br Materials and methods br Results br Discussion Defective

Materials and methods
Results
Discussion Defective regulation and inappropriate activation of the HH signaling pathway is implicated in many cancers. SMO targeting inhibitors were the first to enter clinical trials, however complications with acquired resistance and toxicity, also exemplified in preclinical studies of various tumor models, demonstrated that inhibition of SMO alone does not provide long-term efficacy in constraining tumor progression [[43], [44], [45], [46]]. Additionally, dysregulation of other pathways that crosstalk with HH signaling components, including the tumor suppressor p53, can also contribute to tumor maintenance and resistance. Hence, combinatorial treatments and/or “switching off” the HH signal transduction cascade by direct inhibition of the terminal effectors of the pathway, the GLI factors, may overcome both SMO and “non canonical” HH pathway-driven resistance. In this study we demonstrate that the small molecule RITA downregulates HH signaling irrespective of the p53 status (wild-type or mutant) in the molar volume calculator or p53 depletion via siRNAs. In three different cellular contexts, Daoy MB, Rh36 RMS and MCF7 breast cancer cells, RITA was capable to downregulate RNA expression of typical HH target genes (GLI1, GLI2, PTCH1, PTCH2) [Fig. 1A; Figs. S1A and B] irrespective of p53 depletion. We also provide evidence that the RITA effects are apparently elicited downstream of SMO, since depletion of SMO did not confer major changes in the response of the HH target genes [Fig. 1A; Fig. S1A]. Although since its discovery RITA (NSC652287) [20,47] has generally been considered to exert its tumor suppressive effects in a p53-dependent manner [[22], [23], [24],31,48], in recent years the p53-dependence of the RITA effects was debated, as several studies highlighted RITA-mediated effects without any obvious dependency on the cells' p53 status [[26], [27], [28]], questioning even its selective binding to the p53 protein [25]. In fact, our studies show that in Daoy cells, the p53 protein levels are unchanged upon RITA treatment [Fig. 1C], and the p53 target gene expression levels show only small and sometimes rather inconsistent changes, which in addition are not dependent on p53 depletion [Fig. 1A; Fig. 2A and B; Fig. 3A]. These findings further support the notion that RITA can elicit p53-independent effects. Since ROS-dependent JNK activation has been suggested as a possible mechanism of RITA action [31,32], we studied its potential involvement in the observed HH signaling downregulation. First, ROS accumulation was prevented by different antioxidants (resveratrol and NAC) and second, JNK activity was inhibited by the small molecule SP600125. We found that both resveratrol and NAC were mostly ineffective in reverting the RITA mediated effects on HH signaling gene expression in both Daoy and Rh36 cells [Fig. 2A and B; Figs. S2A–F]. However, JNK inhibition fully rescued the RITA effects [Fig. 3A and B]. Interestingly, the JNK inhibitor did not revert the effect of RITA on p53 target gene expression neither in Daoy [Fig. 3A] nor in Rh36 cells [Fig. 3B], further supporting the claim that the impact of RITA on HH signaling is p53 independent. In fact, these data are in agreement with a dual mechanism of RITA action, a JNK-independent p53 response and a JNK-dependent HH signaling response. Western blot analysis in Daoy cells confirmed that the protein levels of the 54 KDa and 46 KDa JNK isoforms are elevated by activating phosphorylation upon RITA treatment [Fig. 3C]. Recently, a plausible scenario underlying the RITA effects was put forward, suggesting that cells develop resistance to RITA independent of the p53 status, with this resistance not mediated by p53 mutations, but rather by defects in the DNA damage signaling network, which can be induced via both p53-dependent and p53-independent mechanisms [26]. To address this possibility, the cell viability and/or the GLI1/GLI2 RNA expression response of two additional cytotoxic agents (DXR and OXP) was compared to that of RITA. Interestingly, compared to Daoy cells [Fig. 4A,C; Figs. S5A–C], Rh36 cells were much more sensitive to RITA [Fig. 4B, D]. Yet, these two cell lines responded with a similar pattern to both DXR and OXP treatments, with Daoy cells being more sensitive to OXP and DXR. The higher sensitivity of Rh36 cells to RITA treatment might be due to the potentiality of wild-type p53 in inducing a p53 dependent response. In fact, this scenario is consistent with the fact that RITA is a much better inducer of p53 targets (NOXA and BCL2) in wild-type p53 Rh36 cells [Figs. S2A,B,E,F], compared to mutant p53 Daoy cells [Fig. 2A and B; S2C,D]. Further evidence on the differential impact of these drugs is provided by qPCR analyses, which confirmed that DXR does not downregulate GLI1 levels [Figs. S4A and B], in line with a previously reported study [49], and OXP could not mimic the RITA-induced downregulation of GLI1/2 expression [Fig. S4C]. In this context, quite revealing is the finding that DXR was not able to elicit the JNK phosphorylation response observed by RITA [Fig. 3C], with the 46 KDa isoform levels becoming undetectable following this treatment [Fig. S4D], similarly to what is seen when the JNK inhibitor impedes the RITA upregulation of phospho-JNK. Taken together, these results provide evidence that compared to the DNA damaging agents DXR and OXP, RITA has distinct cytotoxic readouts and differential impact on HH signaling. Worth noting is that two known p53 activating drugs, nutlin-3a and APR-246 (PRIMA-1Met) [50], were rather ineffective in downregulating HH signaling target genes [Figs. S4E and F], again highlighting the distinct action of RITA on HH signaling.