It is well documented that SFN can target cancer cells through multiple chemopreventive mechanisms but here we show for the first time that SFN selectively targets benign hyperplasia cells and cancerous prostate cells while leaving the normal prostate cells unaffected. Importantly, SFN selectively induced cell cycle arrest and apoptosis, specifically in hyperplastic and cancer cells. These findings regarding the relative safety of SFN to normal tissues has significant clinical relevance as the use of SFN moves towards use in human clinical trials. In addition, a reduction in HDAC activity and down-regulation of select class I and class II HDAC proteins followed by an increase in acetylation of histone H3 at the P21 promoter and increased acetylation of α-tubulin occurred specifically in the hyperplastic and cancer cells, not normal cells. Together these results highlight the use of dietary SFN as a safe and relatively non-toxic chemopreventive agent that could be readily achieved by simple and affordable incorporation of SFN rich foods in the diet.
The level of HDAC activity within a cell can be altered via direct inhibition of the HDAC enzyme and changes in HDAC protein levels. Previous reports indicate that SFN and its metabolites SFN-Cys and SFN-NAC can directly inhibit HDACs [6
], and in this report we show that SFN can also decrease protein levels of several different HDACs. We also observed that the HDAC inhibition in PrEC cells is more transient than in the other prostate cancer cell lines. The overall metabolite profiles for PrEC and PC3 cells were similar, particularly in the levels of SFN-Cys, the putative active HDAC inhibitor. Although small changes were observed between metabolite ratios in culture media and final metabolite concentrations in cell lysates, it is unlikely that a 2–3 µM difference would exert a significant biological effect that would account for the marked increase in apoptosis and cell cycle arrest in PC3 cells. In contrast, a significant reduction in HDAC protein level was observed in BPH1, LnCap and PC3 cells which produced a greater and more persistent reduction in HDAC activity. It is possible that the differential responses in HDAC activity, HDAC protein levels and ultimately downstream acetylation of histone and non-histone targets may contribute to the differential cell fate responses observed after SFN treatment.
Over-expression of Class I HDACs has been reported in prostate cancer tumors [31
]. Moreover, the effects of the inhibition of Class I HDACs on cycle arrest and apoptosis has been shown previously [32
]. Colon cancer cells treated with 3,3′-Diindolylmethane (DIM), an indole found in cruciferous vegetables, induced proteasome mediated degradation of HDACs 3 and 8, acetylation of histone H3 at the P21
promoter and ultimately induced a G2
/M cell cycle arrest and apoptosis [33
]. Another report showed that siRNA knockdown of HDAC3 in SW480 colon cancer cells increased acetylated H4-K12 at the P21
promoter, induced p21 expression and potentiated butyrate induced cell cycle arrest and growth inhibition [34
]. These reports are consistent with what we observed in our experiments, namely SFN treatment caused a decrease in several class I HDAC proteins, induction of histone acetylation at the P21
promoter and ultimately induction of G2
/M cell cycle arrest and apoptosis. This provides compelling evidence that the changes in class I HDAC proteins and histone acetylation may be responsible the changes in cell fate.
Herein we also provide evidence that HDAC6 is potentially responsible for the selective effects of SFN in cancer cells. Tubulin is a well known deacetylation target for HDAC6 and this process has an impact on microtubule dynamics [11
]. Importantly, tubulin deacetylation increases the turnover rate of microtubules, and over-expression of HDAC6 augments this process [12
]. Decreases in HDAC6 activity [30
] and increases in tubulin acetylation  after SFN treatment has previously been reported by others [35
]. In Gibbs et. al., researchers found that HDAC6 over-expression in LnCap cells abrogated the effects of SFN on HSP90 acetylation and inhibited its association with the androgen receptor (AR) providing further evidence that inhibition of HDAC6 was a specific target for sulforaphane [30
]. Here we confirm the decrease in HDAC6 protein levels in LnCap after SFN treatment reported by Gibbs et. al. and report that HDAC6 protein levels decrease in BPH1 and PC3 cells after SFN treatment. Interestingly, Gibbs et. al. reported that changes in HDAC6 activity ultimately affected the activity of the AR but here we show that HDAC6 is also decreased in PC3 cells which are AR negative indicating that the effect of SFN on AR may not be the only target for HDAC6. Importantly, over-expression of HDAC6 rescues PC3 cells from SFN-induced decreases in cell viability suggesting that HDAC6 plays a critical role in mediating its cytotoxicity. Thus, the inhibition of HDAC6 could play a key role in the stabilization of microtubule networks, disruption of tubulin polymerization and ultimately contribute to the mitotic cell cycle arrest observed with SFN treatment [35
]. The link between tubulin acetylation and selectively toxicity towards cancer cells has been reported after treatment with the HDAC6 specific inhibitor tubacin. Treatment with tubacin induced a dose and time dependent increase in α-tubulin acetylation and ultimately cytoxicity in several multiple myeloma cell lines and bone marrow plasma cells but had no effect in normal peripheral blood mononuclear cells [37
]. Autophagy is another cellular response that is partially mediated by HDAC6 because it functions to deliver polyubiquitinated proteins to aggresomes for degradation by binding both the polyubiquitinated proteins and the microtubule motor dynein [28
]. SFN treatment in PC3 and LNCaP prostate cells results in an induction of autophagy and partial inhibition of cytochrome C release and apoptosis [38
]. The decrease in HDAC6 protein we observed in BPH1, LnCap and PC3 cells may divert the cell fate away from survival (autophagy) towards cell death (apoptosis) by decreasing the formation of autophagic aggresomes. Further investigation into the potential roles of α-tubulin acetylation and the reduction in HDAC6 levels in the selectivity of SFN will provide insights into the mechanisms for SFN mediated cancer cell death.
The use of HDAC inhibitors in cancer prevention and treatment has become an intense area of research. These data provide further support for the relevance of SFN as a dietary HDAC inhibitor and chemopreventive agent by showing that SFN can selectively target BPH1, LnCap and PC3 prostate cells while leaving normal PrEC prostate cells virtually unaffected. This selectivity opens the door to a wide range of new scientific questions that will help in understanding the many mechanisms of action for SFN. The data presented here, taken with the previous reports of SFN action, show that SFN can target multiple steps in the carcinogenesis pathway and make it a promising cancer prevention agent.