues n 70 including 11 paracancerous tissues 4 intraductal carcinoma and 55 invasive ductal carcinoma The primary antibodies against ERβ ab288 1200 Abcam CLDN6 V118 1200 Bioworld Technology and beclin1 D40C5 1200 CST respectively and then incubated at 4 C overnight in a humidified container Following washing three times with phosphate buffer saline PBS the section was treated with UltraSensitive SP MouseRabbit IHC Kit ac cording to the manufacturers instructions KIT9710 MXB Biotechnology Fuzhou China 3 3diaminobenzidine DAB was used for color development Immunostaining was evalu ated by two pathologists using a blind protocol design For ERβ CLDN6 and beclin1 each tissue sample was scored ac cording to its staining intensity 0 none 1 weak 2 moderate 3 strong multiplied by the point of the percentage of stained cells positive cells 25 of the cells 1 2650 of the cells 2 5175 of the cells 3 75 of the cells 4 The range of this calculation was 012 The median value of scores was employed to determine the cutoff Cancers with scores above the cutoff value were considered to have high expression of the indicated molecule and vice versa Animal studies An experimental model of lung metastasis was used to in vestigate the effects of CLDN6beclin1 axis on breast can cer metastasis Three groups of five female Balbc nude mice 46 weeks old and 1820 g were maintained in an environment with a standardized barrier system of the Ex perimental Animal Center of Jilin University One group was injected with the MDAMB231NC cells the second group was injected with the MDAMB231CLDN6 cells and the third group was injected with the MDAMB231 CLDN6shbeclin1 cells For each nude mouse 1 106 cells in 100 μl PBS were injected via the tail vein After 4 weeks of observation mice were sacrificed The lungs and major organs were removed and observed by using a fluorescence imager IVIS Spectrum Caliper Life Sciences Then these organs were fixed embedded and sectioned The metasta sis ability was detected by hematoxylin and eosin HE staining Song et al Journal of Experimental Clinical Cancer Research 2019 38354 Page 4 of 18 Statistical analysis All statistical analyses were performed using the SPSS 130 SPSS Inc USA statistical software and GraphPad Prism 70 GraphPad USA The statistical significance was analyzed by oneway ANOVA or Students ttest The data were pre sented as the mean standard deviation SD at least three independent experiments Categorical data were analyzed by Fishers exact test or chisquare test The correlations were analyzed using Pearsons correlation coefficients P 005 was considered statistically significant Results E2 upregulates CLDN6 expression and inhibits the migration and invasion of breast cancer cells Estrogen signaling pathways are classified as genomic and nongenomic Genomic pathways depend on transcriptional modulation of target genes by ERs and nongenomic path ways mediate rapid activation of signaling cascades partly via membranebound G protein coupled estrogen receptor 1 GPER1GPR30 2730 To evaluate a potential func tional link between E2 and CLDN6 in breast cancer cells we treated MCF7 ERαERβGPR30 and MDAMB231 ERαERβGPR30 cells with DMSO or E2 from 5 nM to 100 nM for 24 h The expression of CLDN6 was mea sured using semiquantitative RTPCR and western blot E2 treatment significantly upregulated the mRNA and protein expression of CLDN6 in a dosedependent manner in MCF 7 cells and 50 nM E2 showed the highest upregulation Fig 1a Similar results were observed in MDAMB231 cells Fig 1b However the E2induced effect on CLDN6 was not observed at the mRNA level in SKBR3 ERα ERβGPR30 cells Fig 1c and 50 nM E2 did not exert significant effects on CLDN6 protein levels in SKBR3 cells Fig 1f The results indicated that E2 did not regulate CLDN6 through GPR30 We treated MCF7 and MDA MB231 cells with a nonselective estrogen receptor antagon ist ICI 182780 ICI ICI cotreatment counteracted the E2 induced effects on CLDN6 mRNA and protein levels Fig 1 dg suggesting a direct involvement of the ERs in the regu lation of CLDN6 Since we have previously demonstrated that CLDN6 overexpression led to a lower migration and in vasion of breast cancer cells 20 22 we investigated whether E2induced CLDN6 expression was involved in reducing breast cancer cell migration and invasion In agreement with previous studies E2 treatment caused a reduction of migra tion and invasion in MCF7 cells as well as in MDAMB 231 cells Fig 1 h i E2 regulates CLDN6 expression through ERβ In this study we found that the expression of CLDN6 was enhanced in MCF7 and MDAMB231 cells by E2 More over when MCF7 and MDAMB231 cells were treated with 50 nM E2 the expression of ERβ in both cells was in creased but ERα expression was not changed in MCF7 cells Fig 2 a b Therefore this E2induced effect was not ERα dependent and we focused our attention on ERβ To dem onstrate that the increased expression of CLDN6 was medi ated by ERβ we knocked down ERβ in MDAMB231 cells Three different ERβ short hairpin RNAs shRNAs were tested and ERβ shRNA1 was used in the following experi ments Fig 2c The depletion of ERβ resulted in no increase in CLDN6 expression after E2 treatment in MDAMB231 cells Fig 2d Thus the induction effect of E2 on CLDN6 could be achieved through ERβ To observe the regulatory role of ERβ on CLDN6 MDAMB231 cells were treated with various concentrations of diarylpropionitrile DPN a selective ERβ agonist It was found that the effect of DPN on CLDN6 expression was similar to that of E2 and the most effective concentration of DPN was 100 nM Fig 2e The immunoreactivity of CLDN6 was prominent along the edges of the MDAMB231 cells upon DPN treatment as demon strated by immunofluorescence Fig 2f By TEM we ob served that the tight junctions between cells were prominent in DPN treated MDAMB231 cells Fig 2g Moreover the migration and invasion abilities of MDAMB231 cells treated with DPN were decreased which was consistent with the effect of E2 Fig 2h Next we asked whether the inhibi tory effect of ERβ on migration and invasion of breast cancer cells was associated with CLDN6 We measured the effects of CLDN6 knockdown on the migration and invasion of MDAMB231 cells treated with DPN The results showed that the DPNinduced reduction in migration and invasion could be rescued by the transfection of CLDN6 shRNA in MDAMB231 cells Fig 2i CLDN6 knockdown efficiency was detected via western blot Additional file 1 Figure S1A To further explore the regulation of ERβ on CLDN6 we used PHTPP a selective ERβ antagonist and ERβ shRNA The results showed that depletion of ERβ abolished the DPNinduced CLDN6 expression in MDAMB231 cells Fig 2 j k Conversely overexpression of ERβ induced CLDN6 upregulation in SRBR3 and MDAMB231 cells after treatment with DPN Fig 2l m We also observed that the depletion or upregulation of ERβ did not affect the ex pression of CLDN6 in the absence of DPN Fig 2jm ERβ transfection efficiency was detected via western blot Add itional file 1 Figure S1B These results suggest that ERβ reg ulates CLDN6 expression in a liganddependent manner ERβ regulates CLDN6 expression at the transcriptional level In the genomic pathway ERs act as transcription factors and regulate the transcription of target genes directly by binding to estrogen response elements EREs in the pro moter region of genes or indirectly by interacting with other transcription factors such as stimulating protein 1 Sp1 and activating protein1 AP1 27 29 Nuclear and cytoplasmic proteins were extracted and examined for ERβ expression Our results demonstrated that nuclear Song et al Journal of Experimental Clinical Cancer Research 2019 38354 Page 5 of 18 Fig 1 See legend on next page Song et al Journal of Experimental Clinical Cancer Research 2019 38354 Page 6 of 18 ERβ protein was increased whereas cytoplasmic ERβ pro tein was decreased in DPN treated MDAMB231 cells Fig 3a Immunofluorescence experiments showed that DPN induced the translocation of ERβ from the cytoplasm to the nucleus in MDAMB231 cells Fig 3b Next we ana lyzed the CLDN6 promoter sequence through the UCSC httpgenomeucscedu and JASPAR databases httpjas pardevgeneregnet In the present investigation we were not able to identify a conventional ERE 31 5GGTC AnnnTGACC3 in the CLDN6 promoter However we predicted potential ERβbinding sites halfERE and Sp1 binding sites GCbox through bioinformatics analysis in JASPAR To examine whether ERβ or Sp1 could bind to the CLDN6 promoter region in cells we performed ChIP assays The primers for the ChIP assays included the potential half ERE and Sp1 binding sites Fig 3c The results indicated that ERβ and Sp1 bound to the CLDN6 promoter in DPN treated cells Fig 3d e To further confirm that ERβ regu lated CLDN6 promoter activity we performed dual lucifer ase reporter assays The CLDN6 promoter fragment 2000 See figure on previous page Fig 1 E2 upregulates the expression of CLDN6 in MCF7 and MDAMB231 cells MCF7a MDAMB231b and SKBR3c cells were treated with DMSO or E2 from 5 nM to 100 nM Cells were harvested after 24 h treatment for analysis of gene and protein expression of CLDN6 using semiquantitative RTPCR and western blot respectively Actin served as a loading control Quantification of CLDN6 expression in MCF7 d and MDAMB231 e cells cotreated with ICI by qRTPCR CLDN6 protein expression levels in MCF7 f and MDAMB231 g cells after treatment with E2 or E2ICI CLDN6 protein expression in SKBR3 f cells after 50 nM E2 treatment Wound healing and Transwell assays were used to detect the migration scale bar 200 μm and invasion scale bar 50 μm abilities of MCF7 h and MDAMB231 i cells treated with E2 Data are presented as mean SD The data shown are representative results of three independent experiments P 005 P 001 P 0001 Fig 2 E2 regulates CLDN6 expression via ERβ a Western blot analysis of ERα and ERβ expression in MCF7 cells treated with E2 Actin served as a loading control b Western blot analysis of ERβ in MDAMB231 cells treated with E2 c MDAMB231 cells were transfected with either negative control shNC or three different ERβ shRNAs for 48 h and were then subjected to western blot analysis to detect the protein abundance of ERβ Actin was used as the loading control d ERβ knockdown abolished the CLDN6 expression induced by E2 e MDAMB231 cells were incubated with DPN for 24 h at the indicated concentration CLDN6 gene and protein expression levels were detected by using semiquantitative RTPCR and western blot f Immunofluorescence of CLDN6 red was prominent along the edges of the MDAMB231 cells upon DPN treatment Nuclei were stained with 4 6diamino2phenylindole DAPI blue scale bar 20 μm g Tight junctions white arrowheads between cells were prominent in MDAMB231 cells after DPN treatment as observed by TEM h The migration scale bar 200 μm and invasion scale bar 50 μm abilities of MDAMB231 cells treated with DPN were decreased i CLDN6 knockdown rescued the migration and invasion abilities of MDAMB231 cells after DPN treatment The ERβ antagonist PHTPP 10 μM 24 h j and ERβ knockdown k abolished the DPNinduced CLDN6 expression Overexpression of ERβ induced CLDN6 upregulation in SRBR3 l and MDAMB231 m cells after treatment with DPN Data are presented as mean SD The data shown are representative results of three independent experiments P 005 P 001 P 0001 Song et al Journal of Experimental Clinical Cancer Research 2019 38354 Page 7 of 18 250 bp was cloned into the pGL3basic vector and then cotransfected with renilla luciferase reporter plasmid into MDAMB231 cells Fig 3c Both ERβ and Sp1 are expressed in MDAMB231cells After transfection the cells were treated with 100 nM DPN for 24 h We found that luciferase activity was significantly in creased in DPNtreated pGL3CLDN6 cells compared with the pGL3CLDN6 cells Fig 3f These data pro vide evidence that ERβ regulates CLDN6 at the tran scriptional level ERβ induces autophagy and suppresses the migration and invasion of breast cancer cells Previously we observed the ultrastructure of MDAMB 231 cells treated with DPN by TEM Surprisingly we found that a large number of autophagic vacuoles ap peared in the cytoplasm of DPNtreated cells compared with control cells Fig 4a It is known that the amount of LC3BII represents the number of autophagosomes 32 33 In this study LC3BIIlabeled puncta formation was also observed through fluorescence microscopy in Fig 3 ERβ regulates CLDN6 expression at the transcriptional level a ERβ protein expression in the nuclear and cytoplasmic fractions of DPN treated MDAMB231 cells PCNA served as a nuclear protein control and βtubulin was used as a cytoplasmic protein control b Subcellular localization of ERβ DPN promotes the transportation of ERβ from the cytoplasm to the nucleus white arrowheads in MDAMB231 cells Nuclei were stained with DAPI blue scale bar 50 μm c Schematic diagram of the sequence spanning from 2000 bp to 250 bp relative to the translational start site TSS of the CLDN6 gene promoter containing the halfERENxGCbox motif d ChIP assay The ERβ binding sites 183 168 bp or 155 170 bp of the CLDN6 promoter were detected by PCR in DPNtreated MDAMB231 cells e ChIP assay The Sp1 binding sites 60 47 bp or 192 203 bp of the CLDN6 promoter were detected by PCR in DPNtreated MDAMB231 cells f The luciferase activities of DPNtreated MDAMB231 cells transfected with pGL3CLDN6 and renilla luciferase reporter pRLTK plasmids were detected by dualluciferase reporter assays Data are presented as mean SD The data shown are representative results of three independent experiments P 005 P 001 P 0001 Song et al Journal of Experimental Clinical Cancer Research 2019 38354 Page 8 of 18 DPNtreated cells but fewer puncta appeared in un treated cells Fig 4b We analyzed the protein expres sion level of LC3B by western blot and we found that the ratio of LC3III was significantly increased in MDAMB231 cells treated with DPN Fig 4c To fur ther prove that DPNactivated ERβ induces autophagy we detected LC3B expression in MDAMB231 cells with ERβ knockdown after DPN treatment The western blot results showed that with ERβ knocked down DPN did not induce an increase in the LC3III ratio in MDAMB231 cells Additional file 2 Figure S2A Immunofluorescence assays showed that LC3BIIlabeled puncta were rarely observed in DPNtreated ERβknockdown cells Additional file 2 Figure S2B These results suggest that ERβ might be involved in the induction of autophagy In fact LC3II can accumulate due to increased autophagosome for mation or a block in the autophagosomelysosome fusion process To distinguish between these two possibilities we assayed the LC3II level in the presence of chloroquine CQ which blocks autophagosomelysosome fusion and leads to the accumulation of autophagosomes After treat ment with DPN and CQ the ratio of LC3III was increased 25 times compared with that in MDAMB231 cells treated with DPN alone Fig 4d Moreover when MDAMB231 cells were treated with DPN and 3methyladenine 3MA an inhibitor of early phases of autophagy the DPNinduced Fig 4 ERβ induces autophagy and suppresses the migration and invasion of breast cancer cells a TEM analysis was performed on MDAMB231 cells treated or not treated for 24 h with 100 nM DPN DPNtreated cells displayed several autophagic vacuoles with the characteristic double membrane white arrowheads which were not observed in the control scale bar 5 μm b Immunofluorescence analysis showed that the numbers of LC3II puncta white arrowheads were increased after DPN treatment Nuclei were stained with DAPI Scale bar 20 μm c The expression of LC3B was detected by western blot in DPNtreated or untreated MDAMB231 cells Actin served as a loading control d The expression of LC3B was detected by western blot in MDAMB231 cells cotreated with DPN and CQ 25 μM 24 h or DPN and 3MA 5 mM 24 h CQ and 3MA were pretreated for 2 h before treatment with DPN Wound healing e and Transwell migration f assays showed that the DPN inhibited migration scale bar 200 μm and invasion scale bar 50 μm abilities were attenuated by CQ and 3MA CQ and 3MA were pretreated for 2 h before treatment with DPN Data are presented as mean SD The data shown are representative results of three independent experiments P 005 P 001 P 0001 Song et al Journal of Experimental Clinical Cancer Research 2019 38354 Page 9 of 18 increase was dramatically reduced Fig 4d These results suggest that ERβ is involved in the early steps of autophago some formation To understand the biological significance of ERβinduced autophagy we tested the migration and inva sion of breast cancer cells treated with autophagy inhibitors We found that both CQ and 3MA effectively hindered the effects of DPN on migration and invasion Fig 4e f ERβ induces autophagy via the CLDN6mediated increase in beclin1 To identify the mechanisms underlying the ERβ activation induced effects described above we used western blot to de tect biomarkers of the early stage of autophagy formation The results showed that beclin1 atg5 atg16 and LC3II were significantly increased in MDAMB231 and SKBR3ERβ ERβoverexpressing SKBR3 cells cells after DPN treat ment Fig 5a b The expression of CLDN6 is shown in Additional file 3 Figure S3A In our previous work we found that CLDN6 induced autophagy in MCF7 cells Thus we wanted to detect the relationship between CLDN6 and ERβ induced autophagy We knocked down CLDN6 in DPN treated cells and western blot analyses showed that beclin1 atg5 atg16 and LC3II expression was decreased Fig 5a b The expression of CLDN6 is shown in Additional file 3 Fig ure S3A These results suggested that ERβ induced autoph agy via the regulation of CLDN6 Interestingly the expression of beclin1 atg5 atg16 and LC3II proteins was also increased in MDAMB231CLDN6 SKBR3CLDN6 and MCF7CLDN6 cells CLDN6overexpressing MDA MB231 SKBR3 and MCF7 cells Fig 5c The expression of CLDN6 is shown in Additional file 3 Figure S3B Given that CLDN6 is strongly associated with autophagy in breast cancer cells we wanted to explore the role of CLDN6 in au tophagy The above results showed that CLDN6 overexpres sion or CLDN6 knockdown led to an increase or decrease in beclin1 expression respectively Fig 5a b The key autoph agy regulator beclin1 might be a downstream molecule of CLDN6 Hence we presumed that CLDN6 regulated au tophagy through beclin1 To verify this hypothesis we de pleted beclin1 in DPNtreated MDAMB231 and SKBR3 ERβ cells by using lentiviral vectors Beclin1 knockdown effi ciency was detected via western blot We found that the ex pression levels of atg5 atg16 and LC3II were significantly decreased in both cells Fig 5a b Similar regulatory roles of beclin1 on autophagy were confirmed in MDAMB231 CLDN6 SKBR3CLDN6 and MCF7CLDN6 cells Fig 5c Taken together these results showed that the regulation of CLDN6 on autophagy was beclin1dependent Next we ad dressed the question of how does CLDN6 regulate beclin1 It is noteworthy that UVradiation resistance associated gene UVRAG is a positive regulator of beclin1 34 UVRAG forms a complex with beclin1 and is involved in autophago some formation 34 Theoretically UVRAG can bind with proteins containing Src homology SH3 domains 35 Interestingly ZO1 contains a Cterminal SH3 domain 36 ZO1 is a scaffold protein that directly binds to PDZ motifs on the extreme Cterminus of CLDNs 37 Thus we specu lated that ZO1 and UVRAG might function as bridge mole cules for the CLDN6beclin1 interaction Western blot analyses showed that the expression levels of ZO1 and UVRAG were increased in both DPNtreated and CLDN6 overexpressing cells Fig 5d Because CLDN6 expression of MDAMB231 cells was low we used DPNtreated cells to perform coIP assays The results revealed that ZO1 and UVRAG indeed had binding affinities for CLDN6 and beclin1 in DPNtreated cells Fig 5e f These results indi cated that CLDN6 and ZO1UVRAGbeclin1 formed complexes with other autophagy proteins to regulate autop hagosome formation in breast cancer cells CLDN6 inhibits migration invasion and metastasis of breast cancer through beclin1 in vitro and in vivo When beclin1 in DPNtreated MDAMB231 DPNinduced CLDN6 expression cells and MDAMB231CLDN6 over expressed exogenous CLDN6 cells was silenced the migra tion and invasion abilities of both cells increased and the inhibitory effects of CLDN6 were abrogated Fig 6a b The results suggested that CLDN6 inhibited the migration and invasion of breast cancer cells in vitro through beclin1 To further investigate the antimetastatic effect of CLDN6 in vivo we established lung metastasis mouse models using MDAMB231NC cells control MDAMB231 cells MDAMB231CLDN6 cells CLDN6overexpressing MDA MB231 cells and MDAMB231CLDN6shbeclin1 cells CLDN6overexpressing and beclin1knockdown MDAMB 231 cells Lung metastasis was determined by a comprehen sive analysis of macroscopic observations and hematoxylin and eosin HE staining of lung tissue sections The results showed that there was a higher occurrence of lung metasta sis in the MDAMB231NC and MDAMB231CLDN6 shbeclin1 groups and a lower occurrence of lung metastasis in MDAMB231CLDN6 group Fig 6ce Figure 6c shows the representative HE staining of lung tissues from mice in the three groups Surprisingly we also observed liver metas tasis in the three groups There were 4 cases of liver metasta sis in both the MDAMB231NC and MDAMB231 CLDN6shbeclin1 groups However 1 case was found in the MDAMB231CLDN6 group Fig 6fh Our results showed that CLDN6 inhibited breast cancer metastasis by beclin1 in vivo The clinical correlation analyses of ERβ CLDN6 and beclin1 expression and prognosis in breast cancer patients We next evaluated the expression of ERβ CLDN6 and beclin1 in tumor samples from 70 breast cancer patients by immunohistochemical IHC staining of the tissue microarray The expression of ERβ CLDN6 and beclin1 Song et al Journal of Experimental Clinical Cancer Research 2019 38354 Page 10 of 18 in each IHC sample was classified as either high score 8 or low score 8 using the median of the IHC scores as the cutoff value The relationship between ERβ CLDN6 and beclin1 expression and clinicopathologic features was further evaluated and the results are summarized in Table 1 We found that high expression of beclin1 was significantly associated with smaller tumors P 0026 and that low ex pression of ERβ was significantly associated with early TNM stage TNM P 0023 No association existed between the expression of ERβ CLDN6 and beclin1 and patient age Fig 5 ERβ induces autophagy via CLDN6mediated increase in beclin1 Expression of autophagyrelated proteins was detected by using western blot in DPNtreated cells and in CLDN6knockdown and beclin1knockdown DPNtreated MDAMB231 a and SKBR3ERβ b cells Actin served as a loading control c Expression of autophagyrelated proteins was detected by using western blot in CLDN6overexpressing MDAMB231 SK BR3 and MCF7 cells and CLDN6overexpressing beclin1knockdown cells d Western blot analysis of ZO1 and UVRAG in DPNtreated and CLDN6overexpressing MDAMB231 cells ef CoIP assays were performed to detect the interaction between beclin1 and CLDN6 in MDAMB231 cells treated with DPN Data are presented as mean SD The data shown are representative results of three independent experiments P 005 Song et al Journal of Experimental Clinical Cancer Research 2019 38354 Page 11 of 18 lymph node metastasis or pathological grade Figure 7a shows the representative IHC staining of ERβ CLDN6 and beclin1 in breast tissues Pearson correlation analyses indicated that ERβ and CLDN6 were positively correlated and that the expression of CLDN6 was positively corre lated with beclin1 in breast cancer tissues Fig 7bc There was no correlation between the expression of ERβ and beclin1 P 0243 In addition we further assessed the correlation between ERβ CLDN6 and beclin1 expres sion and the overall survival OS and disease free survival DFS of breast cancer patients in the KaplanMeier plot ter database httpwwwkmplotcom by KaplanMeier analyses The results showed that breast cancer patients with high beclin1 expression had significantly longer OS Fig 6 CLDN6 inhibits migration invasion and metastasis of breast cancer through beclin1 in vitro and in vivo Wound healing and Transwell migration assays showed that beclin1 knockdown drove the migration scale bar 200 μm and invasion scale bar 50 μm of DPNtreated a and CLDN6overexpressing MDAMB231 cells b c IVIS imaging was performed to show lung metastatic sites Representative HE staining of lung sections from the three groups n 5 d The number of mice with lung metastasis was counted in each group e The number of lung nodules in each group f Representative images of liver metastasis and HE staining in the three groups n 5 Black arrowheads indicate the metastatic sites g The number of mice with liver metastasis was counted in each group h The number of liver nodules in each group Data are presented as mean SD The data shown are representative results of three independent experiments P 005 P 001 Song et al Journal of Experimental Clinical Cancer Research 2019 38354 Page 12 of 18 than patients with low beclin1 expression and patients with high ERβ CLDN6 and beclin1 expression levels showed longer DFS than the respective low expression groups Fig 7d Taken together these findings suggest that high expression of ERβ CLDN6 or beclin1 indicates a better prognosis for breast cancer patients Discussion Our previous studies have shown that E2 upregulated CLDN6 expression and hindered MCF7 cell migration and invasion but the molecular mechanisms were still un clear In this study we found that the regulation of CLDN6 by E2 and its biological effects were mediated by ERβ and were not ERαdependent Furthermore the in hibitory effects of ERβ on the migration and invasion of breast cancer cells were mediated through CLDN6in duced autophagy To our knowledge this is the first study to link the inhibitory role of ERβ on migration and inva sion and autophagy modulation in breast cancer cells Estrogen has pivotal roles in the development and progres sion of breast cancer The biological effects of estrogen are mediated by intracellular ERs ERα and ERβ and the cell membrane receptor GPR30 GPR30 is a member of the G protein coupled receptor GPCR superfamily Estrogen me diates rapid nongenomic actions and cell biological functions partly through GPR30 38 39 Our study found that E2 in duced the expression of CLDN6 in MCF7 ERαERβ GPR30 and MDAMB231 ERαERβGPR30 cells but not in SKBR3 ERαERβGPR30 cells Moreover ICI a nonselective estrogen receptor antagonist cotreatment counteracted the E2induced effects These results indicated that the regulation of E2 on CLDN6 was not through GPR30 ERα is a key target of endocrine therapy and induces protooncogene expression to stimulate cell proliferation in breast cancer 40 Recently a few studies have reported that ERα suppresses the migration and invasion of breast cancer cells by upregulating cytoskeleton protein expression 41 43 Our results showed that the expression of CLDN6 was enhanced and migration and invasion were hindered in both MCF7 and MDAMB231 cells treated with E2 Therefore this E2induced effect was not ERαdependent and we wanted to explore the role of ERβ in this process The role of ERβ in cancer cells is still poorly understood Growing evi dence supports that ERβ is a tumor suppressor ERβ has been shown to increase integrin α1β1 levels and inhibit the migration of breast cancer cells 44 In addition ERβ de creased basallike breast cancer cell invasion by promoting degradation of epidermal growth factor receptor EGFR 45 Our study found that the expression of ERβ was in creased in MCF7 and MDAMB231 cells treated with E2 E2 did not induce CLDN6 expression in MDAMB231 cells with ERβ knockdown DPN is a selective ERβ agonist DPN mediated ERβ activation could also upregulate CLDN6 ex pression and reduce migration and invasion of MDAMB 231 cells When ERβ was knocked down in breast cancer cells the expression of CLDN6 was not increased after DPN treatment The effect of DPN on CLDN6 was similar to that of E2 in breast cancer cells These results indicated that ERβ played a pivotal role in the regulation of CLDN6 by E2 and its effect on biological behavior Although many studies have supported the idea that ERβ expression inhibits the migra tion and invasion of breast cancer cells Zoi et al found that ERβ knockdown decreased the expression of matrix metallo proteinases MMPs and promoted mesenchymalepithelial Table 1 The association between ERβ CLDN6 beclin1 expression and clinical characteristics of breast cancer patients Characteristic ERβ p CLDN6 p Beclin1 p High Low High Low High Low Age 0932 0923 0697 57 years 11 22 13 20 8 11 57 years 11 23 13 21 12 13 Pathology grade 0785 0717 0960 III 10 19 11 19 7 11 IIIII 11 18 9 19 8 13 Tumour size 0451 0716 0026 3 cm 13 29 16 25 15 10 3 cm 10 15 9 17 5 14 Lymph node 0314 0088 0899 Negative 8 24 12 20 11 14 Positive 9 15 4 20 6 7 TNM stage 0023 0356 0947 0II 13 37 20 30 14 20 IIIIV 10 8 5 13 4 6 Song et al Journal of Experimental Clinical Cancer Research 2019 38354 Page 13 of 18 transition MET to suppress MDAMB231 cell migration and invasion 7 Hence the precise role of ERβ and its regu latory mechanism in breast cancer still need to be elucidated Our previous study demonstrated that CLDN6 is a breast cancer suppressor gene CLDN6 overexpression suppressed the migration and invasion of breast cancer cells by revers ing epithelialmesenchymal transition EMT 20 22 46 Then we examined whether the ERβinduced effect on bio logical behavior was related to CLDN6 Here our results showed that CLDN6 knockdown rescued the effects of ERβ on the migration and invasion of MDAMB231 cells Fur thermore we investigated the regulatory effect of ERβ on CLDN6 ERβ is a member of the nuclear receptor super family and functions as a hormonedependent transcription factor 1 The classical model of ERβ activation is that after ligand binding to the receptor ligandER complexes dir ectly bind to EREs in the promoter regions of target genes or indirectly interact with other transcription factors Sp1 Fig 7 Clinical correlation analyses between ERβ CLDN6 and beclin1 expression and prognosis in breast cancer patients a Representative images of IHC analysis of ERβ CLDN6 and beclin1 expression from tissue microarray 400 In breast cancer tissues and adjacent tissues ERβ expression was observed in the nucleus and the cytoplasm CLDN6 was expressed in the cytoplasm and the plasma membrane and beclin1 was mainly expressed in the cytoplasm Low indicates proteins with low expression and high indicates proteins with high expression b Correlation between ERβ expression and CLDN6 expression in breast cancer tissue microarray Pearson correlation test n 67 r 04652 P 00001 c Correlation between CLDN6 expression and beclin1 expression in breast cancer tissue microarray Pearson correlation test n 44 r 03677 P 00141 d KaplanMeier analysis of the overall survival and diseasefree survival of breast cancer patients with different ERβ CLDN6 and beclin1 expression levels in the in KaplanMeier plotter database Statistical differences were determined by logrank test Data are presented as mean SD Song et al Journal of Experimental Clinical Cancer Research 2019 38354 Page 14 of 18 or AP1 to activate the transcription of target genes 28 29 When the ligand is absent ERβ can still be activated by growth factor receptors insulinlike growth factor 1 recep tor IGF1R and EGFR which can stimulate protein kinase cascades that phosphorylate and activate the transcriptional activity of ERβ 47 48 In this study we found that the depletion or upregulation of ERβ did not affect the ex pression of CLDN6 in the absence of DPN However CLDN6 expression in breast cancer cells overexpressing ERβ was increased after DPN treatment Therefore we believe that the regulation of CLDN6 by ERβ is a lig anddependent pathway in breast cancer cells Burek et al reported that E2induced CLDN5 upregulation was mediated by binding to the EREs and Sp1 sites of the CLDN5 promoter in brain endothelial cells 49 Fur thermore we analyzed the promoter region of CLDN6 and identified imperfect EREs and potential Sp1 tran scription factor binding sites ChIP assays showed that the regulation of the CLDN6 promoter could take place either directly by interaction with ERβ or indirectly by interaction with Sp1 Dual luciferase reporter assays showed that ERβ regulated CLDN6 promoter activity The exact molecular mechanism and promoter ele ments responsible for ERβ regulation require further investigation We report for the first time to our know ledge that the inhibitory effects of ERβ on migration and invasion are mediated by CLDN6 and that CLDN6 is a target gene of ERβ in breast cancer Autophagy is a basic catabolic process in which un necessary or dysfunctional intracellular materials are de graded and recycled 50 Autophagy is reported to mediate an oncosuppressive role in the tumor initiation step 51 Surprisingly our data indicated that ERβ in duced autophagy in breast cancer cells While previous studies have reported that ERβ triggers autophagy and inhibits cancer cell proliferation 12 52 few studies have investigated the role of ERβinduced autophagy on the migration and invasion of breast cancer cells In our study autophagy inhibitors reversed the inhibitory effect of ERβ on the migration and invasion of breast cancer cells These results indicated that the inhibitory role of ERβ on the migration and invasion of breast cancer cells was associated with ERβinduced autophagy Further more we found that ERβ induced autophagy through CLDN6 CLDN6 is one of the key proteins in the forma tion of tight junctions TJs and plays important roles in maintaining the epithelial barrier polarity and signal de livery There are few studies on the relationship between CLDNs and autophagy Recently it has been reported that the autophagosome marker Atg16L colocalizes with CLDN5 in endocytosed vesicles transported across the cells indicating that the process of tight junction Fig 8 The proposed model for ERβinduced autophagy inhibiting breast cancer cell migration and invasion In this model when ERβ bound with ligands DPN DPNERβ complexes directly bound to the ERE of the CLDN6 promoter and enhanced CLDN6 expression In addition activated ERβ can interact with Sp1 and bind the Sp1 transcriptional regulation domains of the CLDN6 promoter to induce CLDN6 expression ZO1 and UVRAG act as bridge molecules for the CLDN6beclin1 interaction CLDN6 and ZO1UVRAGbeclin1 form complexes and serve as a platform for recruiting other autophagy regulatory proteins atg5 atg16 and LC3II and induce autophagy to suppress the migration and invasion of breast cancer cells Song et al Journal of Experimental Clinical Cancer Research 2019 38354 Page 15 of 18 remodeling involves the regulation of autophagy 53 Zhao Z et al discovered that CLDN1 regulated drug re sistance by promoting autophagy which was mediated by ULK1 phosphorylation in nonsmall cell lung cancer 54 In addition J Kim et al reported that CLDN1 functions as an autophagy stimulator to increase autoph agy flux and accelerate the degradation of SQSTM1p62 55 In our previous investigation we observed by TEM that CLDN6 overexpression resulted in the appearance of numerous autophagic vacuoles Western blot analysis and IF and acridine orange AO staining methods dem onstrated that CLDN6 induced autophagy in breast can cer cells Nevertheless none of the above studies has clarified the regulatory mechanism of CLDNs on autophagy Next we wanted to unveil the mechanism by which the tight junction protein CLDN6 regulates autophagy Vari ous proteins participate in autophagy modulation Beclin1 the first identified mammalian autophagy gene is a hap loinsufficient tumor suppressor and plays an indispensable role in the initiation of autophagy 5658 Strikingly we found that the expression of beclin1 was consistent with that of CLDN6 in breast cancer cells We also found that the expression of CLDN6 was positively correlated with beclin1 expression in breast cancer tissues Moreover si lencing beclin1 reduced autophagy and reversed the in hibitory effect of ERβCLDN6 on the migration and invasion of breast cancer cells and attenuated the inhibi tory effect of CLDN6 on metastasis in vivo Our results are in line with previous evidence that autophagy inhib ition upon beclin1 knockdown stimulates the migration and invasion of GL15 glioma cells 59 Deletion of beclin1 has been shown to promote the invasion and metastasis of breast cancer cells by increasing the phosphorylation of AKT and ERK 60 Furthermore in this study we found that the expression of ZO1 and UVRAG was increased in CLDN6overexpressing cells CoIP assays revealed that ZO1 and UVRAG had binding affinities for CLDN6 and beclin1 Our results demonstrated that the scaffold protein ZO1 and the autophagy regulatory protein UVRAG func tioned as bridge molecules for the CLDN6beclin1 inter action This is the first study to report that CLDN6 and ZO1UVRAGbeclin1 form complexes with other au tophagy proteins to regulate autophagosome formation in breast cancer Conclusion Our study reported new findings that CLDN6 is a target gene of ERβ Mechanistically we demonstrated that the inhibitory effects of ERβ on the migration and invasion of breast cancer cells were mediated by CLDN6 which induced the beclin1dependent autophagic cascade Fig 8 These findings provide fresh insight into the mechanism underlying the inhibitory effects of ERβ on breast cancer Moreover high ERβ CLDN6 or beclin1 expression predicted a favorable prognosis in breast can cer patients ERβ agonists and CLDN6 may be novel therapeutic approaches for the treatment of breast can cer More in vivo experiments are needed to validate these findings Additional files Additional file 1 Figure S1 Western blot analysis of CLDN6 and ERβ expression after transfection with the indicated plasmids A CLDN6 knockdown efficiency was detected in DPNtreated MDAMB231 cells by western blot B Western blot analysis of ERβ expression after transfection with ERβ cDNA in SKBR3 and MDAMB231 cells Data are presented as mean SD The data shown are representative results of three independent experiments P 005 P 001 TIF 305 kb Additional file 2 Figure S2 DPN regulates CLDN6 expression through ERβ A LC3B protein expression was detected by western blot in DPN treated MDAMB231 cells infected with ERβshRNA B The numbers of LC3II puncta white arrowheads were decreased after knocking down ERβ in DPNtreated MDAMB231 cells as observed in immunofluorescence analysis Nuclei were stained with DAPI Scale bar 20 μm Data are presented as mean SD The data shown are representative results of three independent experiments P 001 P 0001 TIF 2134 kb Additional file 3 Figure S3 Western blot analysis of CLDN6 expression after transfection with the indicated plasmids A Western blot analysis of CLDN6 expression in DPNtreated MDAMB231 and SKBR3ERβ cells CLDN6 knockdown efficiency was detected by western blot in DPN treated MDAMB231 and SKBR3ERβ cells B CLDN6 overexpression efficiency was detected in MDAMB231 SKBR3 and MCF7 cells by western blot Data are presented as mean SD The data shown are representative results of three independent experimentsP 005 P 001 TIF 476 kb Abbreviations 3MA 3Methyladenine AKTPKB Protein Kinase B AP1 Activating protein1 ChIP Chromatin immunoprecipitation CQ Chloroquine DAB 3 3 diaminobenzidine DAPI 4 6diamino2phenylindole DFS Disease Free Survival DPN Diarylpropionitrile E2 17βestradiol ERE estrogen response element ERK Extracellular signalregulated kinase GPER1GPR30 G protein coupled estrogen receptor 1 HE Hematoxylin and Eosin ICI ICI 182780 IHC Immunohistochemistry OS Overall Survival PBS Phosphate Buffer Saline shRNAs Short hairpin RNAs Sp1 Stimulating protein 1 TEM Transmission Electron Microscopy TJs Tight Junctions UVRAG UV radiation Resistance Associated Gene ZO1 Zonula Occludens1 Acknowledgements Not applicable Authors contributions CQ and PS conceived and designed the study PS drafted the manuscript CQ guaranteed the integrity of entire study and revised the manuscript YD YL HQ and DQ performed most of experiments YG XW and WX helped to collect the primary data YL and XJ performed data analysis and interpretation YL and YJ proofread the manuscript All authors read and approved the final manuscript Funding This work was supported by the National Natural Science Foundation of China Grant No 8177101731 the Science and Technology Development Projects of Jilin Province Grant No 20180414045GH and 20170623093TC and the Fundamental Research Funds for the Central Universities JLU Availability of data and materials The datasets used andor analyzed during the current study are available from the corresponding author on reasonable request Song et al Journal of Experimental Clinical Cancer Research 2019 38354 Page 16 of 18 Ethics approval and consent to participate All the tissue samples have been obtained with prior written consent from patients All animal studies were carried out with approval from the Animal Research Ethics Committee of the Jilin University Consent for publication Not applicable Competing interests The authors declare that they have no competing interests Received 18 May 2019 Accepted 6 August 2019 References 1 Heldring N Pike A Andersson S Matthews J Cheng G Hartman J et al Estrogen receptors how do they signal and what are their targets Physiol Rev 200787390531 2 Thomas C Gustafsson JA The different roles of ER subtypes in cancer biology and therapy Nat Rev Cancer 2011118597608 3 Huang B Omoto Y Iwase H Yamashita H Toyama T Coombes RC et al Differential expression of estrogen receptor alpha beta1 and beta2 in lobular and ductal breast cancer Proc Natl Acad Sci U S A 2014111519338 4 Clarke R Tyson JJ Dixon JM Endocrine resistance in breast canceran overview and update Mol Cell Endocrinol 2015418Pt 322034 5 Droog M Beelen K Linn S Zwart W Tamoxifen resistance from bench to bedside Eur J Pharmacol 2013717134757 6 Leygue E Murphy LC A bifaceted role of estrogen receptor beta in breast cancer Endocr Relat Cancer 2013203R12739 7 Piperigkou Z Bouris P Onisto M Franchi M Kletsas D Theocharis AD et al Estrogen receptor beta modulates breast cancer cells functional properties signaling and expression of matrix molecules Matrix biology journal of the International Society for Matrix Biology 201656423 8 Honma N Horii R Iwase T Saji S Younes M Takubo K et al Clinical importance of estrogen receptorbeta evaluation in breast cancer patients treated with adjuvant tamoxifen therapy J Clin Oncol 20082622372734 9 Hartman J Lindberg K Morani A Inzunza J Strom A Gustafsson JA Estrogen receptor beta inhibits angiogenesis and growth of T47D breast cancer xenografts Cancer Res 200666231120713 10 Song W Tang L Xu Y Sun Q Yang F Guan X ERbeta1 inhibits metastasis of androgen receptorpositive triplenegative breast cancer by suppressing ZEB1 Journal of experimental clinical cancer research CR 201736175 11 Ruddy SC Lau R Cabrita MA McGregor C McKay BC Murphy LC et al Preferential estrogen receptor beta ligands reduce Bcl2 expression in hormoneresistant breast cancer cells to increase autophagy Mol Cancer Ther 2014137188293 12 Pierdominici M Maselli A Locatelli SL Ciarlo L Careddu G Patrizio M et al Estrogen receptor beta ligation inhibits Hodgkin lymphoma growth by inducing autophagy Oncotarget 201785852235 13 Jmm L Towers CG Thorburn A Targeting autophagy in cancer Nat Rev Cancer 2017179528 14 Osanai M Takasawa A Murata M Sawada N Claudins in cancer bench to bedside Pflugers Arch Eur J Physiol 201746915567 15 Turksen K Troy TC Claudin6 a novel tight junction molecule is developmentally regulated in mouse embryonic epithelium Dev Dyn 2001 2222292300 16 Tsukita S Tanaka H Tamura A The Claudins from tight junctions to biological systems Trends Biochem Sci 201944214152 17 Fanning AS Anderson JM Zonula occludens1 and 2 are cytosolic scaffolds that regulate the assembly of cellular junctions Ann N Y Acad Sci 2009116511320 18 Quan CS Lu SJ Identification of genes preferentially expressed in mammary epithelial cells of Copenhagen rat using subtractive hybridization and microarrays Carcinogenesis 2003241015939 19 Xu X Jin H Liu Y Liu L Wu Q Guo Y et al The expression patterns and correlations of claudin6 methyCpG binding protein 2 DNA methyltransferase 1 histone deacetylase 1 acetylhistone H3 and acetyl histone H4 and their clinicopathological significance in breast invasive ductal carcinomas Diagn Pathol 20127133 20 Wu Q Liu YF Ren Y Xu XM Yu LN Li YL et al Tight junction protein claudin6 downregulates the malignant phenotype of breast carcinoma Eur J Cancer Prev 201019318694 21 Guo YX Lin DJ Zhang MZ Zhang XW Li YR Yang R et al CLDN6induced apoptosis via regulating ASK1p38JNK signaling in breast cancer MCF7 cells Int J Oncol 2016486243544 22 Yang ML Li YR Ruan Y Lu Y Lin DJ Xie YP et al CLDN6 enhances chemoresistance to ADM via AF6ERKs pathway in TNBC cell line MDAMB231 Mol Cell Biochem 20184431216980 23 Liu Y Jin X Li Y Ruan Y Lu Y Yang M et al DNA methylation of claudin6 promotes breast cancer cell migration and invasion by recruiting MeCP2 and deacetylating H3Ac and H4Ac Journal of experimental clinical cancer research CR 2016351120 24 Yang M Li Y Shen X Ruan Y Lu Y Jin X et al CLDN6 promotes chemoresistance through GSTP1 in human breast cancer Journal of experimental clinical cancer research CR 2017361157 25 Shaw LM Tumor Cell Invasion Assays 2005 26 Yang ML Li YR Shen XF Ruan Y Lu Y Jin XS et al CLDN6 promotes chemoresistance through GSTP1 in human breast cancer J Exp Clin Cancer Res 201736157 27 Bjornstrom L Sjoberg M Mechanisms of estrogen receptor signaling convergence of genomic and nongenomic actions on target genes Mol Endocrinol 200519483342 28 Paech K Webb P Kuiper GG Nilsson S Gustafsson J Kushner PJ et al Differential ligand activation of estrogen receptors ERalpha and ERbeta at AP1 sites Science 19972775331150810 29 Porter W Saville B Hoivik D Safe S Functional synergy between the transcription factor Sp1 and the estrogen receptor Mol Endocrinol 19971111156980 30 Revankar CM Cimino DF Sklar LA Arterburn JB Prossnitz ER A transmembrane intracellular estrogen receptor mediates rapid cell signaling Science 20053075715162530 31 KleinHitpass L Ryffel GU Heitlinger E Cato AC A 13 bp palindrome is a functional estrogen responsive element and interacts specifically with estrogen receptor Nucleic Acids Res 198816264763 32 Kabeya Y Mizushima N Uero T Yamamoto A Kirisako T Noda T et al LC3 a mammalian homologue of yeast Apg8p is localized in autophagosome membranes after processing EMBO J 2000192157208 33 Mizushima N Methods for monitoring autophagy Int J Biochem Cell Biol 200436122491502 34 Chengyu L Pinghui F Bonsu K Iris D Dan C ByungHa O et al Autophagic and tumour suppressor activity of a novel Beclin1binding protein UVRAG Nat Cell Biol 20068768899 35 Yoshinori T Domenico C Norimasa M Cualing HD Mei S Yuya S et al Bif1 interacts with Beclin 1 through UVRAG and regulates autophagy and tumorigenesis Nat Cell Biol 2007910114251 36 Balda MS Anderson JM Matter K The SH3 domain of the tight junction protein ZO1 binds to a serine protein kinase that phosphorylates a region Cterminal to this domain FEBS Lett 19963993326 37 Itoh M Furuse M Morita K Kubota K Saitou M Tsukita S Direct binding of three tight junctionassociated MAGUKs ZO1 ZO2 and ZO3 with the COOH termini of claudins J Cell Biol 19991476135163 38 Prossnitz ER Arterburn JB Smith HO Oprea TI Sklar LA Hathaway HJ Estrogen signaling through the transmembrane G proteincoupled receptor GPR30 Annu Rev Physiol 2008701165 39 Liang S Chen Z Jiang G Zhou Y Liu Q Su Q et al Activation of GPER suppresses migration and angiogenesis of triple negative breast cancer via inhibition of NFκBIL6 signals Cancer Lett 201738612 40 Sebastián G Vaqué JP Rocío S Victoria W Vanzulli SI Rubén M et al Estrogen receptor alpha mediates progestininduced mammary tumor growth by interacting with progesterone receptors at the cyclin D1MYC promoters Cancer Res 20127292416 41 Sisci D Middea E Morelli C Lanzino M Aquila S Rizza P et al 17βestradiol enhances α5 integrin subunit gene expression through ERαSp1 interaction and reduces cell motility and invasion of ERαpositive breast cancer cells Breast Cancer Res Treat 201012416377 42 Gao Y Wang Z Hao Q Li W Xu Y Zhang J et al Loss of ERα induces amoeboidlike migration of breast cancer cells by downregulating vinculin Nat Commun 2017814483 43 Padillarodriguez M Parker SS Adams DG Westerling T Puleo JI Watson AW et al The actin cytoskeletal architecture of estrogen receptor positive breast cancer cells suppresses invasion Nat Commun 2018912980 44 Lindberg K Strom A Lock JG Gustafsson JA Haldosen LA Helguero LA Expression of estrogen receptor beta increases integrin alpha 1 and integrin beta 1 levels and enhances adhesion of breast Cancer cells J Cell Physiol 2010222115667 Song et al Journal of Experimental Clinical Cancer Research 2019 38354 Page 17 of 18 45 Thomas C Rajapaksa G Nikolos F Hao RX Katchy A McCollum CW et al ER beta 1 represses basallike breast cancer epithelial to mesenchymal transition by destabilizing EGFR Breast Cancer Res 2012146R148 46 Lu Y Wang LP Li HR Li YR Ruan Y Lin DJ et al SMAD2 Inactivation Inhibits CLDN6 Methylation to Suppress Migration and Invasion of Breast Cancer Cells Int J Mol Sci 2017189 47 Tremblay A Tremblay GB Labrie F Giguere V Ligandindependent recruitment of SRC1 to estrogen receptor beta through phosphorylation of activation function AF1 Mol Cell 1999345139 48 Levin ER Hammes SR Nuclear receptors outside the nucleus extranuclear signalling by steroid receptors Nat Rev Mol Cell Biol 2016171278397 49 Burek M AriasLoza PA Roewer N Forster CY Claudin5 as a novel estrogen target in vascular endothelium Arterioscl Throm Vas 2010302298U317 50 Galluzzi L Pietrocola F BravoSan Pedro JM Amaravadi RK Baehrecke EH Cecconi F et al Autophagy in malignant transformation and cancer progression EMBO J 201534785680 51 Morselli E Galluzzi L Kepp O Marino G Michaud M Vitale I et al Oncosuppressive functions of autophagy Antioxid Redox Signal 20111411225169 52 Carmela G Salvatore P Marta S Paola A Maria Luisa P Ida P et al Estrogen receptor beta ERβ produces autophagy and necroptosis in human seminoma cell line through the binding of the Sp1 on the phosphatase and tensin homolog deleted from chromosome 10 PTEN promoter gene Cell Cycle 20121115291121 53 Gehne N Lamik A Lehmann M Haseloff RF Andjelkovic AV Blasig IE Cross over endocytosis of claudins is mediated by interactions via their extracellular loops PLoS One 2017128e0182106 54 Zhao Z Li J Jiang Y Xu W Li X Jing W CLDN1 increases drug resistance of nonsmall cell lung Cancer by activating autophagy via upregulation of ULK1 phosphorylation Medical science monitor international medical journal of experimental and clinical research 201723290616 55 Kim J Choi S Kim JO Kim KK Autophagymediated upregulation of cytoplasmic claudin 1 stimulates the degradation of SQSTM1p62 under starvation Biochem Biophys Res Commun 2018496115966 56 Liang XH Jackson S Seaman M Brown K Kempkes B Hibshoosh H et al Induction of autophagy and inhibition of tumorigenesis by beclin 1 Nature 19994026762672 57 Kang R Zeh HJ Lotze MT Tang D The Beclin 1 network regulates autophagy and apoptosis Cell Death Differ 201118457180 58 Thongchot S Vidoni C Ferraresi A Loilome W Yongvanit P Namwat N et al Dihydroartemisinin induces apoptosis and autophagydependent cell death in cholangiocarcinoma through a DAPK1BECLIN1 pathway Mol Carcinog 20185712173550 59 Catalano M DAlessandro G Lepore F Corazzari M Caldarola S Valacca C et al Autophagy induction impairs migration and invasion by reversing EMT in glioblastoma cells Mol Oncol 201598161225 60 Rohatgi RA Janusis J Leonard D Bellvé KD Fogarty KE Baehrecke EH et al Beclin 1 regulates growth factor receptor signaling in breast cancer Oncogene 20153442535262 Publishers Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations Song et al Journal of Experimental Clinical Cancer Research 2019 38354 Page 18 of 18