Lovastatin causes FaDu hypopharyngeal carcinoma cell death via AMPK-p63-survivin signaling cascade

Statins are used widely to lower serum cholesterol and the incidence of cardiovascular diseases. Growing evidence shows that statins also exhibit beneficial effects against cancers. In this study, we investigated the molecular mechanisms involved in lovastatin-induced cell death in Fadu hypopharyngeal carcinoma cells. Lovastatin caused cell cycle arrest and apoptosis in FaDu cells. Lovastatin increased p21cip/Waf1 level while the survivin level was decreased in the presence of lovastatin. Survivin siRNA reduced cell viability and induced cell apoptosis in FaDu cells. Lovastatin induced phosphorylation of AMP-activated protein kinase (AMPK), p38 mitogen-activated protein kinase (MAPK) and transcription factor p63. Lovastatin also caused p63 acetylation and increased p63 binding to survivin promoter region in FaDu cells. AMPK-p38MAPK signaling blockade abrogated lovastatin-induced p63 phosphorylation. Lovastatin’s enhancing effect on p63 acetylation was reduced in HDAC3- or HDAC4- transfected cells. Moreover, transfection of cells with AMPK dominant negative mutant (AMPK-DN), HDAC3, HDAC4 or p63 siRNA significantly reduced lovastatin’s effects on p21cip/Waf1 and survivin. Furthermore, lovastatin inhibited subcutaneous FaDu xenografts growth in vivo. Taken together, lovastatin may activate AMPK-p38MAPK-p63-survivin cascade to cause FaDu cell death. This study establishes, at least in part, the signaling cascade by which lovastatin induces hypopharyngeal carcinoma cell death.


Results
Lovastatin arrested cell cycle and induced apoptosis in FaDu cells. MTT assay was employed to determine whether FaDu cell viability is altered in the presence of lovastatin. As shown in Fig. 1a, lovastatin concentration-dependently decreased FaDu cell viability after 24 h exposure. Longer exposure to lovastatin (48 h) further decreased FaDu cell viability (Fig. 1a). To determine whether lovastatin-decreased FaDu cell viability was a result of cell cycle arrest or apoptosis, flowcytometry was used. As shown in Fig. 1b, the percentage of propidium iodide (PI)-stained cells in the S region was significantly decreased in FaDu cells after exposure to lovastatin for 24 h. In addition, lovastatin increased the percentage of PI-stained cells in the G0/G1 region (Fig. 1b). Moreover, 24 h treatment of lovastatin only slightly induced cell apoptosis (sub-G1 region) (Fig. 1b). However, lovastatin significantly induced apoptosis in FaDu cells after 48 h exposure of lovastatin (Fig. 1c). To detect apoptosis in FaDu cells exposed to lovastatin, flowcytometry with PI and annexin V-FITC double-labeling was also employed. As shown in Fig. 1d, lovastatin increased the percentage of early apoptotic cells (annexin V + PI − cells) and advanced apoptotic cells and/or necrotic cells (annexin V + PI + cells) after 48 h exposure. We next determined whether lovastatin activates caspase 3. As shown in Fig. 1e, lovastatin increased the cleaved (active) form of caspase 3 and PARP, a selective caspase 3 substrate. These findings suggest that lovastatin induced apoptosis and inhibited cell proliferation in FaDu cells.
Lovastatin modulated p21 cip/Waf1 , cyclin D1 and survivin expressions in FaDu cells. Since cyclin-dependent kinase (CDK) inhibitor protein, p21 cip/Waf140 , cyclin D1 and survivin 6 play essential role in cell cycle progression or apoptosis. We therefore examined whether lovastatin had any effects on these proteins in FaDu cells. Results from immunoblotting analysis demonstrated that p21 cip/Waf1 (Fig. 2a) was increased, while cycin D1 (Fig. 2b) and survivin (Fig. 2c) were decreased in FaDu cells exposed to lovastatin. We also determined whether lovastatin decreases survivin mRNA. Results from RT-PCR analysis demonstrated that lovastatin significantly decreased survivin mRNA in FaDu cells (Fig. 2d). A survivin siRNA oligonucleotide (survivin siRNA) was employed to determine whether survivin down-regulation induces FaDu cell apoptosis. Survivin siRNA reduced the basal surivvin level in FaDu cells (Fig. 2e). Survivin down-regulation by survivin siRNA mimicked the lovastatin's effects in decreasing cell viability (Fig. 2f). Transfection with survivin siRNA also induced cell apoptosis (Fig. 2gb) while negative control siRNA was without effects (Fig. 2ga). These results suggest that reduced survivin level contributes to lovastatin-induced FaDu cell apoptosis. p63 contributes to lovastatin's actions in FaDu cells. Transcription factor p63 modulates several downstream target genes such as survivin and p21 cip/Waf141 6 , which regulate apoptosis and cell cycle progression. We therefore explored the impact of p63 on lovastatin's actions in FaDu cells. As shown in Fig 3a, lovastatin-increased p21 cip/Waf1 levels were reduced in FaDu cells transfected with p63 siRNA. p63 siRNA also reduced lovastatin's effects on survivin levels (Fig. 3b). In addition, p63 siRNA markedly reduced the basal level of p63 in FaDu cells. Similar to p53, activation of p63 is modulated by its modifications such as acetylation and phosphorylation [42][43][44] . We thus determined the acetylated protein levels in FaDu cells after exposure to lovastatin. As shown in Fig. 3c, the acetylated protein level with molecular weights about 72 kDa (p63 molecular weight: 75 kDa) was increased in cells exposed to lovastatin as determined in immunoblotting using anti-acetylated lysine antibody (Cell Signaling). Resuls from immunoprecipitation analysis further confirmed that lovastatin induced p63 acetylation in FaDu cells (Fig. 3d). We also examined whether lovastatin induces p63 phosphorylation using anti-phosphorylated p63 antibody (Cell Signaling). Lovastatin significantly increased p63 phosphorylation at Ser160 and Ser162 in FaDu cells (Fig. 3e). Moreover, the anti-p40 antibody directed against an N-terminal truncated form of the p63 protein (Δ Np63) is currently replacing anti-p63 antibody as several studies 45,46 . To confirm Scientific RepoRts | 6:25082 | DOI: 10.1038/srep25082 p63 is truly involved in lovastatin's actions in FaDu cells, anti-p40 antibody was used. Results from immunoblotting and siRNA experiments showed that anti-p63 antibody used in this study recognizes the same protein (p63) as anti-p40 antibody does ( Supplementary Fig. 1). These results indicate that lovastatin treatment is capable of modulating p21 cip/Waf1 and survivin and subsequent cellular events through, at least in part, activation of p63.

HDACs inhibition contributes to lovastatin's actions in FaDu cells. Statins including lovastatin may
present as histone deacetylases (HDACs) inhibitors to increase acetylation levels of cellular proteins and subsequent colorectal cancer cell death 20 . We therefore assessed whether HDACs inhibition contributes to lovastatin's actions in FaDu cells. As show in Fig. 4a, expression of a class I HDAC, HDAC3 or a class II HDAC, HDAC4, suppressed lovastatin-induced p63 acetylation. Transfection of cells with HDAC3 or HDAC4 also reduced lovastatin-elevated p21 cip/Waf1 levels (Fig. 4b). In addition, lovastatin-decreased survivin levels were restored in  cells transfected with HDAC3 and HDAC4 (Fig. 4c). We reported previously that p63, similar to p53, might prevent the binding of Sp1 to the promoter region to reduce survivin expression in HT29 cells, a 53-mutant human colorectal cancer cell line 27 . A ChIP experiment was conducted to examine whether lovastatin affects Sp1, p63 or HDAC3 binding to the putative p53/p63 and Sp1 binding sites containing promoter region (− 264 to − 37) of survivin. As shown in Fig. 4d, lovastatin increased p63 binding, while decreases Sp1 and HDAC3 binding to the survivin promoter region (Fig. 4d). These results suggest that HDACs inhibition contributes to lovastatin-induced p63 acetylation and subsequent cellular events in FaDu cells.

NF-κB and STAT3 contribute to survivin repression in lovastatin-stimulated FaDu cells.
In addition to p53, p63 and Sp1, the survivin promoter region (− 300 to − 41) also contains putative NF-κ B and STAT3 binding sites. Several studies showed that transcription factors NF-κ B and STAT3 play important roles in inducing survivin expression 48 . We thus determined whether NF-κ B subunit p65 and STAT3 phosphorylation status, which represent NF-κ B and STAT3 activation, was altered in FaDu cells after exposure to lovastatin. As shown in Fig. 7a, lovastatin reduced p65 phosphorylation in FaDu cells. Results from reporter assays showed that lovastatin reduced NF-κ B-luciferase activities (Fig. 7b). Similarly, lovastatin also suppressed STAT3 phosphorylation (Fig. 7c). Transfection of cells with STAT3 siRNA significantly reduced the basal level of survivin in FaDu cells (Fig. 7d). We next determined whether lovastatin alters the recruitment of p65 or STAT3 to the survivin promoter region (− 300 to − 41). As shown in Fig. 7e, 6 h exposure to lovastatin reduced p65 and STAT3 binding to the survivin promoter region. It is likely that STAT3 and NF-κ B may also account for survivin repression by lovastatin in FaDu cells.
Lovastatin suppressed tumor growth in vivo. We next explored the in vivo effects of lovastatin using a xenograft murine model. After the average tumor size of tumors reached approximately 100 mm 3 , mice were daily administrated by intraperitoneal injections (I.P.) with vehicle or lovastatin (20 mg/kg/day) for 29 days. At the end of the experiment, mice were sarcificed to collect tumor samples. As show in Fig. 7f, lovastatin reduced tumor growth comparing to the vehicle-reated control group. Mice treated with lovastatin had a smaller tumor weight (Fig. 7g). In addition, mouse body weight was not altered after lovastatin treatment (data not shown). Together, these findings suggested that lovastatin is capable of suppressing Fadu xenografts growth in vivo.

Discussion
Growing evidence supports the therapeutic benefits of statins as anti-cancer agents in addition to its anti-inflammatory and anti-proliferative activities in non-cancerous tissues 1,49 . However, the precise anit-tumor mechanisms of statins remains incompletely understood. Keto et al. 50 demonstrated that statins' anti-tumor actions in certain tumors involved HMG-CoA reductase-mevalonate pathway. Geranyl pyrophosphate (GGPP) and farnesyl pyrophosphate (FPP) play critical roles in this pathway and are essential for activating Ras/Rho small G protein family and carcinogenesis 51 . It appears that suppression of Ras/Rho signaling by mevalonate pathway blockade contributes to statins' anti-tumor effects. However, statins' anti-tumor actions may also attribute to its non-lipid effects 20,52 . It is likely that statins' anti-tumor effects accrue from a variety of different mechanisms. Although there have been many studies reported that statins exhibit anti-tumor properties in numerous different human cancer cell lines, few studies have been undertaken to explore the underlying mechanisms by which statins induce HNSCC cell death. We show in the present study that lovastatin causes FaDu human pharyngeal squamous carcinoma cell apoptosis via AMPK-p38MAPK-p63-survivin signaling cascade. HDACs inhibition may also be involved in lovastatin's actions in FaDu cells.
Activated AMPK regulates cell survial and growth by activating the downstream signaling events, including p38MAPK activation 27 and Akt-mammalian target of rapamycin (mTOR) pathway down-regulation 53 . Whether p38MAPK contributes to lovastatin-induced FaDu cell death has not been previously reported. We show in this study that p38MAPK activation was causally related to lovastatin's actions. We also demonstrated that AMPK mediated lovastatin's effects on p38MAPK activation and survivin modulation in FaDu cells. However, whether lovastatin-activated AMPK leads to autophagy in FaDu cells, as suggested in another study 54 , remain to be established. The precise mechanism by which lovastatin induces AMPK phosphorylation in FaDu cells remain unresolved. Increased intracellular AMP/ATP ratio might play a causal role in AMPK activation and cell death 16 . Activation of tumor suppressor LKB1, a putative AMPK kinase, also contributed to lovastatin-induced AMPK activation in SCC cells 16 . Moreover, Kou et al. 55 demonstrated that simvastatin-induced LKB1 phosphorylation is a consequence of mevalonate-Rac1 cascade activation. However, knockdown of Rac1 did not affect simvastatin-induced AMPK phosphorylation in endothelial cells 55 . Together these observations suggest that increased AMP/ATP ratio may contribute to lovastatin-activated AMPK-p38MAPK apoptotic signaling cascade in FaDu cells. Whether lovastatin affects AMP/ATP ratio in FaDu cells exposed remains to be investigated. It is also worthy to clarify whether mevalonate pathway or LKB1-related mechanisms contributes to lovastatin-induced AMPK activation and subsequent cellular events in FaDu cells.
Lovastatin was previously shown to induce SCC cell death 16 . Kaneco et al. 56 further reported that survivin down-regulation contributes to lovastatin-induced colorectal cancer cell cell death. Similarly, we showed that survivin repression by lovastatin led to FaDu cell apoptosis. We further demonstrated the effectiveness of lovastatin in suppressing tumor progression in an in vivo xenograft murine model. The underlying mechanisms by which lovastatin induces survivin down-regulation and apoptosis in FaDu cells remains incompletely understood. We noted that knock-down p63 using p63 siRNA restored lovastatin's effects of decreasing survivin. It appears that p63 is causally related to lovastatin-induced survivin repression. Activation of p63 is regulated by its modifications such as phosphorylation, acetylation and ubiquitination 42,44 . In this study, we showed that AMPK-p38MAPK signaling blockade reduced lovastatin-induced p63 phosphoryaltion. These results support the contention that lovastatin activates the AMPK-p38MAPK-p63 pathway, leading to survivin down-regulation and subsequent cell death in FaDu cells.
Elevated levels of HDAC family members in tumor cells are correlated with poor prognosis in cancer patients 57,58 . Lin et al. 20 reported that statins may induce cancer cell death via HDACs inhibition. In agreement with this, we showed in this study that expression of HDAC3 or HDAC4 significantly reduced lovastatin's actions on p21 cip/Waf and survivin levels. Moreover, HDAC3 and HDAC4 also attenuated lovastatin-increased p63 acetylation. It is likely that p63 modified by phosphorylation and acetylation may contribute to lovastatin-induced p63 activation in FaDu cells. Whether lovastatin-induced p63 acetylation involves other HDAC isoforms remains to be investigated.
Similar to our previous report that p38MAPK-p53-survivin signaling mediated simvastatin-induced HCT116 colorectal cancer cell death 6 , we demonstrated that AMPK-p38MAPK cascade also plays a pivotal role in lovastatin-induced FaDu hypopharyngeal carcinoma cell death. In contrast to p53, we showed that p63, contributes to survivin repression and cell death in p53 mutant FaDu cells. These findings suggest that p38MAPK and p53 family members may play pivotal roles in statins-induced cancer cell death. Moreover, statins was reported to inhibit renal cancer cell proliferation and metastasis through inactivating STAT3 signaling 59 . It also suppresses NF-κ B-dependent anti-apoptotic gene expression to promote cell apoptosis 60 . Consistent with this, we noted in this study that lovastatin suppressed STAT3 and NF-κ B activation and reduced their binding to the survivin promoter region in FaDu cells. Whether lovastatin affects the interactions between p63 and these transcription factors to localize within the survivin promoter needs further investigations.
In conclusion, we show that lovastatin exhibits anti-tumor properties, at least in part, via AMPK-p38MAPK-p63-survivin signaling cascade in FaDu cancer cells. Moreover, lovastatin also suppressed the phosphorylations of ERK1/2 and Akt, the survival signaling molecules that causally related to NF-κ B and STAT3 activation, in FaDu cells (unpublished data). The exact mechanisms of these activities remain to be fully investigated, but together these observations support the therapeutic potential of lovastatin in future oncologic therapy in HNSCC patients.
Immunoblotting. Immunoblotting was conducted as described previously 6 . Cells were lysed using tris (10 mM, pH 7.0), triton X-100 (1%), pepstatin A (0.05 mM), leupeptin (0.2 mM), NaCl (140 mM), MgCl 2 (1 mM) and PMSF (2 mM) containing lysis buffer. Equal amount of each sample was subjected to SDS-PAGE. The protein was then transferred to the nitrocellulose membrane. After blocking for 1 h, target protein was detected by incubating in the specific primary antibody solution for 2 h and in the secondary antibody solution for another 1 h. Target proteins were visualized and quantified using ECL detection kit and densitometer with a scientific imaging system (Biospectrum AC System, UVP, Upland, CA, USA) .

MTT assay.
To determine cell viability, the colorimetric MTT assay was employed as described previously 6 .
Flow cytometric analysis. Flow cytometric analysis with propidium iodide (PI) single staining was performed using FACS Calibur and Cellquest program (BD Biosciences, San Jose, CA, USA) as described previously 6 . The percentage of cell cycle distribution was analyzed using ModFit programs (BD Biosciences, San Jose, CA, USA). The annexin V-FITC and PI double labeling was also employed to detect apoptotic cells. After treatment, cells were incubated for 15 min in the staining buffer (2 μ g/ml annexin V-FITC, 40 μ g/ml PI). The FACSCalibur and Cellquest program was employed to analyze the samples. The FCS Express program (BD Biosciences, San Jose, CA) was used to determine the percentage of stained cells in three quadrants: the lower left (annexin V − PI − ) quadrant, which reprents the viable cells; the lower right (annexin V + PI − ) quadrant, which represents the early apoptotic cells; the upper right (annexin V + PI + ) quadrant, which reprents advanced apoptotic and necrotic cells.

RT-PCR (reverse-transcription polymerase chain reaction) analysis. TRIzol reagent (Thermo
Fisher Scientific, Waltham, MA, USA) and GoScript ™ reverse transcription system (Promega , Madison, WI, USA) were used to isolate total RNA and perform reverse transcription. Primers used to generate 187 bp survivin fragment and 420 bp GAPDH fragment are: survivin sense, 5′-gcc ttt cct taa agg cca tc-3′; survivin anti-sense, 5′-aac cct tcc cag act cca ct-3′; GAPDH sense, 5′-gtc agt ggt gg acct gac ct-3′; GAPDH anti-sense, 5′-agg ggt cta cat ggc aac tg-3′. The PCR reaction with 25 cycles (30 s denature at 94 °C, 30 s annealing at 56 °C, 45 s extension at 72 °C) of amplification was performed. Amplication products were subjected to agarose gel electrophoresis and detected using ethdium bromide staining and ultraviolet illumination. Reporter assay. FaDu cells were transfected for 24 h with NF-κ B-luc and renilla-luc reporter constructs using Turbofect TM transfection reagent (Millipore, Billerica, MA, USA). After treatment with vehicle or lovastatin (30 M) for another 24 h, cells were harvested, and the luciferase activity was examined using a Dual-Glo luciferase assay system kit (Promega, Madison, WI, USA). The renilla luciferase activity represents as the basis for normalization.
Immunoprecipitation. Cells were lysed in 0.5 ml lysis buffer (1 mM PMSF, 1% Triton X-100, 10 μ g/ ml leupeptin, 10 μ g/ml aprotinin, 100 μ M sodium orthovanadate, 20 mM Tris-HCl, pH 7.5, 1 mM MgCl 2 and 125 mM NaCl). After centrifugation for 30 min at 4 °C, the supernatant was removed and incubated with antibodies against IgG or p63 with gentle rotation at 4 °C overnight. To collect the immune complexes, 15 μ l protein A-Magnetic Beads (Millipore) was added at 4 °C for another 2 h. After washed with lysis buffer for three times, the immunoprecipitated complexes were subjected to immunoblotting for assessing acetylation status of p63.

ChIP (Chromatin immunoprecipitation) analysis.
The ChIP analysis was conducted as previously described 6 . The 262-bp and 228-bp survivin promoter fragments between − 302 and − 41 or − 264 and − 37 were amplified using the following primers: sense-1, 5′-GATTACAGGCGTGAGCCACT-3′ and antisense-1, 5′-ATCTGGCGGTTAATGGCGCG-3′; sense-2, 5′-TTCTTTGAAAGCAGTCGAGG-3′; antisense-2, 5′-TCAAATCTGGCGGTTAATGG-3′. The PCR reaction with 30 cycles (30 s denature at 94 °C, 30 s annealing at 56 °C, 45 s extension at 72 °C) of amplification was performed. The PCR products were subjected to agarose gel electrophoresis and detected using ethdium bromide staining and ultraviolet illumination. In vivo xenograft mouse model. 4-week old nude nu/nu mice (BioLasco, Taipei, Taiwan) were used to perform xenograft model. PBS in a volume of 300 μ l containing FaDu cells (5 × 10 6 cells) were injected subcutaneously into the flank of each mouse. Mice were treated with vehicle or lovastatin (20 mg/kg/day) after the tumor size reached approximately 100 mm 3 . Lovastatin was intraperitoneally administered once daily for 25 days. A digital caliper was used to measure the tumor size every day. The formula V (mm 3 ) = [ab 2 ]× 0.52 (a: the length of the tumor; b: the width of the tumor) was used to calculate tumor volume. At the end of the treatment, mice were sacrificed to remove xenografts. These procedures were approved (Permit Number: LAC-2014-0230) by the Taipei Medical University Laboratory Animal Care and Use Committee. The present study was performed in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health and in accordance with the approved guidelines. All surgery was conducted under sodium pentobarbital anesthesia to minimize suffering.

Statistical analysis.
Compiled results represent as the mean ± S.E.M. of at least three independent experiments. To determine the statistical significance of the difference between means, one-way analysis of variance (ANOVA) and the Newman-Keuls test were used, when appropriate. It is considered statistically significant when a p value of < 0.05.