Purified sulforaphane was obtained from LKT labs (St. Paul, MN), curcumin, β-naphthoflavone (BNF), propylene glycol, dimethyl fumarate (DMF), 1,2-dichloro-4-nitrobenzene (DCNB) and 1-chloro-2,4-dinitrobenzene (CDNB) from Sigma-Aldrich Inc. (St. Louis, MO), and AIN 76A diet from Research Diets (New Brunswick, NJ). Male F-344 rats were purchased from Jackson labs (Bar Harbor, ME).
Treatment of animals and tissue collection
Eight-week-old male F344 rats were housed in microisolator cages in groups of 2 or 3 animals per cage. Mean body weight of the rats was 188 ± 4.0 g at the start of the study, and animals had access to AIN 76A diet and water ad libitum over the duration of the study. Animals were randomly divided into 5 groups of 10 animals corresponding to each of the 4 test compounds and a control group that received vehicle alone. All compounds were mixed fresh each day by either dissolving or suspending them in 100 μl propylene glycol at the following doses: sulforaphane 50 mg/Kg/day , curcumin 45 mg/Kg/day , β-naphthoflavone 41 mg/Kg/day  and dimethyl fumarate 37.5 mg/Kg/day . Doses were chosen that had been reported to be non-toxic and effective at inducing phase 2 enzymes in other model systems. Compounds or propylene glycol were administered in a single dose once a day by gavage at doses corrected for the body weight of each animal. The gavage feeding was carried out after the rats received isofluorane inhalation anesthesia and involved minimal trauma. Animals recovered rapidly from this agent and were observed until fully awake in their cages. All rats were monitored for infection or toxicity to prevent suffering, and there were no obvious signs of discomfort, distress, or pain over the duration of the study. One animal in the sulforaphane group died shortly after the first gavage feeding, and another died after the second feeding, both apparently due to aspiration of the dose. Necropsy did not reveal any gross abnormalities of any organs. On the morning of the sixth day, the rats were sacrificed by CO2 asphyxiation approximately 24 hours after the last dose.
The rats were housed at the Animal Care Facility at the Stanford University School of Medicine in compliance with PHS Policies on Humane Care and Use of Laboratory Animals. All work was carried out under Administrative Panel on Laboratory Animal Care approved protocols. All animals were under strict veterinarian care of the Department of Comparative Medicine in compliance with all Federal and State regulations to assure proper and humane treatment.
Collection of tissues and preparation of cytosol
The liver, kidneys, bladder and the prostate tissues were removed, weighed and snap frozen in liquid Nitrogen, and stored in -80°C until processed for the enzyme assays. Cytosols were prepared from the harvested tissues by homogenization in 0.25 M sucrose and centrifugation at 5000 × g for 20 minutes at -4°C. 0.2 volume of 0.1 M CaCl2 in 0.25 M sucrose was added to the supernatant and, after incubation on ice for 30 minutes, samples are centrifuged at 15,000 × g at -4°C for 20 minutes.
Total glutathione transferase enzyme activity was determined using 1,2-dichloro-4-nitrobenzene (DCNB) and GST mu activity was measured using 1-chloro-2,4-dinitrobenzene (CDNB) according to the procedure of Habig et al. . Cytosols (50 μl) were added to 150 μl 0.1 M phosphate buffer pH 6.5 buffer with 1 mM GSH, 1 mM CDNB or DCNB, and 1% BSA, mixed and optical absorbance was read at 340 nm at 30 sec intervals over 5 minutes. Because of high specific activity, liver samples were diluted 5-fold. GST mu activity was not measured in the bladder samples. Quinone reductase activity was determined by the rate of the NADPH-dependent, menadione-coupled reduction of 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide in 96-well microtire plates as described previously [18,19,22]. Enzyme activities were normalized to total cytosolic protein measured according to the Bradford method . All assays and protein measurements were performed in triplicate. Mean enzyme specific activities for tissues from animals in each group were calculated and the fold-induction enzyme specific activities determined by taking a ratio of log-transformed inducer-treated enzyme activities to the controls. The 95% confidence intervals for the fold-induction of enzyme specific activities of the inducer-treated animals compared to controls were calculated on log-transformed data and the results back-transformed to the fold-scale.