The anticancer activity of BA initially showed high potency against melanoma in cell culture and animal models, and subsequent studies show the effectiveness of this compound against multiple tumor types [39, 40, 42]. The low in vivo toxicity of BA coupled with supporting in vitro and in vivo results suggest that this compound or some derivative has potential for clinical applications in cancer chemotherapy. However, BA is a highly lipophilic molecule with limited water solubility and this may decrease in vivo uptake of this compound; therefore, development of specialized formulations/carriers such as liposomes may help to enhance the in vivo efficacy of BA as an anticancer agent . Previous studies in this laboratory showed that BA inhibits prostate cancer cell and tumor growth and this is accompanied by proteasome-dependent degradation of Sp1, Sp3 and Sp4 and several Sp-regulated pro-oncogenic gene products . Several other anticancer agents including tolfenamic acid, curcumin, arsenic trioxide, a nitro-NSAID (GT-094), and two synthetic triterpenoid derivatives, CDDO-Me and CDODA-Me, also induce Sp downregulation in various cancer cell lines via proteasome-dependent and -independent pathways [19, 21, 33–38].
BA inhibits colon cancer cell growth and induces caspase-dependent PARP cleavage in RKO and SW480 colon cancer cells (Figure 1) and these results are consistent with other reports on the effects of BA on colon cancer cell lines [39, 40, 44–46]. Moreover, BA also inhibited tumor growth in athymic nude mice bearing RKO cells as xenografts (Figure 6). We observed that BA decreased expression of Sp1, Sp3 and Sp4 proteins in both RKO and SW480 colon cancer cells and tumors (Figures 2A and 6C) and this was accompanied by parallel decreases in survivin and VEGF (Figures 2A and 2B), and these results are comparable to those observed in LNCaP prostate and KU7 bladder cancer cells treated with BA [20, 32]. Recent RNA interference studies show that p65 (NFκB subunit), EGFR, cyclin D1, and pituitary tumor transforming gene-1 (PTTG-1) are also Sp-regulated genes [32–35], and results in Figure 3C demonstrate that BA decreased expression of these gene products in RKO and SW480 cells. Moreover, knockdown of Sp1, Sp3 and Sp4 (in combination) in RKO colon cancer cells also decreased expression of EGFR, cyclin D1, p65 and PTTG-1, confirming the role of Sp transcription factors in regulating expression of these genes. These results are consistent with the induction of apoptosis by BA since many of these Sp-regulated genes are important for survival pathways.
Previous studies showed that BA-induced downregulation of Sp1, Sp3 and Sp4 was proteasome-dependent in LNCaP cells but proteasome-independent in KU7 bladder cancer cells [20, 32]. Similar variability was observed in RKO and SW480 colon cancer cells (Figure 2D) where BA-induced downregulation of Sp proteins was proteasome-independent and -dependent, respectively. This demonstrates that, for BA and possibly other drugs that downregulate Sp1, Sp3, Sp4 and Sp-regulated genes, the pathways required for this response are variable and dependent not only on tumor type but also cell context within the same tumor. At least two of these pathways, namely induction of proteasome- and caspase-dependent degradation of Sp proteins, involve activation of post-transcriptional processes [20, 21, 37]; however, their mechanisms have not been determined and are currently being investigated in this laboratory.
We have previously reported that the synthetic triterpenoid CDODA-Me and the NO-NSAID GT-094 decrease Sp protein expression in SW480 and RKO colon cancer cells through a transcriptional repression pathway in which miR-27a is decreased and this results in the induction of ZBTB10, a transcriptional repressor [36, 38]. BA decreased luciferase activity in RKO cells transfected with constructs containing several GC-rich promoter inserts (Figures 3B-D) and also decreased expression of miR-27a and induced expression of ZBTB10 in RKO cells (Figures 5A-C). Since overexpression of ZBTB10 and antisense-miR-27a also decreases expression of Sp1, Sp3, Sp4 and Sp-regulated genes in colon cancer cells , the mechanism of action of BA in RKO cells is linked to disruption of miR-27a:ZBTB10 as previously reported for CDODA-Me and GT-094 in colon cancer cells [36, 38].
BA is known to be a mitochondriotoxic drug and decreases the mitochondrial membrane potential in several different cancer cell lines leading to induction of apoptosis [39, 40, 44] and BA also decreased MMP in RKO cells (Figure 4C). Previous studies have demonstrated that at least four agents that are mitochondriotoxic and induce ROS also downregulate Sp proteins; this effect is ROS-dependent and reversible with antioxidants or catalase, and compounds activating this pathway include arsenic trioxide (bladder), curcumin and CDDO-Me (pancreatic), and GT-094 (colon) [33, 37, 38]. Moreover, for GT-094 and CDDO-Me, the mechanism of ROS-dependent downregulation of Sp1, Sp3, and Sp4 involves disruption of miR-27a:ZBTB10 [33, 38]. Results of this study show that BA also induced ROS-downregulated Sp1, Sp3, Sp4 and miR-27a and induced ZBTB10 in RKO cells, and all of these responses were significantly attenuated in cells cotreated with BA plus catalase (Figure 4). Moreover, catalase also reversed the growth inhibitory effects of BA (Figure 4C), further demonstrating the importance of ROS activation for the anticancer activity of this compound in RKO cells. In contrast to previous studies showing that CDODA-Me and GT-094 activated transcriptional repression of Sp proteins in both RKO and SW480 cells [33, 36], BA induced transcriptional repression in RKO cells but activated the proteasome pathway for degradation of Sp proteins in SW480 cells. The mitochondrial or extra-mitochondrial origins of ROS in cancer cells treated with BA and other agents that downregulate Sp transcription factors is currently being investigated.