PEITC (indicated purity 99%), dimethylsulfoxide (DMSO), and propidium iodide were purchased from Sigma-Aldrich Chemical Co. (Milwaukee, WI). Powdered AIN-93M diets with or without PEITC [3 μmol/g diet (17)] were prepared by Research Diets (New Brunswick, NJ) and stored at −20°C until weekly feedings.

LNCaP human prostate cancer cells were obtained from the American Type Culture Collection (Manassas, VA) and grown in RPMI-1640 supplemented with 2 mM L-glutamine, 10% fetal bovine serum (FBS), 100 U/ml penicillin, and 100 μg/ml streptomycin (Invitrogen, Carlsbad, CA) at 37°C in a 5% CO₂ atmosphere.

All experimental protocols were in accordance with the National Institutes of Health guidelines and were approved by the USDA Animal Research Advisory Committee. Male athymic nude mice (BALB/c nu/nu, 20–22 g, 5–6 weeks old; Charles River, Frederick, MD) were individually housed in filter-top cages at the USDA BHNRC animal facility and consumed food and fresh tap water ad libitum. Food consumption and body weights were recorded weekly. After an acclimation period of 1 week, during which mice were fed control AIN-93M diet, the mice were randomized into 2 experimental groups, with 19 animals in the control group and 22 animals in the PEITC treatment group. Two weeks later, LNCaP human prostate cancer cell xenografts were established in the mice by injection s.c. in the flank with LNCaP cells (2×10⁶ cells) in 50 μl of phosphate-buffered saline (PBS) plus 50 μl Matrigel (BD Biosciences, Mansfield, MA). This produced palpable tumors within 4–5 days. Cancer preventive efficacy of the treatments was assessed once a week by measuring tumor volume (cm³) = 0.523 × [length (cm) × width² (cm²)] (18). Mice remained on their respective diets for 7 weeks after cell injection, until tumors reached 2–3 cm³ in volume, and the animals were then sacrificed. A portion of tumor tissue was fixed in 10% neutral buffered formalin for in situ immunohistochemical analysis and quantitation of proliferation, apoptosis, and angiogenesis as previously described (19). A portion of tumor tissue was quickly frozen in liquid nitrogen and stored at −70°C for mRNA and protein analysis as described below.

LNCaP cells (1–5×10⁵ cells/well) were plated in 24-well plates (Corning Life Sciences, Corning, NY) and 24 h later were treated with 0 (vehicle, DMSO), 0.1, 1, or 5 μM PEITC for 0–75 h, with the medium replaced every 24 h. Cell growth was analyzed using the sulforhodamine B (SRB) assay as described previously (22). Results are expressed as mean absorbance plus or minus standard error of the mean (mean ± SE). For cell attachment assays, LNCaP cell were harvested by trypsinization and re-suspend in media containing 0, 0.5, 1, 2.5 or 5 μM of PEITC at 1×10⁵ cells/ml. Cells were immediately plated in 24-well-plates (1×10⁵ cell/well). After overnight incubation, media were removed from the wells and the wells were wash once with PBS, fixed with 10% trichloroacetic acid. Cells attached to culture plate surface were determined by SRB method (22).

LNCaP cells were plated in duplicate at a density of 1×10⁵ cells/well in 6-well plates (Corning), which were incubated for 24 h and subsequently synchronized by culturing without serum for 24 h; results are from two independent experiments. Based on the results from the cell growth assays, the cells were then treated with PEITC at 5 or 10 μM for 24, 48, and 72 h. The cells were fixed and stained with propidium iodide for analysis by flow cytometry (FACSCalibur, Mansfield, MA), and the results were evaluated using FlowJo software (BD Biosciences, San Jose, CA).

LNCaP prostate cancer cells were plated in 6-well plates (1×10⁶ cells/well) and 24 h after plating were switched to Media B [RPMI-1640 medium without phenol red (Invitrogen), 2 mM L-glutamine, 100 U/ml penicillin, 100 μg/ml streptomycin, and 10% charcoal dextran-treated FBS (Hyclone, Logan, UT)] to minimize the effect of serum hormones. Twenty-four hours later, the medium was replaced with fresh medium containing 1 nM dihydrotestosterone (DHT) with or without 0–5 μM PEITC. The medium was changed daily to fresh medium with test compounds, and after 48 h cells were harvested for total RNA isolation using the TRIzol method (Invitrogen) as described previously (22). For determination of mRNA expression in tumor samples, tumor total RNA was isolated using the TRIzol (Invitrogen) method following the manufacturer's protocol. Gene expression was quantified by the TaqMan real-time RT-PCR method as described previously (22). TaqMan gene expression assay primers and probes for human prostate specific antigen (PSA) and glyceraldehydes-3-phosphate dehydrogenase (G3PDH), proliferating cell nuclear antigen (PCNA), Ki-67, integrin β1 (ITGB1), integrin α2 (ITGA2), integrin α5 (ITGA5), integrin α6 (ITGA6), mouse and human vascular endothelial growth factor (VEGF) A, mouse PECAM-1 and mouse and human TATA binding protein were purchased from Applied Biosystems (Foster City, CA). G3PDH was used as a housekeeping gene for calculation of relative expression levels in vitro and mouse or human TATA binding protein was used for tumor samples.

Tumor tissue from 4 mice in each dietary treatment group was lysed with T-PER Tissue Protein Extraction Reagent (Pierce, Rockford, IL) supplemented with protease inhibitors (Complete Mini Protease Inhibitor Cocktail Tablets (Roche Applied Science, Indianapolis, IN), 50 mmol/l sodium fluoride, and 1 mmol/l sodium orthovanadate) and homogenized. Total cell lysate proteins (30 μg) were separated using 4–20% or 8–16% pre-cast denaturing polyacrylamide Tris-glycine gels (Invitrogen) and transferred by iBlot machine (Invitrogen) onto PVDF membranes (Invitrogen). The membrane was probed or with antibodies (all from Santa Cruz Biotechnology, Santa Cruz, CA) against human caspase-3 (clone H-277), human VEGF (clone 147), diluted 1:200 overnight at 4°C, and developed using the WesternBreeze Chemiluminescent Immunodetection Kit (Invitrogen) according to the manufacturer's instructions. Two blots were run for each antibody. Blots were stripped and re-probed with anti-β-actin (Chemicon International, Inc., Temecula, CA) as a loading control. Immunoreactive bands were visualized by X-ray film (Kodak X-Omat MR-1) and digitized using a flatbed scanner (Hewlett-Packard model 4850, Hewlett-Packard, Palo Alto, CA) according to the manufacturer's protocol. The intensity of the band was determined using ImageJ software (http://rsb.info.nih.gov/ij/).

In our in vivo experiments StatView (SAS Institute, Cary, NC) software was used for statistical analysis. Repeated measures multivariate analysis of variance with Wilks' λ as the test statistic was used to compare differences between control and PEITC groups in time-to-appearance of palpable tumor and rate of tumor growth between control and PEITC groups over time. The 19–22 animals used per group gave >97% power to detect effects and provided ≥85% power to detect differences as small as 35%. For our in vitro experiments, the GraphPad PRISM 4 (GraphPad Software Inc. San Diego, CA) was used. The unpaired Student's t-test was used to compare experiments with two groups. ANOVA followed by Bonferroni's multiple comparison post hoc test was used to examine differences in PSA mRNA levels between DHT-treated and DHT-non-treated groups.

Dietary administration of PEITC (3 μmol/g diet) did not significantly affect body weight or food consumption of treated mice compared to control mice (Fig. 1). However, during the final weeks of the experiment mice in both the control and treated groups experienced a slight decrease in body weight and food consumption as the LNCaP prostate cancer cell xenografts grew in size in both the control and treated groups. Each mouse in the treatment group consumed ~100–150 mg/kg body weight/day of PEITC. Four mice from each group died prior to study termination. Overall, administration of PEITC to athymic nude mice did not present any observable toxicity as determined by pathological analysis of kidney and liver from the mice (data not shown).

Fig. 2 illustrates the temporal effects of dietary PEITC on tumor growth in mice fed either control or PEITC diet. We observed a significant delay in tumor growth in mice consuming PEITC compared to control diet. The difference was observed at 6 and 7 weeks after injection of LNCaP prostate tumor cells.

We evaluated whether the reduced rate of tumor growth with PEITC treatment was due to induction of apoptosis or a decrease in proliferation in the tumor. As illustrated in Fig. 3A, dietary treatment of mice with PEITC did not lead to changes in apoptosis index in the tumor as determined by TUNEL assay, nor did it affect the intensity of cellular proliferation marker PCNA (Fig. 3B). In addition to cell proliferation and apoptosis, we examined the effects of PEITC on androgen-dependent pathways, as androgen is a known risk factor for prostate cancer and is associated with increased proliferation and decreased apoptosis. We found the androgen responsive gene PSA protein levels in the tumor were not affected by PEITC as determined by immunohistochemistry (Fig. 4A). Additional analyses using real-time PCR and Western blot analyses were performed on tumor samples to confirm the results of immunohistochemical analysis. We also observed no differences in mRNA levels of PSA (Fig 4B), the proliferation markers PCNA (Fig. 5A) or Ki-67 (Fig. 5B) between control (n=4) and PEITC-treated (n=4) tumor samples. We found no change in pro-apoptosis protein Bax mRNA levels (Fig. 5C). We detected pro, but not processed (cleaved) caspase-3 in the tumor samples by Western blot analysis, and densitometric quantitation of the bands revealed no significant differences between control and PEITC-treated mice (Fig. 6A).

In addition to cellular proliferation and apoptosis, the tumor microenvironment may also affect prostate tumor growth. Therefore, we examined the effects of PEITC on microvessel density. As illustrated in Fig. 7, treatment with PEITC resulted in significantly lower microvessel density, as indicated by PECAM-1/CD31 staining intensity in the tumor. We were not able to detect the mouse PECAM-1 using Western blot analysis. However, real-time PCR analysis showed no difference in mouse PECAM-1 mRNA expression between control (n=4) and PEITC-treated (n=4) tumor samples (Fig. 5D). We also found no difference in mouse or human VEGF mRNA expression between control (n=4) and PEITC-treated (n=4) tumor samples (Fig. 5E and F) or in mouse or human VEGF expression using immunohistochemical detection (data not shown). There were no differences in human VEGF protein expression between control and PEITC-treated tumor samples as determined by Western blot analyses (Fig. 6B).

Given the small response of PEITC on angiogenesis we further asked what other pathways may account for inhibition of tumor growth using LNCaP cell culture model. Consistent with the lack of effects on cell proliferation and apoptosis in vivo, PEITC exerted minimal effects on LNCaP cells in culture. Concentrations of PEITC ≤5 μM had no effect on the growth of the LNCaP cells in vitro (Fig. 8A). In addition, cell cycle analysis showed no effect of PEITC on cell cycle progression or appearance of sub-G0 DNA, an indicator of apoptosis. Also consistent with in vivo results, PEITC did not affect DHT-induced increase in PSA mRNA levels in vitro (Fig. 8B). Interestingly, we found as illustrated in Fig. 9A, in the presence of PEITC there was a significant inhibition of LNCaP cell plating efficiency. This effect occurred at relatively low (0.5 μM) PEITC concentration. The integrin family proteins are known to play important roles in cell adhesion/attachment and prostate cancer development (23). We therefore asked whether alteration in integrins may correlate with a decreased in LNCaP cell plating efficiency. We found mRNA expression of integrins ITGB1, ITGA2 and ITGA6 was significantly inhibited in the PEITC (5 μM) treated LNCaP that attached to cell culture plate as compared to vehicle control (Fig. 9B). By contrast, ITGA5 mRNA expression was not affected by PEITC treatment (Fig. 9B).