FXYD5/DYSADHERIN Expression in Prostate Cancer. Association with EMT Markers and Clinico- Pathological Parameters

Abstract

Purpose: FXYD5/Dysadherin (FXYD5/Dys) is associated with cancer progression and aggressiveness. This study evaluated FXYD5/Dysadherin (FXYD5/Dys) expression in Prostate cancer (PCa) cell lines and tumors, and its association with Epithelial to Mesenchymal Transition (EMT) markers and Clinico-pathological parameters.
Methods: Expression of FXYD5/Dys and EMT markers was analysed in PCa cell cultures, PC3-derived mice tumors and human tumors using GEO and TCGA datasets, and in urine cell sediment of Benign Prostatic Hyperplasia (BPH) and PCa patients.
Results: Medium/high metastatic potential cell lines (C4-2B, DU-145, PC3) expressed higher (P<0.05) levels of FXYD5/Dys mRNA and protein, lower (P<0.05) E-cadherin, and higher (P<0.05) Zeb1, Slug, N-cadherin, and Vimentin mRNA levels than low metastatic potential cell lines (LNCaP, 22Rv1, C4-2). PC3 cells treatment with FXYD5/Dys siRNA resulted in reversed expression of EMT markers and higher (P<0.05) cell-cell adhesion. Gene expression analysis done on a dataset from an in vivo PCa model of subcutaneous tumors with low (LMP) and High (HMP) Metastatic Potential (GSE7930) revealed higher (P<0.05) FXYD5/Dys mRNA levels and an EMT profile in the HMP tumors. PCa TCGA dataset analysis revealed a higher (P<0.05) FXYD5/Dys expression and a lower (P<0.05) free-disease survival, higher (P<0.05) Gleason, Stage, and Grade Group, and an EMT profile. A multiple regression analysis identified FXYD5/Dys as a high tumour Stage predictor. Urine sediment FXYD5/Dys mRNA levels were higher (P<0.05) in PCa than in BPH-diagnosed men.
Conclusions: FXYD5/Dys high expression was associated with an EMT profile, disease progression, and aggressiveness in PCa.

Keywords:Male reproductive health, Prostate cancer, EMT, FXYD5/Dysadherin, Biomarkers, Urine, E-cadherin

Introduction

Prostate cancer (PCa) is the second most frequently diagnosed cancer in men, with over 1.4 million cases reported in 2022 [1]. While a large proportion of PCa is diagnosed in early stages and have slow evolution, its course varies from indolent to metastatic [2]. Understanding its molecular basis is highly relevant to identify entities associated with tumor progression and aggressiveness. Expression of the cell-cell adhesion protein epithelial cadherin (Ecadherin) is highly relevant in maintaining prostate epithelial integrity, and preventing carcinogenesis [3]. A decrease in E-cadherin expression and function(s) is a key event in Epithelial to Mesenchymal Transition (EMT), a process that stands out among the molecular mechanisms involved in tumor progression and metastasis in numerous solid tumors [4]. The E-cadherin decrease triggers several changes, among them the gain of mesenchymal markers such as Vimentin [5]. Also, an increase in Neural Cadherin (N-cadherin) [6], an event regulated by transcriptional repressors, i.e. Twist, Snail, Zeb1 and Zeb2 that elicit EMT and negatively modulate E-cadherin expression [7]. The EMT process results in the acquisition of a mesenchymal phenotype and the development of a progressive motile and invasive behaviour [4]. In PCa, several studies have identified a decrease in E-cadherin expression and other EMT-related changes, and highlighted the relevance of the EMT process in tumor progression, increased stemness properties, drug resistance, and metastasis [8,9].

Among E-cadherin expression negative modulators, FXYD5/ Dysadherin (FXYD5/Dys) has been reported in several tumors. FXYD5/Dys, a member of the FXYD protein family that contain a conserved F-X-Y-D motif in the transmembrane domain, is a type I transmembrane glycoprotein initially identified as a cancer-related protein and associated with decreased E-cadherin expression and reduced cell-cell adhesion [10]. Increased expression of FXYD5/ Dys has been associated with tumor progression and poor outcome of cancer patients in numerous tumors [11,12]. The FXYD5/Dys negative impact on cell-cell adhesion is associated to steric hindrance, actin cytoskeleton competition, and post-transcriptional negative regulation of E-cadherin [13]. Also, an increased expression of FXYD5/Dys has been associated to activation of TGF-β and NF-κB pathways [14], and with expression and secretion of Chemokine (CC Motif)-Ligand 2 or CCL2, a cytokine associated with tumorigenesis and metastasis [15], and previously reported as a PCa progression marker [16,17]. The expression of FXYD5/Dys and its association with tumor progression and aggressiveness in PCa has not been reported.

The expression of FXYD5/Dys and EMT markers was evaluated in six PCa cell lines, in datasets of low and high aggressiveness PCa tumors developed in mice from PC3 cells, and in a large cohort of PCa patients from The Cancer Genome Atlas (TCGA) with different clinicopathological parameters (Gleason score, Tumor Stage, Grade Group (GG) and Disease-Free Survival). Moreover, FXYD5/Dys expression was modulated in PC3 cells to evaluate its impact upon cell behaviour. Finally, FXYD5/Dys mRNA levels in the urine cell sediment of PCa and BPH patients were determined.

Materials and Methods

Bioinformatics

Messenger RNA expression levels were determined from a gene expression dataset (GSE7930; GEO) of tumors developed in mice by injection of PC3 cells, as previously reported [18]. In addition, gene expression analysis of 497 PCa samples of the Prostate Cancer TCGA (PRAD version 16.08.2016; TCGA), downloaded from the UCSC Xena web tool, was done.

Patient Urine Samples

Urine samples were obtained from patients subjected to prostate biopsy because of increased serum PSA levels (>4.0 ng/ml) and/or an abnormal diagnostic Digital Rectal Exam, attending the Department of Urology of the Vall d’Hebron Hospital in Barcelona, Spain. Eight samples from patients diagnosed with Bening Prostatic Hyperplasia (BPH) and 18 with PCa (4 Gleason 6; 12 Gleason 7; 1 Gleason 8 and 1 Gleason 9-10) were included in the evaluation. The Vall d’Hebron Hospital IRB approved the protocol; all patients signed a written informed consent. Procedures for urine sample collection and processing were previously reported [19].

Antibodies

Primary antibodies: Anti-E-cadherin (610181 (BD), HECD-1 (Thermo), H-108 (Santa Cruz Biotec, SCBT)); anti-N-cadherin (H- 63, SCBT); anti-Vimentin (Clone V9, Dako); anti-β-catenin (610153, BD); anti-GAPDH (14C10, Cell Signaling). Anti FXYD5/Dys NCC-M53 (kindly provided by Dr. S. Hirohashi) and L-15 (sc-30603, goat polyclonal, towards the N-terminal end; SCBT), and FL-178 (sc-98246, rabbit polyclonal, towards residues 1-178; SCBT). In control experiments, primary antibodies were replaced by purified mouse, rabbit or goat IgGs (Sigma-Aldrich).

Secondary Antibodies: Cy3-labelled anti-mouse or anti-rabbit (Sigma- Aldrich) and FITC-labelled anti-mouse (Vector) for immunocytochemistry protocols; anti-mouse, anti-rabbit or anti-goat (Vector; Sigma-Aldrich) IgGs coupled to HRP for Western immunoblotting procedures.

Cell Cultures

Six PCa cell lines selected as in vitro models (LNCaP, 22Rv1, C4- 2, C4-2B, DU-145 and PC3) were cultured under standard conditions (ATCC; Manassas, VI, USA). Cell lines were cultured in RPMI 1640 culture medium in the presence of 10% heat-inactivated fetal bovine serum, 100 UI/mL penicillin and 100 mg/mL streptomycin, and incubated at 37°C in a humidified atmosphere of 5% CO2 in air (standard conditions). Low-passage cell lines were used in all cases; in addition, quality control tests were done by monitoring cell morphology and growth rates, as well as mycoplasma detection with Hoechst 33342 (Sigma Aldrich). In a set of experiments, PC3 cells were treated with 10 ng/mL rec-TGF-β1 (TONBO Biosciences) [14].

Transient Transfection Assay

PC3 cells transient transfection with a specific commercial siRNA was done to knockdown endogenous FXYD5/Dys expression, as previously reported [14]. Scramble (37007, SCBT) and specific (45745, SCBT) siRNAs (100 pmoles/mL) were transfected into PC3 cells according to the manufacturer’s instructions.

Hanging Drop Assay

To promote cellular aggregates formation, PCa cells were cultured under independent-anchorage conditions, as previously reported [20].

Migration and Invasion Assays

Both procedures were done on PC3 siRNA control and FXYD5/ Dys cells as previously described [14].

Protein Extraction/Western Immunoblotting Protocols

Cell lines protein extracts were prepared and processed for Western immunoblotting as previously reported [20]. Membranes were developed with specific antibodies towards E-cadherin (610181, 0.25 μg/mL), N-cadherin (2 μg/mL), Vimentin (2 μg/mL), FXYD5/Dys (NCC-M53, 1 μg/mL) and GAPDH (2 μg/mL). The use of the antibodies was previously validated [20]. Immunodetection was specific, since no signal was found when the primary antibody was replaced with normal IgG (data not shown).

Immunocytochemistry and Immunohistochemistry

Cells grown on glass coverslips were subjected to immunocytochemical analysis to assess E-cadherin (HECD-1 6 μg/mL), β-catenin (4 μg/mL), N-cadherin (2 μg/mL), Vimentin (2 μg/mL) and FXYD5/Dys (FL-178 4 μg/mL) immunolocalization [20]. Nuclear cell staining was done with Hoechst 33342. To assess filamentous actin (F-actin), cells were stained with AlexaFluor 488 Phalloidin (Thermo). Immunohistochemical analysis of FXYD5/Dys (L-15 4 μg/mL) and E-cadherin (H-108 2 μg/mL) was done in tissue sections of subcutaneous PC3-cell tumors, using a commercial kit (immunoCruz ABC, SCBT).

RNA Extraction and RT-PCR Procedures

TRIzol^TM (Thermo) RNA isolation, cDNA synthesis and qRT-PCR analysis was performed as previously reported [20]. Negative controls (omitting RNA or reverse transcriptase) were included. RNA from urine cell sediments recovered after prostatic massage was purified and FXYD5/Dys expression levels using GAPDH as housekeeping control gene were assessed.

Statistical Analysis

Experiments were carried out at least three times. Data is presented as mean ± Standard Error of the Mean (SEM). Cell lines studies were analysed with Kruskal Wallis test followed by orthogonal contrast analysis to compare results between low (LNCaP/22Rv1/ C4-2) and high (C4-2B/DU-145/PC3) aggressiveness PCa cells. The Mann-Whitney test was applied to compare two groups’ results, and the Kruskal Wallis test followed by the Dunn test for multiple comparisons was applied for do three group analyses. The Spearman correlation coefficient (ρ, rho) was used to perform correlation analyses for several markers. FXYD5/Dys expression and patient Recurrence Free Survival (RFS) were assessed by means of the Kaplan Meier method. FXYD5/Dys expression cut-off value (low/ high) was determined using the web http://molpath.charite.de/ cutoff. A multiple logistic regression analysis to identify PCa pathologic Stage predictor variables was done. Analyses were performed on GraphPad Prism version 5 (GraphPad Software) and R (version 3.4.4).

Results

Expression of FXYD5/Dys and EMT markers in PCa cell lines

First, expression of FXYD5/Dys and EMT markers in six PCa cell lines, grouped in low (LNCaP, 22Rv1, C4-2) and high (C4-2B, DU- 145, PC3) aggressiveness (based on their metastatic potential) was evaluated. qRT- PCR analysis revealed differences among cells, finding lower E-cadherin mRNA levels in more aggressive cells (P<0.01 between subgroups; Figure 1.A, left). In agreement with these findings, lower levels of the 120 KDa full-length cell-cell adhesion protein were found in C4-2B, DU-145 and PC3 cells (P<0.001; Figure 1.A, right). The 38-KDa E-cadherin CTF1 digestion fragment was only detected in the low aggressiveness PCa cell group. Expression analysis of transcriptional repressors revealed differences among cells in Slug, Zeb1 and Zeb2 mRNA levels (P<0.001), with higher levels in the highly aggressive cells (P<0.001) (Figure 1.B). No significant changes in Twist and Snail were found (data not shown). N-cadherin and Vimentin mRNA levels were higher (P<0.001) in aggressive cells, and proteins were only immunodetected in these cells (Figure 1.C). ESRP1 and ESRP2, RNA binding proteins responsible for a splicing regulated program essential for epithelial cells morphology and function, depicted lower expression levels in highly aggressive cells (ESRP1: P<0.05, ESRP2: P<0.001) (Figure 1.D).

Figure 1a:

Figure 1. A. Left. E-cadherin mRNA expression assessed by quantitative real time PCR (qRT-PCR). 2^-ΔCt values were calculated as described in Materials and Methods, and GAPDH was selected as housekeeping gene. Statistical analysis done using Kruskal Wallis test, followed by orthogonal contrasts done to compare mRNA expression between low (LNCaP, 22Rv1, C4-2) and high (C4-2B, DU-145, PC3) aggressiveness PCa cell lines. **P<0.01. Right: Immunodetection of E-cadherin protein after SDS-PAGE and Western immunoblotting analysis. (Anti E-cadherin 610181: 0.25 μg/mL). GAPDH selected as loading control (Anti GAPDH 14C10: 2 μg/mL). Numbers represent results of densitometric analysis of E-cadherin signal; intensity determined using the Image J program, and normalized using loading control signal intensity. B. mRNA expression analysis of Slug, Zeb1 and Zeb2 transcriptional represors assessed using qRT-PCR (Kruskal Wallis test, ***P<0.001). C. Top: qRT-PCR mRNA expression analysis of N-cadherin (left) and Vimentin (right) mRNA level by qRT-PCR. ***P<0.001. Bottom: Inmunodetection of N-cadherin and Vimentin proteins after SDS-PAGE and Western immunoblotting analysis (Anti N-cadherin H-63: 2 μg/mL; Anti Vimentin Clon V9: 2 μg/mL). Loading control done with GAPDH. (see A.). D. qRT- PCR mRNA expression analysis of splicing factors ESRP1 and ESRP2 (*P<0.05; **P<0.01).

Evaluation of FXYD5/Dys mRNA expression revealed higher levels in C4-2B, DU-145 and PC3 cells (P<0.001) (Figure 1.E, left). While the FXYD5/Dys protein was only detected in these cells, a ~55 KDa protein form was immunodetected in C4-2B and DU-145 cells and a ~50 KDa form in PC3 cells (Figure 1.E, right), probably caused by differences in the O-linked glycosidic branches. Immunolocalization analysis of FXYD5/Dys in PC3 cells revealed a strong signal in the cytoplasm and plasma membrane, suggesting an active protein expression. Cells also showed a low E- cadherin cell signal, and a strong N-cadherin and Vimentin signal (Figure 1.F). Staining was specific, with a lack of signal in cells incubated with control IgG. PC3 cell tumors generated in nude mice showed a strong specific FXYD5/Dys signal and low E-cadherin levels, consistent with findings in cells grown in monolayers (Figure 1.G).

Figure 1b:

Figure 1. E. Left: qRT-PCR mRNA expression analysis of FXYD5/ Dys (***P<0.0001). Right: Inmunodetection of FXYD5/Dys protein after SDS-PAGE and Western immunoblotting analysis (Anti FXYD5/Dys NCC-M53: 1 μg/mL). Loading control was done with GAPDH (see A). F. Fluorescence immunocytochemistry and confocal laser microscopy analysis of FXYD5/Dys in PC3 cells (Anti FXYD5/ Dys FL-178, 4 μg/mL), adherent complex (E-cadherin, Beta-catenin and filamentous (F) actin) and mesenchymal markers (N-cadherin and Vimentin) in PC3 cells. Negative controls are included; cell nuclei visualized using Hoechst 33342. Bar: 50 μm (20x magnification) and 20 μm (40x magnification). G. Colorimetric immunohistochemistry assay of FXYD5/Dys (Anti FXYD5/Dys L-15, 4 μg/mL) and E-cadherin (H-108, 2 μg/mL) in paraffin sections of tumors generated by injection of PC3 in nude mice. Contrast staining was performed using hematoxylin. Controls were done in which the primary antibody was replaced by IgG (goat for FXYD5/Dys detection; rabbit for E-cadherin detection) added at the same concentration. Bar: 50 μm (20x magnification).

Next, a PC3 cell-silencing protocol was done, leading to 80% decrease in FXYD5/Dys mRNA levels and undetectable protein levels (Figure 1.H). Since FXYD5/Dys pro-metastatic role has been attributed to decreased cell-cell adhesion, the hanging drop method was applied to treated cells, and cell aggregates (area >20,000 (pixels) 2) were estimated, finding that siRNA-FXYD5/Dys cells formed larger aggregates (Figure 1.I). Evaluation of cell migration (Figure 1.J) and invasion (Figure 1.K) revealed reduced (P<0.01) properties in PC3 siRNA-FXYD5/Dys cells. In agreement with these observations, mRNA levels of two invasion/metastasis-associated molecular markers, MMP9 and EZH2 (Figure 1.L) were lower (P<0.05) in PC3 siRNA FXYD5/Dys cells.

Figure 1. H. Left: Quantitative analysis of FXYD5/Dys mRNA expression by qRT-PCT in PC3 cells treated with Scramble siRNA (Control) and FXYD5/Dys siRNA (siRNA FXYD5/Dys) (Mann Whitney test, **P<0.01). Right: Immunodetection of FXYD5/Dys protein after SDS-PAGE and Western immunoblotting analysis in PC3 siRNA Control and PC3 siRNA FXYD5/Dys cells (Anti FXYD5/ Dys NCC-M53: 1 μg/mL). Loading control was done with GAPDH (Anti GAPDH 14C10: 2 μg/mL). I. Evaluation of the cell-cell adhesion capacity of PC3 siRNA Control and PC3 siRNA FXYD5/Dys cells by means of the Hanging Drop method. Representative images of cells grown in suspension for 24 h are shown. J. Evaluation of the migration behavior of PC3 siRNA Control and PC3 siRNA FXYD5/ Dys cells by the Transwell™ Migration test. Left: The number of cells that were able to migrate through the transwell after 8 h incubation was measured and plotted (Mann Whitney test, **P<0.01). Right: A representative image is shown. K. Left: Evaluation of the invasive capacity of PC3 siRNA Control and PC3 siRNA FXYD5/Dys cells by the Transwell™ Invasion Test, done in the presence of an extracellular matrix. The number of cells that were able to invade was measured and plotted (Mann Whitney test, **P<0.01). Right: A representative image is shown. L. Quantitative analysis of the EZH2 (Left) and MMP9 (right) mRNA expression by qRT-PCR in PC3 cells transiently transfected with a scramble siRNA (siRNA Control) or with an FXYD5/Dys specific siRNA (siRNA FXYD5/Dys) (Mann Whitney test, *P<0.05).

Figure 1:Expression analysis of EMT markers and FXYD5/Dys in established human PCa cell lines.

Gene Expression Analysis of PC3-Derived Tumors with Low and High Metastatic Potential

The association between FXYD5/Dys expression and EMT was assessed in a gene expression dataset from an in vivo PCa progression model of PC3 cells-derived mouse subcutaneous tumors with low (LMP) and high (HMP) metastatic potential [18]. In line with results shown above, higher (P<0.05) FXYD5/Dys and lower (P<0.05) E-cadherin expression was determined in HMP tumors. Neither N-cadherin nor Vimentin mRNA levels show difference between LMP and HMP tumors. Similar expression levels were determined for ESRP1and ESRP2 in LMP and HMP tumors. On the other hand, increased (P<0.05) levels of Twist, Snail and Zeb1 transcripts were found in HMP PC3-tumors, but no differences were found for Zeb2 and Slug (Table 1).

Table 1:Expression of FXYD5/Dys and EMT markers expression in an in vivo PCa progression model of PC3 cells-derived mouse subcutaneous tumors with low (LMP) and high (HMP) metastatic potential.

Note*: Results are expressed as mean ± SEM. Statistical analysis done by means of Mann Whitney test.

PCa Global Expression Analysis of FXYD5/Dys and its Associations with EMT Markers

Next, to evaluate FXYD5/Dys expression in association with clinicopathological parameters, specifically Gleason score, Tumor Stage, and GG, TCGA RNAseq gene expression dataset was analysed. Results of E-cadherin expression levels from the dataset also were retrieved. As result, higher (P=0.0270) FXYD5/Dys and lower (P<0.0001) E-cadherin mRNA levels were found in Gleason 8-10 tumors when compared to Gleason 6-7 ones. Moreover, higher (P=0.0015) FXYD5/Dys and lower E- cadherin (P=0.0076) transcript levels were found in Stage T3-T4 tumors compared to Stage T2 tumors. Lastly, GG5 tumors depicted higher (P=0.0117) FXYD5/Dys levels than GG2-4 tumors. E-cadherin expression was lower (P<0.0001) among GG tumors (GG1 vs GG5; GG2-4 vs GG5) (Table 2.A). To evaluate the association between FXYD5/Dys and the tumor EMT profile, TCGA samples were grouped based on Low or High FXYD5/Dys expression and EMT markers evaluated in each group. High FXYD5/Dys expression tumors showed low E-cadherin levels, and depicted a negative correlation (r=-0.4556; P<0.00001; Figure 2.A). In addition, high FXYD5/Dys samples depicted lower levels of ESRP1 and ESRP2 and higher levels of Snail, Slug, Zeb1 and Zeb2, as well as higher levels of N-cadherin and Vimentin (P<0.0001) in the high DXYD5/Dys subgroup (Table 2B). Next, a PCa survival analysis (Kaplan Meier curves and Log-Rank test) revealed a higher Recurrence Free Survival (RFS) in PCa patients with low FXYD5/Dys mRNA tumor levels than in patients with high FXYD5/Dys levels (HR: 0.4981; P=0.03115); 46 diseases were listed in 329 (0.1398) cases of high FXYD5/Dys expression, and 6 of 107 (0.056) showing low FXYD5/Dys levels (Figure 2.B).

Figure 2. A. Association between FXYD5/Dys and E-cadherin mRNA expression in PCa. Correlation analysis of E-cadherin and FXYD5/Dys mRNA levels (r=-0.4556, ***P<0.0001). Data was extracted from TCGA, cohort Prostate Cancer TCGA (PRAD) dataset. B. Kaplan-Meier survival curves were constructed to evaluate the impact of FXYD5/Dys, upon PCa patient survival (*P <0.05, Log Rank test). C. Detection of FXYD5/Dys mRNA in urine cellular sediment from Bening Prostatic Hyperplasia (BPH) and PCa samples. Quantitative analysis of FXYD5/Dys mRNA levels were assessed in 8 BPH and 18 PCa urine cellular sediment samples by qRT-PCR (*P<0.05; unpaired t test).

Figure 2:FXYD5/Dys, E-cadherin, and patient survival.

Table 2:FXYD5/Dys and E-cadherin expression in PCa samples. Gleason, Stage and GG.

FXYD5/Dys, Cytokines and Stem Cell Markers

The Transforming Growth Factor-beta 1 (TGF-β1) was previously shown to induce EMT-related changes in PCa [21]. Moreover, studies from our group showed an association between FXYD5/ Dys and TGF-β1 expression [14]. Expression of TGF-β1 was higher (Low=-0.8219±0.05630, High=0.2520±0.03407; P<0.0001) in the TCGA high FXYD5/Dys subgroup (Figure 3.A). In addition, a positive correlation (r=0.7718, P<0.0001) was found between FXYD5/ Dys and TGF-β1 expression in the TCGA cohort (Figure 3.B). Moreover, PC3 cells treatment with TGF-β1 resulted in a ~40% increase in FXYD5/Dys protein expression (P<0.05; Figure 3.C).

Figure 3:FXYD5/Dys, TGF-β1 and CCL2.

FXYD5/Dys expression has been related with increased metastatic capacity [10-12;22]. The pro-metastatic role may be associated both to its antiadhesive role and to the induction of the expression and secretion of the CC-Motif-Chemokine Ligand 2, CCL2 [23]. Assessment of CCL2 expression in Low and High FXYD5/Dys tumors revealed a higher expression in the latter (Low=-1.069±0.1449, High=0.3357±0.07753; P<0.0001) Figure 3.D, left). In addition, a positive correlation between CCL2 and FXYD5/Dys mRNA levels was found (r=0.4724, p<0.0001; Figure 3.D, right). Since FXYD5/Dys expression also promotes a stem cell like phenotype [24], expression of Oct4, Nanog, CD133 and CD44 markers was analysed, finding higher (P<0.0001) levels of CD133 and Oct4 in FXYD5/Dys high- expression tumors (CD133: Low=- 1.291±0.1953, High=0.1598±0.1048; Oct4: Low=-0.5891±0.1077, High=0.1642±0.0714).

Figure 3. A. Quantitative analysis of FXYD5/Dys mRNA expression and its association with expression of TGF-β1 done with a dataset from TCGA. Samples were grouped based on FXYD5/Dys mRNA expression levels (Low and High), and subjected to analysis for each gene (Mann Whitney test; ***P<0.0001). B. Correlation analysis of FXYD5/Dys mRNA levels and TGF-β1 (r= 0.7718, ***P<0.0001). Data was extracted from TCGA. C. PC3 cells were treated with TGF-β1 or vehicle (Control) as previously reported. Left: Immunodetection of FXYD5/Dys (Anti FXYD5/Dys NCC-M53: 1 μg/mL). Loading control was done with GAPDH (Anti GAPDH 14C10: 2 μg/mL). Numbers in image are the result of densitometric analysis of FXYD5/Dys signal; intensity was determined using the Image J program, and normalized using the signal intensity of the loading control (*P<0.05). D. Quantitative analysis of FXYD5/Dys mRNA levels and association with CCL2 expression. Samples were grouped based on FXYD5/Dys mRNA expression levels (Low and High) (Mann Whitney test; ***P<0.0001). E. Correlation analysis of CCCL2 and FXYD5/Dys mRNA levels (r= 0.4724, ***P<0.0001). Data was extracted from TCGA.

Regression Analysis and PCa FXYD5/Dys Predictive Value

A correlation analysis done with FXYD5/Dys and 20 PCa genes evaluated in this study revealed the highest associations between FXYD5/Dys, Vimentin and TGF-β1. A logistic regression analysis done with 21 genes as independent variables and pathologic Tumor Stage (dichotomized in Low (T2a-c, T3a) and High (T3b, T4) Stage), identified FXYD5/Dys as a predictive variable in 2 of 3 validated models of 200 random samples resampled from the TCGA dataset. A 74% sensitivity and 63% specificity were determined.

FXYD5/Dys mRNA quantification in PCa patient urine

The identification of biomarkers in biological fluids for cancer diagnosis and prognosis is of great interest, since these molecules may contribute to patient management in a simple non-invasive fashion. In particular, urine is a source of cells from the prostate when collected after a prostate massage [25]. FXYD5/Dys transcript analysis was carried out by means of real-time PCR in cDNA from PCa patients urine RNA cell sediment (n=18) or BPH (n=8), collected after prostate massage. Q-PCR analysis of the aforementioned samples revealed a higher (P=0.0230) expression of FXYD5/ Dys in the cell pellet of tumor samples compared to those in BPH (BPH=-0.02513±0.00499, PCa=0.05499±0.00910) (Figure 2.C.).

Discussion

The lack of effective markers to distinguish PCa aggressive tumors from those without clinical relevance at the time of diagnosis, highlights the need to delve into its molecular mechanisms. E-cadherin is a key cell-cell adhesion molecule to ensure prostate epithelial integrity [3]. In association with E-cadherin loss as an EMT hallmark, previous studies reported evidence on EMT in PCa [8,9]. On the other hand, FXYD5/Dys has been reported as a negative modulator of E-cadherin expression and function (s), and a marker of aggressiveness in several solid tumors [10,11]. Despite its possible relevance in PCa progression and aggressiveness, no reports on FXYD5/Dys expression in PCa have been published until now. In the present report, evaluation of FXYD5/Dys expression in PCa cell lines revealed higher levels of both mRNA and protein in most aggressive cells, as well as in PC3-derived tumors that depicted lowest E-cadherin levels. High FXYD5/Dys levels were associated with high expression of E- cadherin transcriptional represors Snail, Slug, Zeb1, Zeb2, as well as mesenchymal markers N-cadherin and Vimentin, revealing a strong association between increased FXYD5/ Dys expression and a tumor EMT profile, in agreement with studies in other tissues from our group [14,20] and others [26]. Transcript expression analysis of TCGA samples with worst clinicopathological features revealed higher FXYD5/Dys levels in tumors with high Gleason score, high Tumor Stage and high GG. Moreover, a negative association between FXYD5/Dys and E-cadherin expression was found, and high FXYD5/Dys expression patients had lower RFS, in line with reports in other tumors [11,12].

FXYD5/Dys silenced-PC3 cells had higher cell-cell adhesion and lower cell migration/invasion, highlighting its association with cell tumor progression, as shown in hepatocarcinoma, endometrial and recently colon cancer [7,14,27]. Moreover, FXYD5/ Dys silenced-PC3 cells showed a decrease in MMP9 expression, in line with recent findings in colorectal cancer [28]. TGF-β1-treated cells showed higher FXYD5/Dys expression, and high FXYD5/Dys tumors depicted higher TGF-β1 levels. Findings are in line with studies in endometrial and ovarian cancer [14,26]. FXYD5/Dys was previously proposed to promote a stem cell phenotype. In line with these findings, a positive association was found between FXYD5/ Dys, CD133 and Oct4. Moreover, FXYD5/Dys expression was previously associated with CCL2 expression and secretion [14,15]. PCa progression and aggressiveness are associated with the CCL2 cytokine [16]; interestingly, CCL2 showed high expression levels in TCGA dataset samples with elevated FXYD5/Dys expression. Since FXYD5/Dys was shown to induce CCL2 expression by increasing TNFα receptors, this would suggest an inflammatory environment, as reported in other pathologies. There is growing evidence of chronic inflammation contribution to PCa development/progression [29].

Finally, to initiate studies toward FXYD5/Dys evaluation in patients with PCa using a non-invasive method, urine sediment analysis was subjected to qRT-PCR to determine whether levels detected differ from those determined in BPH patients. As result, higher FXYD5/Dys mRNA expression was found in PCa when compared to those diagnosed with BPH. A large cohort will help confirming these findings, and show its potential use in the active surveillance management of PCa patients. This the first report on FXYD5/Dys expression, its relationship with EMT markers and clinico-pathological parameters of tumor progression/aggressiveness in PCa.

Conflict of Interest

Authors have no conflicts of interest to declare. Human samples were obtained from Vall d´ Hebron Hospital (CEI C), PRCAG 144/2011. Animal experiments were conducted following national guidelines and relevant national laws on animals’ protection. Protocol for PC3 mice tumors was CICUAL 019/2017.

Authors’ Contributions

MVM and MVL were involved in study conception and design.
MVM, RB, MJB, MR, JM, JR, MR, CM, and MVL participated in sample/data collection.
MVM, RB, and MVL participated in data analysis.
MVL drafted the manuscript, and all authors read and accepted the final version.
MVL was responsible of coordinating and supervising the entire project, and obtaining funds that supported this study.

Funding

Manuscript preparation was done with the support of a grant from Agencia Nacional de Promotion Scientific y Technology de Argentina [Grant PICTSU-1072], National Council for Scientific Research and Technicist (CONICET) [Grant PIP887], Instituto Nacional del Cancer [Grant INC 2016-2017], to MVL. Authors are thankful to Fundacion R. Baron and Fundación Williams for institutional (IBYME) support.

Acknowledgments

Authors want to thank Q.F. S. Méndez por her technical assistance. Authors sincerely thank Dr. A. De Siervi for facilitating validated cell lines to initiate studies. Also, authors thank Prof. Dr. S. Hirohashi, for providing NCC-M53 antibody.

References

1. Gandaglia G, Leni R, Bray F, Fleshner N, Freedland SJ, et al. (2021) Epidemiology and Prevention of Prostate Cancer. Eur Urol Oncol 4(6): 877-892.
2. Wilt TJ, Brawer MK, Barry MJ, Jones KM, Kwon Y, et al. (2009) The Prostate cancer Intervention Versus Observation Trial: VA/NCI/AHRQ Cooperative Studies Program #407 (PIVOT): design and baseline results of a randomized controlled trial comparing radical prostatectomy to watchful waiting for men with clinically localized prostate cancer. Contemp Clin Trials 30(1): 81-87.
3. Wang X, Dong B, Zhang K, Ji Z, Cheng C, et al. (2018) E-cadherin bridges cell polarity and spindle orientation to ensure prostate epithelial integrity and prevent carcinogenesis in vivo. PLoS Genet 14(8): e1007609.
4. Thiery JP (2002) Epithelial-mesenchymal transitions in tumour progression. Nat Rev Cancer 2(6): 442-254.
5. Usman S, Waseem NH, Nguyen TKN, Mohsin S, Jamal A, et al. (2021) Vimentin Is at the Heart of Epithelial Mesenchymal Transition (EMT) Mediated Metastasis. Cancers (Basel) 13(19): 4985.
6. Hazan RB, Qiao R, Keren R, Badano I, Suyama K (2004) Cadherin switch in tumor progression. Ann N Y Acad Sci 1014: 155-163.
7. Debnath P, Huirem RS, Dutta P, Palchaudhuri S (2022) Epithelial- mesenchymal transition and its transcription factors. Biosci Rep 42(1): BSR20211754.
8. Goncharov AP, Vashakidze N, Kharaishvili G (2024) Epithelial- Mesenchymal Transition: A Fundamental Cellular and Microenvironmental Process in Benign and Malignant Prostate Pathologies. Biomedicines 12(2): 418.
9. Gogola S, Rejzer M, Bahmad HF, Abou-Kheir W, Omarzai Y, et al. (2023) Epithelial-to-Mesenchymal Transition-Related Markers in Prostate Cancer: From Bench to Bedside. Cancers (Basel) 15(8): 2309.
10. Ino Y, Gotoh M, Sakamoto M, Tsukagoshi K, Hirohashi S (2002) Dysadherin, a cancer-associated cell membrane glycoprotein, down- regulates E-cadherin and promotes metastasis. Proc Natl Acad Sci U S A 99(1): 365-370.
11. Niinivirta A, Salo T, Åström P, Juurikka K, Risteli M (2022) Prognostic value of dysadherin in cancer: A systematic review and meta-analysis. Front Oncol 12: 945992.
12. Bai Y, Li L, Li J, Lu X (2024) Association analysis of FXYD5 with prognosis and immunological characteristics across pan-cancer. Heliyon 10(9): e30727.
13. Nam JS, Hirohashi S, Wakefield LM (2007) Dysadherin: a new player in cancer progression. Cancer Lett 255: 161-169.
14. Besso MJ, Rosso M, Lapyckyj L, Moiola CP, Matos ML, et al. (2019) FXYD5/Dysadherin, a Biomarker of Endometrial Cancer Myometrial Invasion and Aggressiveness: Its Relationship With TGF-β1 and NF-κB Pathways. Front Oncol 9: 1306.
15. Nam JS, Kang MJ, Suchar AM, Shimamura T, Kohn EA, et al. (2006) Chemokine (C-C motif) ligand 2 mediates the prometastatic effect of dysadherin in human breast cancer cells. Cancer Res 66: 7176-84.
16. Izumi K, Mizokami A, Lin HP, Ho HM, Iwamoto H, et al. (2016) Serum chemokine (CC motif) ligand 2 level as a diagnostic, predictive, and prognostic biomarker for prostate cancer. Oncotarget 7: 8389-8398.
17. Marsland M, Jiang CC, Faulkner S, Steigler A, McEwan K, et al. (2024) CCL2/CCR2 Expression in Locally Advanced Prostate Cancer and Patient Long-Term Outcome: 10-Year Results from the TROG 03.04 RADAR Trial. Cancers (Basel) 16(16): 2794.
18. Wong SY, Haack H, Kissil JL, Barry M, Bronson RT, et al. (2007) Protein 4.1B suppresses prostate cancer progression and metastasis. Proc Natl Acad Sci U S A 104(31): 12784-12789.
19. Rigau M, Morote J, Mir MC, Ballesteros C, Ortega I, Sanchez A, Colás E, Garcia M, Ruiz A, Abal M, Planas J, Reventós J, Doll A (2010) PSGR and PCA3 as biomarkers for the detection of prostate cancer in urine. Prostate 70(16): 1760-1767.
20. Matos ML, Lapyckyj L, Rosso M, Besso MJ, Mencucci MV, et al. (2017) Identification of a Novel Human E-Cadherin Splice Variant and Assessment of Its Effects Upon EMT-Related Events. J Cell Physiol 232(6): 1368-1386.
21. Al Azayzih A, Gao F, Somanath PR (2015) P21 activated kinase-1 mediates transforming growth factor β1-induced prostate cancer cell epithelial to mesenchymal transition. Biochim Biophys Acta 1853(5): 1229-1239.
22. Lubarski Gotliv I, Dey K, Kuznetsov Y, Kalchenco V, Asher C, et al. (2017) FXYD5 (dysadherin) may mediate metastatic progression through regulation of the β-Na+-K+-ATPase subunit in the 4T1 mouse breast cancer model. Am J Physiol Cell Physiol 313(1): C108-C117.
23. Adekoya TO, Richardson RM (2020) Cytokines and Chemokines as Mediators of Prostate Cancer Metastasis. Int J Mol Sci 21(12): 4449.
24. Park JR, Kim RJ, Lee YK, Kim SR, Roh KJ, et al. (2011) Dysadherin can enhance tumorigenesis by conferring properties of stem-like cells to hepatocellular carcinoma cells. J Hepatol 54(1): 122-131.
25. Skarmoutsos A, Skarmoutsos I, Katafigiotis I, Tataki E, Giagini A, et al. (2018). Detecting Novel Urine Biomarkers for the Early Diagnosis of Prostate Cancer: Platelet Derived Growth Factor-BB as a Possible New Target. Curr Urol 12(1): 13-19.
26. Bai Y, Li LD, Li J, Chen RF, Yu HL, et al. (2020) A FXYD5/TGF-β/SMAD positive feedback loop drives epithelial-to-mesenchymal transition and promotes tumor growth and metastasis in ovarian cancer. Int J Oncol 56(1): 301-314.
27. Lee CJ, Jang TY, Kim JH, Lim S, Lee S, et al. (2024) The dysadherin/FAK axis promotes individual cell migration in colon cancer. Int J Biol Sci 20(7): 2356-2369.
28. Lee CJ, Jang TY, Jeon SE, Yun HJ, Cho YH, (2024) The dysadherin/MMP9 axis modifies the extracellular matrix to accelerate colorectal cancer progression. Nat Commun 15(1): 10422.
29. Stark T, Livas L, Kyprianou N (2015) Inflammation in prostate cancer progression and therapeutic targeting. Transl Androl Urol 4(4): 455-463.