- Top of page
- RESULTS AND DISCUSSION
- LITERATURE CITED
RECK is expressed in the rat ventral prostate. The amount of mRNA increased after castration. In situ hybridization and immunohistochemistry demonstrated a transition from epithelial to stromal expression. This demonstrates that stromal cells upregulate RECK expression to regulate matrix metalloproteinases activity responsible for extracellular matrix (ECM) changes occurring after castration. Anat Rec, 2010. © 2010 Wiley-Liss, Inc.
The dependence of the prostate gland on androgen makes it a unique system for the study of both epithelial cell apoptosis and stromal/extracellular matrix remodeling, since androgen deprivation by surgical or chemical castration results in marked organ atrophy (Isaacs, 1994; Bruni-Cardoso et al., 2009). Stromal remodeling takes place in response to or in coordination with epithelial shrinkage and involves modifications of the smooth-muscle-cell phenotype (Antonioli et al., 2004, 2007) and reorganization of the basement membrane (Carvalho and Line, 1996; Ilio et al., 2000), collagen fibers (Vilamaior et al., 2000), and glycosaminoglycans (Augusto et al., 2008).
The involvement of matrix metalloproteinases (MMP) in both prostate development (Bruni-Cardoso et al., 2008) and stromal remodeling after castration has been demonstrated (Wilson et al., 1991; Powell et al., 1996; Bruni-Cardoso et al., 2009). Also, MMP-7 expression is downregulated by androgen, and castration results in increased expression of this MMP (Powell et al., 1996). MMP-7 cleaves Fas and contributes to the induction of epithelial cell apoptosis (Powell et al., 1999). Similarly, MMP-7 is involved in disruption of tumor cell–cell adhesion, increasing the propensity of prostate cancer cells for invasion (Davies et al., 2001). Castration promotes a transition from apical to basal secretion of the enzyme that shifts the targeting of the enzyme from the luminal secretion to the basement membrane (Felisbino et al., 2007).
MMP function is regulated by a sequence of activation events and a number of inhibitors. It was shown before that TIMP-2 mRNA exhibited a 3.3-fold increase 7 days after castration (Desai et al., 2004). More recently, it was demonstrated that TIMP-2 is restricted to the proximal ductal segments of the rat ventral prostate (Delella et al., 2007), which may imply that additional factors are required for MMP regulation in this organ. Considering that RECK, one such inhibitor, is important in determining tumor aggressivity and in regulating angiogenesis (Oh et al., 2001) and that reduction and regrowth of blood vessels are involved in the prostate remodeling in response to androgen deprivation and reposition (Shabsigh et al., 1999), we decided to investigate whether RECK is expressed in the rat ventral prostate, how its expression and location are affected by castration, and whether RECK could complement TIMP-2 for the regulation of MMPs in the intermediate and distal ductal regions.
Three-month-old male Wistar rats were purchased from CEMIB-UNICAMP. The rats were divided into two groups: (i) intact control rats and (ii) castrated animals that were sacrificed 7, 14, and 21 days after surgery. Surgical castration was done under ketamine (80 mg/kg) and xylazine (10 mg/kg) anesthesia. All rats were sacrificed by cervical dislocation, and the ventral prostate was dissected and fixed as below or frozen in liquid nitrogen and stored at −70°C until used. Total RNA extraction was performed using Trizol reagent (Invitrogen Life Technologies, Carlsbad, CA). Samples were homogenized in 1 mL of Trizol using a Polytron homogenizer (Kinematica, Lucerne, Switzerland), and the subsequent procedures were carried out according to the manufacturer's recommendations. RNA samples were reverse-transcribed with 200 U SuperScript III and Oligo (dT)12–18 Primer (Invitrogen Life Technologies) according to the supplier's instructions. To perform a semiquantitative determination of RECK expression, a RT-PCR reaction was optimized for simultaneous amplification of RECK and β-actin, which was used as the reference. All primers were synthesized by Invitrogen Life Technologies (São Paulo, SP, Brazil). For β-actin, the forward primer was 5′- TCACCCACACTGTGCCCATCTACGA -3′ and the reverse primer was 5′- CAGCGGAACCGCTCATTGCCAATGG -3′; for RECK, the forward primer was 5′- CCTCAGTGAGCACAGTTCAGA -3′ and the reverse primer was 5′- GCAGCGCACACACTGCTGTA -3′. PCR thermal cycling conditions were (i) an initial step at 94°C for 5 min, (ii) 30 cycles of 94°C for 1 min, (iii) 60°C for 1 min, (iv) 74°C for 1 min 30 sec, and (v) a final cycle of 7 min at 74°C. The reaction was performed in a 13 μL final volume containing 78 ng cDNA, 0.39 U Taq DNA Polymerase (Promega, Madison, WI), 2 mM MgCl2, and 400 nM of β-actin and RECK primers in the same reaction tube. The reaction product was electrophoresed in a 2% agarose gel and stained with ethidium bromide. Gel images were captured with a Kodak DC120 Camera, using the Kodak Digital Science 1.0 software (Eastman Kodak Company, New Haven, CT). Analyses were performed using the Scion Image Beta 4.0.2 for Windows (Scion Corporation, Frederick, MA). All samples were assayed in triplicate.
For real-time reverse-transcriptase polymerase-chain-reaction (RT-PCR), the total RNA was extracted using the Illustra RNAspin Mini (25–0500-71, GE Healthcare, UK) and reverse-transcribed as described above. Real-time quantitative RT-PCR was performed using TaqMan Universal PCR Master Mix (Applied Biosystems, Foster City, CA) in the Applied Biosystems 7300. Inventoried assay (Primers and FAM-conjugated probe) for phosphoglycerate kinase-1 (PGK-1) (Rn_00821429) (chosen after testing 10 endogenous controls, including β-actin) and customized RECK assay (primers: forward -5′ GTG CGT CTA TTA GTC CAC AGT TGA TAC -3′ and reverse 5′ AGT TGG GTT TCT CAT TGG ATA CG -3′ and the TaqMan MGB Probe: 6FAMCTG TGT GAA CAA TTA CAC CCAMGBNFQ) were purchased from Applied Biosystems. cDNA (20 ng) was used in each reaction, according to universal cycling conditions for the TaqMan system. The results were normalized using the CT (threshold cycle) values of the internal control PGK-1 on the same plate. To quantify and acquire the fold increase of RECK, the mathematical model 2−ΔΔCT was utilized, considering the control group as the calibrator. The efficiencies of the RECK and PGK-1 assays were calculated through the equation: E =10(−1/slope), with resulting values of 0.95 and 0.97 for RECK and PGK-1, respectively. All reactions were performed in triplicate on the same plate. Results from five animals for each experimental time point were analyzed by ANOVA followed by posthoc Tukey's test.
RECK RNAm was localized by in situ hybridization (ISH), according to Emson and Gait (1992), using the RECK probe FITC -5′-GTTCTGTTGGCCTGTTGTTAAAGTTTGTAC -3′ (Invitrogen Life Technologies, Carlsbad, CA). An alkaline phosphatase-conjugated antifluorescein antibody was used to locate hybridization. A sense probe was employed as the negative control. The alkaline phosphatase activity was developed using nitroblue tetrazolium as substrate. The sections were counterstained with methyl green.
Immunohistochemistry was performed on paraffin-embedded, paraformaldehyde-fixed tissues as described previously (Augusto et al., 2008), using a goat polyclonal antibody against RECK (diluted 1:100) (Santa Cruz Biotechnology, Santa Cruz, CA). The tissue-bound primary antibody was detected with the ABC kit (NCL-ABC, Novocastra, Newcastle-upon-Tyne, UK). The sections were counterstained with methyl green.
RESULTS AND DISCUSSION
- Top of page
- RESULTS AND DISCUSSION
- LITERATURE CITED
RECK was demonstrated to inhibit MMP-2 and MMP-9 activation after secretion (Oh et al., 2001). We have been interested in the stromal remodeling taking place in the rat ventral prostate after castration, and hypothesized that RECK could be involved in the regulation of MMPs in this organ, especially after the demonstration of a restricted distribution of TIMP-2 (Delella et al., 2007) and marked extracellular remodeling involving MMP activity (Vilamaior et al., 2000; Bruni-Cardoso et al., 2009). The semiquantitative analysis of RECK expression, using β-actin as the reference (Fig. 1A), revealed an increase in the levels of RECK mRNA in the prostate of control and castrated rats. The use of real-time PCR confirmed that the expression levels of RECK increased less than twofold and were therefore considered “normal” (Hu et al., 2006), but showed statistically significant increases 14 and 21 days after castration (Fig. 1B).
Figure 1. Relative expression of RECK mRNA in the ventral prostate of control and castrated rats, as determined by RT-PCR using β-actin as an internal control in duplex reactions (A) and fold-increase as determined by real-time RT-PCR, using PGK-1 as the internal control (n = 5) (B). Ct, control, noncastrated rats. Castrated rats were killed 7, 14, or 21 days after surgery. ANOVA followed by Tukey's multicomparison test. NS, non significant.
Download figure to PowerPoint
This was consistent with an increased activity of different MMPs in response to castration, but particularly of the gelatinases MMP-2 and MMP-9, which are involved not only with ECM processing but also likely with inducing epithelial cell apoptosis by anoikis in the rat ventral prostate (Bruni-Cardoso et al., 2009).
We then examined which cells were responsible for expressing RECK. By using ISH and immunohistochemistry, we observed that RECK is expressed in epithelial cells of the prostate of control, castrated, and testosterone-treated castrated rats. mRNA staining was intense in the epithelium of the prostate of control rats (Fig. 2A). Castration resulted in epithelial-cell atrophy and marked reduction in the ISH staining for RECK mRNA in the epithelium, in contrast to increased labeling of the smooth-muscle cells and fibroblasts in the stroma (Figs. 2B–E). It is possible that this decreased expression corresponds to an overall reduction in transcription and not a direct effect of androgen on RECK regulation. Testosterone treatment of castrated rats restored the aspects observed in noncastrated rats, with respect to RECK mRNA distribution (Fig. 2E). The negative control (Fig. 2F) showed an absence of labeling within cells and the existence of nonspecific labeling of elastic fibers around blood vessels and in stromal cells. Immunohistochemistry demonstrated that RECK is localized on the basolateral surface of epithelial cells and the surface of smooth-muscle cells and fibroblasts (Figs. 2G–J). In the proximal ductal regions, the staining for RECK was not as conspicuous as in the intermediate and distal regions, because the cells are shorter and the basolateral membrane is less regular. Some concentration of RECK was found at the cell–cell interface (Fig. 2H), and acquired a circular profile apically (Fig. 2J). We believe these circular structures are found at the level of cell–cell junctions but cannot associate a specific structure with them at the moment. Basal cells were not stained (Fig. 2I). Isolated cells in the epithelium showed intense cytoplasmic labeling (Fig. 2K). No specific pattern of regional distribution of RECK was observed. Castration caused a marked reduction in the reaction product in the epithelial cells, with considerable increases in the amount of RECK in both smooth-muscle cells and fibroblasts (Figs. 2L–N). The proximal regions also showed reduced staining (Fig. 2M, inset). Testosterone treatment only partially restored the control distribution of RECK and equally had no effect on RECK expression (not shown).
Figure 2. In situ hybridization (ISH) (A–F) and immunohistochemistry (G–N) localization of RECK mRNA and RECK respectively, in the rat ventral prostate. (A) Control, noncastrated rats. An intense reaction for RECK mRNA was found in the epithelial cells (Ep), occupying the perinuclear and apical cytoplasmic region facing the ductal lumen (L). Little reaction was also observed in the cell nucleus. Stromal cells showed a comparatively weaker reaction. Endothelial cells (EC) as well as lymphatic endothelial cells (LEC) were negative for RECK mRNA. A nonspecific reaction was observed for elastin around the blood vessels and in thin elastic fibers in the stroma. Castration at 7, and 21 days (B–D, respectively) resulted in epithelial cell atrophy but did not result in total inhibition of RECK mRNA expression. A comparatively stronger reaction was observed for smooth-muscle cells (SMC) after castration. A more intense reaction was also observed in the fibroblasts. (E) Testosterone administration to castrated rats after 14 days of androgen ablation partially restored the epithelial polarity and RECK mRNA expression. On the other hand, stromal cell expression was reduced. (F) Negative control, using a sense oligonucleotide probe. (G–N) Immunohistochemistry located RECK on the basolateral surface of epithelial cells (arrowheads) and on the surface of smooth-muscle cells (arrows) and fibroblasts. (H) Cross section of the epithelial cells, showing the presence of RECK and the cell surface. (I) Basal cells (double arrowheads) showed no prominent staining for RECK. (J) The arrowheads indicate circular structures stained for RECK at the cell–cell contacts in the apical regions. (K) Individual cells in the epithelium, possibly neuro-endocrine cells, showed intense staining. Androgen deprivation for 7 (L), 14 (M), or 21 days (N), resulted in marked reduction in the staining of the epithelium, while the staining of the stromal cells, particularly the smooth-muscle cells (arrows) remained intense. The inset in (M) is a detail, showing the staining for RECK at the surface of epithelial cells in the proximal ductal region. Bars = 10 μm.
Download figure to PowerPoint
This balance between the expression of RECK by epithelial and stromal cells in control and castrated rats indicates that RECK is differentially regulated in these cells with respect to androgen stimulation. Because epithelial cell death is at least in part dependent on MMP activity (Powell et al., 1999; Bruni-Cardoso et al., 2009), as is the case for the mammary gland (Lund et al., 1996; Wiseman and Werb 2002), it is reasonable to assume a reduced expression of RECK in these cells. Given the marked remodeling of the stromal compartment after castration, it is also reasonable to expect that stromal cells upregulate RECK expression to locally regulate the ECM changes in the stroma, and perhaps to preserve the integrity of the smooth-muscle-cell basement membrane (Antonioli et al., 2004; 2007) and the signaling cascades resulting from this cell-matrix interaction. It has been demonstrated for cardiac myocytes that disruption of the integrin-mediated signaling by MMP-2 results in apoptosis (Menon et al., 2006) and that MMP-2 and MMP-9 are related to the detachment of apoptotic melanoma cells (Pereira et al., 2005).
It became apparent from the present results that RECK is superimposed on TIMP-2 in the proximal regions, and is also found in the intermediate and distal ductal regions, where TIMP-2 is absent (Delella et al., 2007). Together, these results demonstrate that the regression of rat ventral prostate in response to castration is a complex phenomenon, involving the coordinated action of both epithelial and stromal cells. The changes in RECK expression and RECK location demonstrated here show the existence of compartmentalized mechanisms of regulation of the enzymes involved in organ remodeling.