E-cadherin is down-regulated in benign prostate hyperplasia and required for tight junction formation and permeability barrier in prostatic epithelial cell monolayer

Abstract

BACKGROUND:

We previously reported the presence of prostate specific antigen (PSA) in the stromal compartment of benign prostate hyperplasia (BPH). Since PSA is expressed exclusively by prostatic luminal epithelial cells, PSA in the BPH stroma suggests increased tissue permeability and the compromise of epithelial barrier integrity. E-cadherin, an important adherens junction component and tight junction regulator, is known to exhibit downregulation in BPH. These observations suggest that the prostate epithelial barrier is disrupted in BPH and E-cadherin downregulation may increase epithelial barrier permeability.

METHODS:

The ultra-structure of cellular junctions in BPH specimens was observed using transmission electron microscopy (TEM) and E-cadherin immunostaining analysis was performed on BPH and normal adjacent specimens from BPH patients. In vitro cell line studies using benign prostatic epithelial cell lines were performed to determine the impact of siRNA knockdown of E-cadherin on transepithelial electrical resistance (TEER) and diffusion of FITC-dextran in trans-well assays.

RESULTS:

The number of kiss points in tight junctions was reduced in BPH epithelial cells as compared to the normal adjacent prostate. Immunostaining confirmed E-cadherin down-regulation and revealed a discontinuous E-cadherin staining pattern in BPH specimens. E-cadherin knockdown increased monolayer permeability and disrupted tight junction formation without affecting cell density.

CONCLUSIONS:

Our results indicate that tight junctions are compromised in BPH and loss of E-cadherin is potentially an important underlying mechanism, suggesting targeting E-cadherin loss could be a potential approach to prevent or treat BPH.

INTRODUCTION:

Benign prostatic hyperplasia (BPH) is a common disease condition in aging men. BPH prevalence in men increases with age from 8% at 31–40, to 40–50% at 51–60, to over 80% in men over 80 years of age (1). Although it is not life threatening, BPH symptoms significantly impact quality of life (2). Furthermore, BPH treatment costs society more than $4 billion annually (3). As male life expectancy increases, the number of men affected by BPH is expected to increase. Thus, there is an urgent need for new and more effective approaches for BPH prevention and treatment, which will require a thorough understanding of the molecular and cellular mechanisms underlying BPH development and progression.

Androgens are involved in BPH development and progression (4–6), and increased androgen receptor (AR) signaling in BPH tissues has been reported by our lab and others (5,7). Our recent study showed the expression of prostate specific antigen (PSA) and kallikrein 2 (KLK2) in the surrounding stroma of BPH nodules (8). However, according to the literature, PSA and KLK2 are expressed by prostatic luminal epithelial cells and secreted to the glandular cavity because of polarity and prostatic luminal epithelial barrier in prostate (9). Our further in situ hybridization experiments verified that stromal cells did not express PSA since PSA mRNA was not detected in the BPH stromal compartment (8). How this stromal PSA travels from the luminal epithelial cells to the stroma is unknown and is intriguing to explore.

One major function of the mucosal epithelium is to form a selectively permeable barrier which is mediated by the tight junctions and subjacent adherens junctions of the apical junctional complex (Reviewed in (10)). Transport of proteins, lipids, ions and small molecules between the apical and basolateral cell surfaces is controlled and maintained by tight junctions (Reviewed in (11)), while adherens junctions form strong adhesion between cells. Adherens junctions develop first and are required for the assembly of tight junctions (12). Adherens junctions are composed of cadherins, and tight junctions are composed of a complex of transmembrane proteins, scaffolding proteins and regulatory molecules (10). Disruption of the mucosal epithelial barrier has been associated with the development and progression of many diseases, including pancreatitis, asthma, cystic fibrosis, Crohn’s disease and cancer (11,13–17). The human prostate is composed of a series of glandular ducts lined by a mucosal layer of luminal secretory cells and an underlying basal cell layer (18). Loss of epithelial barrier integrity in the prostate could result in the leakage of PSA and other secreted proteins into the stromal compartment, and thus promote BPH pathogenesis and/or progression. However, the exact status of tight junctions in BPH remains unclear.

In addition to the classic tight junction proteins claudins, occludins, and zonula occludens (ZOs), the adherens junction protein E-cadherin also plays a pivotal role in tight junction development and maintenance. The absence or downregulation of E-cadherin leads to improper formation of tight junctions, resulting in permeability increase and dysfunction of epithelium barriers (19–21). Downregulation of E-cadherin in BPH specimens was observed by our group and others (8,22). However, it is not clear if E-cadherin down-regulation affects luminal epithelial tight junction formation as well as prostatic luminal epithelial permeability. Arenas et al., reported downregulation of E-, N- and P-cadherin, and alpha-, beta- and gamma-catenin protein expression in BPH compared to normal prostate tissues (23). Gap junction proteins connexin 43 and connexin 32 were increased in BPH compared to normal prostate (24). Connexin 32 is localized to basal cells, while connexin 43 was localized to the luminal epithelial cells, potentially due to increased basal and stromal proliferation and increased luminal metabolism and communication, respectively (24). While these observations strongly suggest alteration of cell junctions in BPH, direct evidence showing a compromised epithelial barrier in BPH is lacking.

MATERIALS AND METHODS:

Reagents, antibodies and cell culture

Benign prostatic epithelial cell lines BHPrE1 (25) and BPH-1 (26) were gifts from Dr. Simon Hayward (Northshore University HealthSystem, USA). The HEK-293 cell line was purchased from American Type Culture Collection (Manassas, VA, USA). Corning DMEM (Dulbecco’s Modified Eagle’s Medium/Hams F-12 50/50 mix (10–090-CVR, Corning Inc., Corning, NY, USA), RPMI-1640 (10–041, Gibco, Waltham, MA, USA) culture medium, 100x penicillin and streptomycin (30–002-CI, Gibco), 100x L-glutamine (25030081, Gibco). For experiments utilizing transwell inserts, 12 mm Transwell® with 0.4µm Pore Polyester Membrane Inserts (3460, Corning) were used. Fetal bovine serum (FBS) was from Atlanta Biologicals (Flowery Branch, GA, USA), proteinase and phosphatase inhibitors cocktail, FITC-dextran (46945) and MTT (M2003) were from Sigma-Aldrich (St. Louis, MO, USA). cDNA reverse reagents (RR037A) and SYBR advantage qPCR premix (639676) were from Takara (Kusatsu, Tokyo, Japan). RNeasy Mini Kit was from Qiagen (74104, Hilden, Germany). Primary antibodies are listed in Table 1 and secondary anti-rabbit and anti-mouse antibodies were from Santa Cruz (Dallas, TX, USA). PVDF membrane and ECL reagents were from Bio-Rad Laboratories (Hercules, CA, USA). DharmaFECT 1 transfection reagent was from GE Healthcare Life Science (Marlborough, MA, USA).

Table 1.

Antibodies used for western blot (WB) and immunohistochemical staining (IHC)

Antibody	Source	Clone, Cat.#	Description	Application, Dilution
E-cadherin	Cell Signaling Technology	3195	rabbit monoclonal	WB, 1:1000
E-cadherin	Invitrogen	4A2C7	mouse monoclonal	IHC, 1:200
β-catenin	Cell Signaling Technology	2146	rabbit monoclonal	WB, 1:1000
N-cadherin	Cell Signaling Technology	14215	mouse monoclonal	WB, 1:1000
TJP1	Invitrogen	61–7300	rabbit polyclonal	WB, 1:1000
TJP2	Invitrogen	71–1400	rabbit polyclonal	WB, 1:1000
TJP3	Invitrogen	36–4000	rabbit polyclonal	WB, 1:500
Claudin 1	Santa Cruz	sc-81796	mouse monoclonal	WB, 1:1000
GAPDH	Santa Cruz	sc-0411	mouse monoclonal	WB, 1:10000

BPH-1 cells were cultured in RPMI-1640 medium supplemented with 10% FBS, 100 U/ml penicillin, 100 μg/ml streptomycin and 29.2 μg/ml L-glutamine (26). The BHPrE1 cell line was maintained in DMEM/F12 containing 5% fetal bovine serum, 1 µg/ml insulin-transferrin-selenium-X (51500056, Invitrogen), 0.4% bovine pituitary extract (13028014, Gibco), 3 ng/ml epidermal growth factor (S0155, Gibco), 29.2 μg/ml L-glutamine, and 1% antibiotic-antimycotic mix (15240112, Gibco) (25). HEK-293 cells were cultured in DMEM medium supplemented with 10% FBS, 100 U/ml penicillin, 100 μg/ml streptomycin and 29.2 μg/ml L-glutamine. Cells were cultured in a 37°C incubator with 5% CO₂ and 95% humidity. Culture medium was replaced every other day or according to experimental designs.

BPH clinical specimens

Collection and use of the clinical specimens used for this study was approved by the IRB of the University of Pittsburgh. Areas of BPH were sampled from five radical prostatectomy specimens from patients who were naïve to androgen therapy. A frozen sample of either normal-adjacent or BPH was procured adjacent to the sample submitted for clinical histologic assessment, and a frozen section of each research tissue specimen was histologically assessed by a board-certified genitourinary pathologist (R. Dhir or A. Parwani) to identify normal adjacent and BPH areas, and to confirm the tissues were free of cancer. Normal adjacent tissues were taken from either the transition or central zone of the prostate.

Transmission electron microscopy (TEM)

Specimens were identified by a board-certified genitourinary pathologist as BPH or normal adjacent tissues and fixed in 4°C cold 2.5% glutaraldehyde in 0.01 M PBS, pH 7.3 for two hours. Fixed specimens were then rinsed in PBS, post-fixed in 1% osmium tetroxide with 1% potassium ferricyanide, dehydrated through a graded series of ethanol (30% - 90% and 100% - Ethanol 200 Proof) and embedded in Polybed 812. Semi-thin (300 nm) sections were cut on Reichart Ultracut, stained with 0.5% toluidine blue and examined under a light microscope. Ultrathin sections (65 nm) were stained with 2% uranyl acetate and Reynold’s lead citrate and examined on JEOL 1011 transmission electron microscope (JEOL, Peabody, MA, USA).

Immunohistochemistry (IHC)

IHC was performed on five-micron sections of formalin-fixed paraffin-embedded (FFPE) prostate tissues. Sections were de-paraffinized and dehydrated through a graded series of ethanol. Heat induced epitope retrieval was performed using pH 6 10 mmol/L citrate buffer, followed by 5 min rinsing in TBS buffer. Primary antibodies are listed in Table 1. Slides were counterstained in hematoxylin and cover-slipped. Immunostained sections were imaged with a Leica DM LB microscope (Leica Microsystems Inc, Bannockburn, IL, USA) equipped with an Imaging Source NII 770 camera (The Imaging Source Europe GmbH, Bremen, Germany) and NIS-Elements Documentation v 4.6 software (Nikon Instruments, Inc., Mellville, NY, USA). All tissues were examined by board-certified genitourinary pathologists (R. Dhir or A. Parwani) using light microscopy.

Immunostaining image analysis

Immunostained slides were evaluated quantitatively using the H-Score method by assessing protein expression as a function of staining intensity, where no staining=0, weak=1, moderate=2, strong=3, times the percentage of cells exhibiting each level of intensity as previously (27). A minimum of 5 fields from each section was analyzed and only patients with both BPH and normal adjacent tissues were scored. An average score for each tissue type was calculated for each patient for a total of 8 patients. All scores were reviewed and confirmed by a board-certified genitourinary pathologist (R.Dhir).

siRNA transfection

siRNAs were synthesized by Dharmacon (Lafayette, CO, USA) and transfected into cells using DharmaFECT 1 (Dharmacon) transfection reagent in accordance with the manufacturer’s protocol. Briefly, culture medium was replaced by serum-free medium 1 h before transfection. siRNAs and transfection reagents were diluted in serum-free medium separately for 5 mins and then mixed together and incubated at room temperature for 15 min. The transfection complex was then diluted in culture medium and mixed well by shaking back and forth several times. Cultured medium was replaced by fresh complete medium 24 h later. siRNAs sequences were listed in Table 2.

Table 2.

Sequence of siRNAs

siRNA name	Sequence
siE-cadherin #1	5’-GGGUUAAGCACAACAGCAA-3’
siE-cadherin #2	5’-CAGACAAAGACCAGGACUA-3’

Cell treatments for in vitro permeability assays

Cells were seeded into 6-well plates at a density of 300,000 cells/well suspended in 2 ml complete culture medium and knockdown of E-cadherin was performed the next day. After 48 h, cells were digested by 0.25% trypsin and cell number was calculated using a Beckman Z2 coulter counter (Brea, CA, USA). Inserts were seeded with 100,000 cells suspended in 500 μl medium, the lower chamber was filled with 1 ml culture medium. Inserts were processed in triplicate. Remaining cells were seeded onto 6-well plates and mRNA was isolated the next day. The day when cells were seeded to inserts was counted as Day 0. Culture medium was replaced with fresh media every day. From Day 3, transepithelial electrical resistance (TEER) was checked every day while FITC-dextran transwell permeability assay was performed every other day. To maintain high knockdown efficiency, E-cadherin knockdown was repeated on Day 4 in inserts. On Day 8, for each treatment, one insert was fixed for TEM, one for mRNA purification and one for protein lysis. Cell density was determined by counting the total number of cells in 9 non-overlapping images taken from each insert and from at least 3 independent experimental replicates for each group to insure that cell number was similar across all treatments.

FITC-dextran transwell permeability assay

Medium in both inserts and lower chambers was aspirated, then the lower chambers were filled with 1 ml complete medium while the inserts were filled with 500 μl complete medium in the presence of 50 μg/ml FITC-dextran. After 24 h incubation in cell culture incubator, fluorescence of the medium in the lower chamber was measured by a SpectraMax M2 Microplate Reader (Molecular Devices, San Jose, CA, USA) by multipoint with depth check with excitation at 485 nm and emission at 535 nm.

Transepithelial electrical resistance (TEER) measurement assay

Medium in both inserts and lower chambers was replaced by fresh complete culture medium, 1 ml in lower chamber and 500 μl in inserts respectively. Inserts in 12-well plate were incubated at 37°C for 30 min. The electrode was sterilized in 75% ethanol for 10 min and then neutralized in sterilized PBS at room temperature for 10 min. TEER for each insert was measured at three points (12, 4 and 8 o’clock positions) by Millicell® ERS-2 voltohmmeter (MERS00002, Millipore, Billerica, MA, USA). TEER values were recorded when the measurement became stable (R1). TEER of inserts without cells was used as the blank control (R2). The formula used to calculate TEER was as following:

$$\text{TEER}\hspace{0.17em}=\hspace{0.17em}\text{R}1-\text{R}2$$

Western blot (WB)

Cell lysis was prepared in 1x RIPA lysis buffer containing proteinase and phosphatase inhibitors cocktail and boiled with 4x SDS loading buffer. Protein concentration was quantified by BCA quantification kit (Pierce, Waltham, MA, USA) and 30 μg total protein for each sample was loaded into 7–12% SDS-PAGE gel for electrophoresis. Proteins were then transferred to PVDF membrane at 350 mA constant current for 2 h on ice. After 1 h blocking with 5% skim milk in TBST at room temperature, membranes were incubated with primary antibodies (see Table 1) for 2 h at room temperature or overnight at 4°C. Membranes were washed with TBST 3 times and then incubated with secondary antibodies for 1 h at room temperature. After 3 times TBST washing, membranes were immersed into ECL mix for 5 mins. Protein bands were detected by Bio-Rad ChemiDoc (Bio-Rad, Hercules, CA, USA) system and quantified by ImageJ software (NIH, USA).

mRNA isolation and quantitative real-time polymerase chain reaction (qPCR)

Protocols used for isolation of mRNA from cultured cells, cDNA reversing and qPCR were described elsewhere (28). Briefly, mRNA was isolated by RNeasy Mini Kit (Qiagen, Hilden, Germany) and then reverse transcribed to cDNA using Takara reverse transcription reagents. Reaction solution which consisted of primers, cDNA and SYBR advantage qPCR premix was made and analyzed using Applied Biosystems StepOnePlus Real-Time PCR Systems (Applied Biosystems, Foster City, CA, USA). Each sample was duplicated. Primer sequences were listed in Table 3.

Table 3.

Primer sequences used in qPCR

	Forward	Reverse
GAPDH	5’-CGACCACTTTGTCAAGCTCA-3’	5’-AGGGGAGATTCAGTGTGGTG-3’
E-cadherin	5’-CGAGAGCTACACGTTCACGG-3’	5’-GGGTGTCGAGGGAAAAATAGG-3’

MTT assay

When seeding cells to inserts for in vitro permeability studies (see above), an aliquot of cells was also seeded into 96-well plates (10,000 cells/well) and cultured for the indicated time. Cells were incubated with 0.5 mg/ml of MTT at 37°C for 4 h, then medium was aspirated and precipitates were solubilized in 150 μl DMSO. OD value was read by M2 micro-plate reader at the wavelength 490 nm.

Statistical methods

All graphs were generated by GraphPad Prism 6 software (GraphPad Software, Inc. La Jolla, CA, USA). GraphPad Prism 6 or SAS, version 9.4 (SAS, Cay, NC, USA) were used to perform all statistical analyses. One-way ANOVA, and ad hoc multiple comparison tests were utilized to determine statistical comparisons between or among groups. Data were presented as mean ± standard deviation. A P value <0.05 was considered to be statistically significant.

RESULTS:

The number of tight junctions was decreased in BPH

Epithelial barrier integrity is maintained predominantly by tight junctions. The enhanced permeability of BPH tissues could be due to alterations in tight junction structure and/or function in BPH. Thus, transmission electron microscope (TEM) was utilized to observe the ultra-structures of luminal epithelial cell junctions in BPH and normal adjacent tissues. We encountered technical difficulties in obtaining high quality tissue suitable for TEM imaging from matched pairs of BPH and normal adjacent tissues, but were able to obtain good data from two individual patients. The number of tight junction ‘kiss points’ or fusion points between two adjacent cell membranes (11), were counted in TEM images from BPH and matched normal adjacent tissues from these two patients. Compared to normal prostatic tissues, BPH luminal epithelial cells from two patients displayed a statistically significant decrease in tight junction kiss points (Figure 1). More clinical BPH specimens will need to be analyzed to determine the frequency and extent of kiss point reduction in BPH specimens.

Figure 1. TEM micrographs representative of normal and BPH epithelial cell-cell junctions.

(A) Junctions from normal tissues show multiple contact points between cells at the apical membrane (left panels, arrows). BPH epithelial cells do not show these same contact points (line, area along the membrane) and present with uninterrupted spaces between the adjacent cell membranes. (B) Quantification of the number of kiss points per tight junction between cells at the apical membrane from Patient 1 and Patient 2.

E-cadherin is down-regulated and displays a discontinuous pattern in BPH

Adherens junctions are important for tight junction formation; they bring together neighboring cells and provide the basis for tight junction assembly. Among all adherens junction proteins, E-cadherin is the most critical for providing this function (19). E-cadherin immunostaining was less intense in BPH tissue relative to that in normal adjacent prostatic tissue in all of the patients analyzed (8/8) (Figure 2A, B). One typical characteristic of membrane proteins is their continuous expression pattern along the interface between cells (29), and E-cadherin staining exhibited this pattern in normal adjacent tissues (Figure 2A). However, in BPH, and particularly in atrophic glands, E-cadherin staining was discontinuous at the interface between both luminal-luminal epithelial cells and luminal-basal epithelial cells, as indicated by red arrows (Figure 2C).

Figure 2. E-cadherin is down-regulated and displays a discontinuous pattern in BPH.

(A) Representative IHC images of E-cadherin in normal adjacent prostate tissue (Normal) and BPH. (B) H-Score of E-cadherin staining in normal and BPH tissues from 8 different patients. (C) Red arrows pointed to the discontinuous expression pattern of E-cadherin in an atrophic BPH gland.

Epithelial barrier is formed in BHPrE1 and BPH-1 monolayer

To determine if benign prostatic epithelial cells form barriers in vitro, we performed TEER and FITC-dextran transwell assays, two well-established methods for studying permeability (30,31). HEK-293 cells grown as a monolayer do not form an epithelial barrier and do not form tight junctions (32,33). Thus, we utilized HEK-293 cells as a negative control. BHPrE1 and BPH-1 benign prostatic epithelial cell lines had higher TEER and less permeability to FITC-dextran diffusion than HEK-293 cells, suggesting epithelial barriers were formed in these two cell lines (Figure 3A, B). Although there was no statistically significant difference in TEER at Day 4 between BHPrE1 and BPH-1 (P=0.2655), TEER was significantly higher in BHPrE1 than BPH-1 on Days 6 and 8 (Figure 3A). BHPrE1 also had a significantly lower permeability to FITC-dextran diffusion than BPH-1 cells (Figure 3B). To further confirm this phenotype, the ultrastructure of junctions in monolayers was examined by TEM. As shown in Figure 3C, tight junctions were not found in HEK-293, but existed in both BHPrE1 and BPH-1 monolayers as indicated by red arrows. Taken together, these data suggest that epithelial barriers are formed in BHPrE1 and BPH-1 monolayers because of the existence of tight junctions. We also evaluated the expression of classic tight junction related proteins and adherens junction related proteins in HEK-293, BPH-1, and BHPrE1 cells (Figure 3D). E-cadherin expression was readily detectable in both BPH-1 and BHPrE1 cells. Therefore, these two cell lines can be potentially used for in vitro studies to test the role of E-cadherin in prostatic epithelial monolayer permeability.

Figure 3. Epithelial barrier is formed in BPH-1 and BHPrE1 monolayer.

Monolayer permeability was measured by TEER (A) and FITC-dextran transwell permeability assay (B) every other day and compared to that of HEK-293 cells. On day 8, inserts were fixed and the ultrastructure of tight junctions was observed by TEM (C). TEM magnification was optimized to show ultrastructure. Original magnification for HEK-293 cells was 20k ×, inset 80k ×; BHPrE1 cells 40k ×, inset 120k ×; BPH-1 cells 60k ×, inset 120k × (k=×1,000). The day cells seeded to inserts was counted as day 0. (D) Expression of adherens and tight junction proteins in BPH-1 and BHPrE-1 cells. Expression of E-cadherin, β-catenin, N-cadherin, TJP1, TJP2, TJP3 and claudin 1 was checked by WB using lysis from inserts. ^&P<0.0001.

E-cadherin knockdown increases permeability and decreased the formation of tight junctions in BHPrE1 and BPH-1 monolayers

To elucidate the role of E-cadherin in prostate epithelial barrier, we tested the effect of E-cadherin knockdown by two different siRNAs on the permeability of BHPrE1 and BPH-1 monolayers. Knockdown was confirmed by qPCR and western blot analysis (Figure 4A, B and Supplemental Figure S1A, B). E-cadherin knockdown decreased TEER (Figure 4C and Supplemental Figure S1C) and increased FITC diffusion (Figure 4D and Supplemental Figure S1D) in both BHPrE1 and BPH-1 using two different siRNAs, indicating that the permeability was increased after E-cadherin knockdown. The ultrastructure of tight junctions assessed by TEM revealed that E-cadherin knockdown significantly decreased the formation of tight junctions in both BHPrE1 and BPH-1 (Figure 4E, F, and Supplemental Figure S1E, F). The average number of tight junction kiss points in BHPrE1 and BPH-1 cells was 3.9 and 1.9, respectively. The average number of tight junction kiss points in BHPrE1 and BPH-1 cells following E-cadherin knockdown was 0.22 and 0.11, respectively.

Figure 4. E-cadherin knockdown increased permeability and decreased the formation of tight junctions in BHPrE1 monolayers.

Cells were seeded to 6-well plate (300,000 cells/well) overnight followed by knockdown of E-cadherin. Two days after siRNA knockdown, cells were digested and seeded to inserts (100,000 cells/well). (A) Knockdown efficiency of E-cadherin was examined by qPCR and (B) WB. (C) Monolayer permeability was checked by TEER daily, and (D) FITC-dextran transwell permeability assay every other day. Cells in inserts were harvested at Day 8. (E) Representative TEM images of junctions using samples from inserts with/without E-cadherin knockdown treatments. Red arrows indicate tight junctions. TEM magnification was optimized to show ultrastructure. Original magnification for siCon was 80k ×, inset 250k ×; siE-cad1 40k ×, inset 120k ×; siE-cad2 40k ×, inset 120k × (k=×1,000). (F) Quantification of the number of kiss points of tight junctions between cells at the apical membrane. *P<0.05, ^@P<0.01, ^#P<0.001, ^&P<0.0001.

To exclude the possibility that the increased permeability was because of decreased cell density, images of the monolayers in each insert were taken and cell number was counted. E-cadherin knockdown in BHPrE1 cells did not affect monolayer integrity (Supplemental Figure S2A) or cell density (Supplemental Figure S2B). In addition, E-cadherin knockdown had no detectable effect on cell viability as determined by MTT assay (Supplemental Figure S2C). Similar results were found in BPH-1 cells (Supplemental Figure S3).

DISCUSSION:

The detection of PSA (which is exclusively expressed by prostate luminal epithelial cells) in the stromal compartment of benign prostate hyperplasia (BPH) tissue specimens (8) suggests that prostatic epithelial barrier integrity may be compromised in BPH. BPH tissue displayed a decrease in the number of tight junctions in BPH specimens (see Figure 1), suggesting increased permeability in BPH tissues. E-cadherin expression was decreased and displayed a discontinuous pattern in BPH specimens (see Figure 2). E-cadherin knockdown inhibited tight junction formation in prostatic epithelial cells in vitro. These results suggest that E-cadherin down-regulation may lead to reduced tight junction formation, contributing to PSA and other secretory proteins from the glandular lumen leakage into the stromal compartment in BPH specimens.

Benign prostate epithelial cell lines BHPrE1 and BPH-1 were shown to be capable of forming an epithelial barrier as characterized by increased TEER and decreased permeability to FITC-dextran diffusion in vitro, while HEK-293 was not (see Figure 3). Knockdown of E-cadherin in BHPrE1 and BPH-1 cell lines induced an increase in permeability accompanied by a reduction in tight junctions (see Figures 4 and Supplemental Figures S1) while not impacting the monolayer or cellular proliferation (see Supplemental Figures S2 and S3). The BHPrE1 cell line was derived from benign human prostate (25); while BPH-1 was derived from a patient “undergoing transurethral resection of the prostate for urinary obstruction consistent with BPH.(26)” BHPrE1 had a significantly higher TEER and lower permeability to FITC-dextran diffusion than BPH-1 cells (see Figure 3A, B) and expressed higher levels of E-cadherin (see Figure 3D). While both cell lines have been shown capable of forming benign, non-tumorigenic prostate glandular tissue in vivo (25,34), BPH-1 cells may behave more similarly to BPH epithelial cells in vitro than BHPrE1 with respect to TEER, FITC-dextran permeability and E-cadherin expression. In summary, our in vitro experiments suggest that the down-regulation of E-cadherin observed in BPH tissues could contribute to an increased epithelial permeability and thus could contribute to the leakage of proteins from the glandular lumen into the stromal compartment of BPH.

Downregulation of E-cadherin in BPH specimens was previously observed by others and our group (8,22) and further confirmed in this study (see Figure 2). Besides decreased expression, it is also notable that E-cadherin displayed a discontinuous pattern in BPH, indicating alterations in prostatic epithelial cell-cell contacts and perhaps deficits in its membrane protein trafficking. Discontinuous and weaker E-cadherin staining patterns have been reported by others in epithelial choroid plexus cell lines used to study the blood-cerebrospinal fluid barrier (35); and discontinuous adherens junctions have been associated with cell border retraction in endothelial cells (36–38). Our in vitro studies revealed that E-cadherin knockdown increased monolayer permeability and disrupted tight junction formation without affecting cell density. The etiology of E-cadherin downregulation in prostate and its role in BPH pathogenesis remains to be elucidated, however, the assembly and seal of tight junctions depend on cell-cell contacts through a cascade of signaling reactions including two different G-proteins, PLC, PKC, and calmodulin (39). These contacts between neighboring cells are initiated by E-cadherin, the major transmembrane protein of the adherens junction(35). E-cadherin depletion was shown to disrupt the establishment of tight junctions in Madin-Darby canine kidney (MDCK) cells (40), and E-cadherin knockout in the murine uterine epithelium inhibited the expression of cellular junction proteins claudin, occludin and tight junction protein ZO-1 (TJP1) (41). During prostate development, E-cadherin gene transcription is repressed by site-specific DNA methylation, stimulating prostate bud outgrowth and maturation and differentiation of the prostatic bud epithelium (42). In murine urogenital sinus, blocking of E-cadherin function increases the number of prostatic buds undergoing morphogenesis and accelerates prostate lumen formation (42). Loss of E-cadherin is evident in both prostate cancer and BPH, however, the infiltration of PSA protein into the surrounding stroma is only evident in BPH (8). The stromal cells from BPH tissues may be characterized by properties that allow the infiltration and retention of secreted proteins that are not evident in prostate cancer stromal cells. Others have shown that BPH stromal cells are more fibroblastic, while prostate cancer stromal cells are more myofibroblastic (43). Additionally, BPH and prostate cancer typically occur in different anatomic locations (44), which may also play a role in the presence of secreted proteins in BPH stroma. Further studies will be required to fully elucidate the mechanisms and impact of E-cadherin down-regulation on prostate luminal epithelium permeability and its role in BPH pathogenesis.

Loss of prostate luminal epithelial barrier integrity could result in the leakage of PSA and other secreted proteins in prostatic fluid into the stromal compartment and contributing to BPH pathogenesis. PSA has previously been shown to induce BPH-derived stromal cell proliferation (45), and PSA can degrade extra-cellular matrix proteins including fibronectin and laminin (46). PSA is thought to contribute to seminal clot liquefaction by proteolysis of semenogelins I and II resulting in the release of spermatazoa (47) and KLK2 has been reported to have trypsin-like activity and to be able to cleave and activate pro-PSA (48,49). PSA and possibly other luminal epithelial secreted proteins could have a significant impact on the stromal microenvironment potentially by degrading proteins in the extracellular matrix and eliciting an inflammatory response and loss of epithelial barrier integrity may be an important step in BPH pathogenesis.

Taken together, BPH tissues displayed increased permeability and decreased E-cadherin. Our cell line studies suggest that loss of E-cadherin is potentially an important underlying mechanism leading to increased epithelial barrier permeability. E-cadherin down-regulation could result in the chronic leaking of secreted prostatic proteins into the basement membrane and underlying stroma, potentially contributing to the development and/or progression of BPH. Novel therapeutic approaches to prevent or alleviate E-cadherin down-regulation in the prostate may be developed in the future for BPH prevention and/or treatment.

Supplementary Material

Supp figS1-3

Figure S1. E-cadherin knockdown increases permeability and decreased the formation of tight junctions in BPH-1 monolayers. Cells were seeded to 6-well plate (300,000 cells/well) overnight followed by knockdown of E-cadherin. Two days after siRNA knockdown, cells were digested and seeded to inserts (100,000 cells/well). Cells in inserts were harvested at Day 8, knockdown efficiency of E-cadherin was examined by qPCR (A) and WB (B). Monolayer permeability was measured by TEER daily (C) and FITC-dextran transwell permeability assay every other day (D). (E) Representative TEM images of junctions using samples from inserts with/without E-cadherin knockdown treatments. Red arrows pointed to tight junctions. TEM magnification was optimized to show ultrastructure. Original magnification for siCon was 20k ×, inset 200k ×; siE-cad1 20k ×, inset 60k ×; siE-cad2 25k ×, inset 100k × (k=×1,000). (F) Quantification of the number of kiss points per tight junction between cells at the apical membrane. *P<0.05, ^@P<0.01, ^&P<0.0001.

Figure S2. E-cadherin knockdown has no influence on BHPrE1 density. (A) Representative photos of monolayers in inserts. (B) Quantification of cell number in each 40X field relative to siCon. (C) MTT assay was performed 4, 6 and 8 days after seeding to check cell viability. When seeding cells to inserts in (A), an aliquot of cells from each group were seeded to 96-well plate (10,000 cells/well) for MTT assay. OD, optical density.

Figure S3. E-cadherin knockdown has no influence on BPH-1 density. (A) Representative photos of monolayers in inserts. (B) Quantification of cell number in each 40X field relative to siCon. (C) MTT assay was performed 4, 6 and 8 days after seeding to check cell viability. When seeding cells to inserts in (A), an aliquot of cells from each group were seeded to 96-well plate (10,000 cells/well) for MTT assay. OD, optical density.

ABBREVIATIONS:

PSA

prostate specific antigen

BPH

benign prostate hyperplasia

TEM

transmission electron microscopy

IHC

immunohistochemistry

TEER

transepithelial electrical resistance

qPCR

quantitative real-time polymerase chain reaction

WB

western-blot

AR

androgen receptor

DMEM

Dulbecco’s modified Eagle’s medium

FBS

fetal bovine serum

PBS

phosphate-buffered saline

PFA

paraformaldehyde.