Previous reports showed that PCPH is mutated or deregulated in some human tumors, suggesting its participation in malignant progression. Immunohistochemical analyses showed that PCPH is not expressed in normal prostate, but its expression increases along cancer progression stages, being detectable in benign prostatic hyperplasia, highly expressed in prostatic intraepithelial neoplasia, and remaining at high levels in prostate carcinoma. Experiments designed to investigate the contribution of PCPH to the malignant phenotype of prostate cancer cells showed that PCPH overexpression in PC-3 cells, which express nearly undetectable PCPH levels, increased collagen I expression and enhanced invasiveness, whereas shRNA-mediated PCPH knockdown in LNCaP cells, which express high PCPH levels, down-regulated collagen I expression and decreased invasiveness. PCPH regulated invasiveness and collagen I expression by a mechanism involving protein kinase Cδ (PKCδ): (a) PCPH knockdown in LNCaP cells decreased PKCδ levels relative to control cells; (b) PKCδ knockdown in LNCaP cells recapitulated all changes caused by PCPH knockdown; and (c) forced expression of PKCδ in cells with knocked down PCPH reverted all changes provoked by PCPH down-regulation and rescued the original phenotype of LNCaP cells. These results strongly suggested that the expression level and/or mutational status of PCPH contributes to determine the invasiveness of prostate cancer cells through a mechanism involving PKCδ. Data from immunohistochemical analyses in serial sections of normal, premalignant, and malignant prostate specimens underscored the clinical significance of our findings by showing remarkably similar patterns of expression for PCPH and PKCδ, thus strongly suggesting their likely coregulation in human tumors. [Cancer Res 2007;67(22):10859–68]
Prostate cancer remains the second most commonly diagnosed cancer and the second cause of cancer death among men in the United States and other western countries. The American Cancer Society estimates that nearly 219,000 new prostate cancer cases will be diagnosed in the United States in 2007 and that, although efforts in screening facilitate earlier diagnosis, this year ∼27,000 men will die of this disease (1). Prostate cancer progression proceeds through defined states, including benign prostatic hyperplasia, prostatic intraepithelial neoplasia (PIN), prostate carcinoma in situ, and invasive and metastatic cancer (2). Despite its high morbidity and mortality, the etiology of prostate cancer remains obscure, and although standard therapies initially cause tumor regression, tumors eventually relapse and develop into hormone-refractory disease (3). Therefore, identifying novel molecular pathways involved in prostate cancer initiation and malignant progression and developing new therapies to specifically target early molecular carcinogenesis effectors are essential for future therapeutic gains.
Molecular analyses showed that malignant transformation of Syrian hamster embryo cells with 3-methylcholanthrene was caused by the activation of a novel oncogene (4, 5), initially termed cph and later renamed as mt-PCPH to distinguish it from the normal PCPH proto-oncogene. PCPH is highly conserved in vertebrates, being expressed in a broad range of tissues and developmental stages (6, 7), suggesting that it may play an important cellular function(s) in high eukaryotic organisms. Activation of mt-PCPH involved a single base-pair deletion within the coding region of the proto-oncogene that shifted the normal open reading frame and caused the early translation termination of the mutated protein. The mt-PCPH oncoprotein is a truncated form of the normal protein, incorporating its first 216 amino acids fused to a rather hydrophobic COOH-terminal tail of 33 residues not present in normal PCPH (8).
Functionally, mt-PCPH synergized with the H-ras oncogene in NIH 3T3 transformation (5, 8) and, unlike normal PCPH, induced a sustained activation of extracellular signal–regulated kinase (ERK)-1 in mammalian cells (9) as an important component of its transforming activity. Biochemical and functional analyses indicated that the participation of mt-PCPH in cellular mechanisms of stress response also contributes to its transforming ability. Additional experiments showed that PCPH is identical to CD39L4 (later renamed ENTPD5), that PCPH proteins have ATP diphosphohydrolase (apyrase) activity (10), and that the apyrase activity enhanced the resistance to various stress stimuli elicited by mt-PCPH (11). These results predicted that PCPH and mt-PCPH may interact functionally with the mammalian target of rapamycin (mTOR) kinase, a known ATP sensor (12) involved in controlling cell growth and apoptosis (13, 14). Indeed, we showed that the survival-promoting function of the mt-PCPH oncoprotein is mediated by its ability to antagonize proapoptotic mTOR signaling activated by exposure to ionizing radiation (15).
Mutational activation and apoptosis suppression in response to stress are not the only mechanisms by which PCPH contributes to malignant development. Analyses of human and animal cell lines and solid tumors indicated that deregulated expression of the PCPH protein is frequently associated with tumor progression. Alterations in PCPH expression detected in rat mammary tumors (16) and in tumor cell lines of diverse tissue origin (17) were consistent with aberrations of either PCPH transcript splicing or posttranslational processing of the PCPH protein, resulting in loss of the normal PCPH protein and the presence of aberrant PCPH polypeptides. Analyses of human laryngeal, breast, and testicular tumors and tumor cell lines yielded similar results (18–20). We now report results from experiments to evaluate the PCPH expression in human prostate cancers, its possible participation in initiation and/or malignant progression, and its involvement in determining the malignant phenotype of prostate cancer cells. Data indicate that PCPH is deregulated in prostate carcinoma; identify PCPH as a novel, very early marker for prostate cancer progression that could be potentially useful for diagnostic purposes; and recognize protein kinase Cδ (PKCδ) as a downstream mediator of the invasiveness-promoting activity of PCPH in prostate cancer cells.
Materials and Methods
Immunohistochemistry. Prostate samples were collected from 63 patients diagnosed by the Pathology Service of the Hospital “Príncipe de Asturias” of Alcalá de Henares, Madrid, Spain. Tissues were fixed in 10% (v/v) formaldehyde in PBS (pH 7.4) for 24 h, dehydrated, and embedded in paraffin. Sections (5 μm) were processed following the avidin-biotin complex (ABC) method (21). Deparaffinized sections were hydrated and incubated for 30 min in 0.3% H2O2 diluted in methanol to reduce endogenous peroxidase activity. For antigen retrieval, sections were incubated with 0.1 mol/L citrate buffer (pH 6) for 2 min in a conventional pressure cooker. After rinsing in TBS, slides were incubated with normal donkey serum, at a 1:5 dilution in TBS (TBS/NDS) for 30 min, to prevent nonspecific binding of the primary antibody. Then, primary antibodies against PCPH (9, 17) or anti-PKCδ were applied at 1:4,000 or 1:300 dilution, respectively, in TBS/NDS at 4 °C overnight. Afterwards, sections were washed twice in TBS and incubated for 1 h with swine anti-rabbit or rabbit anti-goat biotinylated immunoglobulin (DAKO), at 1:500 dilution, as the secondary antibodies. Sections were then incubated with streptavidin-biotin-peroxidase (strpABC-HRP) or streptavidin-biotin-alkaline phosphatase (strpABC-AP; DAKO) complexes at 1:400 dilution for 1 h at room temperature. Peroxidase and alkaline phosphatase activities were detected using the 3,3′-diaminobenzidine (DAB) plus substrate kit and the AP-red solution (both from Zymed Laboratories, Inc.). Sections were lightly counterstained with hematoxylin, and DAB-stained sections were dehydrated and mounted in DePex (Probus) and Aquatex (Merck & Co., Inc.).
The rabbit polyclonal anti-PCPH antiserum, produced by BioSynthesis, Inc., was previously described (18–20). Sections of normal human tissues (larynx, testes, skin, liver) that express PCPH (6, 9, 18) were used as positive controls. The quality and specificity of the PKCδ monoclonal antibody (BD Biosciences) were previously characterized in diverse experimental systems (22, 23), including immunohistochemical analyses of early prostate cancer specimens (24). Positive controls for each anti-PKCδ batch were done with rat brain lysates, as indicated by the manufacturers. In negative control experiments, serial sections of each specimen were processed as described, but the corresponding primary antibody was omitted. Immunoreactivity of each focus of interest was semiquantitatively graded by two independent observers as negative , weakly positive , moderately positive , or strongly positive .
Cell culture and reagents. RWPE-1, LNCaP, and PC-3 cells were obtained from the American Type Culture Collection. C4-2 cells were a gift from Robert Bakin (Georgetown University, Washington, DC). RWEP-1 cells were cultured in keratinocyte serum-free medium (Invitrogen). All other cells were maintained in RPMI (Invitrogen) supplemented with 10% fetal bovine serum, penicillin, and streptomycin and then cultured at 37°C in a humid atmosphere containing 5% CO2. Monoclonal anti–glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was from Abcam. Oligonucleotide primers were from Bio-Synthesis or Invitrogen. The PKCδ expression vector was purchased from Origene. Rottlerin, Gö6976, and the PKCβ inhibitor were purchased from Calbiochem/EMD Biosciences, Inc. Plasmids for the expression of normal PCPH or the mt-PCPH oncoprotein were described previously (9). All other general reagents were from Sigma-Aldrich.
Colony formation assays. Anchorage-independent proliferation was quantified by standard soft-agar assays. Cells (104) were suspended in 1.5 mL of 0.4% (w/v) Noble agar (Difco) and overlaid on 0.6% (w/v) agar. After 2 weeks of incubation at 37 °C, colonies were visualized by staining with nitrotetrazolium blue chloride (1 mg/mL), and those >0.5 μm in diameter were counted. All experiments were done in triplicate and repeated at least thrice.
Invasion assays. Cell invasion assays were done using BioCoat Matrigel chambers (BD Biosciences). Briefly, 104 cells suspended in 0.5 mL of serum-free medium were seeded in the top chamber, and 0.5 mL of serum-containing medium was placed in the bottom chamber. Invasion was allowed to proceed for 48 h in the case of LNCaP cells and for 20 h for PC-3 cells. After that, the noninvading cells were scrubbed away, and the membranes were fixed in 100% methanol for 5 min, stained with 1% crystal violet, and invading cells were manually counted in three different fields. All experiments were done in triplicate and repeated at least thrice.
Selection of cells stably expressing shRNA constructs. Duplex small interfering RNA (siRNA) oligonucleotides were purchased from Ambion. PCPH sequences targeted by siRNA corresponded to nucleotide positions 353–371, 375–399, and 799–817 in the PCPH mRNA (GenBank no. AF136572). The presence of these same sequences in the mt-PCPH mRNA (7, 8) allowed the knockdown of either the normal PCPH or the truncated mt-PCPH. DNA templates encoding PCPH-targeted shRNAs were synthesized, annealed, and cloned into pSilencer 4.1-CMV neo (Ambion). LNCaP cells were transfected with the expression plasmid using Oligofectamine (Invitrogen) and selected with 400 μg/mL neomycin (Geneticin, Invitrogen). HuSH PKCδ shRNA (29-mer) and control plasmids, purchased from Origene, were transfected into LNCaP and PC-3 cells using Lipofectamine (Invitrogen) according to the manufacturer's protocols, followed by selection with 1 ng/mL puromycin (MP Biomedicals). Antibiotic-resistant pools and individual clones were isolated and maintained in the presence of neomycin or puromycin.
Immunoblot analysis. Methods for preparation of total cellular extracts in the presence of a protease inhibitors cocktail, SDS-PAGE electrophoresis of cellular proteins (50 μg), and transfer onto polyvinylidene difluoride membranes were as previously described (15, 18, 20). Membranes were incubated with anti-PCPH, anti-PKCδ, or anti-GAPDH; washed with Tween 20 in PBS; incubated with peroxidase-conjugated secondary antibody; and the signal was then detected with a chemiluminescence-based system (Pierce). For each protein tested, Western blot analyses were repeated at least thrice.
Reverse transcription-PCR. Total RNA (3 μg), extracted using the RNeasy Mini Kit (Qiagen), was used for cDNA synthesis with SuperScript (Invitrogen). PCR primers for PCPH, COL1A1, COL1A2, PKCδ, and GAPDH were designed using the Oligo 6.0 software program (National Bioscience). A 1,448-bp PCPH fragment was amplified using the primers 5′-GGTGTGCGAGCAGGATTG-3′ (forward) and 5′-GGAGATGCCCAGAGACTG-3′ (reverse). The primer set 5′-CAAAGAAGGCGGCAAAGGTCCCCGTGGTGAG-3′ (forward) and 5′-GGAGAACCGTCTCGTCCAGGGGAACCTTCG-3′ (reverse) was used to amplify a 359-bp COL1A1 fragment, whereas the primer set 5′-GACCTCCAGGTGTAAGCGGT-3′ (forward) and 5′-TTCAGGTTGGGCCCGGATAC-3′ (reverse) was used to amplifty a 348-bp fragment of COL1A2. For PKCδ, primers 5′-CAGCAAGGGCATCATTTACAG-3′ (forward) and 5′-TCCGGTCACTCCCAGCCTCTT-3′ (reverse) were used to amplify a 390-bp fragment, and primers 5′-CGGGAAACTGTGGCGTGATG-3′ (forward) and 5′-GGAGGAGTGGGTGTCGCTGTTG-3′ (reverse) were used to amplify GAPDH. For each set of primers, the number of cycles was adjusted so that the reaction end points fell within the exponential phase of product amplification, thus providing a semiquantitative estimate of relative mRNA abundance. For each relevant transcript, reverse transcription-PCR (RT-PCR) determinations were carried out at least thrice, and relative transcript quantifications were done by densitometric analysis using a FluorChem 8000 image analyzer from Alpha Innotech.
Statistical analysis. For assays requiring statistical analysis, ANOVA or Student's t test was used to assess the significance of differences between groups or individual variables, respectively. P ≤ 0.05 was regarded as significant.
PCPH expression pattern in prostate cancer progression. To evaluate PCPH expression at different stages of human prostate cancer development, we carried out immunohistochemistry studies on samples from 63 patients corresponding to different stages of the disease progression. As shown in Fig. 1, relative to positive control sections from human skin (Fig. 1A,, a), PCPH expression was not detected in normal prostate epithelial cells (Fig. 1A,, b) and was slightly expressed in samples of benign prostatic hyperplasia (Fig. 1A,, c). However, PCPH expression was elevated in samples of both PIN and prostate carcinoma, being especially high in PIN areas (Fig. 1A,, d) and remaining heterogeneously high in carcinoma specimens (Fig. 1A,, e and f). PCPH expression was found predominantly in the cytoplasmic compartment in all cases. Figure 1A (bottom) summarizes these results and provides a relative quantification of PCPH expression among the prostate samples tested. Taken together, these data indicate that PCPH expression is deregulated in human prostate cancer and strongly suggest that PCPH may be an active participant in the initiation and/or progression of prostate tumors.
Generation of PCPH-knockdown and PCPH-overexpressing cell lines. To establish a suitable cellular system to investigate a possible role of PCPH in prostate cancer progression, we examined PCPH expression in several human prostate cell lines including RWPE-1, nonneoplastic human prostatic epithelial cells immortalized with human papillomavirus 18 (25); LNCaP, an androgen-dependent prostate cancer cell line; C4-2, an androgen-independent cell line derived from LNCaP; and PC-3, an androgen-independent prostate cancer cell line (26, 27). As expected, PCPH expression was not detected in RWPE-1 cells by RT-PCR or immunoblot analysis (Fig. 1B). However, PCPH was highly expressed in LNCaP cells. PCPH expression was also detected in C4-2 cells, but at lower levels than in LNCaP cells. RT-PCR results were confirmed by Western blot (Fig. 1B , bottom). In contrast, PCPH mRNA was present at very low levels in PC-3 cells, but the PCPH protein was undetectable. Consequently, we selected LNCaP cells to study the effects of PCPH down-regulation and PC-3 cells to study the effects of ectopic expression of PCPH.
To knock down PCPH expression in LNCaP cells, siRNA oligonucleotides were designed against three nonoverlapping target regions of the PCPH mRNA. Among them, siPCPH2 caused the most efficient knockdown of the endogenous PCPH transcript (Fig. 1C). The corresponding duplex oligonucleotide was cloned into pSilencer 4.1-CMV neo, and the construct (shPCPH) was transfected into LNCaP cells to stably down-regulate PCPH using a sequence scrambled shRNA showing no homology to any known human sequences as the nonspecific control. Results showed that PCPH expression was indeed almost completely knocked down in the transfected cells at both mRNA and protein levels, whereas no effect was observed in scrambled shRNA–transfected cells (Fig. 1D,, left). As expected from the presence of the targeted sequences in both PCPH and mt-PCPH transcripts, shPCPH expression down-regulated both the normal 47-kDa PCPH polypeptide and the 27-kDa polypeptide (Fig. 1D , left), which is also expressed in LNCaP cells and was previously identified as the molecular mass of the truncated mt-PCPH oncoprotein (6, 8).
Because PCPH expression was undetectable in PC-3 cells, these cells were transfected with recombinant pcDNA-3.1–based constructs designed to express either the normal PCPH protein or the mt-PCPH oncoprotein, as well as with control empty vector DNA, and individual cellular clones and pooled populations stably expressing either protein were isolated (Fig. 1D , right).
Expression of mt-PCPH enhances the invasiveness of prostate cancer cell lines. To study the possible involvement of PCPH in determining some of the malignant properties of prostate cancer cells, we first investigated if PCPH knockdown or forced expression modified the anchorage-independent growth of prostate cancer cell lines. The ability of shPCPH-expressing LNCaP cells to grow in semisolid medium was not significantly different from that of control scrambled shRNA–transfected cells (Fig. 2A). In contrast, relative to empty vector–transfected controls, the soft agar growth of PC-3 cells was markedly reduced by expression of normal PCPH and almost completely abolished by expression of mt-PCPH oncoprotein (Fig. 2B). These results suggested that the PCPH expression status of human prostate tumor cells may contribute to determine their ability to grown under anchorage-independent conditions.
The effect of PCPH expression on cell invasiveness was next evaluated using Boyden chamber invasion assays. Results showed that shRNA-mediated PCPH knockdown resulted in a marked (∼70%) inhibition of the invasiveness of LNCaP cells (Fig. 2C). Conversely, relative to empty vector–transfected controls, PC-3 cells expressing PCPH showed increased invasiveness (by ∼62%), and mt-PCPH expression caused PC-3 cells to become significantly more invasive (by ∼2.25-fold; Fig. 2D). Additional invasion assays were done with C4-2 cells, which express intermediate PCPH levels relative to LNCaP and PC-3 cells (Fig. 1B), after transfection with constructs for the expression of PCPH or mt-PCPH. Similar to the PC-3 case, results showed that, relative to empty vector–transfected controls, overexpression of PCPH or mt-PCPH increased the invasiveness of C4-2 cells by about 1.5- and 3.8-fold, respectively. These results clearly showed that the PCPH expression status is an important determinant of the invasive ability of human prostate cancer cell lines.
Morphologic changes associated with PCPH down-regulation in LNCaP cells are prevented by type I collagen. Microscopic examination of individual clones and pooled populations revealed that PCPH knockdown was consistently associated with morphologic changes in LNCaP cells. Indeed, when cultured on plastic, scrambled shRNA–transfected control cells grew in a monolayer spreading over the entire surface of the plates, whereas shPCPH-transfected LNCaP cells did not colonize the entire plate, grew upward, and formed large aggregates (Fig. 3A), indicating that PCPH down-regulation had an effect on cell-to-cell contact and/or adhesion. On the contrary, PC-3 cells transfected with either PCPH or mt-PCPH did not show any obvious changes in morphology or growth pattern (data not shown). Interestingly, when LNCaP cells were cultured on different extracellular matrix components (collagen I, poly-d-lysine, collagen IV, fibronectin, laminin), we observed that only collagen I prevented the appearance of the morphologic and growth pattern changes provoked by PCPH knockdown (Fig. 3A). Collagen I, the most abundant member of the collagen family, is composed of two identical α1 polypeptide chains and a similar, but distinct, polypeptide chain designated α2 (28, 29). These polypeptides are encoded by the COL1A1 and COL1A2 genes, respectively (29). Interestingly, these genes have been reported to be overexpressed in metastatic prostate tumors (30, 31). The fact that PCPH down-regulation caused morphologic and growth pattern changes and that they were prevented by collagen I suggested that PCPH could affect collagen I expression. To explore such a possibility, we used semiquantitative RT-PCR to determine the mRNA levels of COL1A1 and COL1A2 in prostate cancer cells expressing various levels of PCPH. Results showed that, relative to scrambled shRNA–transfected control cells, PCPH knockdown in LNCaP cells dramatically decreased the expression of COL1A1 and COL1A2 (Fig. 3B,, left), and that, conversely, expression of PCPH and especially mt-PCPH up-regulated the expression of both COL1A1 and COL1A2 genes in PC-3 (Fig. 3B , right). These data indicated that PCPH regulates the levels of type I collagen in prostate cancer cells and suggested that collagen I regulation may be a component of the mechanism by which PCPH modulates their invasive ability.
PCPH regulates the levels of PKCδ in prostate cancer cells. It has been reported that cell spreading and invasiveness are regulated in human tumor cells by a transforming growth factor-β signaling network that involves PKCδ (32, 33). Moreover, studies with rottlerin, a PKCδ inhibitor, and dominant-negative PKCδ constructs have shown that PKCδ regulates the expression levels of collagen I genes in several human cell types (34–36). Consequently, we examined possible changes in the expression levels of PKCδ in LNCaP and PC-3 cells expressing different levels of PCPH or mt-PCPH. Results showed that, relative to cells expressing the scrambled control shRNA, PCPH knockdown substantially reduced the levels of PKCδ mRNA and protein in LNCaP cells (Fig. 4A,, left), and that, relative to empty vector–transfected controls, ectopic expression of either PCPH or mt-PCPH increased the levels of both PKCδ mRNA and protein in PC-3 cells (Fig. 4A,, right). Because these results suggested that PCPH expression may regulate PKCδ, and it has been reported that the invasive ability of prostate tumor DU145 and PC-3 cells was prevented by PKCδ knockdown (37, 38), it became important to determine whether PKCδ participated in the invasive process mediated by PCPH expression (Fig. 2C and D). To that end, we transfected LNCaP cells, PC-3 cells, PC-3 cells expressing PCPH, and PC-3 cells expressing mt-PCPH with a shRNA targeted against PKCδ (shPKCδ), and once the efficient PKCδ knockdown in these cells was verified (Fig. 4B), invasion assays were done. Results showed that, similar to the scrambled shRNA control used for PCPH knockdown experiments, the PKCδ scrambled control did not significantly modify the invasiveness of the recipient cells (not shown), whereas PKCδ knockdown greatly reduced the invasiveness of all prostate cancer cell lines tested (Fig. 4C). Down-regulation of PKCδ reduced LNCaP invasion by ∼85% (Fig. 4C,, left). The specific dependence of this effect on PKCδ was shown by the fact that LNCaP invasiveness was also nearly completely prevented by rottlerin, a PKCδ inhibitor, whereas invasion levels were not significantly different from those of controls after treatment with inhibitors specific for PKCα or PKCβ (Fig. 4C,, left). Importantly, a role for PKCδ as a possible mediator of the effect of PCPH on cell invasion was strongly supported by the fact that reconstitution of PKCδ levels by transfection of a PKCδ expression vector into LNCaP cells in which PCPH expression was previously knocked down by shPCPH allowed the recovery of their invasion activity to levels similar to those of scrambled shRNA–transfected control LNCaP cells (Fig. 4C,, left). Additionally, shRNA-mediated down-regulation of PKCδ in PC-3 cells and PC-3 cells expressing PCPH or mt-PCPH, in which PKCδ was up-regulated (Fig. 4A,, right), nearly abolished their invasiveness (Fig. 4C,, right). Importantly, invasion by PC-3 cells as well as their PCPH- and mt-PCPH–expressing derivatives was also inhibited by rottlerin (Fig. 4C , right), but not by PKCα or PKCβ specific inhibitors, to an extent similar to that caused by shPKCδ (data not shown). Taken together, these results showed that PKCδ is involved in the invasive process mediated by PCPH expression in prostate cancer cells.
PKCδ mediates the effects of PCPH expression on prostate cancer cells. To test whether the effects of PCPH on anchorage-independent growth, morphology, and growth pattern in culture and on collagen I expression in prostate cancer cells described above were mediated by PKCδ, we examined those same phenotypic features in cells in which PKCδ expression was knocked down by stable shPKCδ expression. In soft-agar assays, shPKCδ-expressing LNCaP cells showed slightly increased colony formation activity relative to scrambled shRNA–transfected controls (Fig. 4D), yielding a number of colonies within the same range observed in shPCPH-expressing LNCaP cells (Figs. 2A and 4D). Importantly, colony numbers were kept at the levels in scrambled shRNA–transfected controls when PKCδ was reconstituted in PCPH-knockdown LNCaP cells (Fig. 4D,, left). Furthermore, mt-PCPH–expressing PC-3 cells, which did not form colonies in soft agar (Fig. 2B) and expressed high levels of PKCδ (Fig. 4A,, right), efficiently formed colonies in soft agar after transfection with shPKCδ (Fig. 4D , right), suggesting that, although colony numbers did not reach the levels formed by empty vector–transfected control PC-3 cells, up-regulation of PKCδ by mt-PCPH expression was, at least in part, responsible for the changes observed with the anchorage-independent growth. These results showed that PCPH expression contributes to modulate the anchorage-independent growth of prostate cancer cells through the regulation of PKCδ.
In addition, PKCδ knockdown provoked morphologic and growth pattern changes (Fig. 5A) essentially identical to those caused by PCPH knockdown, as described above (Fig. 3A). As in the case of shPCPH-expressing LNCaP cells, shPKCδ-expressing cells did not colonize the entire surface of the culture plates and grew upward, forming dense aggregates. Furthermore, reexpression of PKCδ in shPCPH-expressing LNCaP cells was sufficient to revert their morphology and growth pattern to those characteristic of the original and scrambled shRNA–transfected LNCaP cells. These results strongly supported the notion that the morphology and growth pattern features of prostate cancer cells determined by PCPH expression are mediated by PKCδ.
To investigate whether the regulation of collagen I expression by PCPH was also mediated through PKCδ, we used RT-PCR to determine collagen I mRNA levels in shPKCδ-expressing LNCaP and PC-3 cells and their PCPH- and mt-PCPH–expressing derivatives. Results (Fig. 5B–D) showed that PKCδ regulates collagen I in prostate cancer cells. Relative to scrambled shRNA–transfected controls, PKCδ knockdown in LNCaP cells decreased the levels of collagen I (Fig. 5B,, left). Interestingly, similar results were observed when LNCaP cells were treated with the PKCδ inhibitor rottlerin (Fig. 5B,, right). Additionally, the levels of collagen I expression were dramatically reduced in PC-3 cells by both shRNA-mediated down-regulation of PKCδ and by inhibition of PKCδ with rottlerin (3 μmol/L; Fig. 5C), even in the case of mt-PCPH–expressing PC-3 cells, in which PKCδ and collagen I levels are up-regulated (Fig. 5C). Moreover, and most importantly, when PKCδ expression levels were reconstituted in shPCPH-expressing LNCaP cells, collagen I expression returned to levels similar to those in scrambled shRNA–transfected control LNCaP cells (Fig. 5D), thus providing strong evidence supporting that PCPH regulation of type I collagen is also mediated through PKCδ. Taken together, these data strongly support the notion that PKCδ is a key mediator of PCPH functions related to cell morphology, growth, and invasiveness in human prostate cancer cells.
PKCδ expression in prostate cancer. If the functional interaction established in prostate cancer cells in culture between PCPH and PKCδ had any clinical relevance, one would expect a high degree of parallelism between the expression patterns of PCPH and PKCδ at the various stages of prostate cancer progression. As a first approximation to test this prediction, we examined the expression of PKCδ in prostate cells lines in which we already established the expression pattern of PCPH. Results (Fig. 6A) showed a pattern of PKCδ expression similar to that of PCPH (Fig. 1B), with PKCδ not being expressed in immortalized RPWE-1 but being slightly expressed by PC-3 cells and highly expressed by LNCaP and C4-2 cells. Immunohistochemical analyses were done next to study PKCδ and PCPH expression in serial sections of the same normal, premalignant, and malignant prostate tissue specimens previously used for PCPH evaluation (Fig. 1A). As expected, the pattern of expression was essentially the same as that observed for PCPH (Fig. 6B,, 1, 3, 5, and 7). PKCδ was not detectable in normal prostate epithelial cells (Fig. 6B,, 2), but high PKCδ expression was detected in low-grade (Fig. 6B,, 4) and high-grade (Fig. 6B,, 6) PIN samples and remained also high in prostate carcinoma specimens (Fig. 6B,, 8). Figure 6C shows the quantitative evaluation of the immunohistochemistry data to summarize the similarity between the patterns of expression of PCPH and PKCδ during prostate cancer progression among all human samples tested. These results confirmed our prediction of a close association between the expression of PCPH and PKCδ in samples representative of the various phases of prostate cancer development, and provided strong evidence suggesting their coregulation in human tumors and supporting the clinical relevance of the PCPH-PKCδ interaction.
The study described here represents the first report on the involvement of PCPH, a known proto-oncogene, in the development of human prostate cancer. Our results, which are consistent between prostate cell lines and clinical samples, conclusively show that PCPH is deregulated during prostate carcinoma progression. The absence of PCPH expression in normal prostate and its presence in benign prostatic hyperplasia specimens and later remarkable increase in PIN samples clearly indicate that PCPH is likely to be functionally influential from the early stages of prostate carcinoma development. This is important because identification of novel molecular markers absent in normal tissue but present at the earliest tumor stages may be particularly useful for diagnostic purposes. Our finding that PCPH expression levels are highest in PIN samples is particularly interesting because clinical and experimental evidence supports a precursor relationship between high-grade PIN and prostate carcinoma (2). In addition, our results confirm microarray analysis data from a previous report identifying PCPH (also called ENTPD5) as one of the genes overexpressed in PIN relative to normal prostate cells (39).
PCPH has previously been associated with malignant development on the basis of its activity as an oncogene (4, 5), its aberrant expression in animal and human tumors (16–20), and its antiapoptotic activity (8, 15). However, the involvement of PCPH in the regulation of tumor cell invasiveness has not been reported to date. Our results conclusively show that the expression status of PCPH is a key determinant of the invasive ability of prostate cancer cells, and that PCPH expression, particularly in the case of the mt-PCPH oncoprotein, modulates the expression of the two collagen I genes, COL1A1 and COL1A2, which are typically overexpressed in metastatic prostate cancer. In this context, the opposing effects of PCPH expression on cell invasiveness and their ability to form colonies under anchorage-independent culture conditions are especially interesting because they seem to indicate that the function of PCPH in prostate carcinoma is more closely implicated with cell motility than with cell proliferation.
The involvement of PKC in PCPH function was already suggested (9) as a possible mechanism by which PCPH, when expressed in several mammalian cell types, could cause a slight but general down-regulation of the Ras-Raf-mitogen-activated protein kinase mitogenic pathway while simultaneously inducing the prolonged activation of ERK1. The findings described here not only corroborate the existence of a PCPH-PKC interaction but also show that PKCδ is the main PKC isoform regulated by PCPH in prostate carcinoma. Our results also agree with data from other laboratories on the role of PKCδ in regulating tumor cell invasiveness (37, 40, 41). Interestingly, PKCδ inactivation by rottlerin, its specific inhibitor, or PKCδ down-regulation by siRNA was reported to decrease the migration and invasiveness of DU145 and PC-3 cells provoked by stimulation of the epidermal growth factor pathway (40, 41). Although sometimes even contradictory functions have been attributed to PKCδ (42–44), including its function in regulating invasion (45–47), it seems clear that, as an essential component of a novel (PCPH-PKC-collagen I) pathway through which PCPH regulates the expression of collagen I, PKCδ acts as a promoter of invasiveness in prostate carcinoma cells. Whether PKCδ also acts as a mediator of the known prosurvival function of PCPH in prostate carcinoma remains to be elucidated. It is important to note that our finding that PKCδ is highly expressed in prostate cancer agrees not only with previous reports on immunohistochemical detection of increased PKCδ levels in early prostate cancer (24) but also with more recent data from gene expression profiling studies of prostate cancer (48–50).
In summary, our data identify PCPH not only as a novel, early marker for prostate cancer progression, which could be potentially useful for diagnostic purposes and early therapeutic intervention, but also as a molecular target for prostate cancer treatment because prevention or blockage of PCPH expression could preclude prostate carcinoma cell invasion, make tumor cells more susceptible to hormonal manipulation, and sensitize them to treatments with ionizing radiation or chemotherapeutic drugs. Additionally, our finding that PCPH promotes prostate carcinoma invasiveness via PKCδ makes available a novel molecular target and a second alternative strategy for the treatment of localized and metastatic prostate cancer.
Note: Present address for O.M. Tirado: Institut d'Investigació Biomèdica de Bellvitge (IDIBELL) and Centre d'Oncologia Molecular (COM), Hospitalet de Llobregat, Barcelona, Spain.
Grant support: U.S. Public Health Service grant RO1-CA64472 from the National Cancer Institute, NIH.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.