ORIGINAL RESEARCH published: 06 December 2019 doi: 10.3389/fonc.2019.01306 FXYD5/Dysadherin, a Biomarker of Endometrial Cancer Myometrial Invasion and Aggressiveness: Its Relationship With TGF-β1 and NF-κB Pathways María José Besso 1 , Marina Rosso 1 , Lara Lapyckyj 1 , Cristian Pablo Moiola 1,2 , María Laura Matos 1 , María Florencia Mercogliano 3 , Roxana Schillaci 3 , Jaume Reventos 2† , Edited by: Eva Colas 2 , Antonio Gil-Moreno 2,4 , Alejandra Wernicke 5 , Roberto Orti 5 and Luisa Lanfrancone, European Institute of Oncology Mónica Hebe Vazquez-Levin 1* (IEO), Italy 1 Laboratorio de Estudios de la Interacción Celular en Reproducción y Cáncer, Instituto de Biología y Medicina Experimental Reviewed by: (IBYME; CONICET-FIBYME), Buenos Aires, Argentina, 2 Biomedical Research Group in Gynecology, Vall d’Hebron Research Gabriele Multhoff, Institute (VHIR), Universitat Autónoma de Barcelona, CIBERONC, Barcelona, Spain, 3 Laboratorio de Mecanismos Technical University of Moleculares de Carcinogénesis, Instituto de Biología y Medicina Experimental (IBYME; CONICET-FIBYME), Buenos Aires, Munich, Germany Argentina, 4 Gynecological Oncology Department, Vall Hebron University Hospital, CIBERONC, Barcelona, Spain, 5 Hospital Robert Friis, Italiano de Buenos Aires, Buenos Aires, Argentina University of Bern, Switzerland *Correspondence: Objective: Endometrial cancer (EC) is the second most common gynecological cancer Mónica Hebe Vazquez-Levin mhvazquez@ibyme.conicet.gov.ar; worldwide. Myometrial invasion (MI) is a key event in EC dissemination. This study aimed mhvazl@gmail.com to evaluate FXYD5/dysadherin (FXYD5/Dys) expression in EC tissue and uterine aspirate † Present address: (UA) biopsies and to assess molecular/functional changes associated with its expression Jaume Reventos, in cellular models. Departament de Ciències Bàsiques, Universitat Internacional de Catalunya, Methods: FXYD5/Dys messenger RNA (mRNA) levels were determined in EC tissue and Barcelona, Spain UA biopsies. FXYD5/Dys expression was evaluated in EC RNAseq data from The Cancer Specialty section: Genome Atlas (TCGA) and GENEVESTIGATOR tools. FXYD5/Dys impact on E-cadherin This article was submitted to expression and cell behavior was assessed in EC Hec1a cells treated with transforming Molecular and Cellular Oncology, growth factor (TGF)-β1, stably transfected with ETV5, and transiently transfected with a section of the journal Frontiers in Oncology FXYD5/Dys small interfering RNA (siRNA) or pcDNA3-FXYD5/Dys plasmid. Received: 12 August 2019 Results: FXYD5/Dys was associated with EC aggressiveness, finding high mRNA Accepted: 11 November 2019 levels in tumors depicting MI > 50%, Grade 3, and intermediate/high risk of recurrence. Published: 06 December 2019 FXYD5/Dys was highly expressed at the tumor invasive front compared to the superficial Citation: Besso MJ, Rosso M, Lapyckyj L, area. Most results were recapitulated in UA biopsies. FXYD5/Dys modulation in Hec1a Moiola CP, Matos ML, cells altered cell migration/adhesion and E-cadherin expression. TGF-β1 treatment of Mercogliano MF, Schillaci R, Reventos J, Colas E, Gil-Moreno A, Hec1a cells induced FXYD5/Dys expression. TCGA-UCEC RNAseq analysis revealed a Wernicke A, Orti R and positive correlation between FXYD5/Dys, TGF-β1, and plasminogen activator inhibitor Vazquez-Levin MH (2019) (PAI)-1 mRNA levels. FXYD5/Dys induced nuclear factor (NF)-κB pathway activation FXYD5/Dysadherin, a Biomarker of Endometrial Cancer Myometrial in Hec1a cells. FXYD5/Dys mRNA levels positively correlated with transcriptional Invasion and Aggressiveness: Its activation of NF-κB p65-regulated genes. Survival analysis revealed patient segregation Relationship With TGF-β1 and NF-κB Pathways. Front. Oncol. 9:1306. into low- and high-risk groups, the latter depicting the highest FXYD5/Dys, PAI-1, doi: 10.3389/fonc.2019.01306 tumor necrosis factor (TNF)-α, and TGF-β1 mRNA levels and shorter survival rates. Frontiers in Oncology | www.frontiersin.org 1 December 2019 | Volume 9 | Article 1306 Besso et al. FXYD5/Dysadherin and Endometrial Cancer Conclusion: FXYD5/Dys is a novel biomarker of EC progression related to TGF-β1 and NF-κB pathways that collectively promote tumor dissemination and result in poor patient prognosis. Keywords: endometrial cancer, E-cadherin, FXYD5, dysadherin, TGF-β1, NF-κB, CCL-2 INTRODUCTION (19). More recently, these evidences were confirmed in a lung cancer model, where FXYD5/Dys expression was associated Endometrial cancer (EC) is the second most common with NF-κB pathway activation, CCL-2, interleukin (IL)-6, and gynecological neoplasm and the fourth most frequent TNF-α upregulation and consequent monocyte recruitment (20). women cancer worldwide. Projections for the coming years Finally, FXYD5/Dys expression has been related to distant show a trend toward an increased number of EC cases (1). metastasis and poor prognosis in various tumors (12). Histopathologically, EC is classified into two subtypes: Type The present study aimed to characterize FXYD5/Dys mRNA I or endometrioid carcinomas (EEC) and Type II or non- levels in EC samples and to evaluate the impact of modulating its EEC (NEEC) (2), with EEC being about 80% of EC cases (3). expression upon cell behavior in established EC cell models. We Myometrial invasion (MI) is a key event in EC dissemination proposed a relationship between FXYD5/Dys expression and EC and a critical factor to define the risk of recurrence. While tumor progression/aggressiveness. Stage IA tumors (International Federation of Gynecology and Obstetrics, FIGO, 2009 classification) display <50% MI, Stages IB–IV display >50% MI (deep MI), the latter associated with MATERIALS AND METHODS poor prognosis (5-years survival rate: Stage IA, 90%, Stages Chemicals IB–IV, 78–21%) (4). Routine EC diagnosis involves preoperative Chemicals were of analytical or tissue culture grade and anatomopathological endometrial biopsy evaluation coupled purchased from Sigma-Aldrich (St. Louis, MO, USA). Molecular to imaging techniques, adjusted at evaluation of the surgical biology and electrophoresis reagents were from Thermo-Fisher piece. Among 75% tumors preoperatively classified as early Scientific (Carlsbad, CA, USA) or Bio-Rad (Hercules, CA, stage EC (FIGO Stage I), ∼20% are surgically reclassified as USA). The following antibodies were used: anti-FXYD5/Dys, advanced-stage EC (FIGO Stages II–IV). Moreover, 30% of EC (a) D-2 (mouse monoclonal) and (b) FL-178 (rabbit polyclonal) cases are diagnosed when the tumor has invaded >50% of the from Santa Cruz Biotechnology (SCBT, Santa Cruz, CA, USA); myometrial wall (5). This classification is critical for therapeutic anti-E-cadherin 610181 (mouse monoclonal; Becton Dickinson management and greatly impacts post-surgery patient morbidity. Biosciences, BD, San Diego, CA, USA); anti-NF-κB p65 (rabbit Currently, there are no established molecular biomarkers to monoclonal; C22B4, Cell Signaling, Danvers, MA, USA); anti- determine deep MI and/or to assist in risk stratification in a phospho-NF-κB p65 (Ser536) (rabbit monoclonal; 93H1, Cell sensitive, objective, and reproducible fashion. Signaling); anti-IκB-α (rabbit polyclonal; #9242, Cell Signaling); Alterations in cell–cell adhesion and decreased E-cadherin anti-β-tubulin (mouse monoclonal; clone D66, Sigma-Aldrich). expression have been reported in EC progression (6–8). Cy3-labeled anti-mouse or anti-rabbit secondary antibodies The expression of E-cadherin and related molecules was (Sigma-Aldrich) were used for immunocytochemistry. previously characterized in an EC cellular model of MI, Horseradish peroxidase-conjugated anti-mouse (Vector generated by overexpression of ETV5 transcription factor in Laboratories Inc., Burlingame, CA, USA) or anti-rabbit (Sigma- Hec1a cells (HGE cells) (9). HGE cells showed lower E- Aldrich) IgG was used as secondary antibodies in Western cadherin levels and epithelial-to-mesenchymal transition (EMT)- immunoblotting assays. related molecular changes, as well as higher migratory/invasive properties than control cells (10). Moreover, HGE cells Plasmids depicted higher expression [messenger RNA (mRNA)/protein] The coding sequence of human FXYD5/Dys was cloned in of FXYD5/dysadherin (FXYD5/Dys) (10). FXYD5/Dys is a Type the pcDNA3 commercial plasmid (Thermo-Fisher Scientific) to I transmembrane glycoprotein, a member of the FXYD protein generate the pcDNA3-FXYD5/Dys plasmid; the whole sequence family (11). High FXYD5/Dys protein expression has been was obtained by PCR amplification from a complementary DNA reported in several solid tumors in association to decreased generated by retrotranscription of total RNA from MDA-MB- cell–cell adhesion (12) and low E-cadherin protein expression 231 human breast cancer cells, followed by digestion with EcoRI (13–16). In addition, other E-cadherin-independent mechanisms and HindIII restriction enzymes, insert purification, and plasmid have been proposed; in particular, an association has been found ligation. Nucleotide sequence analysis was done to confirm between FXYD5/Dys expression, nuclear factor (NF)-κB pathway correct insertion and insert coding sequence. activation, and CCL-2 (NF-κB target gene) upregulation in breast and renal cancer cell lines (17, 18). Moreover, an autocrine loop Patient Samples involving FXYD5/Dys, NF-κB pathway activation, and tumor Endometrial tissue samples were obtained from EC patients necrosis factor (TNF) receptor 1 (TNFR1) has been suggested who underwent surgery before receiving hormonal and/or Frontiers in Oncology | www.frontiersin.org 2 December 2019 | Volume 9 | Article 1306 Besso et al. FXYD5/Dysadherin and Endometrial Cancer chemotherapy treatment at Vall d’Hebron Hospital (Barcelona, indicated, transcript relative expression was calculated using an Spain) and at Hospital Italiano from Buenos Aires (Buenos appropriate reference sample, according to the 2−11Ct method Aires, Argentina). In addition, uterine aspirate (UA) samples (11Ct = 1Ct sample - 1Ct reference). Supplementary Table 4 from EC patients were collected (Vall d’Hebron Hospital) lists the primers used in this study. (21). Both institutional review boards approved the protocol, and a written informed consent was signed by all patients. Fluorescence Immunocytochemistry Samples were classified based on the 2009 FIGO staging Cell monolayers were fixed with 4% paraformaldehyde, treated system. Sample collection and handling were done as previously with 0.1% Triton X-100, blocked with 4% bovine serum albumin described (21). Supplementary Tables 1–3 show patient sample (BSA) in PBS, and placed for 1 h with specific primary antibodies, clinical information. followed by a 1-h incubation with secondary antibodies. Nuclear cell staining was done with Hoechst 33342 (Sigma-Aldrich). Cell Culture Cell preparations were analyzed in a Nikon C1 confocal laser Hec1a and HGE cells were cultured in McCoy’s 5A medium microscope (Tokyo, Japan). Images were evaluated using the (Thermo-Fisher Scientific), supplemented with 10% fetal bovine ImageJ software (Wright Cell Imaging Facility, Toronto, ON, serum (FBS), as previously described (10). Canada) (22). EC Cell Line Treatment With Recombinant Sample Preparation, SDS-PAGE, and Human TGF-β1 Western Immunoblotting Hec1a cells (3 × 105 cells per well of six-well plates) were Cell lysates were prepared in a RIPA (radio-immunoprecipitation cultured overnight, followed by treatment with 10 ng/ml of assay) buffer with protease inhibitors. Protein extracts (20 µg) TGF-β1 (Tonbo Biosciences, San Diego, CA, USA) in a culture were electrophoresed in 10 and 12% SDS-polyacrylamide medium without FBS or vehicle (phosphate buffer saline, PBS). gels and electrotransferred to nitrocellulose membranes After 24 h, cells were processed for total RNA/protein extraction. (Amersham Hybond ECL, GE Healthcare, Buckinghamshire, UK). Membranes were blocked in PBS containing 5% non-fat EC Cell Line Transfection Studies milk for 1 h, incubated overnight at 4◦ C with specific primary Downregulation of FXYD5/Dys expression in HGE was achieved antibodies diluted in blocking solution, and for 1 h with by transient transfection of a FXYD5/Dys small interfering RNA secondary antibodies (0.4 µg/ml in a blocking solution). Assays (siRNA) (#45745, three target-specific 19–25-nt siRNAs designed were developed with an ECL Western Blotting Detection Kit to knock down gene expression; SCBT) with Lipofectamine R (ECL, GE Healthcare) chemiluminescence system. Densitometric 2000 (Thermo-Fisher Scientific) following the protocol suggested analysis was done using the ImageJ software (22). by the manufacturer. Briefly, 5 × 105 cells were seeded in each well of a sterile six-well plate and incubated overnight. Hanging Drop Assay Then, cells were transfected with 100 pmol/ml of FXYD5/Dys EC cells were forced to grow under independent anchorage or control scramble siRNA (#37007, SCBT). In addition, Hec1a conditions to promote the formation of cellular aggregates, as cells were transfected with pcDNA3-FXYD5/Dys or pcDNA3 previously reported (22). EC cells were harvested, and 20-µl (control) plasmids. Cells numbering 5 × 105 were transfected drops of 2,000 cells were plated in the lid of a p100 dish and with 1 µg of plasmid DNA with Lipofectamine R 2000. After a cultured for up to 72 h. Images were taken at 24–72 h to evaluate 5-h incubation period, the transfection mixture was replaced by a cell morphology and then analyzed using the ImageJ software. complete culture medium. Cells were subjected to transcript and protein expression analysis as well as to functional tests 24–72 h Wound Healing Assay post-transfection, as specifically indicated. EC cells were subjected to the wound healing assay (22). Briefly, 2 × 105 cells per well were cultured overnight to reach 90% RNA Extraction, cDNA Synthesis, and confluence. Cell monolayers were scratched with a pipette tip, Quantitative Real-Time PCR washed with PBS to remove non-adhered cells, and incubated Procedures were done as previously reported (22), with some for 48 h in a culture medium without FBS. Images were taken at modifications. Total RNA was extracted from tissues and 0–48 h and analyzed using the ImageJ software to determine the UA biopsies and cell lines using standard protocols with percentage of wound healing closure: % wound healing closure TRIzol R (Thermo-Fisher Scientific). Synthesis of complementary = (Area t0 – Area tf )/At0 ) × 100, where Area t0 is the area at the DNA (cDNA) was done with 1–2 µg of total RNA using initial time and Area tf is the area at the final time. the SuperScriptTM III reverse transcriptase enzyme (Thermo- Fisher Scientific). Quantitative evaluation of mRNA levels was GENEVESTIGATOR® and Co-expression performed by real-time PCR (RT-qPCR) using SYBR Green R Analysis PCR Master Mix (Thermo-Fisher Scientific) with the CFX96 GENEVESTIGATORTM (https://genevestigator.com/gv/index. TouchTM unit (Bio-Rad). Transcript expression levels were jsp) integrates manually curated gene expression data from determined as follows: 2−1Ct , where 1Ct = Ct gene under study public repositories to investigate gene regulation at the – Ct housekeeping gene. GAPDH (cell lines and EC tissues) or transcriptional level using a wide variety of experimental POLR2A (UA samples) was used as housekeeping genes. When conditions (23). In this study, the GENEVESTIGATOR R Frontiers in Oncology | www.frontiersin.org 3 December 2019 | Volume 9 | Article 1306 Besso et al. FXYD5/Dysadherin and Endometrial Cancer co-expression tool was used to find co-regulated genes Stage I tumors, higher FXYD5/Dys mRNA levels were found with FXYD5/Dys in three datasets: “samples” (61,060 at the invasive front compared to the superficial section of the samples), “cancer” (634 categories), and “perturbations” (4,760 tumor (P = 0.0013) (Figure 1C). perturbations), analyzed using the HS_AFFY_U133PLUS_2- With regard to tumor grade, FXYD5/Dys mRNA expression 0 array. A co-expression value was calculated by Pearson’s levels were higher in Grade 3 than in Grade 1 and Grade 2 correlation coefficient. tumors (P = 0.0381) (Figure 1D). Based on these findings, the relationship between FXYD5/Dys transcript levels Gene Expression Analysis of EC Samples and the European Society for Medical Oncology (ESMO) To evaluate the relationship between FXYD5/Dys mRNA risk stratification system (24) was assessed, finding higher levels, E-cadherin (mRNA and protein), and NF-κB pathway FXYD5/Dys mRNA levels in intermediate/high-risk tumors than activation, transcriptomic, proteomic, and clinical data from The in low-risk ones (P = 0.0261) (Figure 1E). Cancer Genome Atlas Uterine Corpus Endometrioid Cancer A few years ago, UA biopsies became of interest in study (TCGA-UCEC) were retrieved from the UCSC Xena the evaluation of EC molecular biomarkers. Compared to repository (https://xenabrowser.net/). conventional tissue biopsies, UA are a reliable source for For transcriptomic analysis, RSEM-normalized RNAseq data EC biomarker assessment with high specificity and sensitivity, from Illumina GA (N = 381) and Illumina HiSeq (N = 201) were capable of capturing intra-tumor heterogeneity with a low- downloaded. In addition, replicate-base normalization (RBN)- cost ambulatory sampling method (21, 25, 26). In order to normalized proteomic data from E-cadherin immunodetection evaluate FXYD5/Dys’s potential as an MI preoperative biomarker, using reverse phase protein array (RPPA) were retrieved (N = FXYD5/Dys expression levels were evaluated in preoperative 440 samples). UA biopsies from EC patients. First, a positive significant To analyze the regulatory transcription factor (TF) impact of correlation (r = 0.5321; P = 0.0412) was found between NF-κB p65 on EC gene expression patterns, data from GA (N = FXYD5/Dys mRNA levels detected in UA and in tissue-paired 313), GAV2 (N = 349), and HiSeqV2 (N = 155) platforms were biopsies (Figure 1F). Next, and in line with the results obtained retrieved and compiled. from tissue biopsies, higher FXYD5/Dys mRNA levels were detected in UA biopsies from tumors depicting MI > 50% (P Statistical Analysis = 0.0315) (Figure 1G), Grade 3 (P = 0.0365) (Figure 1H), and All experiments were performed at least in triplicate. Results were intermediate/high risk of recurrence (P = 0.0190) (Figure 1I), expressed as mean ± standard error (SEM). A P < 0.05 was compared to those with MI < 50%, Grades 1 and 2, or considered statistically significant. Variable distribution analysis low risk of recurrence, respectively. ROC analysis revealed was done using Shapiro–Wilk’s normality test. Student’s t-test and FXYD5/Dys as a significant predictor of MI [area under the Mann–Whitney test were used to compare two groups. Analysis curve (AUC) = 0.7545; P = 0.0486, sensitivity = 70%, specificity of variance (ANOVA) and the Kruskal–Wallis test were used = 81.82%], grade (AUC = 0.8077; P = 0.0353, sensitivity = for comparisons involving more than two experimental groups. 83.33%, specificity = 69.23%), and risk of recurrence (AUC = In some cases, the Wilcoxon signed-ranked test was applied. 0.8091; P = 0.0167, sensitivity = 72.73%, specificity = 80%) in Paired human samples were analyzed with paired Student’s t-test. UA. The estimated cutoff value of FXYD5/Dys expression was For correlation analysis, the Spearman correlation coefficient, ρ 0.4178. Supplementary Figure 1 shows ROC curves for the three (rho), was used. ROC (receiver operating characteristic) analysis clinicopathological parameters. was performed to determine the potential of FXYD5/Dys mRNA in preoperative UA as a marker of deep MI, high grade, and intermediate/high risk of recurrence. For survival analysis, FXYD5/Dys Expression and EC Cell–Cell Kaplan–Meier curves were constructed, and differences between Adhesion and Migration them were analyzed by a log-rank test. GraphPad Prism software, Taking into account results presented in the previous section version 5.01 (GraphPad Software, San Diego, CA, USA), was used and that modulation of FXYD5/Dys expression has been to do the analyses. associated with an aggressive cellular behavior in breast and renal cancer cellular models (17, 18, 27), the impact of RESULTS its knockdown upon cell behavior was evaluated. The EC cellular model of MI HGE was used to perform these studies. Expression of FXYD5/Dys in EC Samples FXYD5/Dys downregulation in HGE cells was confirmed and Association With Clinicopathological at mRNA and protein levels by RT-qPCR and Western Parameters immunoblotting/fluorescent immunocytochemistry, respectively FXYD5/Dys transcript levels were determined by RT-qPCR in (Supplementary Figure 2). EC tissue samples. As shown in Figure 1A, Stage IB as well as Since FXYD5/Dys has been reported to promote cell motility Stage II–Stage III tumors showed higher FXYD5/Dys mRNA in vitro in other tumor types (17, 18, 27), the impact of its levels than Stage IA tumors (P < 0.05). In line with these knockdown in HGE cell migratory capacity was evaluated. findings, in samples from the tumor invasive front, increased HGE siRNA FXYD5/Dys cells depicted diminished migration FXYD5/Dys mRNA levels were found compared with paired compared to HGE siRNA CTL cells (P < 0.01) (Figure 2A). In superficial samples (14/20, 70%; P = 0.0123) (Figure 1B). Among line with these findings, HGE siRNA FXYD5/Dys cells showed Frontiers in Oncology | www.frontiersin.org 4 December 2019 | Volume 9 | Article 1306 Besso et al. FXYD5/Dysadherin and Endometrial Cancer FIGURE 1 | FXYD5/Dys mRNA expression and clinicopathological parameters in EC samples. (A–C) RT-qPCR analysis of FXYD5/Dys mRNA levels in (A) EEC samples grouped according to FIGO stage (Stage IA, N = 27; Stage IB, N = 15; and Stages II + III, N = 15) (differences observed between Stage IA and Stage IB tumors and between Stage IA and Stage II–III tumors; * P < 0.05, ANOVA with Bonferroni post-test), (B) paired biopsies from superficial and invasive front of EEC samples (FIGO Stages I–III) (N = 20; P = 0.0123, paired t-test), and (C) paired biopsies from superficial and invasive front of Stage I EEC samples (N = 9; P = 0.0013, paired t-test). (D) FXYD5/Dys mRNA levels in Grade 1, 2, and 3 EEC samples (Stages I–III). EEC tumors classified as Grades 1 and 2 were included in the same group (Grade 1/2). Differences observed between Grade 1/2 tumors (N = 43) and Grade 3 tumors (N = 17) (P = 0.0381; unpaired t-test with Welch’s correction). (E) RT-qPCR analysis of FXYD5/Dys expression in EC samples grouped according to risk of lymph node involvement and recurrence (low risk N = 19, intermediate/high risk N = 32; P = 0.0261, unpaired t-test with Welch’s correction). (F) FXYD5/Dys mRNA levels assessed in UA biopsies. Correlation analysis between mRNA levels detected in uterine aspirate biopsies and paired EC tissue (N = 15; r = 0.5321, Spearman correlation, P = 0.0412). (G–I) RT-qPCR analysis of FXYD5/Dys expression in UA from EEC grouped according to (G) MI depth (MI < 50%, N = 11; MI > 50%, N = 10; P = 0.0315, unpaired t-test, with Welch’s correction), (H) tumor grade (Grades 1 and 2, N = 13; Grade 3, N = 6; P = 0.0365, Mann–Whitney test), and (I) risk of lymph node involvement and recurrence (low risk, N = 10; intermediate/high risk, N = 11; P = 0.0190, unpaired t-test with Welch’s correction). a significant decrease in the promigratory chemokine CCL- Studies done in breast cancer tissues showed a positive 2 mRNA levels when compared to HGE siRNA CTL cells FXYD5/Dys immunostaining where cell–cell contacts had been (P < 0.01) (Figure 2B). lost (11). In the same study, FXYD5/Dys overexpression in Frontiers in Oncology | www.frontiersin.org 5 December 2019 | Volume 9 | Article 1306 Besso et al. FXYD5/Dysadherin and Endometrial Cancer FIGURE 2 | FXYD5/Dys regulation of cell–cell adhesion and cell migration in HGE cells. (A) Left panel: Wound healing assay of HGE siRNA FXYD5/Dys and HGE siRNA CTL cells. Representative images of cells at 0, 24, and 48 h are shown (magnification 40×). Right panel: Free-cell area quantified using ImageJ software. Wound healing closure percentage was plotted at every time point (**P < 0.01, two-way ANOVA, Bonferroni post-test). (B) RT-qPCR analysis of the pro-migratory chemokine CCL-2 expression in HGE siRNA CTL and HGE siRNA FXYD5/Dys cells (**P < 0.01, Wilcoxon signed rank test). (C) Left panel: Hanging drop assay of HGE siRNA FXYD5/Dys and HGE siRNA CTL cells. Representative images of cell aggregates at 72 h are shown (magnification 100×). Right panel: Cell aggregate area was quantified using ImageJ software, and mean cell aggregates areas (expressed as squared pixels) were plotted and compared (*P < 0.05, Mann–Whitney test). (D) Left panel: Fluorescent immunocytochemistry of HGE siRNA FXYD5/Dys and HGE siRNA CTR cell monolayers using anti-E-cadherin monoclonal antibody (610181; 2.5 µg/ml); cell nuclei are visualized using HOECHST 33342. Bar: 20 µm. Right panel: Percentage of cells depicting strong fluorescence intensity quantified using ImageJ software (*P < 0.05, Mann–Whitney test). (E) RT-qPCR analysis of E-cadherin mRNA in HGE siRNA FXYD5/Dys and HGE siRNA CTL cells (***P < 0.0001, Wilcoxon signed rank test). Frontiers in Oncology | www.frontiersin.org 6 December 2019 | Volume 9 | Article 1306 Besso et al. FXYD5/Dysadherin and Endometrial Cancer liver cancer cell lines was found associated with decreased E- and between FXYD5/Dys and PAI-1 mRNA levels are shown cadherin protein levels and decreased Ca2+ -dependent cellular (Figures 4F,G). aggregation. Thus, the impact of FXYD5/Dys knockdown on HGE cell adhesiveness and E-cadherin expression was evaluated. FXYD5/Dys Expression and NF-κB Cellular aggregates formed by HGE siRNA FXYD5/Dys cells Pathway Activation were larger than those in HGE siRNA CTL cells (P < 0.05) FXYD5/Dys has been found to modulate NF-κB pathway (Figure 2C). In line with this results, HGE siRNA FXYD5/Dys activation (17–20). When it is taken into account that NF- cells depicted higher E-cadherin expression than control cells, κB has been suggested to play a key role during EC which was observed at protein (P < 0.05) (Figure 2D) and carcinogenesis (32, 33), a relationship between FXYD5/Dys and mRNA (P < 0.0001) levels (Figure 2E). Altogether, these results NF-κB activation in EC is found. suggest a direct relationship between an HGE cell adhesive and First, RNAseq data analysis from the TCGA-UCEC cohort migratory behavior and FXYD5/Dys expression. revealed a positive correlation between FXYD5/Dys and most The effect of FXYD5/Dys expression upon cell behavior of the NF-κB pathway genes evaluated (Figure 5A). Moreover, and E-cadherin expression was also evaluated in Hec1a cells experiments performed in Hec1a cells revealed increased NF- that express high E-cadherin levels, by transient transfection κB (P < 0.001) and decreased total protein levels of the with the pcDNA3-FXYD5/Dys plasmid (Hec1a pcDNA3- IκB-α inhibitor in Hec1a pcDNA3-FXYD5/Dys cells compared FXYD5/Dys). The transfection procedure effectiveness was to Hec1a-pcDNA3 cells (Figure 5B). These changes were verified by RT-qPCR and Western immunoblotting analysis accompanied by a trend toward higher CCL-2 (P = 0.0654) (Supplementary Figure 3). FXYD5/Dys overexpression resulted expression and increased TNF-α (P = 0.0313) mRNA levels in a decreased E-cadherin expression in Hec1a pcDNA3- (Figures 5C,D), two target genes of the NF-κB pathway. FXYD5/Dys cells, evidenced by lower mRNA and protein levels Moreover, an increase in TNFR1 mRNA levels (P = 0.0362) was (Figures 3A,B). In agreement with these results, an in silico also observed (Figure 5E). In line with these results, a significant analysis of the TCGA-UCEC RNAseq dataset (N = 397) showed correlation was found between FXYD5/Dys mRNA levels and a negative correlation between FXYD5/Dys mRNA and E- transcriptional activation of the NF-κB p65-regulated genes in cadherin mRNA and protein levels (Supplementary Figure 4). EC samples from the TCGA-UCEC study (Figure 5F) and an These findings could be associated to E-cadherin transcriptional increased expression of several NF-κB pathway target genes, repression (28), as suggested by a higher expression of Snail among them CCL-2 and TNF-α (Supplementary Table 5). (P < 0.05) and Slug (P < 0.01) and a trend for Zeb1 Finally, to assess the prognostic significance of FXYD5/Dys transcriptional repressors (Figures 3C–E) detected in Hec1a expression and genes related to TGF-β and NF-κB pathways, a cells overexpressing FXYD5/Dys. With regard to cell behavior, survival analysis was done in the TCGA-UCEC study using the FXYD5/Dys overexpression in Hec1a cells was associated with SurvExpress tool. As a result of this analysis, two risk groups increased cell migration (P < 0.001) (Figure 3F) and decreased (low and high risk) were defined based on the mRNA expression cell–cell adhesion, as evidenced by a decrease in Hec1a pcDNA3- levels of FXYD5/Dys, TGF-β1, PAI-1, and TNF-α (Figure 5G). FXYD5/Dys cell aggregate area (P < 0.001) (Figure 3G). Patients included in the high-risk group were characterized by an increased expression of the four genes and a significant lower (P Relationship Between FXYD5/Dys < 0.05) survival rate (Figure 5H). Expression and TGF-β1 in EC Cells TGF-β-mediated signaling is activated in EC cells (29, 30) DISCUSSION and plays a key role in the initial stages of invasion and metastasis (31). A co-expression analysis performed An increased expression of FXYD5/Dys has been related to using the GENEVESTIGATOR database revealed that TGF- metastasis and poor prognosis in several tumors. In some tumors β1 was found between the transcripts most significantly co- and cell lines, FXYD5/Dys overexpression significantly correlated expressed with FXYD5 in three independent-sample datasets with a decreased E-cadherin expression, leading authors to (Supplementary Figures 5A–C, respectively). To determine the propose that FXYD5/Dys would promote metastasis through relationship between TGF-β1 and FXYD5/Dys expression, Hec1a a negative modulation of E-cadherin expression/functions (13– cells were treated with human recombinant TGF-β1, and 16). In addition, FXYD5/Dys expression has been shown to FXYD5/Dys expression was evaluated. As a result, FXYD5/Dys induce in vitro changes in cell morphology and to in vivo promote mRNA and protein levels increased in Hec1a upon TGF-β1 metastasis in cell lines and tumors that do not express E-cadherin treatment (Figures 4A–C). On the other hand, Hec1a pcDNA3- (17). The present study aimed to characterize FXYD5/Dys FXYD5/Dys cells exhibited increased TGF-β1 mRNA levels expression in EC. RT-qPCR studies in EC tissue samples revealed compared with control cells (P < 0.05) (Figure 4D). Moreover, an association between increased FXYD5/Dys mRNA levels and an RNAseq analysis from the TCGA-UCEC cohort revealed a several EC clinicopathological parameters. Among them, tumors positive correlation between FXYD5/Dys expression and several with deep MI showed higher FXYD5/Dys mRNA levels, even TGF-β signaling genes, including TGF-β1, ZEB1, ZEB2, SNAI1, in Stage I tumors. Moreover, the tumor invasive front had FN1, SERPINE1 (PAI-1), and SMAD3 (Figure 4E). As an higher FXYD5/Dys mRNA levels than the superficial tumor example, correlation analyses between FXYD5/Dys and TGF-β1 section, suggesting an involvement of FXYD5/Dys in tumor Frontiers in Oncology | www.frontiersin.org 7 December 2019 | Volume 9 | Article 1306 Besso et al. FXYD5/Dysadherin and Endometrial Cancer FIGURE 3 | Modulation of FXYD5/Dys expression in Hec1a cells and changes in cell migration. (A) RT-qPCR analysis of E-cadherin mRNA expression in Hec1a cells transiently transfected with pcDNA3 empty plasmid (Hec1a pcDNA3 cells) or pcDNA3-FXYD5/Dys plasmid (Hec1a pcDNA3-FXYD5/Dys cells) (**P < 0.01, Wilcoxon signed rank test). (B) Immunodetection of E-cadherin by Western immunoblotting of Hec1a pcDNA3 and Hec1a pcDNA3-FXYD5/Dys cell protein extracts using anti-E-cadherin monoclonal antibody (610181, BD; 0.125 µg/ml). β-Tubulin (anti-β-tubulin monoclonal antibody D66; 0.5 µg/ml) was used as loading control. (C–E) RT-qPCR analysis of E-cadherin transcriptional repressors Snail (C) (*P < 0.05), Slug (D) (**P < 0.01), and Zeb1 (E) (P = 0.055) (Wilcoxon signed rank test). (F) Left panel: Wound healing assay of Hec1a pcDNA3 and Hec1a pcDNA3-FXYD5/Dys cells. Representative images of cells at 0, 24, and 48 h post-transfection are shown (magnification 40×). Right panel: Free-cell area was quantified using ImageJ software, and the wound healing closure percentage was plotted at every time point (**P < 0.01, ***P < 0.0001, two-way ANOVA, Bonferroni post-test). (G) Left panel: Hanging drop assay of Hec1a pcDNA3 and Hec1a pcDNA3-FXYD5/Dys cells. Representative images of cell aggregates at 48 h are shown (magnification 100×). Right panel: Cell aggregate area was quantified using ImageJ software, and mean cell aggregate areas (expressed as squared pixels) were plotted and compared (**P < 0.01, Mann–Whitney test). dissemination. In addition, FXYD5/Dys expression was higher in low-risk tumors. In addition, FXYD5/Dys mRNA levels detected Grade 3 tumors than in Grade 1 and Grade 2 tumors, as observed in preoperative UA reproduced tissue biopsy results and in other cancers (34, 35). Consequently, intermediate/high- demonstrated the diagnostic potential of FXYD5/Dys transcript risk tumors showed higher FXYD5/Dys mRNA levels than evaluation for MI, grade, and risk of recurrence. Frontiers in Oncology | www.frontiersin.org 8 December 2019 | Volume 9 | Article 1306 Besso et al. FXYD5/Dysadherin and Endometrial Cancer FIGURE 4 | Association between TGF-β1 and FXYD5/Dys expression. (A) RT-qPCR analysis of FXYD5/Dys expression in Hec1a cells treated with 10 ng/ml TGF-β1 for 24 h. Controls were incubated with vehicle (**P < 0.01, Wilcoxon signed rank test). (B) Western immunoblotting of Hec1a protein extracts using FXYD5/Dys monoclonal antibody (sc-166782; 2 µg/ml). β-Tubulin was included as loading control (monoclonal antibody D66; 0.5 µg/ml). (C) Representative images of fluorescent immunocytochemistry analysis of FXYD5/Dys protein expression (anti-FXYD5/Dys polyclonal sc-98246; 2 µg/ml); cell nuclei were stained with HOECHST 33342, bar: 20 µm. (D) RT-qPCR analysis of TGF-β1 mRNA levels in Hec1a pcDNA3-FXYD5/Dys and Hec1a pcDNA3 cells (*P < 0.05, Wilcoxon signed rank test). (E) Correlation analysis between FXYD5/Dys and TGF-β pathway-related genes. A heatmap was built based on Pearson r correlation values for each gene. (F) Correlation analysis between TGF-β1 and FXYD5/Dys mRNA levels in EC samples from the TCGA-UCEC cohort (Pearson correlation, r = 0.4966, N = 444). mRNA levels are expressed as log2 (x + 1) were x is the RSEM normalized expression value. (G) Correlation analysis between PAI-1 and FXYD5/Dys mRNA levels in EC samples from the TCGA-UCEC cohort (Pearson correlation, r = 0.2886, N = 444). mRNA levels are expressed as log2 (x + 1) were x is the RSEM normalized expression value. Results on FXYD5/Dys expression in EC patient samples EC cell models. Firstly, FXYD5/Dys downregulation in HGE suggest its role in EC progression. To further address this cells resulted in lower cell migration and decreased CCL-2 hypothesis, modulation of FXYD5/Dys expression was done in mRNA levels, in agreement with previous findings (17, 18). Frontiers in Oncology | www.frontiersin.org 9 December 2019 | Volume 9 | Article 1306 Besso et al. FXYD5/Dysadherin and Endometrial Cancer FIGURE 5 | FXYD5/Dys expression and NF-κB pathway activation. (A) Correlation analysis between FXYD5/Dys and NF-κB pathway-related genes. A heatmap was built based on Pearson r correlation values for each gene. (B) Evaluation of NF-κB pathway activation by Western immunoblotting. A representative image of p65 total and phosphorylated protein forms and IκB-α total protein form is shown. Antibodies: NF-κB p65 (monoclonal C22B4 # 4764, 1:1,000 dilution), phospho-NF-κB p65 (Ser536) (monoclonal 93H1 # 3033, 1:1,000 dilution), and IκB-α (polyclonal # 9242, 1:1,000 dilution). (C–E) RT-qPCR analysis of CCL-2 (C) (P = 0.0654), TNF-α (D) (P < 0.05), and TNFR1 (E) (P < 0.01) (Wilcoxon signed-rank test) in Hec1a pcDNA3 and Hec1a pcDNA3-FXYD5/Dys cells. (F) Correlation analysis between FXYD5/Dys mRNA levels and transcriptional activation of genes regulated by p65. Regulatory impact of p65 transcription factor on tumor-specific gene expression patterns from EC samples from the TCGA-UCEC study. p65 regulatory impact to gene expression in each tumor is represented as a linear model t-value: a positive t-value indicates p65 upregulates its target genes, and a negative t-value indicates p65 downregulates its target genes. FXYD5/Dys mRNA levels are expressed as log2 (x + 1) were x is the RSEM normalized expression value (Spearman correlation, r = 0.6927, N = 81). (G,H) SurvExpress tool used to evaluate the impact of FXYD5/Dys, PAI-1, TGF-β1, and TNF-α mRNA levels upon EC patient survival. This analysis was performed using gene expression and survival data of EC samples from the TCGA-UCEC study (N = 332). A prognostic index was estimated by beta coefficients multiplied by gene expression values of the four genes included in the analysis. Then, EC samples were divided into “low-risk” and “high-risk” groups according to their prognostic index values. (G) Box plots generated by SurvExpress showed the mRNA expression levels of FXYD5/Dys, PAI-1, TGF-β1, and TNF-α. The P-value from a t-test of the difference between low- and high-risk groups was also calculated for each gene: FXYD5/Dys, P = 3.80e−03; SERPINE1, P = 2.05e−08; TGF-β1, P = 1.47e−24; and TNF, P = 3.79e−10. Low risk was in green, and high risk was in red. (H) Kaplan–Meier survival curves were constructed to evaluate the impact of FXYD5/Dys, PAI-1, TGF-β1, and TNF-α mRNA levels (low- and high-risk groups) upon EC patient survival (log-rank test). Frontiers in Oncology | www.frontiersin.org 10 December 2019 | Volume 9 | Article 1306 Besso et al. FXYD5/Dysadherin and Endometrial Cancer Also, FXYD5/Dys knockdown was associated with an increase in tumor microenvironment and eventually promoting cancer HGE cell–cell adhesiveness, evidenced in the formation of larger progression, through NF-κB pathway activation. cell aggregates and increased E-cadherin expression, suggesting Survival analysis revealed the potential use of the combined an inverse relationship between both proteins. On the other expression of FXYD5/Dys, PAI-1, TGF-β1, and TNF-α hand, FXYD5/Dys overexpression in Hec1a cells resulted in transcripts as a predictor of EC patient survival outcome. increased cell migration and decreased cell adhesive capacity; However, even when the results showed statistical significance, at the molecular level, these changes were associated with a an independent survival analysis should be carried out in an decreased E-cadherin expression, accompanied by increased independent cohort of patients in order to obtain more robust mRNA levels of Snail, Slug, and ZEB1, in agreement with findings. One possible cause of these findings may relate to previous reports in tumor samples (6, 36). The results obtained the disparity in the percentages of death and censored events after Hec1a EC cell modulation of FXYD5/Dys expression are in in the cohort analyzed. The population included 247 EC line with results obtained in EC patient samples showing higher patients, from which 33 (13%) experienced the event (death) FXYD5/Dys transcript levels in invasive EC (MI > 50%) and at and 214 (87%) were classified as censored events. Moreover, the invasive tumor front, as well as in Grade 3 tumors. Altogether, the clinicopathological characteristics of the cohort showed these findings lead us to propose that an increased FXYD5/Dys a predominance of tumors depicting favorable prognostic expression could promote cell migration and dedifferentiation, factors, evidenced by 80% of cases with endometrioid histology, in part through E-cadherin downregulation, in EC cells, thus 66% with FIGO Stage I tumors, 62% with tumors having facilitating MI and tumor dissemination. MI < 50%, and 62% with histologic Grade 1 or 2. Since our A co-expression analysis using three global expression findings demonstrate an association between higher FXYD5/Dys studies identified TGF-β1 as one of the molecules most mRNA levels and advanced FIGO stage, deep MI, and high significantly co-expressed with FXYD5/Dys. Moreover, TGF-β1 histological grade, the survival analysis results could be justified, induced an increase in FXYD5/Dys expression in Hec1a cells, at least in part, by the characteristics of the studied cohort. and FXYD5/Dys overexpression increased TGF-β1 expression. However, additional molecular mechanisms underlying the EC Furthermore, an association was found between elevated pathogenesis (FXYD5/Dys-related and non-related) should not mRNA levels of FXYD5/Dys and several TGF-β genes in be ruled out. Thus, FXYD5/Dys would exert its action through the TCGA-UCEC cohort. Altogether, these results lead us various molecular mechanisms highly associated with EC to propose an autocrine and/or paracrine regulation loop progression, which has a negative impact on patients’ prognosis. between FXYD5/Dys and the TGF-β1 pathway, in which TGF- Altogether, results here presented lead us to propose β1 produced by the tumor microenvironment or by tumor FXYD5/Dys as a biomarker of EC progression and cells could induce FXYD5/Dys expression in the tumor cell. aggressiveness. These findings are the first evidence of FXYD5/Dys could then activate TGF-β1 expression, leading to FXYD5/Dys implication in EC progression. The evaluation sustained pathway activation. of FXYD5/Dys expression may contribute to current tools for EC Our results revealed that overexpression of FXYD5/Dys in management, in addition to deepening the molecular basis of EC. Hec1a cells induced the NF-κB pathway, leading to activation Additional prospective studies will further help in confirming of NF-κB target genes, as well as of TNFR1, in line with the diagnostic/prognostic value of FXYD5/Dys in EC. previous studies (19). In addition, TCGA RNAseq data analysis revealed a correlation between FXYD5/Dys mRNA levels with transcriptional activation of NF-κB p65-regulated genes in DATA AVAILABILITY STATEMENT EC. FXYD5/Dys expression has been associated with NF-κB The datasets generated for this study can be found in pathway activation in cancer models (17), and more recently, it the manuscript. has been shown to consequently induce macrophage/monocyte recruitment mediated by CCL-2/CCR2 (20). Elevated infiltration of macrophages (37, 38) and lymphocytes (39–42) has been ETHICS STATEMENT found in EC tissues compared to normal tissues. Moreover, macrophage infiltration was associated with poor prognosis in EC Both Institutional Ethical Review Boards approved the protocol (37, 38, 43). Also, low preoperative LMR (lymphocyte–monocyte (Vall d’Hebron Hospital Protocol SAF2008_3997; Hospital ratio) values were associated with shorter survival rates (44). Italiano Protocol #1895) and a written informed consent signed Moreover, high monocyte counts were associated with advanced by all patients; the IBYME institutional Ethical Committee stage, deep MI, lymph node metastases, and decreased survival at reviewed and accepted both clinical protocols (CE021-2/2012). the time of recurrence (45). An RNAseq data analysis performed by us on the TCGA-UCEC cohort revealed a positive correlation AUTHOR CONTRIBUTIONS between FXYD5/Dys and CD68 (monocyte/macrophage marker) (P < 0.0001) and CD163 (macrophage activation marker) MB and MV-L were involved in the study conception and design. (P < 0.0001) (data not shown). Altogether, these results MB, MR, LL, CM, MLM, MFM, RS, EC, AG-M, AW, RO, and lead us to propose FXYD5/Dys as a potential inflammation MV-L participated in the data collection. MB, MR, and MV-L mediator in EC, facilitating monocyte recruitment to the participated in the data analysis. JR and MV-L were responsible Frontiers in Oncology | www.frontiersin.org 11 December 2019 | Volume 9 | Article 1306 Besso et al. FXYD5/Dysadherin and Endometrial Cancer for the grants that supported this study. MB and MV-L drafted MV-L from IRSES (International Research Staff Exchange the manuscript. All authors read and accepted the final version. Scheme; Ref. 269285). MV-L was responsible for coordinating and supervising the entire project. ACKNOWLEDGMENTS FUNDING The authors acknowledge Fundación R. Barón and Fundación Williams for institutional (IBYME) support. Studies were supported by grants to MV-L from the Agencia Nacional de Promoción Científica y Tecnológica SUPPLEMENTARY MATERIAL de Argentina (Grant PICTSU-1072), Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET) (Grants The Supplementary Material for this article can be found PIP740 and PIP887), and Instituto Nacional del Cáncer online at: https://www.frontiersin.org/articles/10.3389/fonc. (Grants INC 2014-2015 and INC 2016-2017), and to JR and 2019.01306/full#supplementary-material REFERENCES association with lymphangiogenesis and prognostic significance. Am J Surg Pathol. (2006) 30:185–93. doi: 10.1097/01.pas.0000178090.54147.f8 1. Globocan. International Agency for Research on Cancer (2018). Available 16. Sato H, Ino Y, Miura A, Abe Y, Sakai H, Ito K, et al. Dysadherin: expression online at: https://www.uicc.org/new-global-cancer-data-globocan-2018 and clinical significance in thyroid carcinoma. J Clin Endocrinol Metab. (2003) 2. Bokhman JV. Two pathogenetic types of endometrial carcinoma. Gynecol 88:4407–12. doi: 10.1210/jc.2002-021757 Oncol. (1983) 15:10–7. doi: 10.1016/0090-8258(83)90111-7 17. Nam JS, Kang MJ, Suchar AM, Shimamura T, Kohn EA, Michalowska AM, 3. Colombo N, Creutzberg C, Amant F, Bosse T, González-Martín A, Ledermann et al. Chemokine (C-C motif) ligand 2 (CCL2) mediates the prometastatic J, et al. ESMO-ESGO-ESTRO consensus conference on endometrial cancer: effect of dysadherin in human breast cancer cells. Cancer Res. (2006) 66:7176– diagnosis, treatment and follow-up. Int J Gynecol Cancer. (2016) 26:2–30. 84. doi: 10.1158/0008-5472.CAN-06-0825 doi: 10.1097/IGC.0000000000000609 18. Schuler Y, Lee-Thedieck C, Geiger K, Kaiser T, Ino Y, Aicher WK, et al. 4. Creasman W. Revised FIGO staging for carcinoma of the endometrium. Int J Osteoblast-secreted factors enhance the expression of dysadherin and CCL2- Gynaecol Obstet. (2009) 105:109. doi: 10.1016/j.ijgo.2009.02.010 dependent migration of renal carcinoma cells. Int J Cancer. (2012) 130:288–99. 5. Seracchioli R, Solfrini S, Mabrouk M, Facchini C, Di Donato N, Manuzzi L, doi: 10.1002/ijc.25981 et al. Controversies in surgical staging of endometrial cancer. Obstet Gynecol 19. Lubarski-Gotliv I, Asher C, Dada L, Garty H. FXYD5 has a pro- Int. (2010) 2010:181963. doi: 10.1155/2010/181963 inflammatory role in epithelial cells. J Biol Chem. (2016) 291:11072–82. 6. Abouhashem NS, Ibrahim DA, Mohamed AM. Prognostic implications of doi: 10.1074/jbc.M115.699041 epithelial to mesenchymal transition related proteins (E-cadherin, Snail) and 20. Brazee PL, Soni PN, Tokhtaeva E, Magnani N, Yemelyanov A, Perlman HR, hypoxia inducible factor 1α in endometrioid endometrial carcinoma. Ann et al. FXYD5 is an essential mediator of the inflammatory response during Diagn Pathol. (2016) 22:1–11. doi: 10.1016/j.anndiagpath.2016.01.004 lung injury. Front Immunol. (2017) 8:623. doi: 10.3389/fimmu.2017.00623 7. Bilyk O, Coatham M, Jewer M, Postovit LM. Epithelial-to-mesenchymal 21. Colas E, Perez C, Cabrera S, Pedrola N, Monge M, Castellvi J, et al. Molecular transition in the female reproductive tract: from normal functioning to disease markers of endometrial carcinoma detected in uterine aspirates. Int J Cancer. pathology. Front Oncol. (2017) 7:145. doi: 10.3389/fonc.2017.00145 (2011) 129:2435–44. doi: 10.1002/ijc.25901 8. Makker A, Goel MM. Tumor progression, metastasis, and modulators 22. Matos ML, Lapyckyj L, Rosso M, Besso MJ, Mencucci MV, Briggiler CI, et al. of epithelial-mesenchymal transition in endometrioid endometrial Identification of a novel human E-cadherin splice variant and assessment carcinoma: an update. Endocr Relat Cancer. (2016) 23:R85–111. of its effects upon EMT-related events. J Cell Physiol. (2017) 232:1368–86. doi: 10.1530/ERC-15-0218 doi: 10.1002/jcp.25622 9. Monge M, Colas E, Doll A, Gonzalez M, Gil-Moreno A, Planaguma J, et al. 23. Hruz T, Laule O, Szabo G, Wessendorp F, Bleuler S, Oertle L, et al. ERM/ETV5 up-regulation plays a role during myometrial infiltration through Genevestigator V3: a reference expression database for the meta- matrix metalloproteinase-2 activation in endometrial cancer. Cancer Res. analysis of transcriptomes. Adv Bioinform. (2008) 2008:420747. (2007) 67:6753–9. doi: 10.1158/0008-5472.CAN-06-4487 doi: 10.1155/2008/420747 10. Colas E, Muinelo-Romay L, Alonso-Alconada L, Llaurado M, Monge M, 24. Colombo N, Preti E, Landoni L, Carinelli S, Colombo A, Marini C, et al. ESMO Barbazan J, et al. ETV5 cooperates with LPP as a sensor of extracellular signals Guidelines Working Group. Endometrial cancer: ESMO clinical practice and promotes EMT in endometrial carcinomas. Oncogene. (2012) 31:4778–88. guidelines for diagnosis, treatment and follow-up. Ann Oncol. (2011) 22:vi35– doi: 10.1038/onc.2011.632 9. doi: 10.1093/annonc/mdr374 11. Ino Y, Gotoh M, Sakamoto M, Tsukagoshi K, Hirohashi S. Dysadherin, a 25. Perez-Sanchez C, Colas E, Cabrera S, Falcon O, Sanchez-del-Rio A, cancer-associated cell membrane glycoprotein, down-regulates E-cadherin Garcia E, et al. Molecular diagnosis of endometrial cancer from and promotes metastasis. Proc Natl Acad Sci USA. (2002) 99:365–70. uterine aspirates. Int J Cancer. (2013) 133:2383–91. doi: 10.1002/ijc. doi: 10.1073/pnas.012425299 28243 12. Nam JS, Hirohashi S, Wakefield LM. Dysadherin: a new player in cancer 26. Mota A, Colás E, García-Sanz P, Campoy I, Rojo-Sebastián AS, Gatius S, progression. Cancer Lett. (2007) 255:161–9. doi: 10.1016/j.canlet.2007.02.018 et al. Genetic analysis of uterine aspirates improves the diagnostic value and 13. Nakanishi Y, Akimoto S, Sato Y, Kanai Y, Sakamoto M, Hirohashi captures the intra-tumor heterogeneity of endometrial cancers. Mod Pathol. S. Prognostic significance of dysadherin expression in tongue cancer: (2017) 30:134–45. doi: 10.1038/modpathol.2016.143 immunohistochemical analysis of 91 cases. Appl Immunohistochem Mol 27. Lee YK, Lee SY, Park JR, Kim RJ, Kim SR, Roh KJ, et al. Dysadherin Morphol. (2004) 12:323–8. doi: 10.1097/00129039-200412000-00006 expression promotes the motility and survival of human breast 14. Batistatou A, Scopa CD, Ravazoula P, Nakanishi Y, Peschos D, Agnantis cancer cells by AKT activation. Cancer Sci. (2012) 103:1280–9. NJ, et al. Involvement of dysadherin and E-cadherin in the development of doi: 10.1111/j.1349-7006.2012.02302.x testicular tumours. Br J Cancer. (2005) 93:1382–7. doi: 10.1038/sj.bjc.6602880 28. Berx G, van Roy F. Involvement of members of the cadherin 15. Kyzas PA, Stefanou D, Batistatou A, Agnantis NJ, Nakanishi Y, Hirohashi superfamily in cancer. Cold Spring Harb Perspect Biol. (2009) 1:a003129. S, et al. Dysadherin expression in head and neck squamous cell carcinoma: doi: 10.1101/cshperspect.a003129 Frontiers in Oncology | www.frontiersin.org 12 December 2019 | Volume 9 | Article 1306 Besso et al. FXYD5/Dysadherin and Endometrial Cancer 29. Mhawech-Fauceglia P, Kesterson J, Wang D, Akers S, DuPont NC, Clark K, 39. Yamazawa K, Seki K, Matsui H, Sekiya S. Significance of perivascular lym et al. Expression and clinical significance of the transforming growth factor- phocytic infiltrates in endometrial carcinoma. Cancer. (2001) 91:1777–84. β signalling pathway in endometrial cancer. Histopathology. (2011) 59:63–72. doi: 10.1002/1097-0142(20010501)91:9<1777::aid-cncr1196>3.0.co;2-# doi: 10.1111/j.1365-2559.2011.03892.x 40. Chang WC, Huang SC, Torng PL, Chang DY, Hsu WC, Chiou SH, et al. 30. Bokhari AA, Lee LR, Raboteau D, Hamilton CA, Maxwell GL, Rodriguez Expression of inhibitory natural killer receptors on tumor-infiltrating CD8+ GC, et al. Progesterone inhibits endometrial cancer invasiveness by T lymphocyte lineage in human endometrial carcinoma. Int J Gynecol Cancer. inhibiting the TGFβ pathway. Cancer Prev Res (Phila). (2014) 7:1045–55. (2005) 15:1073–80. doi: 10.1111/j.1525-1438.2005.00264.x doi: 10.1158/1940-6207.CAPR-14-0054 41. Ohno Y, Ohno S, Suzuki N, Kamei T, Inagawa H, Soma G, et al. Role 31. Muinelo-Romay L, Colas E, Barbazan J, Alonso-Alconada L, Alonso-Nocelo of cyclooxygenase-2 in immunomodulation and prognosis of endometrial M, Bouso M, et al. High-risk endometrial carcinoma profiling identifies TGF- carcinoma. Int J Cancer. (2005) 114:696–701. doi: 10.1002/ijc.20777 β1 as a key factor in the initiation of tumor invasion. Mol Cancer Ther. (2011) 42. Ohno S, Ohno Y, Nakada H, Suzuki N, Soma G, Inoue M. Expression 10:1357–66. doi: 10.1158/1535-7163.MCT-10-1019 of Tn and sialyl-Tn antigens in endometrial cancer: its relationship 32. Pallares J, Martínez-Guitarte JL, Dolcet X, Llobet D, Rue M, Palacios J, et al. with tumor-produced cyclooxygenase-2, tumor-infiltrated lymphocytes and Abnormalities in the NF-kappaB family and related proteins in endometrial patient prognosis. Anticancer Res. (2006) 26:4047–53. carcinoma. J Pathol. (2004) 204:569–77. doi: 10.1002/path.1666 43. Yang X, Dong Y, Zhao J, Sun H, Deng Y, Fan J, et al. Increased expression 33. Lee CH, Jeon YT, Kim SH, Song YS. NF-κB as a potential molecular target for of human macrophage metalloelastase (MMP-12) is associated with the cancer therapy. Biofactors. (2007) 29:19–35. doi: 10.1002/biof.5520290103 invasion of endometrial adenocarcinoma. Pathol Res Pract. (2007) 203:499– 34. Batistatou A, Peschos D, Tsanou H, Charalabopoulos A, Nakanishi Y, 505. doi: 10.1016/j.prp.2007.03.008 Hirohashi S, et al. In breast carcinoma dysadherin expression is correlated 44. Eo WK, Kwon S, Koh SB, Kim MJ, Ji YI, Lee JY, et al. The lymphocyte- with invasiveness but not with E-cadherin. Br J Cancer. (2007) 96:1404–8. monocyte ratio predicts patient survival and aggressiveness of endometrial doi: 10.1038/sj.bjc.6603743 cancer. J Cancer. (2016) 7:538–45. doi: 10.7150/jca.14206 35. Shimamura T, Sakamoto M, Ino Y, Sato Y, Shimada K, Kosuge T, et al. 45. Machida H, De Zoysa MY, Takiuchi T, Hom MS, Tierney KE, Matsuo Dysadherin overexpression in pancreatic ductal adenocarcinoma reflects K. Significance of monocyte counts at recurrence on survival outcome of tumor aggressiveness: relationship to e-cadherin expression. J Clin Oncol. women with endometrial cancer. Int J Gynecol Cancer. (2017) 27:302–10. (2003) 21:659–67. doi: 10.1200/JCO.2003.06.179 doi: 10.1097/IGC.0000000000000865 36. Montserrat N, Mozos A, Llobet D, Dolcet X, Pons C, de Herreros AG, et al. Epithelial to mesenchymal transition in early stage Conflict of Interest: The authors declare that the research was conducted in the endometrioid endometrial carcinoma. Hum Pathol. (2012) 43:632–43. absence of any commercial or financial relationships that could be construed as a doi: 10.1016/j.humpath.2011.06.021 potential conflict of interest. 37. Salvesen HB, Akslen LA. Significance of tumour-associated macrophages, vascular endothelial growth factor and thrombospondin- Copyright © 2019 Besso, Rosso, Lapyckyj, Moiola, Matos, Mercogliano, Schillaci, 1 expression for tumour angiogenesis and prognosis in Reventos, Colas, Gil-Moreno, Wernicke, Orti and Vazquez-Levin. This is an open- endometrial carcinomas. Int J Cancer. (1999) 84:538–43. access article distributed under the terms of the Creative Commons Attribution doi: 10.1002/(SICI)1097-0215(19991022)84:5<538::AID-IJC17>3.0.CO;2-B License (CC BY). The use, distribution or reproduction in other forums is permitted, 38. Ohno S, Ohno Y, Suzuki N, Kamei T, Koike K, Inagawa H, et al. provided the original author(s) and the copyright owner(s) are credited and that the Correlation of histological localization of tumor-associated macrophages original publication in this journal is cited, in accordance with accepted academic with clinicopathological features in endometrial cancer. Anticancer Res. practice. No use, distribution or reproduction is permitted which does not comply (2004) 24:3335–42. with these terms. Frontiers in Oncology | www.frontiersin.org 13 December 2019 | Volume 9 | Article 1306