EFFECTS OF BLACK RASPBERRIES AND THEIR CONSTITUENTS ON RAT PROSTATE CARCINOGENESIS AND HUMAN PROSTATE CANCER CELL GROWTH IN VITRO

Abstract

Prostate cancer patients often use dietary supplements, such as black raspberries, which are a rich source of compounds with antioxidant and anticancer activity, particularly on gastrointestinal cancers. Feeding black raspberries inhibited mammary cancer induction in rats and growth of cancer cells in nude mice, indicating systemic bioavailability of bioactive compounds. We tested whether feeding black raspberries and its constituents would inhibit prostate cancer development. However, we did not find preventive effects in two rat prostate carcinogenesis models, even though the berry anthocyanin metabolite protocatechuic acid was detectable in their prostates. Black raspberry extract, the anthocyanin cyanidin-3-rutinoside and protocatechuic acid did not inhibit prostate cancer cell growth in vitro, but ellagic acid and its urolithin A metabolite did at high concentrations. Prostate cancer cell migration was not affected by these agents nor was growth in soft agar, except that ellagic acid reduced colony formation at physiological concentrations and protocatechuic acid at high concentrations. Low bioavailability of bioactive berry compounds and metabolites may limit exposure of tissues such as the prostate, since we found that cyanidin-3-rutinoside was not bioavailable to prostate cancer cells, but its aglycone cyanidin was and inhibited their growth. Thus, black raspberries are unlikely to prevent prostate cancer.

Introduction

Prostate cancer is the most frequently diagnosed male cancer in the US and, despite advances in early detection and treatment modalities, remains the second leading cause of cancer-related death among US men. In 2019, 174,650 new cases of prostate cancer are anticipated to occur in the U.S., accounting for 20% of new cancers in American men (1). Considering the high prevalence and death rates, strategies to reduce incidence and progression of this malignancy are urgently needed.

The use of dietary agents has long been considered an appealing approach to cancer prevention. Numerous laboratory studies have reported anti-cancer activity of various dietary constituents and phytochemicals, supporting the concept that increased consumption of bioactive compounds through dietary modification or supplementation can reduce cancer incidence (2). For supplements based on black raspberries (BRB; Rubus occidentalis) there is an increasing amount of evidence in support of their chemopreventive potential. BRB preparations and extracts can inhibit in vitro cancer cell proliferation, induce apoptosis, and inhibit angiogenesis (3). The most compelling evidence supporting the use of BRBs as a chemopreventive agent has been demonstrated in studies of gastrointestinal (GI) cancers (4). Dietary administration of BRBs inhibited tumor initiation as well as tumor promotion-progression in the esophagus and colon in animal models of carcinogenesis (5-8). Oral administration of BRBs also inhibited mammary carcinogenesis and bladder xenograft tumor growth in animal models (9-12), suggesting that biologically active berry constituents are systemically bioavailable following oral administration and have the ability to reach a distant tumor site. Additionally, BRBs were well tolerated and had protective effects against GI tumors in several human clinical trials (13).

BRBs contain numerous bioactive phytochemicals including anthocyanins, ferulic acid, ellagitannins, β-sitosterol, and quercetin, and several other, nonnutritive compounds, as well as vitamins and minerals (14). The potent anticancer effects of BRBs have been attributed to their high concentration of ellagic acid and anthocyanins (15-17). BRBs are one of the highest dietary sources of anthocyanins (5–20 mg/g dry weight) and ellagic acid (1.5-2 mg/g dry weight) (18,19). Ellagic acid (EA) is a polyphenolic compound derived by hydrolysis in the gastrointestinal tract from ellagitannins found in fruits and nuts such as pomegranates, berries and walnuts (20). Ingested EA is further metabolized in the gut to yield dibenzopyran-6-one derivatives (urolithins), mainly urolithin A (UroA) (21). EA and urolithins have potent antioxidant activity, inhibit cancer cell migration, and reduce cell proliferation in vivo and in vitro (21-23). Anthocyanins are naturally occurring flavonoids that provide blue, purple, and red pigmentation to a variety of fruits and vegetables, including berries, grapes, apples, and purple cabbage (24,25). They have growth inhibitory, anti-inflammatory, and antioxidant activity, which makes anthocyanin-rich foods, like BRBs, an attractive option for the prevention and treatment of cancer (26-28). Cyanidin-3-rutinoside (Cy-3-Rut) is the most abundant BRB anthocyanin, which is deglycosylated to cyanidin and then is further metabolized to protocatechuic acid (PCA), the major anthocyanin metabolite (24,29).

Although BRBs and their constituents have been shown to prevent carcinogenesis at a number of organ sites, their effects on prostate cancer development has not been evaluated. Based on the abundant literature demonstrating anti-cancer activity of BRBs and evidence that BRBs can inhibit tumor development systemically, we hypothesized that BRBs would prevent prostate carcinogenesis and inhibit proliferation of prostate cancer cells, and that these effects are predominately mediated by BRB anthocyanins. The overall goals of this study were to 1) investigate the efficacy of dietary BRBs for prostate cancer chemoprevention, 2) evaluate anticancer effects of BRB extract on prostate cancer cells, and 3) identify BRB constituent compounds responsible for any inhibitory effects identified of BRBs on prostate cancer growth and development. We chose to include Cy-3-Rut because it is the most abundant BRB anthocyanin and EA because of its known anti-cancer effects (21-23).

Materials and Methods

Reagents and Chemicals

Cyanidin-3-rutinoside, cyanidin chloride, 3,3’-diethyloxacarbocyanin iodide, resazurin, sulforhodamine B, Flutamide, and carboxymethylcellulose were purchased from Sigma-Aldrich (St. Louis, MO). Ellagic acid was obtained from Indofine Chemical (Hillsborough, NJ). Protocatechuic acid waspurchased from LKT Labs (St. Paul, MN). Urolithin A was obtained from Santa Cruz (Dallas, TX). Crystalline testosterone, testosterone propionate, and 17β-estradiol were obtained from Steraloids (Newport, RI). Noble agar was purchased from Affymetrix (Cleveland, OH). DMSO, EtOH, formaldehyde, formic acid and trichloroacetic acid were obtained from Thermo Fisher Scientific (Waltham, MA). Ketamine was obtained from Henry Schein Animal Health (Dublin, OH). Xylazine was purchased from Lloyd Laboratories (Shenandoah, IA). N-methyl-N-nitrosourea was obtained from the NCI Carcinogen Repository (MRIGlobal, Kansas City, MO).

Preparation of Black Raspberry Extract

An ethanol/H₂O (80/20) soluble BRB extract was prepared from a single batch of lyophilized black raspberry powder (Berri Products, Corbett, OR) which was stored at −20°C. 200 g BRB powder was combined with 80% EtOH and stirred overnight at room temperature. The resulting mixture was vacuum-filtered. The solvents were removed to the extent possible by rotary evaporation at reduced pressure, yielding a syrup. The remaining solvent was removed under a nitrogen stream. Final weights of extracts were obtained and BRB extract were reconstituted in 100% EtOH and stored at −20°C.

Cell Lines and Cell Culture

22Rv1, LNCaP, VCaP, and PC-3 cells were obtained from American Type Culture Collection (Manassas, VA), used within 20 passages from arrival, and screened for mycoplasma contamination using a MycoAlert™ PLUS mycoplasma detection kit (Lonza Walkersville, Walkersville, MD). C4-2 cells were a generous gift from Leland Chung (Cedars Sinai Medical Center). LAPC-4 cells were a generous gift from Dr. Karen Knudsen who had obtained them from Dr. Charles Sawyers (Memorial Sloan Kettering, New York, NY) and were provided within less than 20 passages of arrival from Dr. Sawyers.

LAPC-4, LNCaP, C4-2, 22Rv1, and PC-3 cells were cultured in RPMI 1640 (Life Technologies, Grand Island, NY) supplemented with 10% FBS, 100 IU/mL penicillin, 100 μg/mL streptomycin. VCaP cells were cultured in DMEM (Life Technologies) supplemented with 10% FBS, 100 IU/mL penicillin, and 100 μg/mL streptomycin. Cells were maintained at 37°C in a humidified 5% CO₂ incubator. Twenty four hours prior to experiments, media was replaced with phenol-red free RPMI or DMEM containing 10% dextran charcoal stripped FBS, 100 IU/mL penicillin, and 100 μg/mL streptomycin.

Proliferation Assays

The sulforhodamine B (SRB) assay was performed to evaluate growth. Cells were seeded on 96-well plates at a density of 2,000 cells per well 24 hours prior to addition of test compounds to culture media. Cells were incubated for 72 hours, then cells were fixed with 1% TCA and stained with 0.057% SRB. Wells were washed with 1% acetic acid and the remaining dye was solubilized in 10 mM Tris. Optical density at 510 nm was measured with the Synergy HTX plate reader (Biotek, Highland Park, VT). Incorporation of 5-ethynyl-2’-deoxyuridine (EdU) into DNA was measured to assess cell proliferation using the Click-iT EdU Microplate Assay (Invitrogen). Cells were labeled with EdU for 2 hours then fixed. Cells were treated with buffer containing CuSO₄ and Oregon Green 488 azide for 25 min, incubated with 5% BSA for 30 min, and then incubated with anti-Oregon Green antibody conjugated to horseradish peroxidase for 30 minutes, followed by incubation with Amplex Red for 15 minutes. Fluorescence intensity (excitation: 560, emission: 590) was measured with SpectraMax M5 spectrophotometer (Molecular Devices, Sunnyvale, CA). Results are expressed as percent proliferation relative to vehicle treated control cells. Each treatment condition was replicated 3 times in a single experiment, which was repeated at least 3 times.

The resazurin reduction assay was used to as an alternative to the SRB assay to measure cell proliferation. Reduction of non-fluorescent resazurin dye to fluorescent compound resorufin can be measured as an indication of cell viability. Cells were plated and treated under the same conditions as the SRB assay. Resazurin solution (10% w/v resazurin in PBS) was added to culture media to a final concentration of 1%, and cells were incubated for 24 hrs at 37°C in a humidified 5% CO₂ incubator. Resorufin fluorescence (excitation 530-560 nm, emission 580-600 nm) was measured with the SpectraMax M5 spectrophotometer (Molecular Devices, Sunnyvale, CA). Culture media containing resazurin without cells was used as a negative control, and 100% reduced resazurin served as a positive control.

Intracellular Levels of BRB Compounds

Cells were plated and cultured to approximately 80% confluence, then treated for 4 hours with BRB extract, cyanidin, cyanidin-3-rutinoside (Cy-3-Rut), or vehicle control. Cells were harvested with trypsin, suspended in PBS and centrifuged. Cells were resuspended in 95% EtOH and 5% formic acid, lysed by 3 freeze/thaw cycles, and samples were stored at −80°C. HPLC-MS/MS was carried out using an Agilent 6410 Triple Quad mass spectrometer (Agilent Technologies; Santa Clara, CA) coupled with an Agilent HPLC 1200 series chromatographic system.

Anchorage Independent Growth

The effect of test compounds on anchorage independent growth of PC-3 cells was assessed by soft agar colony formation assay. The bottom of each well of 12-well culture dish was coated with 1 mL noble agar mixture (RPMI, 10% FBS, 0.6% agar). After the bottom layer solidified, 1 mL top agar medium mixture (RPMI, 10% FBS, 0.3% agar) containing 5 × 10³ cells with varying concentrations of BRB extract, cyanidin-3-rutinoside, ellagic acid or protocatechuic acid was added and incubated at 37°C for 3 weeks. Culture media was replaced 3 times per week. At the end of the incubation period, colonies were stained with crystal violet and the number of colonies was counted using GelCount equipment and software (Oxford Optronix, Abingdon, UK).

Wound Healing Assay

PC-3 cells were used to study cell migration due to their ability to form a monolayer when confluent. Cells were plated in 6 well plates and grown until confluent. A wound was created in the monolayer by scratching with a sterile 200 μL pipette tip. Cells were washed with PBS and fresh media containing treatments compounds was added to wells. Wound distance was measured after 0, 6, 12 and 24 hours. Distance was measured in the same field of view for each time point.

Animal Experiments

Animal care and experiments were conducted in accordance with protocols approved by the Animal Care Committee (ACC) at the University of Illinois at Chicago.

Testosterone plus Estradiol Model of Rat Prostate Carcinogenesis

Twelve week old male Noble (NBL/Crl) rats were obtained from an in-house breeding colony. Prostate cancer was induced as described by Özten et al. (30). While under ketamine-xylazine anesthesia (100 mg/kg ketamine [Henry Schein Animal Health, Dublin, OH] and 5 mg/kg Xylazine [Lloyd Laboratories, Shenandoah, IA]), each rat surgically received two Silastic tubing implants (Dow Corning, ID 0.078 inch; OD 0.125 inch) containing crystalline testosterone tightly packed over 2 cm length and one implant containing crystalline 17β-estradiol tightly packed over 1 cm length. Rats were randomly divided into three groups of 30 animals. One week after hormone implantation, rats were switched from a standard chow obtained from Harlan Teklad (currently Envigo, Madison, WI) to AIN-93M diet (control) or isoenergetic AIN-93M diets containing 5% or 10% lyophilized BRB powder at the expense of the starch component of the AIN-93M diet. The AIN-93M diet was also obtained from Harlan Teklad and was stored at 4°C; the berry powder was mixed into the diet in-house using a Patterson-Kelly mixer and was stored at −20°C until fed freshly three times per week. Rats were euthanized when moribund or surviving for 48 weeks.

MNU plus Androgen Model of Rat Prostate Carcinogenesis

10-12 week old male Wistar-Unilever (WU:CpbHsd) rats were obtained from Harlan Sprague-Dawley (Currently Envigo, Madison, WI). Prostate cancer was induced as described by McCormick et al. (31) with slight modifications. Rats were given a sequential treatment of Flutamide (Sigma Aldrich) suspended 50 mg/ml in a 2% carboxymethylcellulose solution by gavage for 21 days at a dose of 10 mg/kg/day, followed on day 22 by a single injection of 100 mg/kg testosterone propionate (oil suspension, 50 mg/ml), and followed 52-56 hours later by a single intraperitoneal injection of N-methyl-N-nitrosourea (dissolved in a citrate-phosphate buffer pH 4.8 and diluted in saline at a dose of 30 mg/kg). Two weeks later, each rat surgically received (ketamine-xylazine anesthesia) two subcutaneous Silastic tubing implants (see above) containing crystalline testosterone tightly packed over 3 cm length. One week after MNU, rats were started on AIN-93M diet (control) or isoenergetic AIN-93M diets containing 5% or 10% lyophilized BRB powder (see above). Rats were euthanized when moribund or surviving for 56 weeks.

Necropsy and Histology

At necropsy, the rat accessory sex glands were excised with the urinary bladder following euthanasia through exsanguination while under ketamine-xylazine anesthesia or by cervical dislocation after CO2 inhalation and asphyxiation. Accessory sex glands, pituitaries, and all grossly observed lesions in other organs were fixed in 10% neutral buffered formalin. After fixation, the ventral prostate, dorsolateral prostate and anterior prostate plus seminal vesicles were dissected and processed then embedded in paraffin wax. From the ventral prostate and grossly observed tumor masses one section was made. From all other accessory sex gland tissues step sections were prepared at 250 micrometer intervals. All sections were stained with hematoxylin and eosin as described by McCormick et al. (31). All prostate lobes and other accessory sex glands were evaluated histopathologically and the presence, type, and size of all lesions were scored, using previously published criteria (31-33).

Determination of Prostatic Levels of Protocatechuic Acid

Twelve week old male NBL rats were randomized into three groups, each with 5 animals, and fed AIN-93M control diet, or isoenergetic diets containing 5% or 10% lyophilized BRB powder. Animals were euthanized after three weeks. Dorsolateral prostate tissue was removed immediately following euthanasia and flash frozen in liquid nitrogen and stored at −80°C until analysis. Protocatechuic acid levels were measured using LC-MS/MS by Dr. Zhongfa Liu (Ohio State University, Columbus, OH) as previously described (34) following shipping on dry ice.

Data Analysis

All in vitro experiments were performed at least 3 times. Results are expressed as mean ± standard deviation (SD), unless otherwise specified. Comparisons were performed using Student’s t test or one way analysis of variance (ANOVA) followed by a post-hoc test when appropriate. Differences in lesion incidence among the accessory sex glands were analyzed using Fisher’s exact test (for two group comparisons) and X² analysis (for three group comparisons). Statistical analysis was performed with GraphPad. A p value <0.05 was considered significant.

Results

Effects of Black Raspberries on the Proliferation of Prostate Cancer Cells

We used a panel of six prostate cancer cell lines to measure effects of BRB extract, bioactive constituents (Cy-3-Rut, EA), and metabolites (PCA, UroA) on cell growth and viability: LAPC-4, LNCaP, C4-2, 22Rv1, VCaP, and PC-3 cells. These cell lines were selected to evaluate possible differential effects of BRB compounds on cells with varying growth rates and molecular characteristics (see Supplementary Table 1). Cells were treated with BRB extract, Cy-3-Rut, PCA, EA, UroA, or appropriate vehicle controls for 72 hours and proliferation was measured by SRB assay. BRB extract (Figure 1A), Cy-3-Rut (Figure 1B) and PCA (Figure 1C) had no significant effect on proliferation of any of the prostate cancer cell lines. EA at a concentration of ≥10 μM significantly inhibited proliferation of VCaP and PC-3 cells and 30 μM EA also significantly inhibited proliferation of LAPC-4, C4-2, and 22Rv1 cells (Figure 1D). Urolithin A at a concentration of ≥10 μM significantly inhibited proliferation of 22Rv1 and PC-3 cells and 30 μM urolithin A significantly inhibited proliferation of C4-2 cells [the effects on LAPC-4, VCaP, and 22Rv1 cells were not determined] (Figure 1E). Note: The effects of EA and urolithin A on C4-2, PC-3, and 2Rv1 cells were previously reported by us (35).

Figure 1.

Effects of black raspberry extract and constituents on proliferation of prostate cancer cells. Proliferation was measured in six prostate cancer cell lines using the SRB assay after 72 hours treatment with 1–1,000 μg/mL BRB extract (A), 1–100 μM cyanidin-3-rutinoside (B), 1–10,000 ng/mL protocatechuic acid (C), 1-30 μM ellagic acid (D), and 0.1-100 μM urolithin A. Data were and normalized to vehicle control treated cells represented as mean ± SD of 3-4 experiments; * significantly different from vehicle control; p < 0.05. N/A = not analyzed.

BRB Extract, EA, and PCA Reduce Anchorage-Independent Growth of PC-3 Cells, but do not Inhibit their Migration

Anchorage-independent growth is associated with progression and metastasis of solid tumors, and often involves the downregulation of E-cadherin (36,37). Studies have shown that dietary BRBs can upregulate E-cadherin expression in the colon of both humans and mice (38,39). Therefore, we investigated the effects of BRBs on anchorage independent growth of prostate cancer cells using a soft-agar colony formation assay with PC-3 cells, the only cells in our panel that grew efficiently in soft-agar. Treatment with BRB extract significantly reduced the number of colonies compared to control at concentrations of 10 μg/mL and above (Figure 2). EA inhibited colony formation at 1 μM and higher. PCA inhibited colony formation only at the highest concentration tested (10 μg/mL). Cy-3-Rut did not inhibit colony formation at any concentration tested.

Figure 2.

Effects of black raspberry extract and constituents on anchorage independent growth of PC-3 cells. Data represent mean ± SD; * significantly different from vehicle control; p < 0.05.

Cell migration is a critical step in the metastatic process that enables cancer cells to invade surrounding tissue and disseminate from the primary tumor site (40,41). In vitro studies have shown that BRBs extract and EA can inhibit cell migration of vascular endothelial cells (42,43). Thus, we examined the effect of BRB extract, EA and PCA on migration of PC-3 cells using a wound healing assay but did not find significant differences in rate of wound closure compared to vehicle control (Figure 3).

Figure 3.

Black raspberry extract and constituents do not inhibit migration of prostate cancer cells: Migration of PC-3 cells was measured by wound healing assay following 24 hour incubation with 100 μg/mL BRB extract (A), 3 μM EA (B), and 10 μg/mL PCA (C) or vehicle control. Percent wound closure was calculated by normalizing to distance at 6, 12 and 24 hours to distance of wound opening at 0 hours. Data represent average of three experiments ± SD.

Anthocyanin Uptake

Given the lack of growth inhibition by Cy-3-Rut (the predominant BRB anthocyanin) in our prostate cancer cells in vitro, we speculated that anthocyanin uptake could be limited in these cells. Anthocyanins are large, bulky compounds, which require active transport to cross the cell membrane or metabolism to smaller anthocyanidin aglycones that can enter the cell by diffusion (44). To evaluate anthocyanin uptake in prostate cells, we treated 22Rv1 cells with Cy-3-Rut (30 μM), cyanidin (aglycone of Cy-3-Rut; 30 μM), BRB extract (1 mg/mL), or vehicle, and then measured intracellular levels of cyanidin and Cy-3-Rut by HPLC-MS/MS. Since we expected uptake of anthocyanins to be low based on literature reports, we used high concentrations to increase sensitivity of intracellular detection.

We observed high intracellular levels of cyanidin in cells treated with cyanidin compared to vehicle. Intracellular levels of cyanidin were not increased in cells treated with Cy-3-Rut or BRB (Figure 4A). No intracellular levels of Cy-3-Rut above those in vehicle treated cells were detected in cells exposed to Cy-3-Rut or BRB extract (Figure 4B). Levels of cyanidin and Cy-3-Rut in the cell culture media were quantified to verify that compounds were present in media. Cyanidin was as expected detected in cyanidin-supplemented media and Cy-3-Rut was detected in media supplemented with both Cy-3-Rut and BRB extract (Figures 4C & 4D); this confirmed that cells were exposed to similar levels of cyanidin and Cy-3-Rut across treatment groups and differential intracellular levels are not attributable to differences in exposure levels.

Figure 4.

Differential uptake of cyanidin and its aglycone by prostate cancer cells: Intracellular levels of cyanidin (A) and cyanidin-3-rutinoside (B) were measured in 22Rv1 following incubation with vehicle control, cyanidin (30 μM), cy-3-rut (30 μM), or BRB extract (1 mg/mL). Measurement of cyanidin (C) and cyanidin-3-rutinoside (D) in cell culture media. Data represent average of three experiments ± SD; * significantly different from vehicle control; p < 0.001 (ANOVA followed by Tukey-Kramer multiple comparisons test)

Prostate Cancer Cell Growth is Inhibited by Cyanidin

Because we observed cellular uptake of cyanidin, but not Cy-3-Rut, we investigated effects of cyanidin on proliferation of our prostate cancer cells. Due to the intense color of cyanidin, we could not perform an SRB assay to assess proliferation as previously done with Cy-3-Rut. When present in cell culture media, uptake of cyanidin results in a blueish-purple “staining” of cells, which is visibly apparent with the naked eye and interferes with colorimetric readings in the SRB assay. Therefore, we performed a fluorescent-based resazurin reduction assay as an alternative method of measuring cell viability. 22Rv1, C4-2 and PC-3 cells were treated with Cy-3-Rut or cyanidin. Similar to results of the SRB assay, Cy-3-Rut did not inhibit viability at any concentration in any of these cell lines (Figure 5A). By contrast, cyanidin significantly inhibited viability of all cell lines at 30 and 100 μM (Figure 5A). Additionally, we examined the effects of cyanidin and Cy-3-Rut on S-phase activity by measuring incorporation of a thymidine analog into DNA of C4-2 cells. Consistent with results of the resazurin assay, we observed a reduction in EdU incorporation in cells treated with 30 and 100 μM cyanidin, indicating an inhibition of S-phase activity by cyanidin, whereas Cy-3-Rut had no effect at any concentration (Figure 5C).

Figure 5.

Cyanidin but not cyanidin-3-rutinosude inhibits growth of prostate cancer cells: Viability of 22Rv1, C4-2 and PC-3 cells was inhibited by cyanidin but not by Cy-3-Rut in a resazurin reduction assay (A). Likewise, cyanidin rather than Cy-3-Rut reduces S-phase activity in C4-2 cells (B). Data represent mean ± SD (error bars in panel A are omitted to make the figures more clear); * significantly different from vehicle control; p < 0.05.

BRBs are not Protective against Rat Prostate Carcinogenesis

Two rat models of prostate carcinogenesis were used to evaluate the chemopreventive efficacy of dietary BRBs. Rats were fed AIN-93M control diet or AIN-93M supplemented with 5 or 10% lyophilized BRB powder. In the WU rat model in which prostate cancer is induced by MNU and chronic low-dose testosterone there were no statistically significant differences in incidences of neoplastic lesions in any of the accessory sex glands, although there was an apparent, but non-significant, shift from large (>5 mm) to small (<5 mm) tumors (p = 0.065; Table 1). In the Noble (NBL) rat model of induction of prostate cancer by chronic testosterone plus 17β-estradiol treatment we also did not observe any differences among the three groups in total cancer incidence and prostate cancer multiplicity was also not affected by the BRB treatment (Table 2).

Table 1.

Effect of feeding lyophilized BRBs on induction of cancer of the accessory sex glands of WU rats by sequential treatment with MNU and testosterone.

Group	1	2	3
Treatment (% BRB in diet)	0%	5%	10%
Effective number of animals	31	30	32
	Number (%) of rats:
All Accessory Sex Glands Combined (dorsolateral and anterior prostate plus seminal vesicle):
Grossly observed tumors at necropsy	16 (52) a,b	14 (47) a	8 (25) a,b
Adenocarcino(sarco)ma, All (with or without C.I.S.)	23 (74)	ND	23 (77)
Microscopic (< 5 mm)	6 (19)c	ND	12 (38) c
Macroscopic (> 5 mm)	17 (55) c	ND	11 (34) c
Adenocarcinoma + C.I.S.	26 (84)	ND	24 (75)
Carcinoma in situ (C.I.S.) only	3 (10)	ND	1 (3)
Dorsolateral Prostate Region (originating from dorsolateral or anterior prostate or seminal vesicle):
Adenocarcino(sarco)ma, Macroscopic size	5 (16)	ND	6 (19)
Anterior Prostate/Seminal Vesicle Region (originating from anterior prostate or seminal vesicle):
Adenocarcino(sarco)ma, All	6 (19)	ND	3 (9)
Microscopic (< 5 mm)	0	ND	1 (3)
Macroscopic (> 5 mm)	6 (19)	ND	2 (6)
Dorsolateral plus Anterior Prostate (clearly confined to these glands; w or w/o seminal vesicle lesions):
Adenocarcinoma, Microscopic (with or without C.I.S.)	7 (23)	ND	8 (25)
Adenocarcinoma + C.I.S.	10 (32)	ND	11 (34)
Carcinoma in situ (C.I.S.) only	3 (10)	ND	3 (9)
Seminal Vesicle Only (clearly confined to this gland; w or w/o C.I.S. in dorsolateral/ anterior prostate):
Adenocarcinoma, All (with or without C.I.S.)	4 (13)	ND	6 (19)
Microscopic (< 5 mm)	2 (6)	ND	5 (16)
Macroscopic (> 5 mm)	2 (6)	ND	1 (3)

^aTrend: p = 0.031 (X² Test);

^bDifference: p = 0.039 (2-sided Fisher Exact Test);

^cShift from large to small tumors: p = 0.065 (1-sided Fisher Exact Test); p = 1.30 (2-sided Fisher Exact Test); C.I.S. = Carcinoma in situ; ND = Not Determined

Table 2.

Effect of feeding lyophilized BRBs on induction of prostate cancer of NBL rats by treatment estradiol and testosterone.

Group	1	2	3
Treatment (% BRB in diet)	0%	5%	10%
Number of animals	30	30	30
Number of rats with prostate carcinomas (%)	28 (93)	27 (90)	30 (100)
Mean number of carcinomas per cancer-bearing rat ± SD	2.28 ± 1.01	2.12 ± 1.02	2.77 ± 1.17

SD = Standard Deviation

The Anthocyanin Metabolite PCA is Detectable in Rat Prostate Following Dietary Administration of BRBs

The tissue distribution of BRB metabolites has not been comprehensively evaluated by others. Therefore, we determined whether the major anthocyanin metabolite PCA could be identified in prostate tissue following dietary administration to NBL rats of diets containing 0, 5 or 10% lyophilized BRB powder for three weeks. We detected PCA at levels of 3.64 ± 1.35 and 4.60 ± 1.92 ng/g wet tissue in the dorsolateral prostate of rats fed 5% and 10% BRB diet, respectively, whereas PCA was below the detection limit of 0.5 ng/g tissue in rats fed a control diet. This observation demonstrates that at least one BRB metabolite is bioavailable to the rat prostate following oral exposure to BRBs.

Discussion

Black raspberries have been demonstrated to have anti-cancer effects in GI cancers, including many examples of efficacy in vitro and in animal models as well as in human clinical trials. By contrast, there is relatively little known about the effects of BRBs on organs outside of the GI system. In the present study, we provide the first information of BRB effects on prostate carcinogenesis in animal models and several prostate cancer cell lines.

The main finding is that dietary BRBs are not able to interfere with prostate cancer induction and development in two different rat models, despite bioavailability of the anthocyanin metabolite PCA to the rat prostate following dietary BRB administration. However, prostatic levels of PCA obtained through dietary administration of BRBs were not sufficient to inhibit induction of prostate cancer in our study. Although it has been shown that PCA may play a role in cancer prevention via induction of apoptosis, inhibition of inflammation, and reduction of oxidative damage (45), concentrations of PCA necessary to elicit these effects are much higher than what can be achieved through dietary consumption. In our in vitro studies, we tested a wide range of PCA concentrations and did not see any effects on cell proliferation and migration, and inhibition of colony formation only at a high concentration (10 μg/mL), but not at concentrations comparable to those detected in prostates of NBL rats (<5 ng/mL).

The absence of preventive effects of BRBs in our animal experiments is in contrast to prior studies showing efficacy of dietary BRBs in prevention of estrogen-induced mammary carcinogenesis in ACI rats (9-11) and growth of xenografted bladder cancer cells in nude mice (12). In the ACI rat model, diets supplemented with 2.5% BRB powder resulted in a 69% reduction of tumor volume and a 37% reduction in tumor multiplicity, but did not affect tumor incidence (10). A higher BRB dose of 5% did not further reduce tumor volume, but did significantly delay the appearance of tumors (11). Additionally, BRBs reduced expression of CYP1A1 and CYP1B1, enzymes which are involved in converting estradiol to DNA-damaging catechol metabolites, suggesting that the protective effects of BRBs in this model may be attributed to a reduction in genotoxic metabolites via altered estrogen metabolism (9). To date, all mammary tumor prevention studies with BRBs have been carried out in female ACI rats with hormone-induced mammary cancer. However, in a study investigating the effects of EA, the incidence of carcinogen-induced (7,12-dimethyl benz[a]anthracene; DMBA) mammary tumors was not significantly reduced (46). The latter observation may support the notion that BRBs affect mechanisms that are specific for the ACI rat model and are not pertinent to the two prostate carcinogenesis models we used and the DMBA mammary cancer model in Sprague Dawley rats (46). This may be associated with estrogenic activity of urolithins and anthocyanins (47,48). While it has been shown that estradiol exposure plays a role in prostate carcinogenesis in the NBL rat model, additional mechanisms, including inflammation, genotoxicity, and testosterone-induced proliferation, may contribute to the development of prostate cancer in this model (32,49,50). Alternatively, the carcinogenic stimulus of high estradiol levels in NBL rats may have been too strong to allow mild protective effects of BRBs. Studies in the ACI rat model have shown that the magnitude of BRBs efficacy is dependent on the level of estradiol exposure; 2.5% BRBs reduced mammary tumor incidence induced by chronic low-dose estradiol implants, but did not reduce incidence when a higher dose of estradiol was used to induce carcinogenesis (9,10). BRB intervention reduced tumor volume and burden in both approaches, but effects were greater in animals exposed to low-dose estradiol. Thus, differences among animal models in their response to dietary BRBs related to differences in carcinogenic mechanisms and how BRBs can interfere with those, as well as rat strain-specific differences may be responsible for the divergent results of the preventive effects of BRBs on internal sites other than the GI tract.

The effectiveness of BRBs is often attributed to their high anthocyanin content (15-17). We evaluated the effects of Cy-3-Rut, the most abundant BRB anthocyanin, on growth of prostate cancer cell lines using multiple proliferation assays. Contrary to our hypothesis, we found no effect on proliferation of prostate cancer cells by Cy-3-Rut up to 100 μM. This result was observed consistently across cell lines, reducing the likelihood that these findings are an artifact of experimental design or is due to cell line- or assay-specific characteristics. Cy-3-Rut also had no effect on anchorage-independent growth of PC-3 cells. These results suggest that anthocyanins are not effective inhibitors of prostate cancer.

Because of the lack of growth inhibitory effects of Cy-3-Rut we evaluated the idea that low Cy-3-Rut uptake by prostate cells was precluding inhibitory effects on cell growth. Indeed, culturing 22Rv1 cells in media supplemented with BRB extract or Cy-3-Rut did not lead to intracellular accumulation of Cy-3-Rut or its aglycone (cyanidin). But when cells were treated with cyanidin directly, we observed robust intracellular levels of this compound, as well as reduced proliferation of 22Rv1, PC-3 and C4-2 cells at high concentrations. Conceivably, the metabolism of anthocyanins to more bioavailable aglycone metabolites may be critical in their bioavailability and effects on cancer cells. Impaired uptake and activity of anthocyanins, relative to their aglycones, has been previously reported and is thought to be caused by hindrance due to the presence of bulky glycosidic moieties. In one study comparing the incorporation of several anthocyanins into endothelial cells, an inverse association between anthocyanin uptake and the number of glycoside groups was found (51). In another study investigating protective effects of anthocyanins against DNA damage, cyanidin and delphinidin, but not their glycosides, were effective against lipid peroxidation, suggesting that protective effects of anthocyanins are diminished by the presence of bulky sugar moieties (52).

Considering the results of our in vitro anthocyanin uptake experiments, it is not surprising that we did not observe inhibitory effects in vivo. The inability of BRBs to prevent prostate cancer in our study is conceivably due to low tissue bioavailability. Both anthocyanins and their aglycones are very unstable under physiological conditions, rendering them subject to rapid metabolic modification and/or degradation (53,54). In the digestive tract, anthocyanins are deglycosylated either by gut microflora or by epithelial enzymes (such as β-glucosidase and lactase-phlorizin hydrolase) or they can be taken up intact by glucose transporters and metabolized by intracellular β-glucosidases (44,55-59). Following absorption, anthocyanins can undergo further hepatic modification to glucuronidated, methylated, or sulphated forms before entering circulation (60-62). Overall, the extensive metabolism and rapid degradation of anthocyanins are consistent with poor systemic bioavailability. Concordantly, studies have demonstrated that anthocyanins are detectable at only very low plasma levels after ingestion, with phenolic degradation products appearing soon after ingestion (57,59,63). Intact anthocyanins (cyanidin-3-glucoside) have been identified in prostate tissue of animals fed an anthocyanin-enriched diet, but at only very low levels (1–3% of amount administered in the diet) that are likely insufficient to be biologically active (64,65). This, together with our observation of very low intracellular levels of Cy-3-Rut and cyanidin in prostate cancer cells following in vitro exposure to Cy-3-Rut suggesting that prostate cells lack the ability to transport or metabolize anthocyanins. Even if anthocyanins could reach prostate tissue through the circulation, they would likely lack chemopreventive efficacy.

In contrast to the ineffectiveness of Cy-3-Rut, we did observe growth inhibitory activity of EA on prostate cancer cells. EA significantly reduced viability in five of the six cell lines tested, having the greatest impact on the two most aggressive cell lines (22Rv1 and PC-3), suggesting that EA may preferentially target rapidly proliferating cells types. However, inhibitory concentrations of EA (>10 μM) in this study are unlikely achievable in humans through dietary consumption of BRBs or other EA containing foods which are unlikely to exceed 1 μM in the circulation (66,67). On the other hand, in a colony formation assay EA had anchorage-independent growth of PC-3 cells at concentrations as low as 1 μM, perhaps due to effects on mechanisms associated with metastatic and invasive growth. EA also inhibited growth of human pancreatic cancer cells in nude mice treated by gavage with 40 mg/kg/day EA (68). There are reports that EA can reverse epithelial-to-mesenchymal transition by modulating expression of factors involved in cell adhesion, metastasis, and chemotaxis (68-70). BRBs also inhibited growth in soft agar, conceivably due to their EA content. The fact that EA inhibited colony formation at much lower concentrations than its cell growth inhibiting effect may be a result of prolonged exposure to EA in the colony formation assay (3 weeks) compared to the SRB proliferation assay (3 days). PCA did not affect colony formation until very high concentrations that are highly unlikely to be achievable in vivo as indicated by our finding that PCA levels in the rat prostate do no exceed 5 ng/g wet tissue following dietary BRB administration.

Bioavailability and metabolism of EA may also contribute to the inability of BRBs to prevent prostate cancer. EA has greater stability compared to anthocyanins, but very low aqueous solubility (water solubility ≃ 9.7 μg/mL), which limits absorption (71). The solubility of urolithin metabolites is greater than that of EA due to the reduction in the number of hydroxyl groups, which improves absorption but diminishes their antioxidant capacity (21,72,73). Effects of UroA on prostate cancer cell growth were similar to those of EA (35). We were unable to determine prostatic levels of EA and UroA in this study due to limitations in quantity of our rat prostate tissue available, however some information regarding prostate bioavailability of EA and its metabolites has been reported by others. Following oral administration of UroA (0.3 mg/mouse), urolithin metabolites were detected in mouse prostate tissue, but neither EA nor urolithin metabolites were detected in mouse prostate tissue after feeding a diet supplemented with an ellagitannin-enriched pomegranate extract (0.8 mg/mouse/day, containing approximately 0.03 mg EA) (74). In a clinical trial with 63 men consuming either walnuts (35 g/day, containing 210 mg EA) or pomegranate juice (200 mL/day, containing 279 mg EA) for three days, investigators were able to detect EA metabolites in the prostate tissue of only 8 of the 33 (24%) patients who consumed either walnuts or pomegranate juice and at very low levels (up to 2 ng/g UroA glucuronide conjugate) (75).

In conclusion, BRBs did not inhibit the development of rat prostate carcinogenesis and did not inhibit prostate cancer growth at physiological concentrations in vitro. It seems likely that the chemopreventive effects of BRBs observed in studies of GI cancers are attributed to localized absorption and direct contact of BRB compounds with target tissue. BRB constituents that come in contact with these tissues can exert anti-cancer effects at the site of absorption. However, instability and low bioavailability of bioactive BRB compounds and their metabolites dramatically limits circulating levels, so that distant tissues, like the prostate, are not exposed to quantities sufficient for preventing cancer. While the potential of BRBs as preventive and therapeutic agents for GI cancers and possibly breast cancer remains encouraging, our evidence indicates that BRBs are unlikely to be effective for preventing prostate cancer.