Plumbagin inhibits prostate carcinogenesis in intact and castrated PTEN knockout mice via targeting PKCε, Stat3 and epithelial to mesenchymal transition markers

Abstract

Prostate cancer (PCa) continues to remain the most common cancer and the second leading cause of cancer-related deaths in American males. The Pten deletions and/or mutations are frequently observed in both primary prostate cancers and metastatic prostate tissue samples. Pten deletion in prostate epithelium in mice results in prostatic intraepithelial neoplasia (PIN), followed by progression to invasive adenocarcinoma. The Pten conditional knockout mice (Ptenloxp/loxp:PB-Cre4) ((Pten-KO) ) provide a unique preclinical model to evaluate agents for efficacy for both the prevention and treatment of prostate cancer (PCa). We present here for the first time that dietary plumbagin (PL), a medicinal plant-derived naphthoquinone (200 or 500 ppm) inhibits tumor development in intact as well as castrated Pten-KO mice. PL has shown no signs of toxicity at either of these doses. PL treatment resulted in a decrease expression of PKCε, AKT, Stat3 and COX2 compared to the control mice. PL treatment also inhibited the expression of vimentin and slug, the markers of epithelial to mesenchymal transition (EMT) in prostate tumors. In summary, the results indicate that dietary PL inhibits growth of both primary and castration resistant prostate cancer (CRPC) in Pten-KO mice, possibly via inhibition of PKCε, Stat3, AKT, and EMT markers (vimentin and slug), which are linked to the induction and progression of PCa.

Introduction

Prostate cancer (PCa) continues to remain the most common cancer and the second leading cause of cancer-related deaths in American males. The American Cancer Society has predicted that a total of 233,000 new cases of PCa will be diagnosed and 29,480 deaths will occur from it in the United States alone in the year 2014 (1). PCa may be curable in its early stage by surgical or radiation therapy but there are currently no curative treatment options available for advanced or castration resistant prostate cancer (CRPC) (2, 3). The FDA has approved two chemopreventive drugs finasteride and dutasteride, which have been shown to reduce the risk of PCa upto 25 % in large clinical trials but both drugs showed potential side effects (1). Therefore, more effective therapies that can prevent or treat advanced or CRPC are urgently needed. In recent years, chemoprevention by using nutraceuticals has become an ideal strategy to prevent or slow down the various types of cancers (4, 5) including PCa (6, 7).

Plumbagin (PL), a plant-derived quinoid (5-hydroxy-2-methyl-1,4-napthoquinone), isolated from the roots of the medicinal plant Plumbago zeylanica L (also known as Chitrak) (8) has been shown as a chemopreventive and therapeutic agent against various types of cancer, including PCa. PL has also been found in Juglans regia (English Walnut), Juglans cinerea (butternut and white walnut) and Juglans nigra (blacknut) (8). The roots of Plumbago zeylanic have been used in Indian and Chinese systems of medicine for more than 2,500 years for the treatment of various types of ailments (8). It has also been reported for its neuroprotective (9), and cardioprotective activities in mice (10). PL fed in the diet (200 ppm) inhibits azoxymethane-induced intestinal tumors in rats (11). PL inhibits ectopic growth of breast cancer MDA-MB-231 cells (12), non-small cell lung cancer A549 cells (13) and melanoma A375-S2 cells in athymic nude mice (14). It has been reported that PL inhibits osteoclastogenesis induced by breast cancer cells in mice (15). A recent study has also shown that PL inhibits osteoclast formation and breast cancer cell-derived tumors in the bone microenvironment of mice (16). We previously have shown that PL inhibits ultraviolet radiation-induced development of squamous cell carcinomas (17). We have also shown that PL administration inhibits pancreatic cancer cell growth in vitro and in vivo via targeting EGFR, NF-kB and Stat3 signaling pathways (18). Another study has also shown inhibition of pancreatic cancer cell-derived orthotopic xenograft tumors by PL (19). Our laboratory has previously reported that PL administration inhibits human PCa cells DU-145 ectopic xenograft tumors (20). Recently, we have reported that PL administration inhibits growth and metastasis of highly aggressive human PCa cells (PC-3M (21) and prostate carcinogenesis in the transgenic adenocarcinoma of the mouse prostate (TRAMP) (22). We now present in this communication for the first time that dietary administration of PL inhibits prostate tumor growth in an intact as well as in a castrated Pten-KO mouse model possibly via inhibition of of PKCε, Stat3, AKT activation, and epithelial to mesenchymal transition (EMT) markers (Vimentin and Slug).

Material and methods

Chemicals and antibodies

PL (practical grade, purity >95%) was purchased from Sigma-Aldrich. Monoclonal or polyclonal antibodies specific for AKT, β-actin, PKCε, and total Stat3 were purchased from Santa Cruz Biotechnology, (Santa Cruz, CA). Blocking peptide for PKCε, antibodies and mouse IgG were also procured from Santa Cruz Biotechnology. Monoclonal antibodies specific for pAKT, pStat3Tyr705 and pStat3Ser727 were obtained from BD Biosciences (San Jose, CA). Vimentin and Slug antibodies were purchased from Cell Signaling Technology Inc. (Danvers, MA).

LC-MS/MS Assay

Fifty microliters of either the mouse plasma sample or plasma standard were placed in a microfuge tube. Ten microliters of working internal standard (50 ng/mL honokiol) was added in the tube and vortexed for one minute. One mililiter ethyl acetate was added in the tube and further vortexed for 10 minutes. The tube was centrifuged for 10 minutes at 14,000 RPM. The upper organic phase was transferred to a tube and evaporated under N2. The residue was reconstituted with 150 μL of 60% acetonitrile and placed on an autosampler plate. A 7-point plasma standard curve spanning the range 15.62 to 1000-ng/mL was included with each set of samples. The HPLC consisted of a model 1200 binary pump, vacuum degasser, thermostatted column compartment held at 25.0 °C, and a model 1100 thermostatted autosampler held at 25.0 °C, all from Agilent Technologies, Palo Alto, CA. The HPLC was coupled directly to a model API 4000 triple quadrupole mass spectrometer equipped with a Turbo V™ atmospheric pressure ionization source fitted with the electrospray probe from Applied Biosystems/MDS Sciex, Concord, Ontario, Canada. A 150 X 4.6 mm Zorbax Extend C18 5 micron HPLC column (Agilent) was the analytical column. The injection volume was 20 μL. The mobile phase solvents were: A Millipore Type I water and B HPLC grade Acetonitrile. The solvents were mixed 40% A / 60% B and delivered isocratically at 800 μL/minute. Run time was 10 minutes. Mass spec data were obtained in negative ion mode. The multiple reaction monitoring (mrm) transitions were m/z 187 → m/z 159 for PL and m/z 265.3→ m/z 244.1 for the internal standard honokiol. The retention time for PL was approximately 4.8 min to 5.9 min for honokiol. The lower limit of quantitation (LLOQ) for PL was 15.62 ng/mililiter.

Generation of the Ptenloxp/loxp:PB-Cre4 (Pten-KO) mouse

Mice were generated in our laboratory by crossing Pten floxed (loxp/loxp) with Probasin-Cre (PB-Cre4⁺) as described (23). Both of the mice were on the C57/BL6J background. Pten floxed (loxp/loxp) mice from Jackson Labs were screened for the floxed 328 bp band and/or wild type 156 bp band by using the Fwd IMR9554: caa gca ctc tgc gaa ctg ag and Rev IMR9555: aag ttt ttg aag gca aga tgc. Probasin-Cre (PB-Cre4) from the NCI Mouse Repository was screened for the 393 bp transgene by using the following primers: Fwd P021: ctg aag aat ggg aca ggc att g and Rev C031: cat cac tcg ttg cat cga cc. The animals were bred and maintained at the Animal Resources Facility of the University of Wisconsin-Madison. All of the animal protocols were approved by the University’s Research Animal Resources Committee in accordance with the NIH Guideline for the Care and Use of Laboratory Animals.

Study design to determine the effects of PL on the development of PCa in intact Pten-KO mice

A total of 100 intact Pten-KO mice were used to determine the effect of PL on prostate tumor growth. Mice were divided into three groups control (n = 40), PL (200 ppm) (n = 20), and PL (500 ppm) (n = 40). PL treatment was started at 4 weeks of age and continued until the mice were sacrificed. PL was mixed with the powder diet (8604 Harlen Tekland Rodent Diet (Madison, WI) in a food processor for 10 minutes and poured into a glass cup and replaced with a fresh PL-mixed diet at every alternate day. The control group of mice was fed with diet alone. Mice were sacrificed at 15 and 30 weeks and examined for prostate tumor growth.

Study design to determine the effects of PL on the growth of PCa in castrated Pten-KO mice

A total of 26 mice were used to determine the effect of PL on castration resistant prostate cancer (CRPC) in Pten-KO mice. At 10 weeks of age, all of the mice were castrated by removal of both of the testicles following the standard operation procedure. All of the protocols were approved by our University Institutional Animal Care Use Committee (IACUC). Three days after castration, the mice were divided into two groups (10 mice in each group) and were fed with the normal anti-oxidant free diet and the diet containing PL as described previously. In a parallel group, 10 intact Pten-KO mice were given the normal diet and served as control for castrated Pten-KO mice. All of the mice from both groups were sacrificed at 50 weeks, and the prostates were dissected and weighed. Part of the prostate tumor tissues were processed for tissue sectioning.

CT/PET imaging

CT/PET imaging of three untreated and three PL treated Pten-KO mice were performed using a tumor specific radio-pharmaceutical agent (¹²⁴I-NM404). Details of the method for CT/PET imaging were described previously (24).

Histopathological examination

Part of the prostate from the control and PL-treated mice was excised and processed for histology as described previously (25). Dr Weixiong Zhong, MD, pathologist, Department of Pathology, UW-Madison, examined all the slides for histopathology.

Western blot analysis

Portions of the excised PCa tissues from each group of mice were used to prepare whole tissue lysates as described (21). Fifty micrograms of cell lysate were fractionated on 10–15% Criterion precast SDS-polyacrylamide gels (Bio-Rad Laboratories, Hercules CA).

Immunofluorescence

Paraffin-fixed prostate tumor tissue sections (4-μm thick) from control and PL-treated mice were used to determine the expression of vimentin and slug. Detail methods of immunofluorescence are described previously (21). All of the sections were examined with an Olympus Microscope attached with fluorescence detector.

PKCε kinase activity

It was analyzed by the PKC kinase activity kit obtained from ENZO Life Science, Farmingdale, NY. The procedure followed the Vender’s protocol with slight modification. Forty microliters of specific phospho PKCε antibody were used in each assay sample containing 2 μg protein of excised prostate tumor tissue. Three individual prostate tumor tissue samples of each group at 15 and 30 weeks were analyzed. Protein from normal wild type mouse prostate tissue and PKCε transgenic mouse epidermis were used as negative and positive controls, respectively. Data in the bar graph represent the mean±SE. A p value of <0.05 was considered as significant.

Cytokines array analysis

Mouse specific cytokines array analysis was performed in the serum of WT, Pten-KO control and PL-treated Pten-KO groups. It was done by a commercially available facility (Eve Technologies Corporation, Alberta, Canada). In brief, blood was collected from the retro-orbital plexus of WT, Pten-KO control and PL-treated Pten-KO groups. Serum was isolated by centrifugation of the blood at 5000 rpm at 4°C and stored at −80°C °until it was sent for analysis. The serum of 3 mice from each group was sent for analysis. Cytokines concentrations in mice serum were expressed in pg/ml.

Statistical analysis

Prostate and urogenital organs were excised and weighed upon sacrifice. The normality of the prostate weight data was assessed; log-transformed values were found to conform to normality assumptions better than raw values and subsequent analyses used log-transformed measures. Relationships between treatment arms and continuous data such as prostate weights were tested using the Wilcoxon rank-sum test where there were two treatment arms (0 ppm and 500 ppm), and with the Jonkheere-Terpstra test where there were three treatment arms (0 ppm, 200 ppm, and 500 ppm). Association between treatment and dichotomous data such as PIN or carcinoma differentiation was tested with the Cochran-Armitage test for trend. Computations were performed with SAS (1A) and R software (2A) and figures were created with R software (21).

Results

Oral dose finding and bioavailability study of PL in C57BL/6 wild type mice

In this experiment (Fig. 1A), twenty (6 weeks old) wild type (C57BL/6) mice were used. Mice were randomized into five different cohorts (i.e. 4 treatment cohorts and 1 control cohort). Mice were fed with PL (500, 1000, 2000, 4000 ppm) in the diet for 6 weeks and evaluated for weekly body weight change. Body weights of the PL fed mice were compared to the control mice. Body weight loss greater than 15% was considered as visual sign of toxicity in mice. Results indicate that PL fed in diet was tolerable up to 2000 ppm in wild type mice (Fig. 1A). However, mice fed 4000 ppm of PL exhibited visual toxicity, which was observed by loss of body weight (Fig. 1A). We next evaluated the bioavailability of PL in mouse plasma by the LC-MS/MS method as described (26). PL was detected approximately 50 ng/ml in the plasma of mice fed 2000 ppm dose. However, PL was not detected in the plasma of mice given 500 ppm dose (Fig. 1 Ai–ii).

Figure 1. Maximum tolerated dose and bioavailability of PL in C57BL/6 wild type mice.

A. Maximum tolerated dose (MTD) of dietary plumbagin in C57/BL/6 mice. Mice were randomized into five different cohorts (i.e. 4 treatment cohorts and 1 control cohort). Mice were fed with PL (500, 1000, 2000, 4000 ppm) in diet for six weeks. Mice body weight changes were recorded weekly. B. Bioavailability of PL in plasma of C57BL/6 mice as determined by LC-MS/MS. Blood was collected 12 h after stop giving PL in the diet. i) Chromatogram showing 125 ng/mL PL standard at retention time 4.87 min. Honokiol was used as internal standard as indicated by red peak in chromatogram at retention time 5.9 min. ii) Chromatogram showing peak of PL in plasma of mice fed with PL (2000 ppm).

PL treatment inhibits prostate tumor growth in intact Pten-KO mice

Prostate specific conditional Pten-KO mice provide a unique model to define the mechanism of resistance to androgen ablation therapy in a genetically defined model where the initiating oncogenic event is not androgen dependent. These mice have a significantly shortened latency of prostatic intraepithelial neoplasia (PIN) formation, which results in PCa progression to a metastatic stage, mimicking the disease progression in humans (23). We investigated whether dietary administration of PL inhibits prostate tumor development in Pten-KO mice. In this experiment, 100 intact Pten-KO mice were divided into three groups (control (n = 40), PL (200 ppm) (n = 20), and PL (500 ppm) (n = 40). PL treatment was started when the mice were 4 weeks old, the time when hyperplasia began, and were sacrificed at 15 and 30 weeks. Development of PCa in control and PL-treated Pten-KO mice was evaluated by examining prostate tumor volume and weight. PL administration (200 and 500 ppm) resulted in a dose-dependent significant (P<0.001) decrease in the urogenital apparatus (Fig. 2Ai) and prostate tumor weights (Fig. 2Aii) at 15 weeks. We analyzed the prostate tumor volume of 3 control and 3 PL-treated mice by CT/PET imaging, using tumor specific radio-pharmaceutical agent (¹²⁴I-NM404) at 30 weeks (Fig. 2Bi–ii). PL treatment resulted in a significant (P<0.001) decrease in prostate tumor volumes compared to control mice (Fig. 2Bii). PL treated mice showed less uptake of ¹²⁴I-NM404 compared to control mice which were correlated with decreased tumor volumes (Fig. 2Bi). PL treatment also resulted in a dose-dependent significant (P<0.001) decrease in the weight of urogenital apparatus (Fig. 2Ci) and prostate tumor (Fig. 2Cii) at 30 weeks. We next examined the effects of PL on genitourinary apparatus of wild type (WT) littermates of Pten-KO mice. In this experiment, a total of 20 WT mice (4–6 weeks old) were used. Mice were divided into two groups (control (n = 10), and PL (500 ppm) (n = 10). Mice were sacrificed at 15 and 30 weeks for examining genitourinary apparatus. Results demonstrated that PL treatment did not affect the weight of prostate and genitourinary apparatus of wild type mice at 15 and 30 weeks (Fig. 2Di–iii) compared to control. PL treatment (500 ppm) up to 30 weeks did not show any toxicity in Pten-KO mice as confirmed by histopathological analysis of liver, lungs kidney and spleen (Fig. 2Ei–iv).

Figure 2. Effect of PL on the growth of prostate tumors in Pten-KO mice.

A total of 100 intact Pten-KO mice were used to determine the chemopreventive effects of PL administration in the diet. Mice were divided into three groups (Control (n = 40), PL (200 ppm) (n = 20), and PL (500 ppm) (n = 40). PL was mixed in the antioxidant free powdered diet and given to mice at 4 weeks of age and continued until being sacrificed. Prostate tumor development in Pten-KO mice was analyzed by weight and volume of prostate tumors. A. Representative pictures of excised genito-urinary apparatus of control and PL treated mice at 15 weeks (i). Box plot represents weight of excised genito-urinary apparatus (ii) and prostate tumors (iii) at 15 weeks. Data in the box plots represent mean±SE of control (n = 20), PL 200 ppm (n = 10), and PL 500 ppm (n = 20) mice at 15 weeks. B. Hybrid microPET/CT images of Pten-KO mice were acquired 48 h post-intravenous injection of ¹²⁴I-NM404. Representative hybrid microPET/CT images at 30 weeks of mice fed control and PL diet (i). Green arrow indicates the primary prostate tumor. Bar graph indicates prostate tumor volume determined by microPET/CT imaging. Each value in the graph is the mean±SE from 3 mice (ii). C. Representative pictures of the excised genito-urinary apparatus of control and PL treated mice at 30 weeks (i). Box plot represents weights of excised genito-urinary apparatus (ii) and prostate tumors (iii) at 30 weeks. Data in the box plots represent mean±SE of control or PL-fed mice at 30 weeks. D. Effect of PL (500 ppm) in the growth of prostate and UGA of C57/BL6 wild type (WT) mice. A total of 20 mice (4 weeks old) were used. Half of the mice were fed PL (500 ppm). Mice were sacrificed at 15 and 30 weeks and their prostate and UGA weight were recorded. Representative picture of UGA of control and vehicle and PL treated WT mice at 15 (i) and 30 (ii) weeks. Bar graph represents the weight of UGA and prostate of control and PL treated WT mice. Values in bar graph are mean ±SE of 5 mice in each group (iii). B. Representative pictures of H&E stained sections of lung (i), liver (ii), kidney (iii), and spleen (iv) of 30 weeks old Pten-KO mice administered with PL (500 ppm) in the diet.

PL treatment inhibits progression of invasive adenocarcinoma and cystic dilation in the prostate of Pten-KO mice

To determine whether PL treatment inhibits progression of PIN to invasive adenocarcinoma in Pten-KO mice, histopathological analyses were performed in excised prostate tissues of both the control and PL-treated mice as described by Kaplan-Lefko et al. (27). H&E-stained mice prostate sections were microscopically examined and classified as focal PIN, diffuse PIN, small well differentiated carcinoma, large well differentiated carcinoma, and well differentiated carcinoma with cystic dilation. We randomly selected 10 areas of each mouse prostate section for giving the histopathological score. Representative images of H&E-stained sections of control and PL-fed mice prostates are shown in Fig. 3Ai–iii and Fig. 3Bi–iii. At 15 weeks, all of the control mice prostate showed diffuse PIN and large well differentiated carcinoma, whereas they were significantly (P<0.001) decreased in PL-treated mice in a dose-dependent manner (Fig. 3Ai–iii & Fig. 3Ci–iii). Desmoplastic changes were observed in the prostate stroma of control mice at 15 weeks (Fig. 3Ai) which was significantly reduced in the prostate stroma of PL-treated mice (Fig. 3Aii–iii). None of the mouse either from control or PL-treated groups showed well differentiated carcinoma with cystic dilation at 15 weeks. At 30 weeks, both 200 and 500 ppm PL-treated mice displayed significantly decreased incidence of diffuse PIN (P = 0.008), invasive adenocarcinoma (P<0.001), and invasive adenocarcinoma with cystic dilation (P<0.001) compared to control mice (Fig. 3Bi–iii & Fig. 3Di–iii). We also performed histopathological analysis of liver, lungs and lymph nodes of control Pten-KO mice (n=30) and found none of the mice with PCa metastasis (data not shown).

Figure 3. Effects of PL on the progression of invasive adenocarcinoma in Pten-KO mice.

Histopathological analyses of excised prostate tumor tissues at age 15 and 30 weeks. Representative photographs (final magnification proximity 20X and 200X) showing H&E staining of excised prostate tumors from control and PL-treated mice at 15 (A) and 30 (B) weeks. Green and black arrows indicate PIN and invasive adenocarcinoma respectively. Yellow arrows indicate cystic dialation along with adenocarcinoma. C. Histopathological analysis results of excised prostate tumors of PL-treated and untreated mice at 15 and 30 weeks are summarized in bar graphs.

PL treatment inhibits constitutive expressions of AKT, PKCε, Stat3, COX-2 and decreases serum IL-6 level in Pten-KO mice

AKT serine/threonine kinase is one of the primary targets of the PTEN-controlled signaling pathway (28). Thus, AKT phosphorylation serves as a reliable indicator of PTEN loss. We observed an increased expression of pAKTSer473 in the prostate tissues of Pten-KO mice compared to WT as analyzed by Western blot analysis (Fig. 4Ai–ii). At 30 weeks, PL treatment resulted in an inhibition of AKT phosphorylation in prostate tissues of Pten-KO mice compared to control mice (Fig. 4Ai–ii).

Figure 4. Effect of PL on AKT, PKCε and Stat3 activation and the expression levels of COX-2 in excised prostates of WT and Pten-KO mice.

A. Protein levels of pAKTSer473, PKCε, pStat3Tyr705, pStat3Ser727, total Stat3 and COX-2, as determined by Western blot analysis of prostate tissue lysates at 30 weeks. Details are described in materials and methods. Histogram represents the quantification of the blots by using the mean value of each group (ii). B. Effect of PL-administration on PKCε kinase activity in the prostate tissues of WT, control Pten-KO and PL treated Pten-KO mice at 15 and 30 weeks. WT denotes prostate of wild type mice. PC denotes positive control where, 1 μg protein was used from the epidermis of PKCε transgenic mice. C. Serum IL-6 level of WT, Pten-KO, and PL treated Pten-KO mice at 15 and 30 weeks was determined by serum cytokines analysis as described in material and methods. Values in bar graph of Fig. B and Fig. C represent mean±SE of 3 mice in each group.

PKCε is a transforming oncogene and is involved in the induction and progression of various types of cancers (29, 30), including prostate (26, 31–33). We have previously reported that the expression level of PKCε and Stat3 correlates with human PCa aggressiveness (31–34). We determined the effects of PL on the expression of PKCε and Stat3 in excised prostate tissues of control and PL-treated Pten-KO mice. Western blot analysis results demonstrated increased expression of PKCε and Stat3 in the prostate of Pten-KO mice compared to the prostate of wild type (WT) littermates (Fig. 4Ai–ii). PL treatment resulted in an inhibition of PKCε expression in the prostate tissues of Pten-KO mice at 30 weeks compared to the control (Fig. 4Ai–ii). We further analyzed the kinase activity of PKCε in prostate tissues of WT, control, and PL-treated Pten-KO mice. Results demonstrated significant (P<0.01) increased PKCε kinase activity in the prostate tissues of Pten-KO mice compared to WT littermates at 15 and 30 weeks (Fig. 4B). We observed a dose-dependent decrease in PKCε activity in the prostate tissues of PL-treated Pten-KO mice compared to the control (Fig. 4B). We also observed increased Stat3 phosphorylation at both Ser727 and Tyr705 residues in the prostate tissues of Pten-KO mice compared to WT littermates (Fig. 4Ai–ii). PL treatment resulted in inhibition of both Stat3 phosphorylation at both Ser727 and Tyr705 residues compared to the control (Fig. 4Ai–ii). Accumulating evidence has suggested the link of COX-2 in PCa progression (35). A recent study has shown inhibition of PCa bone metastasis by COX-2 inhibitor (36). We observed increased expression of COX-2 in prostate tissues of Pten-KO mice compared to the prostate of WT littermates (Fig. 4Ai–ii). PL treatment resulted in inhibition of COX-2 expression compared to control (Fig. 4Ai–ii). We observed an increased serum IL-6 level in Pten-KO mice compared to WT littermates, which was significantly (P<0.01) reduced in PL-treated mice (Fig. 4C). PL did not show any significant effects in other cytokines (Exotoxin, IL-1α, IL-2, IL-4, IL-5, IL-9, and IL-10) (data not shown).

PL inhibits constitutive expression of vimentin and slug in Pten-KO mice

Vimentin and Slug have been shown to be overexpressed during the EMT in most tumor types including PCa (37–39). They have also been considered as markers of EMT (40). A possibility was explored whether PL administration inhibits EMT in the PCa tissues of Pten-KO mice. We observed an increased expression of both vimentin and slug in the prostate tissues of Pten-KO mice compared to WT littermates (Fig. 5Ai–ii). Western blot results demonstrated inhibition in the expression of both vimentin and slug in the prostate tissues of PL-treated mice (30 weeks) compared to the control (Fig 5Ai–ii). We next performed immunofluorescence analysis of vimentin and slug in prostate tissue sections of control and PL-treated Pten-KO mice and observed decreased expression of both vimentin (Fig. 5Bi–iv) and slug (Fig. 5Bv–viii) in the prostate tissues of PL-treated mice compared to the control.

Figure 5. Effect of PL on the expression of vimentin, and slug in prostate tissues of Pten-KO mice.

A. Western blot analysis showing the expression of vimentin and slug from 30 weeks old Pten-KO prostate tissue lysates from indicated groups (i). Histogram represents the quantification of blots by using the mean value of each group (ii). B. Representative immunofluorescence images of vimentin (Bi–iv) and slug (Bv–vi) expression in the prostate tissues of WT, control and PL treated Pten-KO mice at 15 weeks. Yellow arrows indicate the vimentin and slug expressions.

PL inhibits CRPC in Pten-KO mice

It has been shown that Pten-KO mice develop CRPC. We next determined the effects of PL administration on CRPC. In this experiment, twenty six Pten-KO mice were castrated at 10 weeks and divided into two groups. One group of mice was administered PL (500 ppm) in the diet, and the other group of mice was fed with the control diet. Both groups of mice were sacrificed at 50 weeks. Prostate tumor development was studied in both groups of mice by determining tumor weight and histopathological analysis. We observed that castration of Pten-KO mice inhibited prostate tumor growth as assessed by prostate weight and histopathological analysis (Fig. 6A–C). We observed a significant (p<0.001) decrease in prostate weight in PL administered mice compared to control mice (Fig. 6B). Histopathological analysis of PL treated mice prostate tissues showed a significant (P<0.01) decrease of both invasive adenocarcinoma and cystic dilation compared to untreated castrated mice (Fig. 6C).

Figure 6. Effect of PL on the growth of prostate tumors in castrated Pten-KO mice.

A total of 26 Pten-KO mice were used and divided into two groups. Both groups of mice were castrated at 10 weeks. PL treatment (500 ppm) in the diet was started one week post-castration and continued until 50 weeks. Control groups of mice were fed with normal powder diet. All of the mice were sacrificed at 50 weeks. Development of prostate tumor was examined by prostate tumor weights and histopathological analyses. A. Representative pictures of excised genito-urinary apparatus of control and PL-treated castrated Pten-KO mice at 50 weeks. B. Bar graph represents weights of excised prostates at 50 weeks. Values in the graphs represent mean±SE of control (n = 13) and PL-treated 500 ppm (n = 13) mice. C. Representative photographs showing H&E staining of the excised prostate tissues of control (Ci–ii) and PL-treated (500 ppm) castrated Pten-KO mice.

Discussion

Because of the heterogeneous nature of PCa, it is important to test the pre-clinical efficacy of a chemopreventive agent in various mouse models. PL is a unique plant-derived napthoquinone and has been known for its health benefits against various types of ailments including cancer (9–22). Previous reports from our laboratory have demonstrated that PL treatment inhibits i) >90% human PCa cells DU-145 ectopic xenograft tumors in athymic nude mice (20), ii) the growth and metastasis of highly aggressive human PCa cells (PC-3M) (21) and iii) prostate carcinogenesis in TRAMP mice (22). A recent study has shown that the combination of PL and androgen withdrawal regress mouse PCa cells (PTEN-P2) derived allograft tumors in mice (41). All of these effects have been shown when PL was administered i.p 1 or 2 mg/kg body weight. In this study, we report, for the first time, that PL administration in the antioxidant free powdered diet inhibits tumor development in intact and castrated Pten-KO mice. PL has also been extensively evaluated for its toxic side effects in rodents which includes diarrhea, skin rashes, hepatic (42) and reproductive toxicity (43). The LD50 for these side effects depends upon the administration method. In mice it was 8–65 mg/kg body weight for oral administration and 16 mg/kg body weight for i.p. administration. All of these toxic side effects were dose-related and it is noteworthy that they were not observed at doses (2 mg/kg body weight i.p., and 500 nmoles topical application) reported to elicit chemopreventive and therapeutic effects (17). In our findings, we observed 2000 ppm of PL in the diet is the maximum tolerated dose in mice. These mice showed normal body weight gain and exhibited no signs of toxicity in vital organs. Hsieh et al., have reported a single dose pharmacokinetic study of PL in rats (26). In this study, rats were dosed with either 3 mg/kg i.v or 100 mg/kg orally, demonstrating an area under the curve (AUC) of 18.76 (min μg/mL) for the i.p and 272 (min μg/mL) for the oral formulation (26). In our study, we detected approximately 50 ng/ml in the plasma of mice orally administered with the 2000 ppm dose of PL. These results indicate that PL is moderately bioavailable, when given orally. PL was not detected in the plasma of mice fed with PL (500 ppm). This may be due to low dose of PL administration.

PL administration to Pten-KO showed significant (P<0.001) inhibition of prostate tumor weight, volume, and invasion. Thirty weeks old Pten-KO mice illustrated well differentiated carcinoma along with cystic dilation which was significantly (P<0.001) reduced in PL-administered Pten-KO mice. Our data also indicate no toxicity of PL treatment (500 ppm) in the vital organs of Pten-KO mice. These results indicate the potential chemopreventive effects of PL against prostate carcinogenesis. Long-term use of PL has toxic side-effects including reproductive toxicity (44). Although, PL has reproductive toxicity but the old age patients suffering from primary and invasive PCa may not have this issue. Our data clearly indicate that PL may be used in the prevention and treatment of both primary as well as invasive prostate carcinoma.

It has been shown that PL is a natural multi-targeting agent which targets several signaling pathways associated with the induction and progression of PCa (19–22). Accumulating evidence suggests that PKCε is an oncogene and plays an important role in the induction and progression of various types of cancers (28, 29), including PCa (25, 29–32). Overexpression of PKCε is sufficient to promote conversion of PCa androgen-dependent (AD) LNCaP cells to androgen-independent (AI) variant, which rapidly initiates tumor growth in vivo in both intact and castrated athymic nude mice (45). We have shown previously that the PKCε expression level correlates with the aggressiveness of human PCa (30). We also have shown that genetic loss of PKCε in TRAMP mice prevents development and metastasis of PCa (25). A recent study has suggested that the overexpression of PKCε in the mouse prostate epithelium promotes the development of prostatic intraepithelial neoplasia (PIN) at 16–18 weeks (32). PKCε has also been shown to cross-talks with various signaling pathways in PCa (46–49). We have also shown that PKCε associates with Stat3, and this association increases with PCa development and progression in human and mice (31). We have shown that constitutive inhibition of PKCε inhibits Stat3 phosphorylation in vitro (34) and in vivo (25, 29). These results prompted us to determine the effects of PL administration on the expressions of PKCε, and Stat3 in excised prostate tissues of Pten-KO mice. Our results indicate that PL-administration inhibits PKCε and Stat3 activation in prostate tumors of Pten-KO mice. These results are in accord with our previous published reports where PL treatment showed inhibition of PKCε and Stat3 in PCa (20–22). Development of PCa is not confined to the prostate epithelial cells, but also involves the tumor microenvironment (TME). Multiple signaling pathways, growth factors and cytokines exist between epithelial cells, stromal cells, and the extracellular matrix to support tumor progression from the primary site to regional lymph nodes and distant metastasis (50). Our cytokines array data indicate increased serum IL-6 level in Pten-KO mice compared to WT littermates. These findings are in accord with previous published reports showing an association between IL-6 and poor prognosis of PCa (51–54). Our data showed a significant decrease in the serum IL-6 level in PL-administered Pten-KO mice. These findings correspond to our previous published report where PL treatment inhibited the serum IL-6 level in pancreatic cancer cells derived xenograft mice (19). From these results, we cannot rule out the possibility whether these molecules are the direct molecular targets of PL or the indirect consequence reflecting the suppressed tumor growth in the treatment groups. However, in our previous study, we have shown in vitro that PL treatment of PCa cells (DU-145) for 6 hours inhibits expression of PKCε, and phosphorylation of Stat3 and AKT (20). It may be possible that these molecules are the direct molecular targets of PL in PCa.

EMT is a biological process by which the normal epithelial cell acquires a mesenchymal phenotype (50). This EMT process helps cancer epithelial cell migration from primary tumor to distant metastatic sites. Several regulatory and specific biomarkers, including vimentin and slug have been shown to be modulated during EMT (50). Vimentin expression has been shown in poorly differentiated PCa and bone metastases of PCa (37). Other studies have shown constitutive overexpression of vimentin in highly aggressive androgen independent PCa cell lines (PC-3M), compared to androgen-dependent LNCaP cell lines (39). Further, it has been shown that constitutive inhibition of vimentin inhibits invasion of PC-3M cells (54). Several other studies also supported the view that vimentin is over expressed in PCa and contributes to their invasive and metastatic potentials (37–40). Our results indicate that PL targets vimentin in the PCa tissues of Pten-KO mice. Previous study has shown that PKCε is involved in phosphorylation of vimentin (55). Therefore, it may be possible that PL inhibits vimentin expression via inhibition of PKCε in PCa cells. Slug, a member of the Snail family of zinc-finger transcription factors (17), was identified as a potential oncogene in various types of cancer (18–21). The role of slug in the induction of EMT is well defined in various types of cancer, including PCa (56–58). Slug induces both androgen and non-androgen transactivation of androgen receptor signaling pathways in PCa (59). Our findings indicate that PL inhibits snail expression in the PCa tissues of Pten-KO mice. Overall, these results indicate that PL inhibits the induction of EMT in PCa via inhibition of vimentin and slug expression.

In summary, our results clearly indicate that PL inhibits prostate tumor development in intact and castrated Pten-KO mice. The anti-tumor potential of PL in Pten-KO mice could be partly due to the inhibition of PKCε, IL6/Stat3 signaling pathways, and EMT. These results further provide evidence that PL is a potential chemopreventive agent against PCa. We suggest that PL alone or in in combination with androgen ablation should be tested in clinical trials against CRPC.

Abbreviations

PC

Prostate cancer

PKCε

Protein kinase C epsilon

Stat3

Signal transducers and activators of transcription 3

EMT

Epithelial to mesenchymal transition