Autophagic flux modulation by Wnt/β-catenin pathway inhibition in hepatocellular carcinoma

Autophagy targets cellular components for lysosomal-dependent degradation in which the products of degradation may be recycled for protein synthesis and utilized for energy production. Autophagy also plays a critical role in cell homeostasis and the regulation of many physiological and pathological processes and prompts this investigation of new agents to effect abnormal autophagy in hepatocellular carcinoma (HCC). 2,5-Dichloro-N-(2-methyl-4-nitrophenyl) benzenesulfonamide (FH535) is a synthetic inhibitor of the Wnt/β-catenin pathway that exhibits anti-proliferative and anti-angiogenic effects on different types of cancer cells. The combination of FH535 with sorafenib promotes a synergistic inhibition of HCC and liver cancer stem cell proliferation, mediated in part by the simultaneous disruption of mitochondrial respiration and glycolysis. We demonstrated that FH535 decreased HCC tumor progression in a mouse xenograft model. For the first time, we showed the inhibitory effect of an FH535 derivative, FH535-N, alone and in combination with sorafenib on HCC cell proliferation. Our study revealed the contributing effect of Wnt/β-catenin pathway inhibition by FH535 and its derivative (FH535-N) through disruption of the autophagic flux in HCC cells.


Introduction
Hepatocellular carcinoma (HCC) is the most prevalent, primary malignancy of the liver and one of the leading causes of cancer-related deaths. Current statistics indicate this cancer affects over 700,000 people worldwide and causes an estimated 600,000 deaths annually [1,2]. Despite improvements in prevention, early diagnosis and new treatments, the mortality of patients with HCC continues to rise and, over the past two decades, the incidence of HCC in the a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 To generate in vivo tumors, Huh7 culture cells in mid-log phase growth were collected and resuspended in a 50% mixture of Matrigel (BD Biosciences, USA) in serum-free medium to a final concentration of 6x10 7 cells per mL. A volume of 0.1 mL of the cell suspension was injected subcutaneously in the right flank of each mouse. Mice were weighed and checked for tumor growth every other day. When tumors reached a volume of 100 mm 3 , mice were randomly divided into two groups of 5: vehicle control group and FH535 group (receiving 15 mg of FH535/kg/day from a stock prepared in dimethyl sulfoxide (DMSO) at 21.7 mg/mL and diluted in serum-free medium to a final concentration of 40% DMSO). Vehicle and FH535 were administered by intraperitoneal injection every other day. Tumors were measured using an optical caliper and tumor size was calculated using the formula: 0.5 × length × (width) 2 . Mice were euthanized at the end of the experiment or when reaching humane end-point following AVMA guidelines. Humane end-points included animals with tumors exceeding 20 mm in maximum diameter, with ulcerated tumors, more than 20% body weight loss, impaired mobility, labored breathing or with a body condition score below 2 [27].

Hematoxylin and Eosin (H&E) and immunohistochemistry of explanted tumors
Tumors from the xenograft model were formaldehyde fixed and paraffin-embedded and were used to performed H&E staining and immunohistochemistry of Ki-67 according to standard procedures.

Western blot analyses
Cell lysates were prepared in ice-cold RIPA buffer with freshly added protease inhibitor cocktail (ThermoFisher, USA). Protein concentration was determined using the BCA Protein Assay (ThermoFisher, USA). Cellular proteins (20-40 μg) were separated on SDS-polyacrylamide gel and transferred to PVDF membrane (ThermoFisher, USA). Primary antibodies are described in S1 Table. All primary antibodies were used at 1:1000 dilution dilution with exception of the β-actin antibody at 1:10000 following manufacturer recommendations. Proteins were detected by incubating with horseradish peroxidase-conjugated antibodies (Cell Signaling Technology, USA). Specific bands were visualized with enhanced chemiluminescence reagent (BioRad, USA) and quantified using the ImageJ software (Bethesda, Maryland, USA).

Metabolic analysis
Oxygen Consumption Rates (OCR) were measured on an XF-96 Extracellular Flux Analyzer (Seahorse Bioscience) using the protocol and conditions optimized for HCC cells as previously described by our group [28]. Briefly, the experiments were performed by seeding 10,000 and 20,000 Huh7 cells per well in XFe 96 well-plates 36 h before the experiment. Cells were treated for 24 h with FH535, FH535-N or vehicle control. In OCR experiments the media was supplemented with 25 mM Glucose and 1mM Pyruvate just before the assay. After minimal incubation time (~20-30 min in non-CO 2 37˚C incubator) mitochondrial stress test was initiated. During the assay, 4 different drugs with the following final concentrations were injected to all of the 96 wells: 1) Port A-1 μM Oligomycin A, 2) Port B-1.5 μM Carbonyl cyanide 4-(trifluoromethoxy) phenylhydrazone (FCCP), 3) Port C-200 mM Etomoxir, and 4) Port D-mixture of 1 μM Rotenone and 1 μM Antimycin A. All reagents used in the Seahorse experiments were purchased from Sigma-Aldrich. Analyses of data were performed with Wave 2.2 software (Seahorse Bioscience), Excel (Microsoft Office 2013) and Prism 7.0 (GraphPad Software).

Autophagy assay
Autophagy responses were monitored with the Cyto-ID autophagy detection reagent 2.0 (Enzo Life Sciences, USA) according to the manufacturer's instructions. Huh7 and PLC/PRF/ 5 cells were seeded in 12-well plates and treated the following day with drugs or DMSO vehicle control at the indicated concentrations. CQ was added to the corresponding wells to a final concentration of 50 μM 8 h prior harvesting. After 40 h of drug treatment, cells were collected and assessed for viability with Zombie Violet dye solution (BioLegend, USA) followed by Cyto-ID autophagy detection reagent staining. Flow cytometry analyses were performed in a BD LSRII at the Flow Cytometry and Cell Sorting Shared Resource Facility of the University of Kentucky Markey Cancer Center and data were analyzed with the FlowJo Software version 7.6.5 (Tree Star, USA). The autophagic flux was quantified by subtracting the Cyto-ID MFI value of the sample without CQ from the Cyto-ID MFI value of the sample with CQ for each condition according to the formula: Analysis of Cyto-ID MFI was performed on live cells (Zombie negative stained population).

[ 3 H]-thymidine incorporation assay
Huh7, PLC/PRF/5 and Hep3B cells were plated in 96-well plates at 3000-4000 cells/well, treated with the concentrations indicated of FH535 or FH535-N, as single agents or in combination with sorafenib, and cultured for 72 h. 3 H-thymidine incorporation assay was performed as described previously [24,30].

Apoptosis assay
Apoptosis assay was performed in Huh7 and PLC/PRF/5 cells treated 48 h with DMSO vehicle control or the indicated doses of FH535, FH535-N alone or in combination with sorafenib. Cells were harvested and stained with the APC Annexin V apoptosis detection kit with PI (Bio-Legend, USA) according to the manufacturing instructions followed by flow cytometry analysis. Flow cytometry data was acquired with an LSRII instrument (BD-Biosciences) and analyzed with FlowJo software (Tree Star).

Dual luciferase assay
Cells were plated at 1x10 5 cells/well in 24-well plates and transiently transfected with 490 ng of luciferase-reporter plasmid and 10 ng of phRL-TK per well using the Turbofect transfection reagent (Thermo Scientific, USA). After 5-6 h post-transfection, cells were treated with drug (s) or DMSO vehicle control for 36 h in the presence or absence of LiCl (10 mM). Luciferase assays were performed using the Dual Luciferase Assay System (Promega, USA) according to manufacturer's instructions.

Statistical analysis
Data was reported as mean ± SD of triplicate experiments (except where indicated). Statistical analyses were performed using GraphPad PRISM 7.0. Statistical significance of differences between two groups was analyzed using Student's t-test or ANOVA with post-hoc Tukey HSD accordingly. In all analyses, p<0.05 was considered a statistically significant difference.

FH535 inhibits the growth of xenograft tumors in mice
We previously showed that FH535 decreased the proliferation of different, human HCC cell lines, including Huh7 cells [24]. To validate further the in vivo anti-tumor effect of FH535, we performed a gross-toxicity assay in mice with FH535 doses ranging from 0 to 30 mg/kg. We first demonstrated that intraperitoneal injections up to 15 mg/kg of FH535 for a period of 5-6 weeks did not induce major signs of body distress or toxicity such as weight loss, decreased ambulatory ability, labored respiration or dehydration (Fig 2A). Next, we evaluated the in vivo anti-tumor activity of FH535 in a Huh7 tumor xenograft model. When HCC tumors reached a volume of 100 mm 3 , mice were injected with DMSO vehicle (control group) or 15 mg/kg of FH535 every other day. After only four days of treatment, the tumor volumes of FH535-treated mice were already significantly reduced compared to control group (p<0.05) (Fig 2B and 2C). This result demonstrated the efficacy of the FH535 in vivo on the progression of HCC tumor growth. We also performed Hematoxylin and Eosin staining to assess tumor characteristics and showed that tumors in both groups were poorly differentiated HCC. We evaluated proliferation index using immunohistochemistry with Ki-67 expression, which demonstrated a proliferation index greater than 95% in both groups ( Fig 2D).

FH535 affects autophagy in HCC cells
Increasing evidence for the crosstalk between Wnt/β-catenin and autophagy prompted an evaluation of the linkage between the anti-proliferative effect of FH535 on HCC and autophagy modulation. To investigate this possibility, we first examined LC3 expression levels as a marker of autophagic activity (Fig 3A). Treatment of Huh7 cells with FH535 increased the lipid-bound expression of LC3II levels in a dose-dependent manner in comparison with control-treated cells, a finding that indicated an accumulation of autophagosomes (Fig 3A, left panel). Consistent with the results observed for LC3II, western blot analyses indicated that p62, another autophagy marker, was also increased in FH535-treated cells (Fig 3B). In addition, the knockdown of β-catenin in Huh7 cells increased the expression of both LC3II and p62, which is consistent with the targeting of β-catenin as potential mechanism of action of FH535 (S1 Fig). In this context, β-catenin was reported as a transcriptional repressor of p62 [31], and as expected, our results verified the increase of p62 mRNA levels in the presence of FH535 (Fig 3B and 3C).

FH535 modulates autophagy flux in HCC cells
The accumulation of LC3II and p62 observed in FH535-treated cells were consistent with an effect on autophagy. This effect was attributed to changes in either autophagosome formation, the autophagic flux, or both. To discriminate among these processes, we analyzed the induced accumulation of LC3II protein levels after blocking the autophagosome degradation with CQ. As expected, the addition of CQ enhanced LC3II accumulation in both control and drugtreated cells compared to the corresponding experiments without CQ (Fig 3A, middle panel). However, the progressive increase of LC3II levels in FH535-treated cells alone (Fig 3A, first and second panels) and the reduced accumulation of LC3II in the presence of CQ (first, third and fourth panels) reflected the alteration of the autophagic activity in FH535-treated cells. To further support these findings, the Cyto-ID autophagy detection assay revealed a reduced accumulation of autophagic vesicles in response to FH535 treatment after addition of CQ (Fig 4). Together, these results are consistent with the reduced autophagic flux in FH535-treated cells.

FH535-N on HCC proliferation, apoptosis and β-catenin pathway
We previously described the synthesis of FH535 derivatives from commercially available halogen-substituted arylsulfonyl chlorides and aryl amines [29]. 2,5-Dichloro-N-(4-nitronaphthalen-1-yl)benzenesulfonamide (FH535-N) inhibited cell proliferation of Huh7, PLC and Hep3B  5) and reduced the Wnt/β-catenin transcriptional activity as demonstrated by using a TOP-Flash TCF4-dependent luciferase reporter assay ( Fig 6A) as well as the expression of known downstream Wnt/β-catenin targets genes (Fig 6B and 6C). FH535-N demonstrated significant increased rate of apoptosis in Huh7 and PLC/PRF/5 (Fig 7). In a previous article our group demonstrated the targeting of FH535 on mitochondrial respiration activity. Based on these results we performed a comparative study of the effects of FH535 and the derivative FH535-N on OCR. Our results, showed that both drugs induced similar inhibition of Spare Respiratory Capacity (SRC) and enhanced Proton Leak. These findings indicate similar alteration of the metabolic plasticity and increased oxidative stress of HCC cells treated with FH535 or FH535-N (Fig 8). We analyzed the expression levels of LC3II and p62 in Huh7 cells after treatment with FH535-N in the presence and absence of CQ. Similar to the results observed with FH535, FH535-N also increased LC3II and p62 protein levels (Fig 9A and 9B). Addition of CQ demonstrated accumulation of LC3II with FH535-N and CQ treatment, as expected. However, this accumulation in the combination treatment is reduced at higher doses of FH535-N consistent with the results of FH535-treated cells. (Fig 9A right panel). These results were consistent with a reduction in the autophagic flux in response to FH535-N in a fashion that mirrored the effects described for FH535. In addition, this possibility was supported by the results of FH535-N in Cyto-ID autophagy readouts (Fig 10). Similar results were observed in PLC/PRF/5 cells (Fig 11 and S2 Fig). Overall, our data demonstrated that the antiproliferative effects of FH535 and its derivative, FH535-N, on HCC cells are associated with the regulation of autophagic processes.

FH535 and FH535-N in combination with sorafenib on HCC cell proliferation, apoptosis and autophagy
Our group have previously demonstrated a synergistic effect on cell proliferation using FH535 in combination with sorafenib. Due to these findings, we assessed the effect of drug combination of FH535 or FH535-N with sorafenib on HCC cell proliferation, apoptosis and autophagic flux. We found that FH535 and FH535-N have an additive effect on HCC cell proliferation of Huh7, PLC/PRF/5 and Hep3B in combination with sorafenib ( Fig 5). We have also observed a significant increase in apoptosis measured by Annexin V/PI using the combination treatments (FH535/sorafenib, FH535-N/sorafenib). This effect of the drug combination was more pronounced that the one seen with FH535, FH535-N or sorafenib alone (Fig 7). FH535/sorafenib and FH535-N/sorafenib drug combination produced a significant decreased in the authophagic flux measured by CytoID (Fig 10). We also used HCC cell lines Huh7 and PLC to perform

Discussion
Aberrant activation of the Wnt/β-catenin pathway occurs in numerous malignancies, including HCC [6,[11][12][13][14]. The poor prognosis and disease progression in liver cancer typically involves the upregulation of the Wnt/β-catenin pathway [7], and recent efforts focus on the development of new compounds targeting this and other signaling pathways as effective therapeutic alternatives for advanced HCC. The N-aryl benezenesulfonamides, such as FH535, inhibits the Wnt/β-catenin signaling pathway and the PPARs δ and γ with demonstrated antiproliferative effect against pancreatic cancer, breast cancer, colorectal carcinoma and HCC cells [21,24,32,33]. FH535 also sensitizes and reverses the epithelial-mesenchymal transition phenotype of radio-resistant esophageal cancer cells [34]. In vivo, FH535 effectively suppresses growth and angiogenesis in pancreatic cancer and decreases tumor burden and progression in colorectal cancer [21,33]. We now demonstrate the potent effects of FH535 on HCC tumor progression in vivo using a mouse xenograft model while showing no significant drug toxicity in the host. Although, there is important pre-clinical evidence for the anti-cancer effects of FH535, the mechanism of action of this drug remains poorly understood. We recently demonstrated that FH535 induces changes in mitochondrial membrane potential and overall mitochondrial health in HCC tumor cells [23]. FH535 targets specifically the electron transport chain complexes I and II and results in defective mitochondrial respiration [23]. Since mitochondrial dysfunction and Wnt/β-catenin signaling affect the regulation of the autophagy process [35], this study reports on the anti-tumor effect of FH535 and its derivative FH535-N on HCC cells through the modulation of the autophagic activity.
Compared to untreated-control cells, our results demonstrate that FH535 increased LC3II and p62 levels in HCC cells, a finding that is indicative of autophagosomal accumulation by the increase in autophagosome formation and/or by a defective lysosomal degradative machinery. A modest CQ-induced increase in autophagosome accumulation occurs in FH535-treated cells, together with the reduced ΔLC3II levels in western blots, an additional finding that is consistent with an impaired autophagic flux. This may contribute to the accumulation of dysfunctional mitochondria and account for the increased apoptosis in FH535-treated cells. Future studies will assess the involvement of FH535 on autophagosomal formation on HCC cells as suggested by the enhancement of p62/SQSTM1 gene expression.
Several studies reveal the complex interplay between Wnt/β-catenin signaling and autophagy [31,[36][37][38][39]. The coordinated regulation of Wnt/β-catenin signaling and autophagy processes occurs in different types of cancers, including the cytotoxic effect of resveratrol on breast cancer cells [40] and the reduced gemcitabine-induced apoptosis in human osteosarcoma cells [41]. In this regard, the inhibition of Wnt/β-catenin pathway induces the accumulation of autophagic proteins such as LC3-II, ATG7, Beclin-1 and p62 proteins [38,42]. Reciprocally, induction of autophagy regulates the Wnt/β-catenin pathway by targeting the clearance of β-catenin and other proteins involved in Wnt signaling such as the Dishevelled protein [31,43]. In agreement with these studies, our results show that FH535 treatment induces the accumulation of LC3II and p62 proteins as well as p62/SQSTM1 mRNA and suggests that the effect of FH535 on autophagy links to the inhibition of Wnt/β-catenin signaling. In support of this possibility, the β-catenin knockdown in HCC cells also exhibits a subsequent increase in LC3II and p62 protein levels.
N-Aryl benezenesulfonamides, such FH535 and FH535-N exhibit significant anti-cancer effects [29]. In this study, FH535 and FH535-N produce similar anti-proliferative activity in HCC cells. Furthermore, both compounds target the Wnt/β-catenin signaling pathway as indicated by β-catenin-dependent reporter assays (Fig 6A) as well as the reduced expression of endogenous downstream β-catenin target genes (Fig 6B and 6C). Additionally, FH535-N induces the accumulation of autophagic proteins p62 and LC3II and impairs the autophagic flux in Huh7 cells. These results provide further evidence that FH535-based derivatives warrant additional development as anti-cancer drug candidates for HCC treatment. Moreover, the synergistic effects of FH535 and sorafenib on the inhibition of HCC cell proliferation and survival was associated to the distinct targeting of both drugs on the mitochondrial function and metabolic pathways [23]. Inhibition of autophagy by ATG7 knockdown or CQ treatment sensitizes HCC cells to sorafenib by enhancing apoptosis [18]. Likewise, and consistent with these reports, our findings indicate that the inhibition of autophagic flux by FH535 and FH535-N contributes, at least partially, to the synergistic effects observed using FH535 in combination with sorafenib. We also demonstrate an additive effect of drug combination therapy using FH535 or FH535-N with sorafenib on HCC cell proliferation and apoptosis. Importantly, our findings revealed a significant increase in autophagic disruption caused by these combinatory treatments compared to FH535, FH535-N or sorafenib alone.
In conclusion, our data demonstrate potent anti-tumor effects of FH535 in vivo at dosage levels (15 mg/kg/day) that produce no gross toxicity in the mice. These studies also reveal a contributing mechanism for the anti-tumor action of FH535 with the Wnt/β-catenin-mediated regulation of the autophagy process. Further studies are warranted to assess the efficacy of FH535 and its derivatives either alone or in combination with conventional therapies as rational therapeutic alternatives for HCC treatment.