RFX1-dependent activation of SHP-1 induces autophagy by a novel obatoclax derivative in hepatocellular carcinoma cells.

Obatoclax is a small molecule which targets the Bcl-2 family, and is to treat leukemia, lymphoma and lung carcinoma. Previously, an obatoclax analogue, SC-2001, was found to disrupt the protein-protein interactions of the Bcl-2 family and also repress Bcl-XL and Mcl-1 expression via STAT3 inactivation. Here, we report a novel mechanism of autophagy induction by SC-2001 in liver cancer cells. The findings indicate that SC-2001 induced the autophagy marker LC3-II in four hepatocellular carcinoma (HCC) cells. Autophagosomes induced by SC-2001-treated cells were confirmed by electron microscopy. SC-2001 activated SHP-1, dephosphorylated STAT3 and Mcl-1, and subsequently released free beclin 1. Overexpression of STAT3 and Mcl-1 in PLC5 cells attenuated the induction of SC-2001 on autophagy. Abolishment of SHP-1 by a specific inhibitor aboragated the autophagic effects induced by SC-2001. In addition, it was further revealed that RFX-1, a transcription factor of SHP-1, is a critical regulator in SC-2001-mediated autophagy. Downregulation of RFX-1 by si-RNA protected cells from SC-2001-induced autophagy. Importantly, Huh7 tumor-bearing nude mice treated with SC-2001 showed downregulation of Mcl-1 and p-STAT3 protein expression and upregulation of SHP-1, LC3II, and RFX-1 protein expression. In summary, our data suggest that SC-2001 induces autophagic cell death in a RFX1/SHP-1/STAT3/Mcl-1 signaling cascade.


INTRODUCTION
Autophagy is a cellular catabolic degradation process that results in the autophagosomic-lysosomal degradation of cytosolic proteins and other cellular components [1]. The first step of autophagy is the initiation of vesicle nucleation followed by the formation of an autophagosome. Lipidated LC3-II isoform is a wellknown autophagy marker, and it appears in both inner and outer membrane of autophagosome during the nucleation and expansion of vesicle. The second step is a docking and fusion process in which the autophagolysosome is formed by the interaction of autophagosome, endosomes, as well as the lysosomes. Biomolecules embedded in the autophagolysosome are degraded by acid-dependent enzymes into metabolic fuel [2]. Under stressed conditions, both cell survival or cell death can mediated by autophagy, therefore, the fate of cells is dependent on the type of cellular stress [3][4][5].
Obatoclax is a BH3 mimetic that can interact with BH3 domain belonged to anti-apoptotic Bcl-2 family then blocks the function of anti-apoptotic proteins and release pro-apoptotic proteins. In preclinical studies, obatoclax showed potent anti-tumor activity predominantly in leukemia and myeloma cells [17,18]. The effect of obatoclax-induced apoptosis in lung cancer [19,20] and breast cancer [21], is not as potent as that seen in leukemia. However, obatoclax has been suggested to have a synergistic effect when used with lapatinib, a targeted therapy for ErbB2-positive breast cancer, to induce apoptotic and autophagic cell death in human breast and colon cancer cells [22,23].
We previously demonstrated that SC-2001, a novel obatoclax derivative, shows a better apoptotic efficacy than obatoclax in hepatocellular carcinoma cells [24]. SC-2001 not only inhibits the function of Bcl-2 family but also regulate represses anti-apoptotic protein expression. We further discovered a novel mechanism by which SC-2001 activates SHP-1 phosphatase and reduces the phosphorylation of STAT3 in HCC cells. SC-2001-mediated STAT3 dephoshphorylation inhibits the transcription activities of STAT3 and expression of its downstream genes which induces HCC cell apoptosis.
In our paper, we revealed that SC-2001 showed more potent effect on inducing of autophagy than the lead compound, obatoclax, and has the different drug target. We proved that SC-2001 can disruption of the beclin 1-Mcl-1 interaction and repressed Mcl-1 expression, resulting in free beclin 1 for the induction of autophagy. Further, we found that SC-2001-induced RFX-1 is crucial to SHP-1 activation and in modulating SC2001mediated autophagy. Most important, SC-2001-indcued RFX-1/SHP-1-dependent inhibition in HCC cells was demonstrated in vivo in a xenograft mouse model. Overall, our results demonstrate the involvement of the RFX-1/ SHP-1/STAT-3 axis in SC-2001-mediated autophagy. These findings suggest a novel molecular mechanism of SC-2001-mediated autophagic cell death and highlight the potential of SC-2001 in the development of RFX-1/SHP-1/STAT-3-targeted anticancer therapeutics.

SC-2001 induces autophagic phenomena in HCC cells
To explore the efficacy of autophagy induction caused by obatoclax and SC-2001, we used HCC cells as model to assess the LC3-II expression level. LC3-II activation represents the autophagic activity. As shown in Figure 1A, treatment with SC-2001 resulted in dosedependent induction of LC3-II to higher levels than obatoclax in PLC5 cells. WL-2, an obatoclax derivative without any cytotoxicity in HCC cells, was used as a negative control. SC-2001 also exhibited the same effect in Huh7, HepG2 and SK-Hep-1 cells in a time dependent manner. Next, we applied two autophagy inhibitors to verify that the observed autophagic effect was induced by SC-2001. 3-Methyladenine (3-MA), a widely-used inhibitor of the initial phase of the autophagic process prevented LC3 II formation from LC-3-I in the presence of SC-2001. Bafilomycin A1, which inhibits the docking and fusion process in the late stage of autophagy, resulted in significant accumulation of LC3II in cells treated with SC-2001 ( Figure 1B). Moreover, the SC-2001induced autophagic effect was confirmed by GFP-LC3 staining. As shown in Figure 1C, PLC5 and Huh7 cells with GFP-LC3 transfection showed increased GFP-LC3II dots. In addition, autophagy as a result of SC-2001 treatment was confirmed with acridine orange staining. Conversely, 3-MA reduced acidic vesicular organelles in SC-2001-treated cells. ( Figure 1D). Finally, the ability of inducing autophagy by SC-2001 was verified by electron microscopy which revealed the formation of autophagosomes in PLC 5 and Huh7 cells ( Figure 1E).

Disruption of beclin 1-Mcl-1 complex interaction and ATG5 expression are associated with SC-2001-induced autophagy
Our previous report showed that inhibition of STAT-3 and downregulation Mcl-1 are key factors in SC-2001induced cell apoptosis [24]. To elucidate the biological and pharmacological effect of SC-2001-induced autophagy, the ability of SC-2001-induced disruption of the Mcl-1-beclin 1(ATG6) complex and/or other ATG family members were tested in HCC cells. SC-2001 disrupted Mcl-1 and beclin 1 interaction and further released beclin 1 to initiate autophagy, but did not influence the complex of beclin 1 and Bcl-2 ( Figure 2A). Furthermore, the expression of ATG5 was increased by SC-2001 in concentrate-dependent fashion in HCC cells. On the other hand, SC-2001 showed no influence on ATG3, ATG7 or beclin 1 in HCC cells ( Figure 2B). Genetic knockdown of beclin 1 by siRNA reversed LC3-II production and partially rescued cell viability ( Figure 2C). Also, 3-MA reversed the reduction of SC-2001 induced-colony formation ( Figure 2D). These data suggest that disruption the interaction between beclin

RFX-1 regulates SC-2001-mediated SHP1 transcription in HCC cells
According to the literature, RFX-1 is an important transcription factor of SHP-1 [25]. To investigate whether RFX-1 is involved in SC-2001-induced SHP-1 expression and autophagy, we examined the expression level of RFX-1 in HCC cells with the treatment of SC-2001. As shown in Fig 3A, RFX-1 was upregulated by SC-2001 in a concentration-dependent fashion. Next, to examine the involvement of RFX-1 in SC-2001activated SHP-1 transcription, we applied chromatin IP to demonstrate binding capacity. As shown in Figure 4A, the binding activity of RFX-1 to the SHP-1 promoter was obviously increased. Furthermore, RFX-1 silencing by si-RNA abolished SC-2001 efficacy on SHP-1 and LC3II expression ( Figure 4B). In addition, we verified the effect of the combination of SC-2001 and ectopic expression of RFX-1 in PLC5 cells. The results demonstrated that treatment of RFX-1 overexpressed cells with SC-2001 significantly increased autophagy ( Figure 4C).

SC-2001 inhibits tumor growth via RFX-1/SHP-1/ STAT3-dependent activation of autophagy
To explore whether SC-2001's pharmacological effect on autophagy may be therapeutically relevant, the in vivo consequences of SC-2001 were examined. Mice bearing Huh7 tumors randomized to treatment groups and treated with vehicle or SC-2001 (20 mg/kg/every other day). SC-2001-treatment groups exhibited significantly tumor growth inhibition effect ( Figure 5A). All animals tolerated SC-2001 treatment well. The mice did not show any visible signs of toxicity or body weight loss ( Figure  5B). To confirm the biological mechanism identified in vitro, we next examined the efficacy that SC-2001-induced  Figure 5C). SC-2001 treatment also activates the SHP-1 activity in Huh7 tumors ( Figure 5D). In addition, negative regulators of autophagy including p-STAT3 and Mcl-1 were decreased. Also, IHC staining proved the effect of SC-2001 on autophagy ( Figure 5E). We have now demonstrated that SC-2001 considerably inhibits tumor growth in vivo. Equally important, we discovered that SC-2001 activated a novel RFX-1/SHP-1 autophagic regulation in HCC.

DISCUSSION
SC-2001 was originally derived from obatoclax as a Bcl-2 family inhibitor against human cancers [26]. In our previous study [24], SC-2001 was identified to be SHP-1 agonist, and a negative regulator of STAT3 (which is over-expressed in various cancers), by enhancing SHP-1 expression via transcription activity. SC-2001 blocks STAT3 phosphorylation activity, a process required for survival and proliferation signals, and inactivates STAT3 transcription. Subsequent reduction of several regulatory substrates of STAT3, such as cyclin D and survivin, leads to inhibition of tumor growth mainly through induction of apoptosis. Whether SC-2001 is involved in other types of cell death in cancer cells, such as autophagy, was totally unknown. In our paper, we demonstrated that SC-2001 is a potent autophagy-inducing agent through activation of the RFX-1/SHP-1/STAT3 axis in HCC cells.
Here  It should be noted that in many instances autophagy constitutes a cytoprotective response activated by dying cells in the attempt to cope with stress, and its inhibition accelerates cell death [32]. In this regards, inhibition of autophagy may enhance the efficacy of currently used antineoplastic agents in cancer cells and has been suggested as a potential anti-cancer strategy [33][34][35]. On the other hand, several studies have shown autophagy may, in some circumstances, mediates the cytotoxic or cytostatic effect of anticancer agents, in which blockade of autophagy abolishes the therapeutic actions [36,37]. For example, Turcotte et al [38] reported a novel compound STF-62247 induced autophagic cell death in renal cell carcinoma cells. Interestingly, it has been reported that autophagy mediated the anti-cancer effect of a natural BH3-memimtic, (-)-gossypol, in androgen-resistant prostate cancer cells [39] and in malignant glioma cells [40]. It appears that autophagy plays a paradoxical role in cell-fate decision mechanisms [36,37]. Based on the new, revised Nomenclature Committee of Cell Death (NCCD) classification on cell death in 2012 [32], Galluzzi et al reintroduced the term 'autophagic cell death' to indicate a cell death that is mediated by autophagy and can be suppressed by the inhibition of the autophagic pathway. On the basis of the NCCD 2012 definition, all cases of cell death that exhibit markers of autophagy but cannot be blocked by autophagy inhibition (such as RNAi targeting beclin 1, or inhibitors of Vps34, a class III phosphoinositide 3-kinase (PI3-kinase) should not be classified as autophagic cell death [32]. Accordingly, our current study showed that SC-2001-induced autophagy in association with cytotoxic/cytostatic effects can be suppressed by Beclin 1 (Figure 2C) or by 3-MA, a broadspectrum PI3-kinase inhibitor ( Figure 1B, 1D, and 2D). In addition, we have added supplementary data showing SC-2001 promotes cancer cell death (Supplementary figure 1). Our current data suggest autophagy mediates part of the anti-cancer effects of SC-2001 in HCC cells. Moreover, our previous study has shown that SC-2001 induces apoptosis by modulating SHP-1/pSTAT3 mechanism in HCC cells [24]. Collectively, we suggest that SC-2001 not only induces apoptosis, but is also able to induce autophagic cell death. We have added a comprehensive plot to provide current known drug mechanisms of SC-2001 (Supplementary figure 2). We

Cell Culture
The Huh-7 HCC cell line was obtained from the Health Science Research Resources Bank (HSRRB, JCRB0403). The PLC/PRF/5 (PLC5), HepG2 and SK-Hep1 cell lines were obtained from American Type Culture Collection (ATCC). All cells obtained from HSRRB or ATCC were immediately expanded and frozen such that all cell lines could be restarted every 3 months from a frozen vial of the same batch of cells. No further authentication was conducted in our laboratory. Cells were maintained in DMEM (Sigma-Aldrich, D5648) supplemented with 10% fetal bovine serum (FBS, Hyclone, SH30071.03), 100 units/mL penicillin G and 100 μg/mL streptomycin solution (Sigma-Aldrich, P4333) in a humidified incubator at 37°C in an atmosphere of 5% CO 2 in air. Primary cancer cells from consenting patients were also analyzed. Study protocols were approved by the institutional review board of the institution, and informed consent was obtained in accordance with the Declaration of Helsinki. Human HCC samples were obtained from the patient who underwent tumor resection. The cells were isolated by mechanical mincing and digestion by collagenase.

Colony formation assay
SK-Hep1 cells were seeded in 6-well plates (~1000-5000 cells per well) and subjected to the indicated treatments, with the drug being removed to terminate the treatment. Two weeks later, plates were washed in PBS, fixed with 100% methanol (Sigma-Aldrich, E7023) and stained with a filtered solution of 5% w/v crystal violet (crystal violet, Sigma-Aldrich, C3886). After washing with tap water, the colonies were counted both manually (by eye) and digitally using a ColCount automated colony counter.

Autophagy analysis
The following methods were used to assess druginduced autophagic cell death: (1) Western blot analysis of microtubule-associated protein 1 light chain 3 (LC3 II) as described previously [2,44]; and (2) electron microscopy. Samples were fixed with 2.5% glutaraldehyde solution buffer in PBS at 4°C for 1 h, postfixed in 1% osmium tetroxide solution at 4°C for 3 h, dehydrated in graded concentrations of ethanol and embedded in LR white resin (Sigma-Aldrich, 62661). Ultrathin sections (70 nm) were examined with a JEOL JEM-1400EX electron microscope at 120 Kv. Representative areas were recorded at 20,000, 40,000 magnification; (3) Immunofluorescence of acridine orange (Sigma-Aldrich, A6014). HCC cells were grown on coverslips. After being washed with PBS, cells were treated with 2 µM SC-2001 or SC-2001+3-MA for 48 h, fixed with ice-cold 4% paraformaldehyde (Sigma-Aldrich, P6148) for 30 minutes at room temperature then stained with acridine orange (5 μg/ml) for 5 minutes at room temperature. Cells were examined under a Leica DM2500 fluorescence microscope.

Chromatin Immunoprecipitation (CHIP) Assay
The CHIP assay was performed according to the EZ ChIP chromatin immunoprecipitation (Merck Millipore, 17-371) and EZ-Zyme Chromatin Prep Kit (Merck Millipore, 17-375) manufacturer's protocol. Briefly, after cross-linking with 17.5% paraformaldehyde, PLC5 cells were washed with phosphate-buffered saline and lysed in lysis buffer. The DNA was fragmented to about 200-500 base pairs by the EZ-Zyme. Approximately 5 × 10 6 cells were used per CHIP assay and the resulting DNA fragments were incubated with 2 μg RFX1 antibodies, which were generated from rabbit, or nonspecific rabbit IgG (Cell signaling, #3900). The immunoprecipitated products were washed sequentially with low-salt immune complex wash buffer, high-salt immune complex wash buffer, LiCl immune complex wash buffer, and twice with TE buffer. The chromatin was eluted from the agarose by incubating with elution buffer (1% SDS, 100mM NaHCO 3 ); and the DNA-protein complexes were reversely cross-linked by high-salt solution containing 200 mM NaCl at 65°C for at least 5 h. To eliminate contaminations of proteins and RNAs, the mixture was treated with 10 mg RNase A at 37°C for 30 min and then treated with protease K for 2 h at 45°C. Finally, the precipitated DNA was recovered using the spin column provided in the ChIP kit, and eluted with 50 ml elution buffer. PCR reaction was conducted using Taq DNA polymerase (Invitrogen, 4398881). Two microliters of the precipitated DNA was used as template. The sequence of the primers used in the ChIP assay were as follows: 5′-CCTCTTGCAGGTGTCCTTAAG-3′, and 5′-TGGAAAGGCAGAGGGAATCAG-3′.

TUNEL assay
The TUNEL assay is performed according to the manufacture's instruction (AAT Bioquest, 22844). Briefly, cells were seeded into 96-well plate. After cells were attached to the plate, cells were treated with SC-2001 as indicated concentration for 48 h. After that, cells were fixed with 4 % formaldehyde for 20 minutes at room temperature and then washed with PBS. Each well was incubated with TUNEL reaction buffer provided by commercial kit for 1 h at 37°C. The fluorescence intensity was measured by fluorescence microplate reader at Ex/ Em = 550/590 nm after incubation.

Xenograft tumor growth
Male NCr athymic nude mice (5-7 weeks of age) were obtained from the National Laboratory Animal Center (Taipei, Taiwan). All experimental procedures using these mice were performed in accordance with protocols approved by National Taiwan University. When Huh7 tumors reached 150 mm 3 , mice received SC-2001 (20 mg/kg) orally (every other day). Controls received vehicle.

Statistical analysis
Data are expressed as mean ± SD or SE. Statistical comparisons were based on nonparametric tests and statistical significance was defined at P < 0.05. All statistical analyses were performed using SPSS for Windows version 12.0 software.