Prostate cancer, one of the most common malignancies in men in the United States and in other Western countries is showing an increasing trend.1 Genetic alterations, particularly, loss of phosphatase and tensin homolog (PTEN) associated with mutation and other age-related factors is a significant risk for human prostate cancer development.2 Although chemoprevention is one of the best options for prostate cancer prevention, the influence of phytochemicals inhibiting the process of prostate carcinogenesis is not well understood, partly because of their poor bioavailability and the rate of absorption. In spite of the inconsistent reports on the bioavailability, several phytochemicals have been studied in cell culture and animal studies for their anticancer effects.3–5 Interestingly, curcumin (diferuloylmethane), one of the phenolic compounds derived from the root of the plant Curcuma longa6–8 has been shown to induce antitumor effects in several cancer types both in vitro and in vivo.9, 10 In animal studies, curcumin has been shown to increase the survival of tumor-bearing rodents by inhibiting tumor growth and impeding metastasis.11 However, there is concern about its low bioavailability and poor absorption in the gastrointestinal tract, thereby limiting its clinical use.12 Recent studies have shown that liposomal encapusulation of curcumin may enhance systemic bioavailability.13, 14
Similarly, resveratrol, a phytoalexin found in grapes and red wines has been identified as an antioxidant and antimutagen that induces Phase II drug-metabolizing enzymes, mediates anti-inflammatory effects, reduces cyclooxygenase and inhibits hydroperoxidase functions, such as antipromotion activity.15 The chemopreventive effect of resveratrol is strongly linked to the negative regulation of androgen receptor (AR) and other nuclear transcription factors.4, 16–22 The combination of curcumin with other agents to increase chemopreventive efficacy has been studied by others.23–25 However, the effect and the mechanism of liposomal curcumin in combination with liposomal resveratrol against prostate cancer have not been investigated. Although the pharmacological profiling and bioavailability of resveratrol have been studied in few preclinical models,26, 27 the relevance of the liposomal delivery of resveratrol and its in vivo effects in human is still unclear.28
To enhance the chemopreventive effect of these 2 potential agents against prostate cancer, we tested the hypothesis that liposomes encapsulated forms of resveratrol and curcumin in combination may inhibit prostate cancer by increasing their bioavailability synergistically and enhance the anticancer effects. In this study, we determined the bioavailability of liposome encapsulated curcumin and resveratrol, individually and in combination at the serum and tissue levels; then we evaluated the inhibitory effects of these agents (liposome encapsulated curcumin and resveratrol) against prostate cancer growth and progression. Further, using cell culture studies, we also investigated whether these agents in combination, decreased the rate of tumor cell growth by abrogating a close interaction between PI3K-AKT, AR and mammalian target of rapamycin (mTOR) signaling pathways of prostate cancer involving the loss of PTEN. Findings from this study may predict similar effect by these agents in human clinical studies for prostate cancer prevention or treatment.
Curcumin and resveratrol with a purity of >98% were obtained from LKT Laboratories (St. Paul, MN) and Sigma (St. Louis, MO), respectively.
The animals study conducted for bioavailability and chemopreventive efficacy were in accordance with the approved protocols of the New York University School of Medicine Institutional Animal Care and Use Committee (IACUC).
To improve the bioavailability of curcumin, we used the liposomal preparations of curcumin and resveratrol, here after referred as liposomal (lipo) curcumin or liposomal (lipo) resveratrol. Briefly, the liposomal encapsulation of these curcumin or resveratrol (LKT Laboratories) were prepared with small aliquots of these agents (5 mg of each/kg/bw) by mixing with the liposomal lipid 1, 2-dimyristoyl-rac-glycero-3-phosphocholine (Sigma), as the liposomal agent in a standardized ratio as described earlier.13 By repeated analysis of the ratios ranging from 10:1 to 5:1, we determined the final ratio of 5:1 for optimal encapsulation of curcumin or resveratrol by liposomes. For combination, the lipo-curcumin and lipo-resveratrol (2.5 mg/kg/bw each) were mixed using a 3-way adjuvant mixer. The final products of these preparations were freeze dried overnight in dark, aliquoted and stored at −20°C and warmed up to room temperature just before use. For quality control, liposomal preparations and reproducibility validated following the protocols described earlier.13
In vivo administration of the agents
To determine the bioavailability, male B6C3F1/J mice, 6–8 weeks of age were obtained from Jackson Laboratory (Bar Harbor, ME) and housed in the Department animal facility. The animals were housed in cages with wood chip bedding in a temperature-controlled room (68–72°F) with 12-h light–dark cycle and relative humidity (45–55%) and were permitted free access to diet and drinking water. All the mice were allowed to acclimatize for a week and then randomly assigned to experimental and control groups (n = 3). After overnight fasting, the mice were administered with a total volume of 250 μl of the liposome encapsulated agents individually and in combination by oral gavage, as per the methods recommended for oral administration of natural compounds with appropriate modifications.29–32 Control mice received liposome alone. After the administration of the agents, animals from control and experimental groups were sacrificed at various time points of 0:30, 1:00, 1:30, 2:00, 3:00, 4:00, 6:00 and 12:00 hr by terminal CO2 asphyxiation. At the time of sacrifice, blood was collected by cardiac puncture and serum was obtained by centrifugation at 14,000 rpm and stored at −80°C until further HPLC analysis. Prostate from individual mice were microdissected and rapidly frozen in liquid nitrogen for further HPLC analysis and storage did not exceed 1 week. Previous studies have shown that these agents are stable in tissues under these conditions.30
HPLC analysis of serum and prostate tissues
The serum (200 μl) and prostate tissue (200 mg) were treated with sulfatase + β-glucorinidase for resveratrol before extraction, and for curcumin with methanol/ethyl acetate and the mixtures were centrifuged (2,800g 4°C for 15 min). The organic layers were removed twice, combined and evaporated to dryness under nitrogen. The plasma or the tissue samples were reconstituted in 500 μl of the mobile phase that composed of 55% acetonitrile and 45% citrate buffer (1% w/v citric acid solution adjusted to pH 3.0 using concentrated sodium hydroxide solution) and immediately analyzed using reversed-phase Breeze HPLC system (Waters Corporation, Milford, MA) available at the departmental analytical core facility. The Breeze HPLC, System consist of a Waters 1525 binary pump, 717 plus auto sampler with refrigeration unit, 2487 dual wavelength UV–vis detector and in-line degasser, and the results were calculated by the Breeze software version 3.2. Chromatographic separation were accomplished by injecting the processed serum sample (25 μl) into a Phenomenex Luna 4.6 ×150 mm, 5 μm C18 HPLC column (Phenomenex, Torrance, CA) set at 35°C with a flow rate of 1 ml/min. Resveratrol and liposomal curcumin were detected at wavelengths set at 325 and 428 nm, respectively, and the peak identification was done by HPLC with standard materials. Carbamazepine was used as an internal standard for resveratrol and 4-hydroxybenzophenone was used for curcumin as internal standard.31, 32 Standard curves were made using resveratrol and/or curcumin (Sigma).
In vivo efficacy of lipo-curcumin in combination with lipo-resveratrol
Generation of prostate-specific PTEN knockout (PTEN-KO) mice
Chemopreventive effect lipo-curcumin coadministered with lipo-resveratrol against prostate cancer was determined in a genetically modified prostate-specific PTEN-KO mouse model. This murine PTEN-KO model that was developed by deleting the PTEN tumor suppressor gene, specifically in the prostatic epithelium, is an established prostate cancer progression model, mimics the multistep tumorgenesis of human prostate cancer more closely.33 In this model, conditional homozygous deletion of PTEN in mouse prostate significantly shortens the latency of PINs and promotes their progression to metastatic cancer. These conditional PTEN-KO mice will be generated in-house at the department animal core facility by crossing PTENloxP/loxP mice with mice of the ARR2 probasin-cre transgenic line PB-cre4, wherein the Cre recombinase is under the control of a modified rat prostate-specific probasin promoter34, 35 as previously reported.33, 36 Briefly, the male ARR2Probasin-Cre transgenic mice obtained from the NIH/NCI mouse repository, (mouse.ncifcrf.gov/) were first crossed with the female B6.129S4-PTENtm1Hwu/J mice purchased from JAX Mice Services (Bar Harbor, Maine). Subsequently, cross breeding of the F1 generation offsprings resulted in homozygous deletion of PTEN in the F2 generation. PTEN deletion in these mice has been confirmed by PCR using tail DNA as described earlier.37 We used the primer sequences 5′-TCCCA GAGTTCATACCAGGA-3′ and 5′-GCAATGGCCAGTACTA GTGAAC-3′ for PCR amplification.36 Total and p-AKT (ser 473) activity in the prostate tissue lysate have been measured by Western blot analysis as described earlier.37 Only F2 male offsprings were used for the efficacy studies36 and mice with C57BL6/J background (JAX Mice Services) in which prostates were histologically normal were used as wild type control. All the mice were maintained under controlled conditions (21°C and 50% relative humidity) in a 12-h light/dark cycle and were fed with sterile irradiated food and water.
Four-weeks-old homozygous male PTEN-KO mice were randomly assigned to the experimental and control groups (n = 3). The experimental groups received lipo-curcumin (50 mg/kg/bw), lipo-resveratrol (50 mg/kg/bw) and a combination of both coadministered (25 mg/kg/bw each) by oral gavage in a total volume of 250 μl, 3 times in a week, for a period of 7 weeks. Plain liposomes were administered as lipo-control in a control group of mice. Oral treatment procedures are being used for dietary intervention studies by other investigators.38, 39 Resveratrol and lipo-curcumin doses for the present study were chosen from the dosage regimens used by earlier investigators.40–42 All the animals were inspected at least once daily to monitor their general health status and body weights were recorded twice every week. At the end of the bioassay study (after 7 weeks of treatment), all the mice were sacrificed by CO2-asphyxiation and necropsied. At the time of sacrifice, the total genitourinary (GU) tract was removed and weighed. The whole prostate was examined grossly for malignant lesions and weighed separately. The prostate was micro-dissected, weighed and fixed in 10% neutral-buffered formalin, embedded in paraffin, and 5-μm sections were used for histopathological evaluations. The prostate tumor incidence/burden was determined by examining H&E stained sections showing PIN and adenocarcinomas. The reduction in the total prostate weight, as well as the histopathological changes indicating regression of adenocarcinoma of the dorsolateral (DL) prostate in the treatment groups compared to control were recorded to determine the cancer chemopreventive effects.
In vitro assays using PTEN-CaP8 prostate cancer cells
Cell growth and treatments
To investigate the effects of curcumin in combination with resveratrol on cell growth and the underlying mechanisms, we used murine PTEN-CaP8 prostate cancer cells (kindly provided by Dr. Hong Wu, Department of Medicine, University of California, Los Angeles, CA) that were derived from the PTEN-KO mice of prostate cancer, as reported previously.33, 43 PTEN-Cap8 cells were grown and maintained in DMEM supplemented with 10% fetal bovine serum (Gibco, Invitrogen, Carlsbad, CA), 25 μg/ml bovine pituitary extract, 5 μg/ml bovine insulin, and 6 ng/ml recombinant human epidermal growth factor, as described earlier.43 We used 75% confluent cells grown in 6-well plates for trypan blue exclusion assays and/or cells grown in 96-well plates for MTT assays. For treatments, we used resveratrol (10 μM) or curcumin (10 μM) individually or resveratrol and curcumin (5 μM each) in combination.
Cell growth assays
Standard MTT and trypan blue exclusion (0.2%) assays were performed to assess cell growth inhibition as described earlier.37 Briefly, the PTEN-Cap8 cells grown in 6-well plates were treated with curcumin or resveratrol alone or in combinations for a period of 48 hr. Cells harvested (adherent and floating cells) after trypsinization were recovered by a quick centrifugation. Immediate staining of (10,000) cells/well with trypan blue enabled rapid and clear identification of dead cells with deep blue staining and the living cells with less uneven staining of the cell membrane. Total numbers of viable cells from each treatment were recorded and were compared with the untreated controls. Three independent set of experiments were performed simultaneously for each treatment.
Cell cycle analysis
To determine the effect on cell cycle regulation, treated and control PTEN-Cap8 cells were harvested after 48 hr by trypsinization and fixed in 10% neutral buffer formalin followed by fixing in 80% ethanol for 24 hr. The cells were washed in PBS and resuspended in 1 ml of 5 mg/ml of propidium iodide with 0.1% RNase A (Sigma, St. Louis, MO) in PBS to measure DNA content. After 30-min incubation at room temperature in the dark, the cells were analyzed by flow cytometry with Coulter Epic Elite ESP. Cell cycle analysis was performed as described earlier.37 To determine the significant difference in the DNA content between control and treatments, the analysis was repeated with triplicate samples.
The PTEN-Cap8 cells were treated with curcumin or resveratrol alone or in combination for 48 hr and were harvested and stained with DAPI (4′,6-diaidine-phenylindole dihydrochloride) to detect apoptotic cells as described earlier.44 Subsequently, cells from a parallel set of similar experiments were used for Annexin-V staining (using FITC-conjugated antibody) of the membrane for phosphatidylserine externalization to detect cells in the early phase of apoptosis. An AX70 microscope (Olympus) was used to detect Annexin-V positive cells with morphological changes characteristic of apoptosis and quantified with Image-Pro plus software (Media Cybernetics, Silver Spring, MD).
Transfection assays with AR-siRNA
To determine AR gene silencing effects on the molecular targets of PTEN-CaP8 cells, we used AR gene-specific siRNA duplexes along with HiPerFect transfection reagent purchased from Qiagen (Valencia, CA). Transfection efficiency was assed by immunofluorescence and Western blot analysis for protein expression as described earlier.37 Briefly, ∼106 cells were grown in 6-well plates to reach 70% confluence. The original stock of the siRNA was resuspended in siRNA suspension buffer provided by the manufacturer. The remaining part of the RNA interference (RNAi) experiments were performed as described earlier.37 For all the biochemical and molecular target analysis, we used protein and/or RNA extracted from cells exposed to siRNA. The level of AR knockdown mediated by siRNAs was compared with that of the curcumin effect in combination with resveratrol after 48-hr treatment.
Western blot analysis
Total protein extracted form PTEN-CaP8 cells treated with curcumin and resveratrol were fractionated on a 10% SDS-PAGE as described earlier.37 Briefly, we used 50 μg of each sample protein after extraction with a buffer containing 150 mmol/l NaCl, 10 mmol/l Tris (pH 7.2), 5 mmol/l EDTA, 0.1% Triton X-100, 5% glycerol and 2% SDS in addition to a mixture of protease inhibitors (Boehringer Mannheim, Mannheim, Germany). Fractionated proteins were transferred onto polyvinylidene difluoride membranes. Standard Western blot procedure was carried out using specific monoclonal antibody (1:1,000 dilutions) to detect PTEN, p-Akt (ser-473) AR and D-type cyclins and mTOR (Santa Cruz Biotechnology, Santa Cruz, CA). The level of β-actin expression was used as reference for equal loading. Reactive protein bands were detected using the ECL kit (Amersham, Piscataway, NJ). Quantification of the reactive protein bands were performed by densitometry as described earlier.37
Serum and tissue bioavailability of the agents treated were monitored over time using repeated measures of Tukeys pairwise and time-line comparisons performed with ANOVA.45 Differences in body weight, total prostate weight, and histologically identified mPIN and adenocarcinoma of the PTEN-KO mouse prostate between treatment groups and control were compared using 1-way ANOVA followed by Tukey's multiple comparison procedure.45 Similar statistical tests were performed to determine the significance of differences between treatments pertinent to cell growth, apoptosis and other molecular biomarkers. All statistical analyses were performed using GraphPad Prism 4 software (San Diego, CA).
Enhanced bioavailability of curcumin when coadministered with resveratrol
Lipo-curcumin has been reported to increase its systemic bioavailability in vivo in contrast to plain curcumin.13, 14 In this study, we used liposomal preparations to deliver curcumin in order to increase its bioavailability in combination with resveratrol (coadministered) in male B6C3F1/J mice. The HPLC analysis of curcumin concentrations in the serum and prostate tissue are presented in Figures 1a and 1b. The overall findings indicate an increase in the level of both lipo-resveratrol and lipo-curcumin in the serum and prostate tissue. Interestingly, oral administration of lipo-curcumin alone in mice showed serum concentration of 100 ng/ml after 1.5 hr. However, the mice that received lipo-curcumin coadministered with resveratrol, showed a higher serum concentration of curcumin 252 ng/ml; and this level was stable up to 4 hr in the range between 245–238 ng/ml in contrast to the plain curcumin (Fig. 1a). A similar analysis of the whole prostate tissue extract from these mice showed curcumin level increasing from 21 ng/ml (curcumin alone) to 151 ng/ml (lipo-curcumin) after 3 hr (>7-fold increase, p < 0.001); and this was stable up to 6 hr in the range between 151 and 148 ng/ml (Fig. 1b).
Curcumin in combination with resveratrol modulate body weight and inhibit prostate cancer incidence
An overall weight gain was observed in the control and experimental mice. The experimental mice that were administered with either lipo-curcumin or lipo-resveratrol (50 mg/kg each) and their combination (25 mg/kg) for 7 weeks showed no signs of toxicity, but had a small decrease in total weight gain as compared with the control group (Fig. 2a). Interestingly, our findings reveal that the mean GU tract and prostate weights were remarkably reduced in mice that received lipo-curcumin coadministered with resveratrol (Fig. 2b). A significant decrease in the prostate weight in the combination treatment group was associated with histological changes indicating a decrease in the tumor growth (inhibition of epithelial cell proliferation) in the DL prostate clearly indicate a decrease in the incidence of mPIN lesions and regression of adenocarcinomas (Fig. 2c). Total number of adenocarcinomas determined in the combination treatment group was reduced from 400 to 110 (p < 0.001) (Fig. 2d). It is also important to mention here for the readers that liposome alone has no effect on the tumor development or incidence or any other in vivo effect.
Inhibitory effects of curcumin in combination with resveratrol on cell growth and cell cycle
To investigate the molecular mechanism(s) underlying the in vivo effects of curcumin in combination with resveratrol, we performed in vitro assays using PTEN-CaP8 cells deleted for PTEN. Cells treated with DMSO served as the vehicle control. The percentage of viable cells detected with trypan blue exclusion assay after 48 hr of treatment with curcumin in combination with resveratrol showed a significant decrease in the cell viability (p < 0.001) (Fig. 3a). An overall increase in Annexin-5 staining to detect apoptosis was also confirmed with DAPI staining simultaneously to detect nuclear DNA fragmentation indicated higher number of apoptotic cells. More than 5-fold increase (55–62%) in the rate of apoptosis was significant in cells treated with combination of these 2 agents (Fig. 3b). Interestingly, cell cycle analysis of these cells exposed to curcumin in combination with resveratrol (5 μM of each) revealed a G1 peak associated with a distinct pre-G1-peak indicative of apoptotic cells (Figs. 3c1 and 3c2).
AR-siRNA and combination of curcumin plus resveratrol effects on cell cycle regulatory proteins
To determine whether modulation of AR is one of the mechanism(s) underlying the anticancer effect of curcumin in combination with resveratrol, we compared the AR-siRNA effects with that of the agent simultaneously. We used PTEN-Cap8 cells for RNAi experiments. After 48 hr of transfection with AR-siRNA, we observed a significant decrease in cyclin D1; and this was associated with a negative effect on cell growth. A sharp increase in the expression of the cell cycle regulatory proteins, p21 (Cip1) and p27 (cdk inhibitor) supports the view that the cell cycle regulatory proteins could be the direct targets controlled by AR and could be modulated by these agents of interest. Accordingly, the cells treated with curcumin in combination with resveratrol showed a similar effect as that of siRNA by decreasing the protein levels of AR and cyclin D1, associated with an increase in p21 and p27 (Fig. 4).
Curcumin in combination with resveratrol modulate AR and mTOR signaling
In this study, we further extended our analysis to investigate the effect of curcumin in combination with resveratrol, on AR/mTOR signaling pathways. Findings from Western blot analysis indicate a significant inhibitory effect induced by these agents in combination on p-Akt (ser 473), AR, cyclin D1 and mTOR proteins in PTEN-Cap8 cells with loss of PTEN (Figs. 5a and 5b). Part of these findings are consistent with our earlier reports on the single effect of resveratrol mediating gene–nutrient interactions by modulating p53, and cell cycle regulatory proteins, AR and cyclin D1 associated with growth inhibition and apoptosis in human LNCaP cells.16
Higher consumption of natural dietary factors has been associated with a decreased risk for cancer of GU tract.46, 47 Particularly, the potential effects of phytochemicals such as resveratrol and curcumin against prostate cancer have been already documented by us16, 48 and other investigators.49–52 However, it is not known whether these agents can modulate the genetic risk of prostate cancer and the underlying molecular targets. So far, findings from in vivo assays are not able to predict specific mechanisms because of the inconsistent findings on curcumin bioavailability.30, 53 Interestingly, recent studies are suggesting that higher concentrations of curcumin could be achieved by means of liposomal delivery.13, 14 In this study, we have shown that liposomal delivery of curcumin is effective and that this agent in combination with resveratrol (coadministered) could further increase its bioavailability and stability in the serum and tissues for a longer period, which is enough to induce anti cancer effects in mice. Based on the pharmacokinetics of curcumin in combination with resveratrol, we examined the anticancer activity determined by the rate of mPIN lesions and invasive adenocarcinoma of the DL prostate in PTEN-KO mice. Recently, modulation of prostate cancer genetic risk by dietary polyunsaturated fatty acids has been shown in PTEN-KO mice.36
Dietary administration of curcumin or resveratrol against TRAMP by other investigators has shown a significant decrease in tumor growth.22, 54 However, the doses selected in these studies are either too low (1% curcumin) or too high (625 mg of resveratrol) to compare significant biological effects. We believe that the efficacy of these agents via dietary administration could be enhanced and improved by oral administration, particularly using liposomal preparation and in combination at medium level doses that we have used in the present study. The rationale for using 50 mg of curcumin or resveratrol and half of the dose of each agent for combination in PTEN/KO model is based on its enhanced efficacy within 16-weeks period of time against prostate cancer unlike the studies conducted in TRAMP. Our findings on the enhanced bioavailability of curcumin when coadministered with resveratrol in mice further suggests a clear biological effect, possibly a synergistic interaction between these 2 phytochemicals in promoting anticancer effects. To address the question on whether resveratrol increases the bioavailability of curcumin or it is the stable lipo-curcumin that is inducing the effects, we propose several possibilities. One of the possibilities is attributed to the property of low water solubility of resveratrol; it must be bound to proteins or conjugated to remain at a higher concentration in the serum.55 We also believe that coadministration of resveratrol with lipo-curcumin may partly increase its binding with albumin in the presence of fatty acids along with curcumin.56 Albumin appears to be one of the plasmatic carriers of transporting resveratrol in blood circulation to induce biological effects. In addition, earlier studies have shown that lipo-curcumin alone modulate molecular targets and induce anticancer effects, and inhibited tumor angiogenesis in pancreatic cancer.13, 14 Our attempts to investigate the possible impact of curcumin plus resveratrol for the first time in a genetically modified mouse model for prostate cancer with loss of PTEN revealed a negative regulation of activated p-Akt, cyclin D1, AR and mTOR, suggesting that these agents can target multiple mechanism(s) of prostate carcinogenesis. Although further investigations are needed to understand more on the interactions of natural agents, our preliminary findings clearly predict the clinical relevance of these dietary components in combination to promote anticancer effects among individuals with high risk for prostate cancer due to the loss of PTEN and/or due to activated p-AKT signaling pathways.
The authors thank Drs. Hong Wu, Department of Medicine, University of California, Los Angeles, CA, and Yong Q. Chen, Comprehensive Cancer Center, Wake Forest University School of Medicine, Winston-Salem, NC, for their expert advice on PTEN knockout mice. They acknowledge Dr. Tchou-Wong Kem-Meng for her assistance with the animal study; Mrs. Sally Lasano for her help with histology; and Mr. Al Bowers for his help with the HPLC analysis at the New York University School of Medicine, Department of Environmental Medicine, Tuxedo, NY.