Development of a dietary formulation of the SHetA2 chemoprevention drug for mice

Dietary ingredients based on the American Institute of Nutrition (AIN) diet included casein, corn starch, dextrose, corn oil, alphacel/cellulose, DL-methionine, mineral mix, vitamin mix and choline bitartrate were purchased from Bio-Serv (Flemington, NJ, USA). Kolliphor HS 15 was purchased from Sigma-Aldrich (St. Louis, MO, USA). SHetA2 compound was kindly provided by the United States National Cancer Institute (NCI) Rapid Access to Intervention Development (RAPID) program. The human ovarian cancer SKOV-3 cell line was purchased from American Type Culture Collection (Manassas, VA) and the human ovarian cancer cell line A2780 and the immortalized ovarian surface epithelial cells (IOSE-80) were gifts from Michael Birrer, MD, PhD (Harvard Medical School, Boston, MS, USA). Tissue culture media reagents (RPMI 1640 Medium 199, MCDB105, L-glutamine, sodium pyruvate, HEPES buffer), a protease inhibitor cocktail and dimethylsulfoxide (DMSO) were purchase from MilliporeSigma (St. Louis, MO, USA). Fetal Bovine Serum (FBS) was purchased from Serum Source International (Charlotte, NC, USA). A Cyclin D1 ELISA kit was purchased from Cloud-Clone Corp. (Houston, TX, USA). The 0.2 μm filters were purchased from Agilent (Santa Clara, CA, USA). The m-PER and t-PER protein extraction reagents, bicinchoninic acid (BCA) protein concentration measurement kit and an antibiotic/antimycotic mixture were purchased from Thermo-Fisher (Waltham, MA, USA). the Polyvinylidene difluoride (PVDF) membrane, the Transfer Blot Turbo Transfer Pack containing the PVDF membrane, and the Clarity Western ECL Substrate Chemiluminescence detection system were purchased from Bio-Rad (Hercules, CA, USA). An antibody specific for cyclin D1 was purchased from BD Pharmingen (San Jose, CA, USA). Luminol reagent and antibodies specific for GAPDH and β-actin and secondary antibodies conjugated to horse-radish peroxide (HRP) were purchased from Santa Cruz (Dallas, TX, USA). Luminol reagent was purchased from GE Healthcare Life Sciences (Pittsburgh, PA, USA).

A semi-purified diet based on modified AIN-76A (Casein, 20%; Corn Starch, 52%; Dextrose, 13%; Corn oil, 5.0%; Alphacel/cellulose, 5.0%; DL-Methionine, 0.3%; Mineral mix AIN, 3.5%; Vitamin mix, AIN, 1.0%; and Choline bitartrate, 0.2%) was prepared in the University of Oklahoma Health Sciences Center (OUHSC) Rodent Barrier Facility. The ingredients were mixed thoroughly to assure that all micronutrients were uniformly distributed in the diet. In order to assure that SHetA2 was uniformly distributed in the diet, the agent was pre-mixed with a small quantity of control diet in a food mixer, added to a pre-weighed amount of control diet in a Hobart Mixer and mixed thoroughly for about 45 min. Both control and experimental diets were prepared weekly and stored at 4 °C in the cold room.

The experimental design is illustrated in Fig. 2. C57BL/6 J female mice were bred at OUHSC Rodent Barrier Facility. The research was performed under the animal protocols approved by the OUHSC Institutional Animal Care and Use Committee (IACUC) and in compliance with the National Institutes of Health guide for care and use of Laboratory animals (NIH Publication No. 8023, revised 1978). Experimental animals were housed in ventilated cages under standardized conditions (21 °C, 60% humidity, 12-h light/12-h dark cycle, 20 air changes/h) in the OUHSC rodent barrier facility. Mice were allowed ad libitum access to the respective diets and to automated tap water purified by reverse osmosis. At 6–7 weeks of age, groups of 24 mice were fed control diet or the experimental diets containing 150 ppm, 300 ppm, 500 ppm, 750 ppm, and 1000 ppm SHetA2 for six weeks. The diet was replenished with freshly-made diet twice per week. All animals were weighed once weekly until termination of the study. Organs were collected and weighed at necropsy. Eight animals from each group were euthanized 1, 3 and 6 weeks after initiating the diet administration and necropsied to collect blood, livers, kidneys, intestines, ovaries, fallopian tubes, uterine horns, and cervices. The blood was collected in microfuge tubes and immediately processed into serum using standard protocols. Due to the low volume of serum that can be collected from and individual mouse, each mouse produced a single serum specimen that was used either for one of the individual toxicity endpoints or for the drug level measurement. The tissues and serum were placed in labeled cryovials, snap frozen in liquid nitrogen and stored at -80 °C until further analysis.

The zero control (internal standard, IS only) was prepared by adding a 10 μL aliquot of 5 μg/mL 2, 3-diphenylquinoxaline (internal standard, IS) to a 1.5 ml amber colored Eppendorf tube, and then adding an additional 90 μL of serum. The tube was vortexed for 30 s, 80 μL of chilled acetonitrile was added and then the tube again was vortexed vigorously on multi-shaker for 10 min followed by centrifugation at 16,100 xg for 10 min at 4 °C. The supernatant was filtered through 0.2 μm filter and then injected (70 μL) into an HPLC.

The frozen tissue (cecum) and colon content were rapidly thawed. The tissue was opened with a razor blade or scissors and approximately 50 mg of the content was transferred into 2 ml brown microcentrifuge tubes (VWR, Radnor, PA, USA), and then submerged in an ice bath. One mL of cold acetonitrile was added and vortexed for 10 min on analog multi-tube vortexer (VWR, Radnor, PA, USA). The homogenate was centrifuged at 16000 rcf at 4 °C for 5 min. The supernatant was filtered through Captiva filtration (0.2 μm filter. Agilent, Santa Clara, CA, USA). 5 or 10 μl of filtrate was injected into HPLC. The entire process was carried out under incandescent lights and under 4 ^oC.

Serum, cecum content and colon content reversed-phase HPLC analyses were performed on a Shimadzu HPLC equipped with a LC-10ADvp pump, SCL 10Avp system controller and a SPD-M10Avp PDA detector. All the operations were controlled by Shimadzu EZ start 7.2.1 software in Windows XP system. The PDA detector was set between 190 to 800 nm and analysis was done at 341 nm. HPLC solvents consisted of 65% Acetonitrile and 35% Distilled water. An isocratic method with flow rate 0.18 ml/min used with total run time 40 min (serum samples), and 20 min (cecum and colon content samples). The injection volume was 70 μL for serum samples and 5 or 10 μl for cecum content and colon content samples. A XBridgeTM C18 (Waters, BEH 130, 3.5 μm, 2.1 × 150 mm) column with a Precolumn XBridge (BEH C18 Sentry Guard Cartridge, 130 Å, 3.5 μm, 2.1 mm × 10 mm) was used. Retention times for DQ (IS) is 9.6 min, and SHetA2 is 11.7 min.

HPLC analysis was performed using a Waters Alliance HPLC System with a V_ydac 201 TP C₁₈ 5 μ (250 mm × 2.1 mm) column equipped with a guard column (V_ydac 201 TP, Grace), and a UV detector set at 341 nm. The mobile phase consisted of acetonitrile:water (80:20, v/v).The flow rate was 0.3 mL/min with a retention time 3.65 min and the total run time of each analysis was 7 min at ambient temperature. The injection volume was 10 μL. Empower software was used for the integration of the peak areas.

Animals were examined daily for any external signs of toxicity that would be predicted to shorten the natural life span of the animal, such as roughened coat, chromodacryorrhea, rhinitis, and prostration. Body weights were recorded at weekly intervals. Organ-to-body weight ratios were determined at weeks 1, 2 and 3. Sections of paraffin-embedded organ specimens were stained with hematoxylin and eosin (H&E) for pathology review. Three serum specimens from each of the groups treated for 6 weeks with 500 ppm, 750 ppm and 1000 ppm SHetA2 diets were evaluated for liver and kidney function tests: alanine amino transferase (ALT), aspartate amino transferase (AST), alkaline phosphatase (ALKP) and blood urea nitrogen (BUN) and creatinine (CRE) using an IDEXX Catalyst Dx Chemistry Analyzer.

For SHetA2 extraction, tissues were rapidly thawed and washed thoroughly with sterile saline solution. Tissues were then weighed and homogenized in an acetonitrile:saline mixture (80:20, v/v) using a bead homogenizer. Samples of 2 ml/g of tissue were centrifuged at 14000 rpm for 10 min. Supernatants were collected and loaded onto a Captiva filtration plate and filtered by vacuum. Filtrates were collected and injected directly into the HPLC for analysis of drug levels.

For Western blot and Enzyme linked immunosorbent assay (ELISA) analysis, necropsy cervical tissues were homogenized in ice-cold t-PER protein extraction solution containing a protease inhibitor cocktail and phosphatase inhibitor cocktail using a homogenizer. The homogenates were incubated on ice for 1 h with occasional shaking and then centrifuged at 13,000×g for 15 min at 4 °C. Protein concentrations in supernatants were determined using the BCA assay according manufacturer’s instructions.

Ovarian cancer cell lines (SKOV3 and A2780) were maintained in RPMI 1640 with L-glutamine (supplemented with 1 mmol/L sodium pyruvate, 1 mmol/L HEPES buffer, antibiotic/antimycotic and 10% FBS) and IOSE-80 cultures were maintained Medium 199/ MCDB105 (1:1) supplemented with antibiotic/antimycotic and 15% FBS. Cells were treated with 10 μM SHetA2 dissolved in DMSO or the same volume of DMSO solvent for 24 h prior to extracting proteins from cells with the m-PER reagent. Protein concentrations in the extracts were quantified with the BCA kit.

Equal amounts of proteins from extracts of animal specimens or cell cultures experiments were electrophoresed through a 10% SDS-PAGE gel and transferred to a PVDF membrane for the cell culture extracts. A 12% SDS-PAGE gel and the Turbo-Blot Turbo Transfer Pack was used for the animal tissue extracts. The membranes were blocked for one hour in 5% skim milk and then incubated overnight at 4 °C with the cyclin D1 primary antibody at 1:1000 dilution. The next day, the membranes were rinsed and incubated with the HRP-conjugated secondary antibody for 1 h. The specific signal was detected with Luminol reagent and exposure to X-ray negatives for the membranes containing the cell culture extracts and with Clarity Western ECL Substrate and BioRad ChemiDoc Touch Imaging System for the membranes containing the animal tissue extracts. To confirm equal loading, the membranes were stripped and re-probed the GAPDH antibody at 1:1000 dilution for 2 h at room temperature for cell culture extracts or the β-actin antibody at 1:2000 overnight at 4 °C for the animal tissue extracts.

Total protein was isolated from tissues in ice-cold m-PER protein extraction solution containing protease inhibitor cocktail and phosphatase inhibitor cocktail by using a homogenizer. Equal amounts of protein (5 μg) were added to an ELISA plate by and evaluated following manufacturer’s instructions. The ODs were measured with a Synergy H1 ELISA plate reader at 450 nm. The experiments were performed in duplicate.

SHetA2 in the dietary formulation was found to be stable after exposure to light at room temperature for 8 h, or stored at 4 °C in the dark for two weeks, with no variation from the initial starting material or presence of peaks corresponding to drug degradation products observed (Fig. 3). Throughout the experiment, the diet was prepared twice per week and stored at 4 °C in the dark to assure drug stability.

Based on an estimated average of 5 g of diet consumed per day, the animals assigned to the 150, 250, 500, 750 or 1000 ppm dietary SHetA2 groups consumed 37.5, 62.4, 125, 187 or and 250 mg/kg/day, respectively, over the 1, 3 and 6 week treatment periods. No significant differences were observed in growth rates between the various treatment groups (Fig. 4, Linear regression comparison of slopes, F = 0.944, P = 0.453) or between the organ-to-body weight ratios between the various groups (Table 1. Repeated Measures ANOVA with Dunnett correction for multiple comparisons, Liver: F = 0.59, p = 0.55; Kidney: F = 3.27, P = 0.16; Spleen: F = 1.35, P = 0.37).

Liver and kidney function tests were performed on blood specimens of the three highest dose groups collected after 6 weeks of dietary administration (Table 2). BUN levels were slightly higher than normal for the 125 mg/kg/day group and slightly lower than normal in the 250 mg/kg/day group. Ast and Alkp levels were reduced in the 250 mg/kg/day group.

An experienced pathologist (S.L.) evaluated H&E stained sections from each of the kidney specimens in blinded manner and reported that all specimens appeared normal. Histologies of the kidney and liver specimens from each dose group are shown in Fig. 5. The proximal and distal tubules, glomeruli, arterioles and venules and glomeruli of the kidney were normal. The kidney glomeruli were well-formed with no inflammation, peri-glomerular fluid or hyalinization. The livers appeared normal with no signs of apoptosis, necrosis or inflammation. The control group exhibited some liver steatosis (accumulation of fat) at a level of 2+. An occasional fat cell was present in livers from the treatment groups, however this was within normal limits and none of the treatment group livers from exhibited 2+ steatosis.

Gynecologic tissues were chosen for evaluation in this study, because SHetA2 is being developed currently for prevention of ovarian, cervical and endometrial cancers. After 3 weeks of treatment, SHetA2 levels were undetectable in the serum and combined ovarian and fallopian tube (Ovary/FT) specimens from mice that received the 37.5 and 62.5 mg/kg/day doses and measurable in the 125, 187 and 250 mg/kg/day dose groups (Fig. 6a, b, respectively). The highest serum drug levels were observed in the 187 mg/kg/day dose group, which were significantly higher compared to the drug levels measured in the 125 mg/kg/day dose group (One-way ANOVA, p = 0.032; and Tukeys multiple comparisons p value between 125 and 187 mg/kg/day, p value <0.05). The 250 mg/kg/day dose group exhibited lower drug levels compared to the 187 mg/kg/day dose group, but the difference was not statistically significant (One-way ANOVA with Tukey’s multiple comparison p value >0.05). The drug levels in the Ovary/FT specimens exhibited the same pattern with the highest concentrations in the 187 mg/kg/day group, however there were insufficient numbers of specimens available for statistical analysis and the tissue levels were much higher than the blood levels. The drug significantly accumulated in the cecum and colon contents in a dose-dependent manner, however there were no significant differences in the levels measured between 3 and 6 weeks of treatment (Fig. 6c). A two-way ANOVA demonstrated significant differences between doses (p = 0.001 for caecum contents and p = 0.016 for colon contents), but not between the treatment times (p = 0.066 for the caecum contents and p = 0.839 for colon contents).

SHetA2 has been shown to induce G1 cell cycle arrest in normal and cancer cells, and the mechanism has been shown to involve phosphorylation, ubiquitination and proteolytic degradation of cyclin D1 [10]. Use of cyclin D1 as a pharmacodynamic (PD) endpoint biomarker of SHetA2 chemoprevention activity was supported by the observation of a SHetA2 dose-dependent decrease of polyp cyclin D1 levels in association with reduction of the number and sizes of polyps in the APC^min/+ model of colorectal cancer [16]. The effects of SHetA2 on molecular endpoints in cancer cells are much greater than its effects in healthy cells, which translates into the lack of observed toxicity [6, 7, 15, 18], however differential effects on cyclin D1 specifically have not yet been evaluated. In this study, Western blot analysis demonstrated a dose-dependent reduction of cyclin D1 in human ovarian cancer cell lines, but not in non-cancerous IOSE cell cultures or in cervical tissues from individual animals treated for 6 weeks with the diet containing various SHetA2 doses (Fig. 7). A repeated measures ANOVA with Dunnett correction for multiple comparisons indicated that the cancer cell lines exhibited significantly greater cyclin D1 reduction compared to IOSE (p < 0.05).

To increase the sensitivity of cyclin D1 detection, protein extracts of cervical tissues (N = 3 per group) collected throughout the study were evaluated by ELISA (Fig. 8). With this increased sensitivity, significant reduction of cyclin D1 was observed only at the highest 250 mg/kg/day dose and the longest treatment time of 6 weeks. The sizes of the ovary/FT specimens were insufficient for Western blot or ELISA analysis of cyclin D1.