PC specimens were obtained from the biopsy archive of the Pathological Anatomy Service, Clinic Hospital University of Chile, with the corresponding authorization of the institutional Ethics Committee. All samples were evaluated by an expert pathologist (I.G.). For the immunohistochemical evaluation specimens were grouped as BPH samples, a non-malignant control, and PC samples with high and low histological Gleason grade.

Human PC cell lines (PC3 and LNCaP) were obtained from the American Type Culture Collection (ATCC, Rockville, MD, USA). Cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum and 1% penicillin and streptomycin. Cells were maintained under standard culture conditions at 37°C and 5% CO₂ in a humidified environment.

Primary antibodies were obtained from Abcam (SNAIL, N-cadherin; Cambridge, MA, USA), BD Transduction (E-cadherin; Franklin Lakes, NJ, USA), Santa Cruz Biotechnology (syndecan-1, Santa Cruz, CA, USA) and Contreras et al (24) (syndecan-2). Anti-rabbit secondary fluorescein-conjugated antibody, anti-mouse and anti-rabbit secondary peroxidase-conjugated antibodies were purchased from Jackson ImmunoResearch (West Grove, PA, USA).

The PC TMA was constructed as follows: first, the most representative tumor areas were carefully selected and marked based on the matched hematoxylin and eosin (H&E) stained slides. Altogether, 104 cores (1.5 mm diameter) of test tissue were taken from the donor blocks with a tissue microarrayer (Beecher Instruments, Silver Spring, MD, USA). Sections were stained with H&E and then evaluated by the pathologist. The TMA contained a mixture of tissue so that both benign and malignant samples of different Gleason grades were represented on each block. Sections of 4 μm were obtained with a microtome and transferred to glass slides (SuperFrosts Plus, Menzel-Gläser, Braunschweig, Germany). Finally, colon (3) and tonsil (3) samples were included as positive control for syndecans.

Tumor and control formalin-fixed and paraffin-embedded samples (TMA) were cut into 4-μm sections, mounted, deparaffinized and rehydrated in decreasing concentration of ethanol and distilled water. The sections were washed with phosphate-buffered saline (PBS) and antigen retrieval was performed in a steam bath for 15 min at 90–95°C in 10 mM Citrate Buffer (pH 6.0). Endogenous peroxydase activity was inhibited by incubation in 3% H₂O₂. Later on, the sections were washed and non-specific binding was blocked with 10% normal horse serum solution (Vectastain). Then, sections were incubated with corresponding primary antibodies overnight at 4°C or 1 h at 37°C. Afterwards, samples were incubated with secondary antibody for 30 min at 37°C. Then, samples were washed and incubated with the streptavidin-biotin system (Histostain^®-Plus Bulk Kit). After washing, the sections were incubated for 2 min at room temperature with liquid 3,3’-diaminobenzidine substrate (DAB) (Zymed^®, LAB-SA Detection System and DAB-Plus Substrate Kit) followed by counterstaining with hematoxylin. Finally, samples were dehydrated in ethanol, cleared in xylene, coverslipped and evaluated in a microscope Leica DM 2500 (18,21).

Cells were grown on 6-well tissue culture plates over sterilized glass coverslips until 50–70% confluence was reached. Then, cells were fixed with a solution containing 4% (v/v) paraformaldehyde and sucrose in PBS for 30 min at room temperature and stored in 0.02% (w/v) sodium azide in PBS at 4°C. Before incubation with the antibodies, the coverslips were washed with a 20 mM PBS-glycine solution and then blocked with PBS-glycine (20 mM)-BSA (0.1%). The cells were incubated with the primary antibodies overnight at 4°C or 1 h at 37°C, rinsed with 20 mM PBS-glycine solution three times and incubated with a FITC-conjugated secondary antibody (Jackson ImmunoResearch Fluorescein-Conjugated AffiniPure Goat Anti-Rabbit) away from light for 2 h. Finally, the coverslips were mounted and visualized under a spinning disc microscope (Olympus BX61Wl).

Photographs from immunohistochemistry and immunocytochemistry were digitally processed to obtain the integrated optical density (IOD). The average gray value of each image was used to obtain the IOD. The IOD corresponds to absorption of an optical element per unit distance for a given wavelength. The staining and illumination conditions of the samples were equivalent.

The cell culture medium was aspirated and the cells were washed with PBS, trypsinized and centrifuged at 1,050 × g for 5 min. The resulting pellet was resuspended in a lysis buffer (50 nM Tris-HCl pH 7.4, 0.15 M NaCl, 1% sodium deoxycholate, 1% NP-40, 0.1% SDS, 5 nM EDTA), with a protease inhibitor cocktail (0.01 mg/ml benzamidine, 0.002 mg/ml antipain, 0.005 mg/ml leupeptin, 4 mM phenylmethylsulfonyl fluoride and 1 mM Na₃VO₄, pH 7.4). Later, the cells were scraped and the lysate was collected in a microfuge tube and passed through a syringe to break up the cell aggregates. The cell lysate was cleared by centrifugation at 15,000 × g for 15 min at 4°C, and the supernatant was discarded. The protein pellet was collected for protein quantification by the Bradford method at 570 nM using a Ray Leigh spectrophotometer (UV-1600 model). For western blot analysis, 40 μg of protein were resolved over 10% polyacrylamide gels and electrotransferred onto a nitrocellulose membrane (Bio-Rad, Hercules, CA, USA). A molecular weight standard (Pierce, Rockford, IL, USA) was also resolved for analyzing specific zones of the gels. The efficiency of the process was measured staining the membranes with Ponceau Red reactive. The non-specific sites on membranes were blocked with blocking buffer [TBS-Tween-20 (100 mM Tris-HCl, 0.9% NaCl, 0.1% Tween-20, pH 7.5) - 5% non-fat dry milk] for 1 h at room temperature. Then, membranes were incubated with the corresponding primary antibody in blocking buffer overnight at 4°C, followed by incubation with anti-mouse or anti-rabbit secondary antibody peroxidase conjugated (in blocking buffer) and detected by chemiluminescence (Biological Industries, Beit Haemek, Israel) and autoradiography. The western blot analysis bands were scanned and analyzed using the scientific software program UN-SCAN-IT (Silk Scientific Corporation, Orem, UT, USA).

Data were tabulated and analyzed using SPSS v17.0 software. Normal distribution was tested by Kolmogorov-Smirnov test. Given the distribution of the data, a parametric test (Pearson test) to calculate the correlation index was used. ANOVA (Tukey’s test) was used to compare means. P<0.01 was considered to indicate a statistically significant difference.

From the 98 samples of PC in the TMA (excluding colon and tonsil controls), 4 spots containing prostatic stromal tissue were ruled out. Samples used for analysis were classified into 4 groups: non-tumoral control (BPH), and low, medium and high Gleason grade PC samples. The histological characteristics of the TMA groups stained with H&E are presented in Fig. 1. TMA included 45 BPH and 47 PC spots [9 corresponding to low (grade 1–2), 23 medium (grade 3) and 15 high Gleason grade (grade 4–5)], giving a total of 98 samples.

SNAIL staining is observed mainly in nuclei and shows increased intensity in high Gleason compared to low grade samples (Fig. 2A). H&E dyeing was omitted to avoid interfering with SNAIL nuclear specific staining. Average IOD for each sample showed normal distribution (Kolmogorov-Smirnov P=0.689). Subsequently, the IOD means were compared by ANOVA. Samples with a high Gleason grades show SNAIL-staining IOD means significantly higher (P<0.01) than samples with low Gleason grade and BPH (P<0.0001) (Fig. 2B). Given the normal distribution, the Pearson test established a correlation coefficient of 0.734 between the IOD and the Gleason grade.

The expression and distribution of syndecan 1 show a very heterogeneous pattern within the groups studied. BPH spots show a strong intensity localized mainly in the cytoplasm and membrane of the basal cells. Furthermore, epithelial cells exhibit a preferential localization in the baso-lateral region and approximately 50% of the cytoplasmic localization is detected at variable intensity (weak to moderate). This syndecan 1 distribution is also found in the low Gleason group and, to a lesser extent, in the medium Gleason group. However, in PC spots with high Gleason grade, membrane localization is lost and a granular cytoplasmic localization with low intensity is observed (Fig. 3). The main difference of this marker among the groups is found in its location. For comparison, E-cadherin (a validated epithelial marker) expression and distribution was evaluated in PC TMA samples. This epithelial marker shows an expected membrane location in most samples with intensities varying from moderate to strong. In BPH spots, E-cadherin shows mainly baso-lateral location in gland epithelial cells and was absent in apical membrane. On the other hand, low Gleason grade samples show syndecan 1 intensity and distribution similar to BPH. However, in high Gleason grade spots, a loss of intensity associated to gland architecture disorganization is observed (Fig. 3E). In addition, E-cadherin distribution shows a mixed pattern including cytoplasm location (Fig. 3F). Significant decrease in E-cadherin expression is observed only in medium and high Gleason grade samples (Fig. 3F).

The expression of this marker is highly variable in terms of location and immunostaining intensity. Similar to syndecan 1, syndecan 2 is found in both baso-lateral membrane and granular cytoplasm. BPH specimens show a high intensity in basal cells and basal lamina (Fig. 4A–D). No significant difference in syndecan 2 expression is observed among the different Gleason grade spots (Fig. 4E). However, cell location changes as Gleason grade increases, switching from membrane-cytoplasmic to cytoplasm-nucleus localization. Syndecan 2 is highly expressed in fibroblast, therefore, its presence in stroma served as internal positive control. For comparison, N-cadherin (a validated stromal marker) expression and distribution was evaluated in PC TMA samples. N-cadherin expression in BPH showed a mixed pattern including both membrane and cytoplasm location in epithelial cells. However, the immunostaining intensity, unlike E-cadherin, is weak (Fig. 4Fa–d). The intensity of N-cadherin staining is strong in stroma due to this molecule being highly expressed in fibroblast and mesenchymal tissue. For this reason, it serves also as an internal positive control (Fig. 4A). As expected, this marker is increasing with the disorganization of prostate gland epithelium. In low Gleason samples, N-cadherin is expressed mainly in baso-lateral membrane of epithelial cells while in medium grade spots the immunostaining of this molecule shows a decrease in membrane and an increase in cytoplasm. Furthermore, in high Gleason samples, N-cadherin shows a high expression of the membrane, cytoplasmatic and even nuclear location. N-cadherin expression is different only between BPH and medium/high Gleason grade samples (Fig. 4Fe).

Considering that LNCaP and PC3 cell lines have been widely used as in vitro model for PC, we studied the location of the transcription factor SNAIL in these commercial cell lines using fluorescent immunocytochemistry. Different cellular SNAIL distribution is observed in these cell lines. LNCaP cells (low tumorigenic capacity) show a homogeneous localization in the nucleus and cytoplasm (Fig. 5A). However, PC3 cells (high tumorigenic capacity) show an exclusively nuclear localization (Fig. 5B). Furthermore, the SNAIL staining intensity is very high and occasionally detected at the perinuclear region. Localization is more evident when performing a merge between SNAIL staining (green) and actin microfilaments (red). When comparing IOD, significant differences in the SNAIL expression between the cell lines were found (Fig. 5C). In addition, protein extraction and western blot analysis were performed to compare the SNAIL expression between the cell lines. Results show a higher SNAIL protein expression in PC3 than LNCaP cells (Fig. 5D and E).

Results obtained from syndecans 1 and 2 expression are presented in Fig. 6. In LNCaP and PC3 cells, expression of both syndecans is evident at plasma membrane (Fig. 6B, C, E and F). Comparison of IOD, in LNCaP cells (low tumorigenic capacity) show a higher syndecans expression than PC3 cells (high tumorigenic capacity) (Fig. 6G and H). E-cadherin (control epithelial marker) shows a similar pattern (Fig. 6A and D) and IOD (Fig. 6I).