YAP Inhibition by Nuciferine via AMPK-Mediated Downregulation of HMGCR Sensitizes Pancreatic Cancer Cells to Gemcitabine

Nuciferine, a major aporphine alkaloid constituent of lotus leaves, is a raw material for obesity treatment. Extensive studies have revealed that obesity is associated with pancreatic cancer (PC). However, it has not been clarified whether nuciferine could be used in PC treatment or prevention. Here, we show that nuciferine could enhance the sensitivity of PC cells to gemcitabine in both cultured cells and the xenograft mouse model. The mechanism study demonstrated that nuciferine induced YAP Ser127 phosphorylation [pYAP(Ser127)] through AMPK-mediated 3-hydroxy-3-methyl-glutaryl-coA reductase (HMGCR) downregulation. Remarkably, wild-type YAP overexpression or YAP Ser127 mutant could resist to nuciferine and no longer sensitize PC cells to gemcitabine. Knockdown of AMPK attenuated pYAP(Ser127) induced by nuciferine. Moreover, knockdown of AMPK reversed nuciferine-mediated HMGCR downregulation. Notably, HMGCR inhibiting could restrain YAP by phosphorylation Ser 127, and therefore enhance the efficiency of gemcitabine in PC cells. In line with this consistent, overexpression of HMGCR reduced growth inhibition caused by nuciferine and/or gemcitabine treatment in PC cells. In summary, these results provide an effective supplementary agent and suggest a therapeutic strategy to reduce gemcitabine resistance in PC.


Introduction
Pancreatic cancer (PC) is the seventh leading cause of cancer-related deaths worldwide. Since patients seldom exhibit symptoms until an advanced stage of the disease, PC remains one of the most lethal malignant neoplasms that caused 432,242 new deaths in 2018 worldwide (GLOBOCAN 2018 estimates). Worldwide incidence and mortality of PC correlate with increasing age and is slightly more common in men than in women [1]. Despite the great progress achieved in the last couple of decades, surgical intervention still remains the optimal treatment option. However, due to the late diagnosis and early metastasis, patients with PC usually lose the opportunity to undergo radical surgery. Therefore, identifying new drugs or therapeutic treatments to address this troublesome disease is urgently needed.
AMP-activation protein kinase (AMPK) is an energy sensor and master regulator of metabolism. Several studies have revealed that AMPK is closely related to drug resistance [2,3]. There are some small molecular inhibitors and bioactive compounds extracted from natural products improving multidrug resistance through activation of AMPK [4,5]. Additionally, the combination between AMPK media contained 10% heat-inactivated fetal bovine serum (FBS, GIBCO, USA), 100 U/mL penicillin and 100 µg/mL streptomycin (GIBCO, USA). All cell lines were incubated 37 • C in a 5% CO 2 atmosphere.

MTT Assay
Cell viability was determined by 3-(4 5-Dimethylthiazol-2-yl)-2 5-diphenyltetrazolium bromide (MTT) assay. PANC-1, BxPC-3 and ASPC-1 cells were treated with nuciferine and/or gemcitabine for 72 h at various concentrations. The formazan was dissolved in DMSO and the absorbance was measured by the Universal Microplate Reader (BIO-TEK instruments, Inc., Vermont, MA, USA). The cell viability was calculated by the following formula: (A treated/A control) × 100%. Drug interaction between nuciferine and gemcitabine was assessed using the combination index (CI), and CI was calculated using the Chou-Talalay equation [24].

Colony Formation Assay
PANC-1 cells were plated (200,000 per well) in a six-well plate and incubated overnight. After nuciferine and/or gemcitabine treatment for 24 h, cells were trypsinized and viable cells were counted. Then 1000 viable cells were plated in 100-mm Petri dishes and incubated for 14 days. Colonies were fixed using 4% paraformaldehyde and then stained for 15min using 2% crystal violet. The number of colonies was counted as described previously [25].

Western Blot Assay
Transplanted tumor tissue or cell protein were extracted with lysis buffer. The concentration of protein was determined by BCA assay kit (Beyotime, Shanghai, China). Protein samples were analyzed by SDS-PAGE using a 10%-12% polyacrylamide and transferred onto the NC membranes (Millipore, Billerica, MA, USA). Immune complexes were formed by incubation of primary antibodies for overnight at 4 • C, then followed conjugated second antibody for 1 h at 37 • C. Immunoreactive protein bands were detected with Tanton chemiluminescenc image analysis system (Shanghai, China), and the density of protein band was detected by Image J software.

Real-Time PCR Analysis
After nuciferine treatment for 24 h, PANC-1 cells were lysised using the Trizol Reagent (Thermo Fisher Scientific, Waltham, MA, USA) and total RNA was extracted. cDNAs were synthesized and RT-PCR was performed on the CFX96™ Real-Time PCR Detection System (BIO-RAD, Hercules, CA, USA). The primer sets used in the PCR assay were as follows:

ATP and ADP Quantification Assays
Cellular ATP and ADP quantification were measured using ADP/ATP Ratio Assay Kit (Sigma, St Louis, MO, USA). PANC-1 and ASPC-1 cells plated in 12-well plates were incubated with or without nuciferine for 72 h, and the cultured cells were lysed to release ATP and ADP. ATP and ADP level were determined according to the manufacturer's instructions.

Immunoprecipitation
PANC-1 cells were treated with nuciferine (50 µM) for 24 h. After treatment, the cells were washed and lysed for 15 min on ice and then centrifuged at 12000× g for 10 min and the soluble fraction was collected. Immunoprecipitation analysis AMPK was immunocaptured from total cell extracts using antibodies to AMPK crosslinked to protein A-agarose beads (Santa Cruz, CA, USA). The complexes were analyzed by Western blot and detected with antibody against YAP.

Immunofluorescent (IF) Staining
After treatment with nuciferine (50 µM) for 24 h, PANC-1 cells were washed with cold PBS, fixed with 4% paraformaldehyde for 20 min and permeabilized with 0.2% Triton X-100 for 5 min. After incubated with 5% BSA for 1 h, cells were incubated with anti-YAP antibody overnight at 4 • C. After being washed twice, the cells were incubated with FITC-labeled goat anti-rabbit IgG (H+L) antibody (Jackson ImmunoResearch, PA, USA) for 1 h at 37 • C. In addition, the coverslips were stained with DAPI for 15 min. The images were captured with a confocal scanning microscope (ZEISS LSM800, Jena, Germany).

Small
Interfering RNA (siRNA) Transient Transfection siRNA targeting AMPK, LATS1/2 and HMGCR for knockdown were purchased from GenePharma (Shanghai, China). The siRNAs were delivered using Lipofectamine 3000 (Invitrogen Life Technologies, CA, USA) according to previous researche [26]. After formation of the siRNA-liposome complexes, the mixture was added to cells for 6 h.

Plasmid Extraction and Transfection
The wild-type YAP plasmid, mutant YAP(c.781T→G, encoding p.Ser127Ala) plasmid and HMGCR plasmid were synthesized by Genechem (Shanghai, China). EndoFree Plasmid Midi Kit from Beyotime (Shanghai, China) was chosen for plasmids extraction. For transfection, PANC-1 cells were seeded in 6-well plates at 65% confluency. Then, the plasmid DNA was introduced into the cells using Lipofectamine 3000 following the instructions. were injected subcutaneously into the right flank of mice. After 4 weeks, nude mice with the xenograft tumour sizes of approximately 100 mm 3 were randomly assigned to four groups (n = 4, each group): Vehicle group, Nuc treatment (intraperitoneally [IP] injected with Nuc at a dose of 30 mg/kg, once/every other day), Gem treatment (20 mg/kg by IP injection twice weekly) and the combination treatment of Nuc and Gem (30 mg/kg Nuc once/every other day and 20 mg/kg gemcitabine twice weekly). Tumor volume and mice body weight were measured every three days. The tumor volume was calculated using the formula, V = length × width 2 /2. After therapy was continued for 4 weeks, mice were sacrificed and tumor samples were excised and weighed. The major organ sections were excised for toxic evaluation.

Histological Analysis
The major organ (heart, liver, spleen lung and kindy) sections were fixed in 4% paraformaldehyde solution, and then embedded and sectioned for Hematoxylin-eosin staining. Images were captured using a light microscope (Leica DM6000B, Munich, Germany).

Statistical Analysis
Each experiment was repeated three times, unless otherwise indicated. Data were presented as mean ± SD from triplicate parallel experiments. Statistical analysis were performed using one-way ANOVA.

Nuciferine Attenuates Gemcitabine Resistance of Pancreatic Cancer Cells
We first examined the abilities of nuciferine to suppress tumor growth in PANC-1, BxPC-3 and ASPC-1 cell lines. As observed in Figure 1a, treatment with nuciferine by itself with concentrations up to 50 µM didn't elicit optimal growth inhibitory effects on PC cells as IC 50 values were not obtained. To determine whether nuciferine could enhance the susceptibility of PC cells to gemcitabine, combination treatments were carried out by varying gemcitabine in the presence of nuciferine for 72 h. As shown in Figure 1b, addition of a suboptimal dose of nuciferine lowered the gemcitabine IC 50 from 1120 nM to 402 nM (2.8-fold) in PANC-1 cells, 164 nM to 66 nM (2.48 fold) in BxPC-3 cells and 720 nM to 305 nM (2.36 fold) in ASPC-1 cells. These experiments demonstrated that suboptimal dose of nuciferine could decrease the doses of gemcitabine required to reach the IC 50 for these PC cell lines. Moreover, these experiments showed the differential sensitivity of these three PC lines to gemcitabine, almost a 7-fold difference in sensitivity between PANC-1 cells and BxPC-3 cells. Furthermore, the Chou-Talalay CI indicated that gemcitabine and nuciferine synergistically enhanced cytotoxicity in ASPC-1 cells (CI = 0.85; Figure 1c). On the other hand, the combined treatment of gemcitabine and nuciferine enhanced cellular cytotoxicity additively in PANC-1 cells (CI = 1; Figure 1c). Althrough combination therapy did further enhance cellular cytotoxicity, the CI showed no synergism and additive effect between nuciferine and gemcitabine in BxPC3 cells (CI = 1.39; Figure 1c).
Next, the antiproliferative activity of nuciferine and/or gemcitabine was further evaluated by colony formation assay. In PANC-1 cells, treatment with either compound effectively reduced colony number compared with controls, and the combination of nuciferine and gemcitabine further inhibited the clonogenic capacity (p < 0.05, Figure 1d). Taken together, these results suggest that nuciferine was not only cytotoxic to PC cells but also sensitized PC cells to gemcitabine and suppressed cellular proliferation.

Nuciferine Activates AMPK in Pancreatic Cancer Cells
A growing body of evidence has suggested that activation of AMPK is pivotal to increase the susceptibilities of tumor cells to chemotherapy drugs [27,28]. To determine whether AMPK activation is susceptible to nuciferine enhanced the sensitivity of PC cells to gemcitabine, we examined the effect of nuciferine on the activation of AMPK in PC cells. A dose-dependent study in PANC-1 cells revealed a moderate increase in phosphorylation of AMPK at Thr 172, the active form of AMPK, 48 h after exposure to 50 µM nuciferine and very extensive increase at concentrations of 100 µM (Figure 2a). Time-course analysis of cells exposed to 50 µM nuciferine demonstrated a significant increase in pAMPK (Thr172) as early as 6 h. These events became apparent after 12 h of nuciferine exposure, and reached near maximal levels after 24 h (Figure 2a). To determine whether nuciferine mediated AMPK activation observed in PANC-1 cells also occur in other PC cell lines, parallel studies were carried out in ASPC-1 cells. Also, these cells exhibited comparable degrees of AMPK activation (Figure 2b). Moreover, we measured ATP/ADP ratios in nuciferine -treated PC cells. Our results showed that the nuciferine treatment significantly decreased ATP/ADP ratio (Figure 2c), which is the indicator of activation of AMPK. Notably, treatment combining nuciferine and gemcitabine induced more pAMPK (Thr172) than monotherapy (Figure 2d), suggesting that AMPK activation was involved in nuciferine enhanced the sensitivity of PC cells to gemcitabine.

Nuciferine Induces YAP Ser127 Phosphorylation and Cytoplasmic Retention
AMPK has been shown to oppose the growth-promoting activity of YAP, which promote chemoresistance in PC cells. To examine the role of YAP inhibition in nuciferine-elevated efficiency of gemcitabine, we examined the effect of nuciferine on combination of AMPK protein and YAP. As co-immunoprecipitation assays showed that nuciferine didn't enhance the combination of AMPK kinase subunits, which were identified as YAP-associated proteins, and YAP (Figure 3a). Concordantly, nuciferine treatment didn't increase YAP S61 or S94 phosphorylation (Figure 3b). Activation of AMPK induces an increase in YAP phosphorylation at Ser127 and favors YAP localization to the cytoplasm. As shown in Figure 3b, nuciferine treatment markedly increased YAP Ser 127 phosphorylation, while phosphorylation of Ser 909 and Thr 1079 in LATS1 which is positively correlates with LATS activity [29,30] has no noticeable change. The confocal microscopic image showed that nuciferine treatment decreased nuclear YAP staining (Figure 3c). In this regard, the downstream target genes of YAP (CYR61 and CTGF) were significantly downregulated in nuciferine treated PANC-1 cells (Figure 3d). Thus, nuciferine impair the nuclear translocation of YAP acts through inducing YAP Ser127 phosphorylation.  . (a,b) The whole-cell lysates were prepared and subjected to western blot for pAMPK(Thr172) and AMPK, and relative density between AMPK and pAMPK(Thr172) was determined. (c) Levels of ATP and ADP were detected and the ratio of ATP to ADP was calculated. (d) After treatment with nuciferine and/or gemcitabine, protein level of AMPK and pAMPK(Thr172) was determined. Nuc, nuciferine; Gem, gemcitabine. * p < 0.05 or ** p < 0.01 versus DMSO-treated control cells.
We next explored whether overexpression of wild-type YAP or YAP Ser127 mutant could resist to nuciferine and no longer sensitized PC cells to gemcitabine. When PANC-1 cells were transfected with wild-type YAP plasmid, cell viability in cells with nuciferine and/or gemcitabine treatment was further increased (Figure 3e,f), and transfection of YAP mutant plasmid completely abolished cytotoxicity induced by nuciferine, gemcitabine or their combination (Figure 3e,f). These results further confirmed that nuciferine enhanced the efficacy of gemcitabine through inducing pYAP(Ser127). ). (f) Cells transfected with wild-type YAP or mutant YAP overexpressed plasmid were treated with nuciferine and/or gemcitabine for 72 h and cell viability was determined. Data were showed as mean ± SD. Nuc, nuciferine; Gem, gemcitabine; WT, wide type. * p < 0.05 and ns = not significant.

Nuciferine Inhibits YAP Involving AMPK-Mediated Downregulation of HMGCR
To investigate the role of AMPK in nuciferine-regulated YAP activity, we used siRNAs targeting AMPKa1 and AMPKa2 to knock down AMPK in PANC-1 cells. The siRNAs greatly inhibited pYAP(Ser127) induced by nuciferine, suggesting that AMPK is required to nuciferine-mediated the regulation of YAP (Figure 4a). Consistently, we also observed that AMPK knockdown rescued the effects of nuciferine on YAP nuclear localization (Figure 4b). In addition, nuciferine treatment didn't increase the protein level of AMOTL1 (Figure 4c) which become more stable after phosphorylated by AMPK, thereby promoting the phosphorylation of YAP by LATS on serine 127. Furthermore, knockdown of LATS1/2 didn't attenuate nuciferine-mediated YAP phosphorylation (Figure 4d). These results further indicate that nuciferine induce YAP phosphorylation through AMPK-mediated LATS-independent manner. We then set out to study how nuciferine regulates YAP activity. Previous studies suggested that RHO regulates YAP is largely independent of the LATS1/2 Hippo pathway kinases and mevalonate cascade is required for activation of RHO [16,31]. Therefore, we investigated whether nuciferine-mediated the inhibition of YAP is associated with the down-regulation of HMGCR, the major regulator of the mevalonate pathway. Western blot analysis showed that nuciferine treatment decreased the protein level of HMGCR in PC cells (Figure 4e). Moreover, knockdown of AMPK reversed nuciferine-mediated HMGCR downregulation (Figure 4f). Notably, the HMGCR overexpression obviously attenuated YAP phosphorylation (Figure 4g). These results suggest that nuciferine may inhibit YAP by activating the AMPK-HMGCR axis.

Nuciferine Sensitizes Pancreatic Cancer Cells to Gemcitabine by Down-Regulating HMGCR
To evaluate the role of HMGCR on PC cells resistant to gemcitabine, PANC-1 cells were transfected with siRNA against HMGCR. Knockdown efficiency was confirmed by western blot (Figure 5a). Results showed that proliferative ability was lower in cells transfected with HMGCR siRNA compared with nonspecific RNA groups after gemcitabine treatment (Figure 5b). Furthermore, HMGCR siRNA enhanced the gemcitabine inhibition effect on colony-formation ability (Figure 5c). To gain mechanistic insight into HMGCR-mediated regulation of gemcitabine efficiency, we tested whether YAP is a substrate of HMGCR. Knockdown of HMGCR increased phosphorylation of YAP Ser 127 (Figure 5d). In line with this consistent, treatment with HMGCR inhibitor simvastatin inhibits YAP by phosphorylation Ser 127 in PC cells (Figure 5e). These events indicate that inhibiting HMGCR could restrain YAP by phosphorylation Ser 127, and therefore enhance the efficiency of gemcitabine in PC cells. As expected, overexpression of HMGCR reduced the growth inhibition caused by nuciferine and/or gemcitabine treatment in PC cells (Figure 5f). We propose that nuciferine-mediated HMGCR downregulation increases sensitivity of PC cells to gemcitabine.

Nuciferine Enhances PANC-1 Cells Sensitivity to Gemcitabine in Vivo
Finally, we evaluated whether nuciferine attenuates gemcitabine resistance of PC in vivo. We generated a subcutaneous xenograft model of PANC-1 cells in immunodeficient mice and then treated them with nuciferine and/or gemcitabine for 4 weeks (Figure 6a). Administration of nuciferine showed a similar effect as gemcitabine, with a remarkable reduction of the xenograft tumor volume and tumor weight compared with the control mice (Figure 6b-d), while combined treatment with nuciferine and gemcitabine further inhibited tumor growth (Figure 6b-d). Importantly, nuciferine and/or gemcitabine treatment for 4 weeks was well tolerated, without significant weight loss (Figure 6e). Histopathological analyses of vital organs (heart, liver, spleen, lung and kidney) also revealed that nuciferine and/or gemcitabine treatment didn't result in the toxicity (Figure 6f). We then detected the expression of pAMPK(Thr172)/HMGCR/pYAP(Ser127) in the xenograft tumors and found that combined nuciferine and gemcitabine treatment up-regulated the expression of pAMPK(Thr172) and pYAP(Ser127), while down-regulated levels of HMGCR compared with gemcitabine treatment (Figure 6g). Taken together, these data support our in vitro findings that nuciferine enhance the PC cells chemosensitivity to gemcitabine by inducing pYAP(Ser127) through AMPK-mediated downregulation of HMGCR.

Discussion
PC chemotherapy is frequently impeded by drug resistance. Recent work indicates that YAP activation induces PC cell chemoresistance and suggests that "switching off" of the Hippo-YAP pathway could help to prevent resistance to PC therapies [32]. The present study verified that treatment with nuciferine effectively attenuates gemcitabine resistance of PC via inhibition of YAP activity. Then, a key question arises, as to the mechanism for nuciferine-mediated inhibition of YAP.
AMPK is an energy sensor and master regulator of metabolism and AMPK activation occurs when the intracellular ATP/ADP ratio decreases. In our study, we found that nuciferine activated AMPK by decreasing cellular ATP/ADP levels in PC cells. Activated AMPK suppresses ATP consumption processes by inhibiting de novo synthesis of glucose, protein, and lipids and diverts cellular metabolism into energy-producing. In view of the essential role of cellular energy and metabolites for survival and growth, YAP is inhibited when energy level is low. Indeed, activation of AMPK induces an increase in YAP phosphorylation, which favors YAP localization to the cytoplasm, and finally decreases expression of YAP target genes [33]. Our research revealed that nuciferine induced YAP Ser127 phosphorylation and cytoplasmic retention and inhibited its transcriptional activity. However, cellular energy stress can inhibit YAP by both AMPK-dependent and independent mechanisms. We used siRNAs targeting AMPKa1 and AMPKa2 to confirm the mechanism of nuciferine-regulated YAP inhibition. AMPK knockdown greatly attenuated YAP Ser127 phosphorylation induced by nuciferine. Moreover, we observed that AMPK knockdown rescued the effects of nuciferine on YAP nuclear localization. These data indicate that AMPK is required to mediate YAP regulation in response to nuciferine treatment. A critical question then arises regarding the mechanism by which AMPK-mediated YAP inhibition occurs during nuciferine treatment.
Previous work indicated that AMPK directly binds to YAP and phosphorylates it. However, nuciferine did not enhance the combination of AMPK protein and YAP. Concordantly, nuciferine treatment did not increase YAP S61 or S94 phosphorylation. These findings argue against the possibility that nuciferine-mediated phosphorylation of YAP through AMPK directly binding to YAP. In response to unfavourable growth conditions, Lats1/2 is phosphorylated and activated, which phosphorylate and inhibit YAP. AMOTL1 could bind LATS1/2 and promote its kinase activity and YAP phosphorylation. De Ran et al. reported that AMPK can stabilize AMOTL1 to stimulate LATS activity, thereby promoting the phosphorylation of YAP on Ser 127 [19]. Our data indicated that nuciferine failed to decrease AMOTL1 protein levels and knockdown of LATS1/2 did not attenuate nuciferine-mediated YAP phosphorylation. These results arguing strongly against the possibility that nuciferine phosphorylates YAP protein through AMPK-mediated LATS1/2-dependent manner.
AMPK stimulation leads to the inhibition of HMGCR [34], which is the major regulators of the mevalonate pathway. Mevalonate metabolism are required for RHO function, which in turn is required for YAP activity [16,35,36]. Moreover, RHO regulates YAP is largely independent of the LATS1/2 Hippo pathway kinases. In our studies, we found that nuciferine decreased HMGCR protein levels and overexpression of HMGCR attenuated nuciferine-mediated YAP Ser 127 phosphorylation. Such findings are accordant with the studies which indicated that inhibition the mevalonate pathway with HMGCR inhibitor causes inhibition of YAP activity [37].
Increasing evidence has revealed that YAP plays a crucial role in regulating PC chemoresistance to gemcitabine [32,38]. We indicated that progressive decrease of pYAP(Ser127) is critical for YAP-mediated gemcitabine resistance. Moreover, our results showed that knockdown of HMGCR enhances the sensitivity of gemcitabine in PC cells, which is consistent with previous report that HMGCR inhibitor simvastatin increased the efficacy of gemcitabine in PC cells [39]. Thus, HMGCR inhibitor-mediated gemcitabine sensitivity is likely related to YAP Ser127 site phosphorylation. In support of this notion, inhibition of HMGCR increased pYAP(Ser127) in PANC-1 cells. Remarkably, overexpression of HMGCR rescued nuciferine-mediated gemcitabine sensitivity in PC cells. Thus, downregulation of HMGCR might serve as an important role for nuciferine to enhance gemcitabine efficiency by promoting pYAP(Ser127).

Conclusions
The current study demonstrates that progressive decrease of pYAP(Ser127) is critical for YAP-mediated gemcitabine resistance. Nuciferine could attenuate gemcitabine resistance occurs in association with increase of pYAP(Ser127) through activation AMPK-mediated HMGCR downregulation in PC cells (Figure 7). This finding reveals a previously unrecognised role of nuciferine in antitumor action, via inhibition of YAP.

Conflicts of Interest:
The authors declare no conflict of interest.