We have previously identified compounds effective at producing modest increases in NQO1 enzymatic activity in human prostate cells in vitro [18,19] and selected 4 compounds for testing whether they could produce induction of phase 2 enzyme activity in vivo. We selected sulforaphane, dimethyl fumarate and cucumin since they were among the most potent NQO1 inducing agents in prostate cells in vitro, have been reported to be monofunctional inducers (i.e. induce phase 2 enzymes primarily), and have been administered to animals without toxicity previously [11,13,20]. β-naphthoflavone, a bifunctional (phase 1 and 2) enzyme inducing compound, was selected because of its documented ability to induce phase 2 enzyme activity in rodent tissues in vivo and for comparison to the other compounds since it increased NQO1 activity to a lesser degree in the prostate cells in vitro .
We have demonstrated that orally administered agents can produce modest increases in phase 2 enzyme activity in prostate tissues in vivo. We have shown previously that sulforaphane, curcumin, dimethyl fumarate and, to a lesser degree, β-naphthoflavone will induce modest increases NQO1 enzymatic activity in prostate cancer cells in vitro [18,19]. Effective induction in vivo depends on candidate phase 2 enzyme inducing compounds being absorbed from the gastrointestinal tract, and those compounds or their active metabolites reaching the prostate, being absorbed from the circulation and acting in prostate cells in the context of their physiological environment. Our finding of even modest induction of phase 2 enzyme activity implies that each of these pharmacokinetic constraints can be overcome, and suggests that phase 2 enzyme induction by orally administered agents could represent a possible prostate cancer prevention strategy. However, whether the modest increases in phase 2 enzyme activity induced by sulforaphane, dimethyl fumarate, cucumin and β-naphthoflavone are sufficient to prevent prostate cancer is unknown and remains to be tested.
The F344 rat will develop prostate adenocarcinoma after chronic administration of 2-amino-1-methyl-6-phenylimidazo [4,5-b]pyridine (PhIP) and is one of the few carcinogen-induced animal models of prostate cancer . We selected this strain of rats to test to the possibility of phase 2 enzyme induction in the prostate as a prelude to future experiments designed to test whether phase 2 enzyme induction in the prostate could prevent PhIP-induced prostate cancers. The degree of increase of NQO1, total GST and GST-mu enzymatic activities in the prostate tissues we observed was modest, and lower than that reported in other model systems where phase 2 enzyme inducing compounds have been documented to prevent carcinogenesis . However, in man, prostate cancer develops over decades, raising the possibility that chronic, low-level phase 2 enzyme induction might be sufficient to prevent the disease. Furthermore, modest induction of phase 2 enzymes (NQO1 and total GST), virtually identical to those reported in the present study, have been observed in the liver tissues of F344 rats treated with sulforaphane and sulforaphane nitrile derived from cruciferous vegetables . Cruciferous vegetables will decrease the incidence preneoplastic lesions in the colon and liver when fed simultaneously with the carcinogen 2-amino-3methylimidazo [4,5-f]quinoline (IQ) to F344 rats . Therefore, even relatively modest induction of phase 2 enzymatic activity can be sufficient to protect against carcinogenesis. Whether similar protection against prostatic carcinoma will occur requires further testing. The finding that consumption of cruciferous vegetables has been associated with a decreased risk of subsequent prostate cancer diagnosis, coupled with the ability of orally administered sulforaphane to induce phase 2 enzyme activity in the prostate, suggests that phase 2 enzyme induction within the prostate is a potential prostate cancer preventive strategy and sulforaphane is a candidate preventive agent [28,29].
In agreement with previous observations, we found that each compound showed differing efficacy at inducing phase 2 enzyme activity in different tissue types. The kidney, for instance, showed little induction of the glutathione transferases, while the GSTs were readily induced in the liver, bladder and prostate. Prochaska et al. have reported that the induction patterns of derivatives of tert-butyl-4-hydroxyanisole (BHA) varied in their efficacy of phase 2 enzyme induction, differed in the spectrum enzymes they each induced and differed in their effectiveness between the liver, esophagus, forestomach, colon, kidney and lung . Similarly, Spencer et al. found that in CD-1 mice, dimethyl fumarate induced NQO1 enzymatic activity in the forestomach, small intestine, kidneys and lungs, but failed to induce NQO1 activity in the liver, similar to our findings in the F344 rat . They also found that the patterns of induction of total GST, GST-mu, and NQO1 enzymatic activities differed between compounds and by tissue type. Van Lieshout et al. have also described differences in phase 2 enzyme responsiveness in the tissues of Wistar rats after treatment oltipraz, α-tocopherol, β-carotene, and phenethyl isothiocyanate .
The reasons for the differences in the responsiveness of phase 2 enzymes between tissues are currently unknown, but likely are a reflection of tissue-specific expression of transcriptional regulators or enzyme cofactors. The difference in responsiveness between tissues does have important implications in the design and interpretation of preventive intervention trials involving phase 2 enzyme induction. For instance, cancers that arise from the oral ingestion of carcinogens, such as the 9,10-dimethyl-1,2-benzanthracene rat model of breast cancer or aflatoxin-induced hepatocellular carcinomas in man, might best be prevented by oral ingestion of agents that will induce phase 2 enzymes and inactivate these carcinogens in the gut and liver [13,30,31]. However, accumulating evidence suggests that for prostate cancer induction of phase 2 enzymes within the prostate might best protect against carcinogenesis.
No environmental carcinogens have been identified as causing human prostate cancer. Accumulating evidence implicates endogenous oxidative damage as one important contributor to prostate carcinogenesis . Prostate cancer increases with age and may be related to inflammatory conditions of the prostate such as prostatitis . Androgens are a known requisite to prostate cancer development. Ripple et al. have demonstrated that treatment of the prostate cancer cell line LNCaP with androgens produces a burst of oxidative stress in these cells with generation of reactive oxygen species, increased lipid peroxidation and a depletion of intracellular glutathione stores [33-35]. Furthermore, Malins et al. have described progressive alterations in DNA structure between normal, BPH and cancerous prostate tissues due to oxidative damage to the DNA template be hydroxyl free radical [36-38]. Two genes recently identified as conferring increased risk to prostate cancer in families (RNASEL and MSR2) participate in the response to infection and inflammation [39,40]. Mice engineered to not express RNASEL, for instance, are more susceptible to overwhelming bacterial infections . Finally, most compounds thus far implicated as prostate cancer preventive agents act as potent antioxidants including lycopene, selenium (essential to glutathione peroxidase activity), and vitamin E [42-44].
The early and near universal loss of expression of the phase 2 enzyme GSTP1 likely renders prostate cells susceptible to local oxidative damage and transformation. GSTP1 knock-out mice treated with the polycyclic aromatic hydrocarbon 7,12-dimethylbenz anthracene and the tumor promoting agent 12-O-tetradecanoylphorbol-13-acetate show increased numbers and earlier onset of skin papillomas demonstrating that loss of expression of a single GST can contribute to carcinogenesis . Since prostate cancer arises with a long latency in the context of local oxidative damage coupled with an intrinsic defect in carcinogen defenses, local induction of phase 2 enzymatic activity, even to a modest degree, could be a promising preventive strategy. Since prostate cancer develops over decades, chronic, low-level, local induction of carcinogen defenses, possibly through diet-derived agents such as sulforaphane, could represent a modest, non-toxic intervention strategy for prevention of prostate cancer, particularly for individuals at risk for the disease.
One notable finding was the significant induction of total GST and NQO1 enzymatic activities in bladder tissues of the F344 rats. Several environmental carcinogens have been linked to bladder cancer including polyaromatic hydrocarbons in tobacco smoke and aniline dyes . Epidemiological studies have demonstrated that consumption of cruciferous vegetables is associated with a decreased risk of bladder cancer . Sulforaphane levels peak in the serum 1–2 hours after ingestion and are cleared relatively rapidly by excretion into the urine . The substantial phase 2 enzyme induction of the bladder tissues could be due to the presence of sulforaphane or its active metabolites at relatively high concentrations over prolonged time periods while they are retained in the bladder. Munday and Munday have found similar induction of NQO1 and GST activity in the bladder tissues of female Sprague-Dawley rats after oral feedings of sulforaphane and several other isothiocyanates derived from cruciferous vegetables . Together, these data strongly suggest that sulforaphane and other isothiocyanates could represent promising candidate bladder cancer preventive agents.
Our study has several shortcomings. We arbitrarily selected a single daily dosing schedule based on prior studies in the literature. It is possible that other dosing schedules, perhaps different for each compound, could produce greater phase 2 enzyme induction . In addition, measurement of phase 2 enzyme activity occurred 24 hours following the last dose of each compound. The serum half-life of sulforaphane is between 1–2 hours and it is possible that measurement of phase 2 enzyme activity at times less than 24 hours would reveal greater induction of enzymatic activity . Finally, all animals were given isofluorane anesthesia at the time of gavage feeding, and the anesthesia could have altered phase 2 enzyme activity in the tissues. However, since both the inducer compound treated animals and controls were treated identically, the relative levels of phase 2 enzyme activity should not have been affected.