Despite increasing knowledge on oral and esophageal squamous cell carcinoma (OSCC and ESCC), specific medicines against both have not yet been developed. Here, we aimed to find novel anticancer drugs through functional cell-based screening of an FDA-approved drug library against OSCC and ESCC. Pitavastatin, an HMGCR inhibitor, emerged as an anticancer drug that inhibits tumor growth by downregulating AKT and ERK signals in OSCC and ESCC cells. One of the mechanisms by which pitavastatin inhibits cell growth might be the suppression of MET signaling through immature MET due to dysfunction of the Golgi apparatus. Moreover, the sensitivity of tumor growth to pitavastatin might be correlated with GGPS1 expression levels. In vivo therapeutic models revealed that the combination of pitavastatin with capmatinib, a MET-specific inhibitor, dramatically reduced tumor growth. Our findings suggest that GGPS1 expression could be a biomarker in cancer therapy with pitavastatin, and the combination of pitavastatin with capmatinib might be a promising therapeutic strategy in OSCC and ESCC.

Implications:

This study provides new insight into the mechanism of pitavastatin as an anticancer drug and suggests that the combination of pitavastatin with capmatinib is a useful therapeutic strategy in OSCC and ESCC.

Oral and esophageal carcinomas tend to easily metastasize to lymph nodes, resulting in poor prognosis. Esophageal cancer is the sixth most common cause of cancer-related death in the world (1). More than 90% of oral and esophageal carcinomas are diagnosed as squamous cell carcinomas (OSCCs, ESCCs) in Asian countries, including Japan. Recent studies revealed that multiple genetic alterations occurring sequentially in a cell lineage underlie the carcinogenesis of OSCC and ESCC, and these genetic alterations are almost identical in both tumors (2, 3). Unfortunately, despite increasing knowledge on OSCC and ESCC, specific medicines against both have not yet been developed. Thus, new drugs are required to improve the prognosis and quality of life in patients with OSCC and ESCC, and the concept of drug repurposing (DR) is helpful for the development of new drugs. DR is a widely used strategy that seeks to identify new medical indications for drugs that are already approved for the treatment of an original disease(s) (4). DR can reduce the time and cost of drug development because the pharmacology, formulation, safety, and toxicity profiles are already established (5). Thus, in this study, we aimed to identify new anticancer drugs in OSCC and ESCC by functional cell–based screening using an FDA-approved drug library.

The mevalonate pathway produces a variety of isoprenoids, such as cholesterol, and vitamin D, through enzymatic steps, resulting in the maintenance of membrane trafficking, cell migration, and homeostasis (6). Dysregulation of enzymes and regulators in the mevalonate pathway contributes to malignancies (6). In particular, the relationship between the mevalonate pathway and mutant p53 in cancer is well known (7). Mutant p53 can interact with SREBPs, key regulators of the mevalonate pathway, and promote the aberrant transcription of genes associated with the mevalonate pathway, resulting in dysregulation of membrane trafficking, cell migration, and cell growth (8–10). Moreover, abnormal functions of enzymes relevant to the mevalonate pathway, such as HMGCR, FDPS, and GGPS1, have been reported by many studies in several cancers (6, 11). It has been suggested that the mevalonate pathway could be targeted in cancer therapy, and statins, HMGCR inhibitors, also have potential as anticancer drugs (6). Statins potently reduce cholesterol levels, so they are used in the clinical setting in patients with dyslipidemia (12). However, statins have not been approved as anticancer drugs in the clinical setting because the anticancer mechanisms of statins are not fully understood, and useful biomarkers do not exist for the stratification of statin sensitivity in patients with cancer. Recent studies have demonstrated that statins can reduce tumor growth in vivo in xenograft models of several cancers, and circumstantial evidence of the anticancer functions of statins has accumulated gradually (6). In this study, pitavastatin emerged as an anticancer drug candidate from a functional cell-based screening using an FDA-approved drug library. Previous studies have revealed that pitavastatin induces apoptosis through the activation of caspases in vitro and inhibits tumor growth in in vivo xenograft models in OSCC, glioblastoma, breast, liver, colon, ovarian, and pancreatic cancers (13–19). However, the detailed mechanisms by which pitavastatin inhibits cell growth remain unclear. Here, we show that pitavastatin inhibits tumor growth through the downregulation of AKT and ERK signals through immature MET by dysfunction of the Golgi apparatus, and the combination of pitavastatin with capmatinib, a MET-specific inhibitor, might be a useful strategy for cancer therapy in OSCC and ESCC. Moreover, our study revealed that the expression of GGPS1, an enzyme that synthesizes geranylgeranyl pyrophosphate (GGPP) from farnesyl pyrophosphate (FPP), might be correlated with the sensitivity to pitavastatin, suggesting that GGPS1 could be a useful biomarker for the stratification between responders and nonresponders for cancer therapy with pitavastatin.

Cell culture and RNA from normal tissues

A total of 43 ESCC cell lines were used, of which 31 belonged to the KYSE series established from surgically resected tumors, and 12 were TE series lines provided by the Cell Resource Center for Biomedical Research, Institute of Development, Aging and Cancer, Tohoku University (Miyagi, Japan; ref. 20). A total of 23 OSCC cell lines were established from surgically resected tumors at Tokyo Medical and Dental University or purchased from the Japanese Collection of Research Bioresources. The highly metastatic cell line HOC313-LM was previously established from HOC313 cells by our laboratory (21). ESCC and OSCC cells were cultured in RPMI or DMEM supplemented with 10% FBS in a humidified atmosphere with 5% CO2 at 37°C. All cell lines were authenticated by monitoring cell morphology. Cells were routinely checked for Mycoplasma contamination using TaKaRa PCR Mycoplasma Detection Set (TaKaRa) and were cultured for no more than 20 passages from the validated stocks. RNA from normal esophageal, tongue, and throat tissues was purchased from BioChain. This study was performed following ethical rule in Tokyo Medical and Dental University (ethical approval number: G2019-013C).

Cell growth screening assay with FDA-approved drug library

The FDA-approved drug library (SCREEN-WELL FDA-approved drug library V2) was purchased from Enzo Life Sciences. HOC313-LM cells were seeded into 96-well plates (3.5 × 104 cells/well). The next day, each drug (766 drugs, 1 μmol/L) was added to each well. After 24 hours, cells were stained with crystal violet solution. After that, stained cells were dissolved in 2% SDS, and their density was measured by an absorption spectrometer (560 nm).

Cell growth assay

Cells were stained with crystal violet (CV) solution, then dissolved in 2% SDS, and their density was measured by an absorption spectrometer (560 nm).

Drug treatment and cell growth assay

Pitavastatin calcium (163-24861), rosuvastatin (187-03361) were purchased from Fujifilm Wako Chemicals. Capmatinib (INCB28060) was purchased from Selleck Chemicals. Digeranyl bisphosphonate (DGBP) was purchased from MedChemExpress. HGF (#100-39-10UG) was purchased from PeproTech. Simvastatin (S6196), fluvastatin (SML0038), crizotinib (PZ0191), cisplatin (479306), mevalonate, IPP, GPP, FPP, cholesterol, and GGPP were purchased from Sigma-Aldrich. MGCD-265 (20097) was purchased from Cayman Chemical. Z-VAD-FMK (3188-v) was purchased from Peptide Institute, Inc. (Osaka, Japan). The number of viable cells at 24–72 hours after drug treatment was assessed by crystal violet staining. For HGF treatment, cells were treated with pitavastatin for 72 hours, after which the medium was replaced with serum-free medium. After 1 hour, HGF was added to the serum-free medium, and treated cells were harvested at specified time points (0, 15, 30, and 60 minutes). For Z-VAD-FMK treatment, cells were treated with pitavastatin and/or Z-VAD-FMK for 72 hours. After that, cells were harvested and stained by the MEBCYTO Apoptosis Kit (MBL) for flow cytometry analysis.

Establishment of pitavastatin-resistant cells in HOC313-LM cells

HOC313-LM cells were seeded into 6-well plates (15 × 104 cells/well). Next day, cells were treated with pitavastatin (1 μmol/L). After that, medium containing pitavastatin (1 μmol/L) was changed every three days for two weeks. After two weeks, the concentration of pitavastatin was increased from 1 μmol/L to 2 μmol/L and finally we confirmed the resistance to pitavastatin by cell growth assay.

Flow cytometry analysis for measurement of apoptotic cells

Apoptotic cells were stained with the MEBCYTO Apoptosis Kit, and cell population analysis was performed using an Accuri Flow Cytometer (BD Biosciences). The procedures were carried out according to the manufacturer's instructions.

Western blotting, immunohistochemistry and immunofluorescence microscopy analyses

The protein concentration of the cell lysate was determined using a Protein Assay Kit (Bio-Rad), and samples were separated on SDS polyacrylamide gels for Western blotting analysis. The following primary antibodies were used for Western blotting, IHC, and immunofluorescence microscopy analyses: anti-phospho-Met (#3077), anti-Met (#8198), anti-cleaved PARP (#9541), anti-phospho-ERK (#4370), anti-ERK (#4695), anti-phospho-AKT (#9271), anti-AKT (#9272), anti-phospho-H2AX (#9718), and anti-CDK6 (#3136) antibodies were purchased from Cell Signaling Technology. Anti-β-actin (A5441), anti-GGPS1 (sc-271679), anti-phospho-Rb (S780) (sc-12901), anti-Rb (sc-50), anti-GM130 (ab169276), anti-GBF1 (612116), and anti-Arf1 (ARF01) antibodies were purchased from Sigma-Aldrich, Santa Cruz Biotechnology, Abcam, BD Biosciences, or Cytoskeleton, Inc. Western blotting, IHC, and immunofluorescence microscopy analyses were performed as described elsewhere (20).

Phospho-RTK array analysis

The Proteome Profiler Human Phospho-RTK Array Kit was purchased from R&D Systems and used according to the manufacturer's instructions. Proteins were extracted from HOC313-LM or HOC313-LM-Pita-R cells treated with pitavastatin (1 μmol/L) or control solvent for 72 hours.

In ovo and in vivo tumor growth assays

Eggs were purchased from Kanto Company. Ten days after incubation at 37°C, we transplanted 3 × 106 HOC313-LM or 2 × 106 KYSE200 cells onto the chorioallantoic membrane (CAM) of the chicken embryo. Two days after transplantation, we confirmed tumor formation (approximately 2 mm × 2 mm) and treated the tumors daily with drugs. On the 16th day, we euthanized the chickens and observed the tumor volume.

Six- to eight-week-old female CB17/Icr-Prkdcscid/CrlCrlj (SCID) mice were purchased from Charles River Laboratories. HOC313-LM cells (1 × 107 cells) in 100 μL of Matrigel (BD Biosciences) were injected into the left abdominal wall of SCID mice. A week later, we confirmed tumor formation (approximately 5 mm × 5 mm) and randomly grouped transplanted mice in order to avoid tumor size variability. After that, we began drug treatment via daily intraperitoneal injection (i.p.) and measured tumor sizes every 2 days. Tumor size was calculated at the indicated times after injection using the formula [tumor size (mm3) = [(length) × (width)2]/2]. Mice were sacrificed on day 24, and tumors were assessed for volume and immediately processed for section preparation (fixed with formalin and embedded in paraffin). These animal studies were performed following animal rule in Tokyo Medical and Dental University (approval number: A2020-059C, A2020-105A).

qRT-PCR

qRT-PCR was carried out using a ViiA 7 Real-Time PCR System (Thermo Fisher Scientific) with the KAPA SYBR Fast qPCR Kit (Kapa Biosystems) according to the manufacturer's instructions. Gene expression values were evaluated as ratios based on the differences in cycle threshold values between the genes of interest and an internal reference gene (GAPDH), which served as a normalization factor for the amount of RNA isolated from the cancer cells. Primer information:

  • GGPS1-forward primer ACTCAAGAAACAGTCCAAAGA,

  • GGPS1-reverse primer TCTGTAGCTTGTCCTCTGGAA,

  • MET- forward primer GACTTCTTCAACAAGATCGTCA,

  • MET-reverse primer CAAAGCTGTGGTAAACTCTGTT,

  • HMGCR- forward primer TACTGGTAACAATAAGATCTGT,

  • HMGCR-reverse primer CGTAAATTCTGGAACTGGA,

  • GAPDH-forward primer CGACCACTTTGTCAAGCTCA,

  • GAPDH-reverse primer AGGGGTCTACATGGCAACTG.

Expression construct

Full-length human GGPS1 cDNA was obtained by RT-PCR. The PCR product was inserted into the pCDH-CMV-MCS-EF1-RFP+Puro vector (System Biosciences). Lentivirus was prepared using HEK293TN cells and the pPACK Packaging Kit (System Biosciences) according to the manufacturer's instructions. The virus was precipitated by PEG-it (System Biosciences) according to the manufacturer's instructions. Cells were infected with lentivirus containing either an empty vector (as a control) or GGPS1 expression vector using TransDux (System Biosciences). Primer information:

  • GGPS1-forward primer TTGCTAGCATGGAGAAGACTCAAGAAACA,

  • GGPS1-reverse primer TTGGATCCTTATTCATTTTCTTCTTTGAACAT.

Gene expression array analysis and pathway analysis

Gene expression profiles were analyzed using SurePrint G3 Gene Expression Microarrays Ver. 3 (Agilent Technologies) according to the manufacturer's instructions. The raw data were analyzed with GeneSpring GX14.9 software (Agilent Technologies). Pathway analysis using WikiPathways was carried out with normalized expression array data by GeneSpring GX14.9. KEGG pathway analysis was carried out by DAVID tools (https://david.ncifcrf.gov/summary.jsp). GSEA (version 4.0.3) was carried out with normalized expression array data, and we then selected “c6: oncogenic signatures gene sets”.

siRNA transfection

siRNA-MET (s8700, s8702), -GGPS1 (s18107, s18108) and negative control siRNA (si-NC) were purchased from Thermo Fisher Scientific. Each siRNA (20 nmol/L) was transfected into cells by Lipofectamine RNAiMAX (Thermo Fisher Scientific) according to the manufacturer's instructions.

Arf1-GTP pull down assay

The Arf1 Activation Assay Biochem Kit (BK-032S) was purchased from Cytoskeleton, Inc. and used according to the manufacturer's instructions. Proteins were extracted from HOC313-LM cells treated with pitavastatin (1 μmol/L) or control solvent for 48 hours with and without mevalonate or GGPP. Quantitative analysis was conducted using ImageJ software.

Analysis of the cancer therapeutics response portal

Correlations between GGPS1 expression and resistance to simvastatin and lovastatin in nonhematopoietic cancer cells were downloaded from the Cancer Therapeutics Response Portal website (https://portals.broadinstitute.org/ctrp.v2.1/).

Kaplan–Meier analysis using TCGA-head and neck cancer data (PanCancer Atlas)

We downloaded the clinical data of head and neck cancer and the expression of the mevalonate pathway-related genes and MET from TCGA (https://www.cbioportal.org/). We modified the clinical data (removing patients: within a one month after surgical therapy). After that, we performed Kaplan–Meier analysis for evaluating five years survival rate using each gene expression data.

Statistical analysis and calculation of the combination index

Statistical analysis was carried out by GraphPad Prism 8. Differences between two groups were tested with Student t test, and two-way ANOVA was applied to analyze tumor growth. P < 0.05 was considered statistically significant. The combination index (CI) was calculated by CalcuSyn software (BIOSOFT).

Data deposition

The microarray data from this publication have been submitted to the GEO database (http://www.ncbi.nlm.nih.gov/geo/) and assigned the identifier “GSE145624”.

Pitavastatin emerged as an anticancer drug candidate from an FDA-approved drug library

To find novel anticancer drugs for OSCC and ESCC treatment, we performed a functional cell-based screening with HOC313-LM, a highly metastatic cell line derived from HOC313, which is a parent OSCC cell line (21), using an FDA-approved drug library containing 766 drugs (Enzo Life Sciences). We added each drug to the cells and measured the cell survival rate (Fig. 1A; Supplementary Table S1). Significant cell growth suppression was observed with 15 drugs. Among them, pitavastatin, a potent competitive inhibitor of 3-hydroxy-3-methyl-glutaryl-CoA (HMG-CoA) reductase (HMGCR), showed the most significant reduction in the cell growth ratio. The other 14 were well-known anticancer drugs and an inhibitor of microtubule polymerization, colchicine (Supplementary Table S1). We focused on the mechanisms by which pitavastatin inhibits the growth of cancer cells. First, we validated the cell growth inhibition by pitavastatin in HOC313-LM cells and observed that pitavastatin could inhibit cell growth in a dose-dependent manner (Fig. 1B). We also checked the status of AKT and ERK because the activation of these protein directly relates to cell growth and survival (Fig. 1C). As a result, the expression of AKT and ERK was downregulated by pitavastatin treatment. We next evaluated the effect of pitavastatin on cell growth in 43 ESCC and 23 OSCC cell lines and then separated them into low and high sensitivity groups (Supplementary Fig. S1A). In the high sensitivity group (HOC313-LM, KYSE150, and 200), AKT and ERK proteins were downregulated by pitavastatin, whereas pitavastatin did not affect the expression of those proteins in the low sensitivity group (KYSE70, 510, 1190, and 1440; Fig. 1C; Supplementary Fig. S1B). We also evaluated whether pitavastatin-induced cell growth inhibition was due to apoptosis using flow cytometry (Fig. 1D; Supplementary Fig. S1C). ZVAD, an apoptosis inhibitor, could significantly inhibit apoptosis under pitavastatin treatment in HOC313-LM, KYSE150, and KYSE200 cells, indicating that pitavastatin could induce apoptosis. We then performed in ovo therapeutic experiments [chick chorioallantoic membrane (CAM) assay] with two high-sensitivity cell lines, HOC313-LM and KYSE200, because in ovo experiments can evaluate the antitumor effects of drugs faster than mouse experiments. Pitavastatin significantly inhibited tumor growth in ovo (Fig. 1E; Supplementary Fig. S1D). Moreover, mouse therapeutic experiments with HOC313-LM cells also revealed that pitavastatin inhibited tumor growth without any adverse events, suggesting that pitavastatin might be a useful drug for cancer therapy (Fig. 1F and G). In addition, using KEGG pathway analysis, WikiPathways analysis and GSEA, we analyzed gene expression data obtained from gene expression array analyses in three cell lines, HOC313-LM, KYSE150, and KYSE200, after treatment with pitavastatin. These analyses revealed that pitavastatin could downregulate cell-cycle–related pathways such as Rb signals (Supplementary Fig. S2A and S2B). In addition, to confirm the results of pathway analyses, we evaluated the expression of phospho-Rb, total Rb, CDK6, and phospho-H2AX (Supplementary Fig. S2C). The expression of phospho-Rb, total Rb, and CDK6 was decreased by pitavastatin treatment, whereas the expression of phospho-H2AX was induced by pitavastatin treatment. These phenomena were consistent with some previous reports on statins (18, 22). Thus, we searched for other mechanisms by which pitavastatin inhibited cell growth.

Figure 1.

Pitavastatin inhibits tumor growth through the downregulation of ERK and AKT signals in vitro and in vivo. A, The result of cell-based screening with HOC313-LM cells using an FDA-approved drug library (766 drugs; top). Top 15 list of identified hits in the screen of the FDA-approved drug library (bottom). B, The number of viable cells 72 hours after treatment with pitavastatin (0.1, 0.25, 0.5, 1, or 2 μmol/L) or control solvent (equivalent to 2 μmol/L) in HOC313-LM cells was assessed by CV staining assay. Each data point represents the mean of three experiments (bars, SD). C, Western blotting of the indicated protein in HOC313-LM cells treated with pitavastatin (0.1, 0.5, or 1 μmol/L) for 72 hours. cPARP, cleaved PARP. D, The apoptosis analysis by flow cytometry analysis in HOC313-LM cells treated with pitavastatin (1 μmol/L) and/or Z-VAD-FMK (50 μmol/L) for 72 hours. E, Schematic of in ovo experiment with HOC313-LM cells (left). Images of isolated tumors [control (PBS with acetic acid): n = 24, pitavastatin: n = 24; middle]. Scale bar, 1 cm. Tumor weight was measured in each in ovo tumor (right). F, Tumor growth curves of subcutaneous HOC313-LM tumors treated with pitavastatin (1 or 2 mg/kg) or control solvent (left). Tumor size (mm3) = [(Length) × (Width)2]/2. G, Images of isolated tumors [control (PBS with acetic acid): n = 4, pitavastatin (1 mg/kg): n = 5, (2 mg/kg): n = 4; left). Scale bar, 1 cm. Tumor weight was measured in each mouse (right). Student t test was used for statistical analysis.

Figure 1.

Pitavastatin inhibits tumor growth through the downregulation of ERK and AKT signals in vitro and in vivo. A, The result of cell-based screening with HOC313-LM cells using an FDA-approved drug library (766 drugs; top). Top 15 list of identified hits in the screen of the FDA-approved drug library (bottom). B, The number of viable cells 72 hours after treatment with pitavastatin (0.1, 0.25, 0.5, 1, or 2 μmol/L) or control solvent (equivalent to 2 μmol/L) in HOC313-LM cells was assessed by CV staining assay. Each data point represents the mean of three experiments (bars, SD). C, Western blotting of the indicated protein in HOC313-LM cells treated with pitavastatin (0.1, 0.5, or 1 μmol/L) for 72 hours. cPARP, cleaved PARP. D, The apoptosis analysis by flow cytometry analysis in HOC313-LM cells treated with pitavastatin (1 μmol/L) and/or Z-VAD-FMK (50 μmol/L) for 72 hours. E, Schematic of in ovo experiment with HOC313-LM cells (left). Images of isolated tumors [control (PBS with acetic acid): n = 24, pitavastatin: n = 24; middle]. Scale bar, 1 cm. Tumor weight was measured in each in ovo tumor (right). F, Tumor growth curves of subcutaneous HOC313-LM tumors treated with pitavastatin (1 or 2 mg/kg) or control solvent (left). Tumor size (mm3) = [(Length) × (Width)2]/2. G, Images of isolated tumors [control (PBS with acetic acid): n = 4, pitavastatin (1 mg/kg): n = 5, (2 mg/kg): n = 4; left). Scale bar, 1 cm. Tumor weight was measured in each mouse (right). Student t test was used for statistical analysis.

Close modal

GGPS1 expression might attenuate cell growth inhibition by pitavastatin in OSCC and ESCC

To reveal how pitavastatin induced cell growth inhibition, rescue experiments against pitavastatin were performed using each of the metabolites in the mevalonate pathway (Fig. 2AD; Supplementary Fig. S3A). The addition of mevalonate or GGPP alone reduced the effect of pitavastatin on cell growth, but IPP, GPP, FPP, and cholesterol did not. This result was consistent with some previous reports on statins and their cell growth suppressions (14, 23–26), suggesting that GGPP might have an important role in cell growth and that GGPS1, a catabolic enzyme synthesizing GGPP from FPP, might be a target for cancer therapy.

Figure 2.

Tumor growth inhibition by pitavastatin is canceled by treatment with mevalonate or GGPP. A, Schematic of the mevalonate pathway. B–D, Rescue experiments using metabolites related to the mevalonate pathway. In these experiments, HOC313-LM cells were treated with pitavastatin (1 μmol/L) or a combination [pitavastatin (1 μmol/L) and capmatinib (0.5 μmol/L)] with and without each metabolite [mevalonate (500 μmol/L), IPP (5 μmol/L), GPP (5 μmol/L), FPP (5 μmol/L), cholesterol (50 μmol/L), or GGPP (5 μmol/L)] for 72 hours. Cell growth results from the CV staining assay (B) and the apoptosis analysis by flow cytometry analysis (C) in HOC313-LM cells. Each data point represents the mean of two or three experiments (bars, SD). P value was calculated by Student t test and indicates NC versus each sample. D, Western blotting of the indicated protein in treated HOC313-LM cells. E, The number of viable cells 72 hours after treatment with pitavastatin (0.1, 0.25, 0.5, 1, or 2 μmol/L) or control solvent (equivalent to 2 μmol/L) in HOC313-LM and HOC313-LM-Pita-R cells was assessed by CV staining assay. Each data point represents the mean of three experiments (bars, SD). F, Western blotting of the indicated protein in HOC313-LM-Pita-R cells treated with pitavastatin (0.1, 0.5, and 1 μmol/L) for 72 hours. G, GGPS1 protein expression level in HOC313-LN and -Pita-R cells. H, The number of viable cells 72 hours after treatment with pitavastatin (0.1, 0.25, 0.5, 1, or 2 μmol/L) or control solvent (equivalent to 2 μmol/L) in GGPS1-overexpressing cell lines and control vector–transfected cell line (RFP) in HOC313-LM was assessed by CV staining assay. Each data point represents the mean of three experiments (bars, SD). I, Western blotting of the indicated protein in GGPS1-overexpressing cell lines and a control vector–transfected cell line (RFP) in HOC313-LM cells treated with pitavastatin (1 μmol/L) for 72 hours. The amount of cPARP was quantified by ImageJ software. J, Schematic of the in ovo experiment with HOC313-LM cells. Images of isolated tumors [control (PBS with acetic acid): n = 7, pitavastatin: n = 6, pitavastatin plus GGPP (5 μmol/L)]. Scale bar, 1 cm. Tumor weight was measured in each in ovo tumor (right). Student t test was used for statistical analysis.

Figure 2.

Tumor growth inhibition by pitavastatin is canceled by treatment with mevalonate or GGPP. A, Schematic of the mevalonate pathway. B–D, Rescue experiments using metabolites related to the mevalonate pathway. In these experiments, HOC313-LM cells were treated with pitavastatin (1 μmol/L) or a combination [pitavastatin (1 μmol/L) and capmatinib (0.5 μmol/L)] with and without each metabolite [mevalonate (500 μmol/L), IPP (5 μmol/L), GPP (5 μmol/L), FPP (5 μmol/L), cholesterol (50 μmol/L), or GGPP (5 μmol/L)] for 72 hours. Cell growth results from the CV staining assay (B) and the apoptosis analysis by flow cytometry analysis (C) in HOC313-LM cells. Each data point represents the mean of two or three experiments (bars, SD). P value was calculated by Student t test and indicates NC versus each sample. D, Western blotting of the indicated protein in treated HOC313-LM cells. E, The number of viable cells 72 hours after treatment with pitavastatin (0.1, 0.25, 0.5, 1, or 2 μmol/L) or control solvent (equivalent to 2 μmol/L) in HOC313-LM and HOC313-LM-Pita-R cells was assessed by CV staining assay. Each data point represents the mean of three experiments (bars, SD). F, Western blotting of the indicated protein in HOC313-LM-Pita-R cells treated with pitavastatin (0.1, 0.5, and 1 μmol/L) for 72 hours. G, GGPS1 protein expression level in HOC313-LN and -Pita-R cells. H, The number of viable cells 72 hours after treatment with pitavastatin (0.1, 0.25, 0.5, 1, or 2 μmol/L) or control solvent (equivalent to 2 μmol/L) in GGPS1-overexpressing cell lines and control vector–transfected cell line (RFP) in HOC313-LM was assessed by CV staining assay. Each data point represents the mean of three experiments (bars, SD). I, Western blotting of the indicated protein in GGPS1-overexpressing cell lines and a control vector–transfected cell line (RFP) in HOC313-LM cells treated with pitavastatin (1 μmol/L) for 72 hours. The amount of cPARP was quantified by ImageJ software. J, Schematic of the in ovo experiment with HOC313-LM cells. Images of isolated tumors [control (PBS with acetic acid): n = 7, pitavastatin: n = 6, pitavastatin plus GGPP (5 μmol/L)]. Scale bar, 1 cm. Tumor weight was measured in each in ovo tumor (right). Student t test was used for statistical analysis.

Close modal

We next established a pitavastatin-resistant cell line (HOC313-LM-Pita-R) to reveal the detailed mechanisms underlying cell growth inhibition by pitavastatin. The cell growth inhibition rate and the downregulation of AKT and ERK by pitavastatin in HOC313-LM-Pita-R cells were lower than in the parent cells (Fig. 2E and F). We examined the expression of GGPS1 in HOC313-LM and HOC313-Pita-R cells (Fig. 2G). The expression of GGPS1 was higher in HOC313-LM-Pita-R cells than in the parent cells, suggesting that GGPS1 contributed to resistance to pitavastatin in cell growth. Thus, we established GGPS1-overexpressing cells from HOC313-LM cells by infection with a lentivirus vector containing full-length sequences of GGPS1 and evaluated the sensitivity to pitavastatin in these cells (Fig. 2H and I). The GGPS1-overexpressing cells showed resistance to pitavastatin. In addition, in ovo experiments demonstrated that GGPP could rescue tumor growth under pitavastatin treatment (Fig. 2J). To determine whether GGPS1 knockdown inhibits cell growth, we performed a cell growth assay and Western blotting analysis in HOC313-LM and KYSE150 cells treated with GGPS1-specific siRNAs (Fig. 3A and B; Supplementary Fig. S3B and S3C). GGPS1 knockdown did inhibit cell growth, and this phenomenon was rescued by the addition of GGPP to these cell lines. We next performed gene expression array analysis and found that the gene transcription pattern in GGPS1 knockdown cells was very similar to that in pitavastatin-treated cells (Fig. 3C). Analyses using the KEGG pathway analysis, WikiPathways analysis, and GSEA revealed that GGPS1 knockdown also suppressed cell-cycle–related pathways, including Rb signals (Fig. 3C; Supplementary Fig. S3D). Moreover, GGPS1 expression in OSCC and ESCC cell lines tended to be correlated with pitavastatin sensitivity (Fig. 3D; Supplementary Fig. S4A), whereas HMGCR expression was not correlated with pitavastatin sensitivity (Supplementary Fig. S4A and S4B). Interestingly, The Cancer Therapeutics Response Portal showed that GGPS1 expression contributed to resistance to other statins, such as simvastatin and lovastatin (Supplementary Fig. S4C). Indeed, the sensitivity to pitavastatin was enhanced by GGPS1 knockdown in KYSE220, TE14, which have high expression of GGPS1, and HOC313-LM-Pita-R cells (Fig. 3E and F). In addition, HOC313-LM-Pita-R cells also showed resistance to simvastatin (Supplementary Fig. S4D). These results indicated that GGPS1 expression might be important for the prediction of sensitivity to statins, including pitavastatin. We also evaluated the effect of the GGPS1 inhibitor digeranyl bisphosphonate (DGBP), on cell growth. Consistent with the GGPS1 knockdown phenotype, DGBP inhibited cell growth in HOC313-LM, KYSE150, and KYSE200 cells (Supplementary Fig. S5A and S5B). Moreover, we performed in ovo experiments, and DGBP treatment inhibited tumor growth as well as pitavastatin treatment (Supplementary Fig. S5C).

Figure 3.

GGPS1 expression is correlated with sensitivity to pitavastatin. A, The number of viable cells 72 or 120 hours after transfection of si-GGPS1 #1, #2 or si-NC with and without GGPP (5 μmol/L) in HOC313-LM cells was assessed by CV staining assay with the mean ± SD (bars) of triplicate experiments. B, Western blotting of the indicated protein in HOC313-LM cells transfected with each siRNA with and without GGPP for 120 hours. C, Gene expression array analysis of HOC313-LM after transfection of si-GGPS1 #2 or si-NC or treatment with pitavastatin (1 μmol/L) or control solvent. Results shown as heatmap (left). The right tables show the results of KEGG pathway analysis and WikiPathways analysis using 664 downregulated probes. D, The relationship between GGPS1 expression level and sensitivity to pitavastatin within ESCC and OSCC cell lines. E, The number of viable cells 72 hours after transfection of si-GGPS1 #1, #2, or si-NC with and without pitavastatin (0.25, 0.5, 1, 2, 4, or 8 μmol/L) in KYSE220, TE14, and HOC313-LM-Pita-R cells was assessed by CV staining assay with mean ± SD (bars) for triplicate. F, Western blotting of the indicated protein in KYSE220, TE14, and HOC313-LM-Pita-R cells transfected with each siRNA with and without pitavastatin (2 or 8 μmol/L). Student t test was used for statistical analysis.

Figure 3.

GGPS1 expression is correlated with sensitivity to pitavastatin. A, The number of viable cells 72 or 120 hours after transfection of si-GGPS1 #1, #2 or si-NC with and without GGPP (5 μmol/L) in HOC313-LM cells was assessed by CV staining assay with the mean ± SD (bars) of triplicate experiments. B, Western blotting of the indicated protein in HOC313-LM cells transfected with each siRNA with and without GGPP for 120 hours. C, Gene expression array analysis of HOC313-LM after transfection of si-GGPS1 #2 or si-NC or treatment with pitavastatin (1 μmol/L) or control solvent. Results shown as heatmap (left). The right tables show the results of KEGG pathway analysis and WikiPathways analysis using 664 downregulated probes. D, The relationship between GGPS1 expression level and sensitivity to pitavastatin within ESCC and OSCC cell lines. E, The number of viable cells 72 hours after transfection of si-GGPS1 #1, #2, or si-NC with and without pitavastatin (0.25, 0.5, 1, 2, 4, or 8 μmol/L) in KYSE220, TE14, and HOC313-LM-Pita-R cells was assessed by CV staining assay with mean ± SD (bars) for triplicate. F, Western blotting of the indicated protein in KYSE220, TE14, and HOC313-LM-Pita-R cells transfected with each siRNA with and without pitavastatin (2 or 8 μmol/L). Student t test was used for statistical analysis.

Close modal

To determine whether the mevalonate pathway–related genes including GGPS1 predict the prognosis of patients with head and neck cancer, we performed Kaplan–Meier analysis using each gene expression data from TCGA. However, the expression of all of the mevalonate pathway–related genes would not be a biomarker for prognosis of patients with head and neck cancer (Supplementary Fig. S6).

Pitavastatin might regulate the MET signaling pathway

To reveal the detailed mechanisms by which pitavastatin inhibits tumor growth, we evaluated the phosphorylation status of receptor tyrosine kinases (RTKs) using a phospho-receptor tyrosine kinase array (Fig. 4A). MET phosphorylation was decreased by treatment with pitavastatin in HOC313-LM cells, whereas this change was not observed in HOC313-LM-Pita-R cells. MET has been well known as an oncogene and contributes to cell growth, survival, and motility in cancer (27) and MET expression was associated with poor prognosis in patients with head and neck cancer (Supplementary Fig. S6). Given the decreased phosphorylation of MET, the inhibition of cell growth by pitavastatin might be related to the MET signaling pathway. Indeed, pitavastatin reduced the phosphorylation of MET in a dose-dependent manner (Fig. 4B), and this downregulation mechanism was not manifested at the transcriptional level (Fig. 4C). In addition, MET expression did not correlate with pitavastatin sensitivity in OSCC and ESCC cell lines (Supplementary Fig. S4A and S4B). Interestingly, our studies on phospho-MET and MET in HOC313-LM and HOC313-LM-Pita-R cells treated with pitavastatin suggested that pitavastatin might reduce cell growth by inhibiting MET maturation (Fig. 4B). The inhibition of MET maturation by pitavastatin could be rescued by the addition of mevalonate or GGPP (Fig. 4D). Moreover, GGPS1 knockdown also produced immature MET, and this phenomenon was rescued by the addition of GGPP (Fig. 4E; Supplementary Fig. S3C). In addition, GGPS1 knockdown not only increased the sensitivity to pitavastatin but also the amount of immature MET induced by treatment with pitavastatin in KYSE220 and TE14 cells (Supplementary Fig. S7A). Because immature (or uncleaved) MET is located on the Golgi apparatus (28), we examined the localization of MET after pitavastatin treatment or GGPS1 knockdown with and without mevalonate or GGPP by immunofluorescence staining with anti-MET, anti-GBF1, and anti-GM130 antibodies (Fig. 4F and G; Supplementary Fig. S7B and S7C). GBF1 and GM130 are located on the cis-Golgi apparatus and regulate vesicular trafficking (29, 30). GBF1 is also known as a guanine nucleotide exchange factor (GEF) of Arf (29). Immunofluorescent analyses revealed that MET was located on the cellular membrane and Golgi apparatus under normal conditions, whereas pitavastatin treatment or GGPS1 knockdown translocated MET from the cellular membrane to the cytoplasm and/or Golgi apparatus. In addition, GBF1 was observed in the cytoplasm after pitavastatin treatment or GGPS1 knockdown, suggesting that GBF1 might not function as a GEF. Thus, we checked the activation of Arf1 under pitavastatin treatment by the Arf1-GTP pull-down assay (Fig. 4H). The results showed that pitavastatin treatment decreased the amount of Arf1-GTP, and adding mevalonate or GGPP canceled the effect of pitavastatin. Furthermore, we found that immature MET after pitavastatin treatment did not respond to hepatocyte growth factor (HGF) stimulation (Fig. 4I; Supplementary Fig. S7D). Therefore, pitavastatin might inhibit the MET signaling pathway due to the prevention of MET maturation through dysfunction of the Golgi apparatus. To evaluate whether the mevalonate pathway was involved in producing mature MET, some cell lines were treated with pitavastatin, other statins (simvastatin, rosuvastatin or fluvastatin), DGBP, or cytotoxic agents (CDDP or crizotinib; Supplementary Fig. S8A and S8B). The inhibition of the mevalonate pathway by statins could hamper MET maturation, whereas cell death caused by cytotoxic agents did not affect MET maturation. Taken together, the evidence suggests that the mevalonate pathway might regulate MET maturation through the function of the Golgi apparatus.

Figure 4.

Pitavastatin or knockdown of GGPS1 suppresses MET signaling through the inhibition of MET maturation. A, The results of the human phospho-RKT array. B, Phospho-MET and MET levels in HOC313-LM and HOC313-LM -Pita-R cells after treatment with pitavastatin (0.1, 0.5, or 1 μmol/L) for 72 hours. C, Expression levels of MET mRNA in HOC313-LM and HOC313-LM-Pita-R cells determined by qRT-PCR, showing the mean ± SD (bars) of duplicate experiments. D, Rescue experiments using metabolites related to the mevalonate pathway. HOC313-LM cells were treated with pitavastatin (1 μmol/L) with and without each metabolite [mevalonate (500 μmol/L), IPP (5 μmol/L), GPP (5 μmol/L), FPP (5 μmol/L), cholesterol (50 μmol/L), or GGPP (5 μmol/L)] for 72 hours. Western blotting of the indicated protein in treated HOC313-LM cells. E, HOC313-LM was transfected with each siRNA (NC, GGPS1 #1 or #2) with and without GGPP for 72 hours. Western blotting of the indicated protein in treated HOC313-LM cells. HOC313-LM cells treated with pitavastatin (1 μmol/L; F) or transfected with si-GGPS1 #2 (G) with and without mevalonate or GGPP were stained for MET (green) and GBF1 (red). Scale bar, 100 μmol/L. H, Arf1-GTP pull-down assay was performed with HOC313-LM cells treated with pitavastatin with and without mevalonate or GGPP. Each data point represent the mean of two experiments (bars, SD). I, Western blotting of the indicated protein in HOC313-LM cells treated with HGF (100 ng/mL) after treatment with pitavastatin (0.5 or 1 μmol/L) or control solvent for 48 hours. Student t test was used for statistical analysis.

Figure 4.

Pitavastatin or knockdown of GGPS1 suppresses MET signaling through the inhibition of MET maturation. A, The results of the human phospho-RKT array. B, Phospho-MET and MET levels in HOC313-LM and HOC313-LM -Pita-R cells after treatment with pitavastatin (0.1, 0.5, or 1 μmol/L) for 72 hours. C, Expression levels of MET mRNA in HOC313-LM and HOC313-LM-Pita-R cells determined by qRT-PCR, showing the mean ± SD (bars) of duplicate experiments. D, Rescue experiments using metabolites related to the mevalonate pathway. HOC313-LM cells were treated with pitavastatin (1 μmol/L) with and without each metabolite [mevalonate (500 μmol/L), IPP (5 μmol/L), GPP (5 μmol/L), FPP (5 μmol/L), cholesterol (50 μmol/L), or GGPP (5 μmol/L)] for 72 hours. Western blotting of the indicated protein in treated HOC313-LM cells. E, HOC313-LM was transfected with each siRNA (NC, GGPS1 #1 or #2) with and without GGPP for 72 hours. Western blotting of the indicated protein in treated HOC313-LM cells. HOC313-LM cells treated with pitavastatin (1 μmol/L; F) or transfected with si-GGPS1 #2 (G) with and without mevalonate or GGPP were stained for MET (green) and GBF1 (red). Scale bar, 100 μmol/L. H, Arf1-GTP pull-down assay was performed with HOC313-LM cells treated with pitavastatin with and without mevalonate or GGPP. Each data point represent the mean of two experiments (bars, SD). I, Western blotting of the indicated protein in HOC313-LM cells treated with HGF (100 ng/mL) after treatment with pitavastatin (0.5 or 1 μmol/L) or control solvent for 48 hours. Student t test was used for statistical analysis.

Close modal

The combination of pitavastatin and capmatinib enhanced tumor growth inhibition

We next performed a cell growth assay using the combination of pitavastatin and capmatinib, a MET-specific inhibitor, because we hypothesized that the complete inhibition of MET signaling by concurrent treatment with both drugs might enhance cell growth inhibition. Treatment with capmatinib alone did not affect cell growth in HOC313-LM, HOC313-LM -Pita-R, KYSE150, and KYSE200 cells, whereas the combination of pitavastatin and capmatinib synergistically enhanced cell growth inhibition compared to treatment with pitavastatin alone (Fig. 5AC; Supplementary Fig. S9A–S9C). Interestingly, treatment with either pitavastatin or capmatinib alone did not completely inhibit MET signaling activated by HGF, but the combination of pitavastatin and capmatinib could induce complete suppression of MET signaling by inhibiting the maturation and phosphorylation of MET (Fig. 5D). Moreover, we examined the effects of combinations of other statins (simvastatin, rosuvastatin, or fluvastatin) and capmatinib on cell growth (Supplementary Fig. S10A). Although these combinations showed synergistic cell growth inhibition, the combination of pitavastatin and capmatinib was the most effective. We also assessed the effects of combinations of pitavastatin and other MET inhibitors (crizotinib or MGCD-265) on cell growth (Supplementary Fig. S10B). However, in these combinations, cell growth inhibition was additive but not synergistic. These results suggested that pitavastatin and capmatinib were the best combination for cell growth inhibition. To evaluate whether the inhibition of MET alters the effect of pitavastatin on cell growth, we performed knockdown experiments using MET-specific siRNAs in HOC313-LM cells (Fig. 5E and F). The results showed that MET knockdown alone did not affect cell growth, whereas the inhibition of cell growth by pitavastatin could be enhanced by MET knockdown, similar to the combined treatment with pitavastatin and capmatinib. In addition, the effects of this combination on cell growth were canceled by the addition of mevalonate or GGPP (Supplementary Fig. S11A and S11B). Supplementation with mevalonate or GGPP also reduced the induction of immature MET by the combination (Supplementary Fig. S11C). In ovo therapeutic experiments were carried out with HOC313-LM and KYSE200 cells using pitavastatin and capmatinib (Supplementary Fig. S11D). Consistent with the in vitro experiments, the combination of pitavastatin and capmatinib inhibited tumor formation in ovo more effectively than pitavastatin alone. Moreover, in vivo mouse therapeutic experiments with HOC313-LM cells showed that the combination of pitavastatin and capmatinib inhibited tumor growth compared to treatment with pitavastatin alone and without any adverse events (Fig. 6A; Supplementary Fig. S11E). We then evaluated the phosphorylation status of MET in tumor specimens from the mice (Fig. 6B). The combination of pitavastatin and capmatinib significantly inhibited the phosphorylation of MET compared with control samples. These results suggest that the complete suppression of MET signaling by the combination of pitavastatin and capmatinib might be a useful strategy for cancer therapy to treat OSCC and ESCC (Fig. 6C). Moreover, our findings indicate that the mevalonate pathway might regulate the function of the Golgi apparatus through the activation of Arf1 or GBF1 (Fig. 6D).

Figure 5.

Combined pitavastatin and MET inhibition synergistically inhibits cell growth in vitro. A, The number of viable cells 72 hours after treatment with pitavastatin (0.1, 0.25, 0.5, 1, or 2 μmol/L) and/or capmatinib (0.1, 0.5 or 1 μmol/L) in HOC313-LM cells was assessed by CV staining assay (left). Isobologram analysis of combination treatment with pitavastatin and capmatinib (right). The line designates the combination index (CI), where CI = 1 indicates an additive effect, CI <1 indicates synergism, and CI >1 represents antagonism. B, Apoptosis analysis by flow cytometry analysis in HOC313-LM cells. Each data point represents the mean of three experiments (bars, SD). C, Western blotting of the indicated protein in HOC313-LM cells treated with pitavastatin (1 μmol/L) and/or capmatinib (0.5 μmol/L) for 72 hours. D, Western blotting of the indicated protein in HOC313-LM cells treated with HGF (100 ng/mL) after treatment with pitavastatin (0.5 μmol/L), capmatinib (0.5 μmol/L), combination (each 0.5 μmol/L), or control solvent for 48 hours. E, The number of viable cells 72 hours after treatment with pitavastatin (0.5 μmol/L) and/or knockdown of MET by si-MET #1 or #2 in HOC313-LM cells was assessed by CV staining assay. Each data point represents the mean of three experiments (bars, SD). F, Western blotting of the indicated protein in HOC313-LM cells treated with pitavastatin (0.5 μmol/L) and/or knockdown of MET by si-MET #1 or #2 for 72 hours. Student t test was used for statistical analysis.

Figure 5.

Combined pitavastatin and MET inhibition synergistically inhibits cell growth in vitro. A, The number of viable cells 72 hours after treatment with pitavastatin (0.1, 0.25, 0.5, 1, or 2 μmol/L) and/or capmatinib (0.1, 0.5 or 1 μmol/L) in HOC313-LM cells was assessed by CV staining assay (left). Isobologram analysis of combination treatment with pitavastatin and capmatinib (right). The line designates the combination index (CI), where CI = 1 indicates an additive effect, CI <1 indicates synergism, and CI >1 represents antagonism. B, Apoptosis analysis by flow cytometry analysis in HOC313-LM cells. Each data point represents the mean of three experiments (bars, SD). C, Western blotting of the indicated protein in HOC313-LM cells treated with pitavastatin (1 μmol/L) and/or capmatinib (0.5 μmol/L) for 72 hours. D, Western blotting of the indicated protein in HOC313-LM cells treated with HGF (100 ng/mL) after treatment with pitavastatin (0.5 μmol/L), capmatinib (0.5 μmol/L), combination (each 0.5 μmol/L), or control solvent for 48 hours. E, The number of viable cells 72 hours after treatment with pitavastatin (0.5 μmol/L) and/or knockdown of MET by si-MET #1 or #2 in HOC313-LM cells was assessed by CV staining assay. Each data point represents the mean of three experiments (bars, SD). F, Western blotting of the indicated protein in HOC313-LM cells treated with pitavastatin (0.5 μmol/L) and/or knockdown of MET by si-MET #1 or #2 for 72 hours. Student t test was used for statistical analysis.

Close modal
Figure 6.

Combination of pitavastatin and capmatinib enhances tumor growth inhibition in vivo. A, Tumor growth curves of subcutaneous HOC313-LM tumors treated with pitavastatin (2 mg/kg) and/or capmatinib [1 mg/kg; control (PBS with acetic acid with acetic acid: n = 10), pitavastatin (2 mg/kg n = 9), capmatinib (1 mg/kg: n = 10), combination: n = 10; left]. Images of isolated tumors (middle). Scale bar, 1 cm. Tumor weight was measured in each mouse (right). Two-way ANOVA was used for statistical analysis. B, Representative results of IHC analysis of phosphor-MET in tumor specimens from an in vivo mouse experiment (left). Scale bar, 200 μm. The phospho-MET–positive area in each tumor specimen was quantified by the naked eye (right). Student t test was used for statistical analysis. C, The relationship between the mevalonate pathway and each survival or growth pathway. D, Induction of dysfunction of the Golgi apparatus by inhibition of the mevalonate pathway.

Figure 6.

Combination of pitavastatin and capmatinib enhances tumor growth inhibition in vivo. A, Tumor growth curves of subcutaneous HOC313-LM tumors treated with pitavastatin (2 mg/kg) and/or capmatinib [1 mg/kg; control (PBS with acetic acid with acetic acid: n = 10), pitavastatin (2 mg/kg n = 9), capmatinib (1 mg/kg: n = 10), combination: n = 10; left]. Images of isolated tumors (middle). Scale bar, 1 cm. Tumor weight was measured in each mouse (right). Two-way ANOVA was used for statistical analysis. B, Representative results of IHC analysis of phosphor-MET in tumor specimens from an in vivo mouse experiment (left). Scale bar, 200 μm. The phospho-MET–positive area in each tumor specimen was quantified by the naked eye (right). Student t test was used for statistical analysis. C, The relationship between the mevalonate pathway and each survival or growth pathway. D, Induction of dysfunction of the Golgi apparatus by inhibition of the mevalonate pathway.

Close modal

In this study, we identified pitavastatin as an anticancer drug candidate from a functional cell-based screening using an FDA-approved drug library in OSCC and ESCC. We also revealed that pitavastatin inhibited tumor growth due to the downregulation of AKT and ERK signals via immature MET caused by dysfunction of the Golgi apparatus and that GGPS1 expression was important for determining sensitivity to cell growth inhibition by pitavastatin and other statins. In addition, we found that the combination of pitavastatin and capmatinib enhanced tumor growth inhibition in vitro, in ovo with CAM, and in vivo in mouse models, suggesting that this combination would be a useful strategy for cancer therapy in OSCC and ESCC. Importantly, one of the mechanisms by which pitavastatin inhibits tumor growth might be the suppression of MET signaling through immature MET caused by dysfunction of the Golgi apparatus.

Statins target the mevalonate pathway by inhibiting HMGCR, resulting in a reduction in isoprenoids such as cholesterol in the body. Epidemiologic research indicates that statins decrease the incidences of some site-specific cancers, and many studies reveal anticancer effect of statins in some types of cancer (6, 31). Statins are expected to be applicable for drug repurposing (DR) in various cancers, and useful molecular biomarkers are needed to stratify responders and nonresponders to statins among patients with cancer, leading to the establishment of precision medicine. Some previous studies in cancer have indicated that the HMGCR expression level is associated with resistance to statins because statins activate the SREBP/HMGCR/LDLR feedback loop (32). Indeed, knockdown of SREBP, HMGCR, or LDLR prevents the upregulation of HMGCR by statins and improves the sensitivity of cancer cell growth to statins (33–35). In this study, we found that GGPS1 expression might be correlated with sensitivity to pitavastatin in OSCC and ESCC cell lines, but the expression of HMGCR, a direct target of pitavastatin, is not. In addition, knockdown or overexpression of GGPS1 changed the sensitivity of cell growth to pitavastatin. These results are supported by rescue experiments with metabolites related to the mevalonate pathway. We then confirmed that the function of pitavastatin in cell growth was canceled by the addition of mevalonate or GGPP, as reported previously (14, 23–26). On the other hand, it has been considered that other metabolites in the mevalonate pathway could not rescue cell growth inhibition by pitavastatin because of lack of isopentenyl pyrophosphate (IPP), an intermediate in the mevalonate pathway (26). IPP is needed the synthesizing GGP, FPP, and GGPP but statins inhibit the formation of IPP from mevalonate by blocking HMGCR (26). Thus, adding GGP or FPP could not rescue the cell growth inhibition due to lack of IPP under pitavastatin treatment. Whereas adding mevalonate could rescue the cell growth inhibition by pitavastatin because mevalonate could restore IPP for the downstream conversion of FPP into GGPP. However, it remained unclear why adding IPP did not rescue the cell growth inhibition by pitavastatin treatment. Because GGPP is produced from FPP catalyzed by GGPS1, GGPS1 might be thought of as a key factor in sensitivity to pitavastatin. Moreover, DGBP, a GGPS1 inhibitor, also inhibited tumor growth in OSCC and ESCC cell lines as well as other types of cancer in previous studies (36, 37). Taken together, the evidence suggests that GGPS1 could be a biomarker for sensitivity to pitavastatin and a useful therapeutic target in cancer therapy.

Although previous studies have shown anticancer effects of pitavastatin in many types of cancer (13–19), the detailed mechanisms by which pitavastatin inhibits cell growth remain unknown. Some studies also indicate that statins, including pitavastatin, inhibit cell growth by autophagy, but there is not enough evidence to explain cell growth inhibition (14). In this study, we revealed that cell growth inhibition by pitavastatin was caused by the downregulation of AKT and ERK signals, and a part of this phenomenon might depend on inhibiting MET maturation through dysfunction of the Golgi apparatus. It has been reported that immature MET is induced by inhibition of the function of the Golgi apparatus by M-COPA (2-methylcoprophilinamide, also called “AMF-26”), a known inhibitor of the Golgi system (38), suggesting that the mevalonate pathway might maintain the Golgi system. In general, MET is cleaved by enzymes such as Furin, resulting in the production of a heterodimer that consists of disulfide bond-linked α and β subunits (39). Processed MET dimerizes on the cell membrane, and HGF can bind to its sema domain. After that, MET stimulation by HGF activates the PI3K/AKT, RAS, and Rho/Rock signaling pathways, promoting cell proliferation and motility (27, 40). Taken together, our findings suggest that pitavastatin inhibits the translocation of MET from the Golgi apparatus to the cell membrane by inhibiting its maturation so that HGF cannot bind to MET on the cell membrane, resulting in the inactivation of MET signaling. Moreover, we hypothesized that pitavastatin treatment alone cannot completely inhibit MET signaling, and we evaluated the effects of the combination of pitavastatin and capmatinib. Capmatinib is an orally available, highly potent and selective inhibitor of MET (41). It has been reported that capmatinib has anticancer effects in head and neck, gastric, colon, papillary renal cancers, non–small cell lung carcinoma (NSCLC), hepatocellular carcinoma, and glioblastoma (41, 42). Recently, capmatinib has been approved by the FDA (https://www.fda.gov/) for adult patients with metastatic NSCLC whose tumors have MET exon 14 skipping mutations. As expected, the combination of pitavastatin and capmatinib enhanced tumor growth inhibition in vivo compared with treatment with pitavastatin alone through the complete suppression of MET signaling. In addition, our studies revealed that other statins also showed synergistic effects on cell growth inhibition in combination with capmatinib in vitro. On the other hand, the combination of pitavastatin and other MET inhibitors (crizotinib or MGCD-265) did not show synergistic effects on cell growth inhibition in vitro. One of the reasons is that crizotinib and MGCD-265 are multikinase inhibitors but not specific MET inhibitors. In fact, we obtained data supporting that the anticancer effects of pitavastatin were enhanced by MET knockdown using MET-specific siRNA. Moreover, the combination of pitavastatin and capmatinib could enhance the antitumor effect compared with pitavastatin alone in vivo mouse model. Although treatment with pitavastatin or the combination of pitavastatin and capmatinib was effective for tumor growth inhibition in the early phase, it could not completely inhibit the tumor growth most likely due to the acquisition of drug resistance or other factors (Figs. 1F and 6A). However, because we could obviously observe the delaying of tumor growth by this combination compared with control or pitavastatin alone, this combination might be applicable as a novel cancer therapy in OSCC and ESCC. Furthermore, this combination might be a useful therapeutic strategy for patients with cancer characterized by MET alterations such as amplification, exon 14 skipping mutation, or MET activation via HGF.

We revealed that the suppression of the MET signaling pathway via Golgi apparatus dysfunction caused by pitavastatin was important for tumor growth inhibition, although it might remain the possibility it might also be that pitavastatin inhibits AKT and ERK signals through other pathways. The mevalonate pathway regulates many cellular processes, such as cell survival, migration, and membrane trafficking, through different signals. In particular, the mevalonate pathway modulates the prenylation of Ras- and Rho-GTPases, and prenylated Ras and Rho are tethered to cell membranes, resulting in the activation of each signal pathway for cell survival and migration (43). Furthermore, GGPS1 increases the prenylation of Ras proteins, leading to the activation of ERK (44). A part of the mechanism by which pitavastatin downregulates AKT and ERK signals might depend on inhibition of the prenylation of some proteins, such as Ras. Taken together, our results suggested that one of the mechanisms by which pitavastatin inhibits cell growth might be caused by the downregulation of AKT and ERK signals through the inhibition of MET maturation due to dysfunction of the Golgi apparatus, and insufficient prenylation of some proteins caused by pitavastatin might also affect cell growth.

In conclusion, our findings show that pitavastatin has potential as an anticancer drug in OSCC and ESCC and that the mevalonate pathway might modulate MET maturation via the Golgi apparatus. This novel knowledge would contribute not only to understanding the detailed mechanisms by which statins inhibit tumor growth but also to revealing the biological significance of the mevalonate pathway.

No disclosures were reported.

B. Xu: Formal analysis, validation, investigation, writing–review and editing. T. Muramatsu: Conceptualization, resources, data curation, formal analysis, supervision, funding acquisition, validation, investigation, methodology, writing–original draft, project administration, writing–review and editing. J. Inazawa: Conceptualization, resources, supervision, funding acquisition, writing–original draft, project administration, writing–review and editing.

This work was supported by KAKENHI [15H05908, to J. Inazawa), 18K15236 (to T. Muramatsu)] from the Ministry of Education, Culture, Sports, Science, and Technology (MEXT). The funders had no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. This study was also partly supported by Nanken-Kyoten, TMDU.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Bray
F
,
Ferlay
J
,
Soerjomataram
I
,
Siegel
RL
,
Torre
LA
,
Jemal
A
. 
Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries
.
CA Cancer J Clin
2018
;
68
:
394
424
.
2.
Pickering
CR
,
Zhang
J
,
Yoo
SY
,
Bengtsson
L
,
Moorthy
S
,
Neskey
DM
, et al
Frederick
, 
Integrative genomic characterization of oral squamous cell carcinoma identifies frequent somatic drivers
.
Cancer Discov
2013
;
3
:
770
81
.
3.
Song
Y
,
Li
L
,
Ou
Y
,
Gao
Z
,
Li
E
,
Li
X
, et al
Identification of genomic alterations in oesophageal squamous cell cancer
.
Nature
2014
;
509
:
91
5
.
4.
Pushpakom
S
,
Iorio
F
,
Eyers
PA
,
Escott
KJ
,
Hopper
S
,
Wells
A
, et al
Drug repurposing: progress, challenges and recommendations
.
Nat Rev Drug Discov
2019
;
18
:
41
58
.
5.
Huang
R
,
Southall
N
,
Wang
Y
,
Yasgar
A
,
Shinn
P
,
Jadhav
A
, et al
The NCGC pharmaceutical collection: a comprehensive resource of clinically approved drugs enabling repurposing and chemical genomics
.
Sci Transl Med
2011
;
3
:
80ps16
.
6.
Göbel
A
,
Rauner
M
,
Hofbauer
LC
,
Rachner
TD
. 
Cholesterol and beyond - The role of the mevalonate pathway in cancer biology
.
Biochim Biophys Acta Rev Cancer
2020
;
1873
:
188351
.
7.
Freed-Pastor
WA
,
Mizuno
H
,
Zhao
X
,
Langerød
A
,
Moon
SH
,
Rodriguez-Barrueco
R
, et al
Mutant p53 disrupts mammary tissue architecture via the mevalonate pathway
.
Cell
2012
;
148
:
244
58
.
8.
Hashimoto
A
,
Oikawa
T
,
Hashimoto
S
,
Sugino
H
,
Yoshikawa
A
,
Otsuka
Y
, et al
P53- and mevalonate pathway-driven malignancies require Arf6 for metastasis and drug resistance
.
J Cell Biol
2016
;
213
:
81
95
.
9.
Ingallina
E
,
Sorrentino
G
,
Bertolio
R
,
Lisek
K
,
Zannini
A
,
Azzolin
L
, et al
Mechanical cues control mutant p53 stability through a mevalonate-RhoA axis
.
Nat Cell Biol
2018
;
20
:
28
35
.
10.
Moon
SH
,
Huang
CH
,
Houlihan
SL
,
Regunath
K
,
Freed-Pastor
WA
,
Morris
JP
 IV
, et al
p53 represses the mevalonate pathway to mediate tumor suppression
.
Cell
2019
;
176
:
564
80
.
11.
Wang
X
,
Xu
W
,
Zhan
P
,
Xu
T
,
Jin
J
,
Miu
Y
, et al
Overexpression of geranylgeranyl diphosphate synthase contributes to tumour metastasis and correlates with poor prognosis of lung adenocarcinoma
.
J Cell Mol Med
2018
;
22
:
2177
89
.
12.
Davies
JT
,
Delfino
SF
,
Feinberg
CE
,
Johnson
MF
,
Nappi
VL
,
Olinger
JT
, et al
Current and emerging uses of statins in clinical therapeutics: a review
.
Lipid Insights
2016
;
9
:
13
29
.
13.
Lee
N
,
Tilija Pun
N
,
Jang
WJ
,
Bae
JW
,
Jeong
CH
. 
Pitavastatin induces apoptosis in oral squamous cell carcinoma through activation of FOXO3a
.
J Cell Mol Med
2020
;
24
:
7055
66
.
14.
Jiang
P
,
Mukthavaram
R
,
Chao
Y
,
Nomura
N
,
Bharati
IS
,
Fogal
V
, et al
In vitro and in vivo anticancer effects of mevalonate pathway modulation on human cancer cells
.
Br J Cancer
2014
;
111
:
1562
71
.
15.
You
HY
,
Zhang
WJ
,
Xie
XM
,
Zheng
ZH
,
Zhu
HL
,
Jiang
FZ
. 
Pitavastatin suppressed liver cancer cells in vitro and in vivo
.
Onco Targets Ther
2016
;
9
:
5383
8
.
16.
Zhang
ZY
,
Zheng
SH
,
Yang
WG
,
Yang
C
,
Yuan
WT
. 
Targeting colon cancer stem cells with novel blood cholesterol drug pitavastatin
.
Eur Rev Med Pharmacol Sci
2017
;
21
:
1226
33
.
17.
de Wolf
E
,
Abdullah
MI
,
Jones
SM
,
Menezes
K
,
Moss
DM
,
Drijfhout
FP
, et al
Dietary geranylgeraniol can limit the activity of pitavastatin as a potential treatment for drug-resistant ovarian cancer
.
Sci Rep
2017
;
7
:
5410
.
18.
Villarino
N
,
Signaevskaia
L
,
van Niekerk
J
,
Medal
R
,
Kim
H
,
Lahmy
R
, et al
A screen for inducers of bHLH activity identifies pitavastatin as a regulator of p21, Rb phosphorylation and E2F target gene expression in pancreatic cancer
.
Oncotarget
2017
;
8
:
53154
67
.
19.
Wang
L
,
Wang
Y
,
Chen
A
,
Teli
M
,
Kondo
R
,
Jalali
A
, et al
Pitavastatin slows tumor progression and alters urine-derived volatile organic compounds through the mevalonate pathway
.
FASEB J
2019
;
33
:
13710
21
.
20.
Muramatsu
T
,
Imoto
I
,
Matsui
T
,
Kozaki
K
,
Haruki
S
,
Sudol
M
, et al
YAP is a candidate oncogene for esophageal squamous cell carcinoma
.
Carcinogenesis
2011
;
32
:
389
98
.
21.
Muramatsu
T
,
Kozaki
KI
,
Imoto
S
,
Yamaguchi
R
,
Tsuda
H
,
Kawano
T
, et al
The hypusine cascade promotes cancer progression and metastasis through the regulation of RhoA in squamous cell carcinoma
.
Oncogene
2016
;
35
:
5304
16
.
22.
Li
J
,
Liu
J
,
Liang
Z
,
He
F
,
Yang
L
,
Li
P
, et al
Simvastatin and Atorvastatin inhibit DNA replication licensing factor MCM7 and effectively suppress RB-deficient tumors growth
.
Cell Death Dis
2017
;
8
:
e2673
.
23.
Crosbie
J
,
Magnussen
M
,
Dornbier
R
,
Iannone
A
,
Steele
TA
. 
Statins inhibit proliferation and cytotoxicity of a human leukemic natural killer cell line
.
Biomark Res
2013
;
1
:
33
.
24.
Greenaway
JB
,
Virtanen
C
,
Osz
K
,
Revay
T
,
Hardy
D
,
Shepherd
T
, et al
Ovarian tumour growth is characterized by mevalonate pathway gene signature in an orthotopic, syngeneic model of epithelial ovarian cancer
.
Oncotarget
2016
;
7
:
47343
65
.
25.
Kany
S
,
Woschek
M
,
Kneip
N
,
Sturm
R
,
Kalbitz
M
,
Hanschen
M
, et al
Simvastatin exerts anticancer effects in osteosarcoma cell lines via geranylgeranylation and c-Jun activation
.
Int J Oncol
2018
;
52
:
1285
94
.
26.
Demierre
MF
,
Higgins
PD
,
Gruber
SB
,
Hawk
E
,
Lippman
SM
. 
Statins and cancer prevention
.
Nat Rev Cancer
2005
;
5
:
930
42
.
27.
Comoglio
PM
,
Trusolino
L
,
Boccaccio
C
. 
Known and novel roles of the MET oncogene in cancer: a coherent approach to targeted therapy
.
Nat Rev Cancer
2018
;
18
:
341
58
.
28.
Frazier
NM
,
Brand
T
,
Gordan
JD
,
Grandis
J
,
Jura
N
. 
Overexpression-mediated activation of MET in the Golgi promotes HER3/ERBB3 phosphorylation
.
Oncogene
2019
;
38
:
1936
50
.
29.
Kaczmarek
B
,
Verbavatz
JM
,
Jackson
CL
. 
GBF1 and Arf1 function in vesicular trafficking, lipid homoeostasis and organelle dynamics
.
Biol Cell
2017
;
109
:
391
9
.
30.
Nakamura
N
. 
Emerging new roles of GM130, a cis-Golgi matrix protein, in higher order cell functions
.
J Pharmacol Sci
2010
;
112
:
255
64
.
31.
Jeong
GH
,
Lee
KH
,
Kim
JY
,
Eisenhut
M
,
Kronbichler
A
,
van der Vliet
HJ
, et al
Effect of statin on cancer incidence: an umbrella systematic review and meta-analysis
.
J Clin Med
2019
;
8
:
819
.
32.
Ma
S
,
Sun
W
,
Gao
L
,
Liu
S
. 
Therapeutic targets of hypercholesterolemia: HMGCR and LDLR
.
Diabetes Metab Syndr Obes
2019
;
12
:
1543
53
.
33.
Pandyra
AA
,
Mullen
PJ
,
Goard
CA
,
Ericson
E
,
Sharma
P
,
Kalkat
M
, et al
Genome-wide RNAi analysis reveals that simultaneous inhibition of specific mevalonate pathway genes potentiates tumor cell death
.
Oncotarget
2015
;
6
:
26909
21
.
34.
Göbel
A
,
Breining
D
,
Rauner
M
,
Hofbauer
LC
,
Rachner
TD
. 
Induction of 3-hydroxy-3-methylglutaryl-CoA reductase mediates statin resistance in breast cancer cells
.
Cell Death Dis
2019
;
10
:
91
.
35.
Furuya
Y
,
Sekine
Y
,
Kato
H
,
Miyazawa
Y
,
Koike
H
,
Suzuki
K
. 
Low-density lipoprotein receptors play an important role in the inhibition of prostate cancer cell proliferation by statins
.
Prostate Int
2016
;
4
:
56
60
.
36.
Weissenrieder
JS
,
Reilly
JE
,
Neighbors
JD
,
Hohl
RJ
. 
Inhibiting geranylgeranyl diphosphate synthesis reduces nuclear androgen receptor signaling and neuroendocrine differentiation in prostate cancer cell models
.
Prostate
2019
;
79
:
21
30
.
37.
Agabiti
SS
,
Li
J
,
Dong
W
,
Poe
MM
,
Wiemer
AJ
. 
Regulation of the Notch-ATM-abl axis by geranylgeranyl diphosphate synthase inhibition
.
Cell Death Dis
2019
;
10
:
733
.
38.
Ohashi
Y
,
Okamura
M
,
Hirosawa
A
,
Tamaki
N
,
Akatsuka
A
,
Wu
KM
, et al
M-COPA, a Golgi disruptor, inhibits cell surface expression of MET protein and exhibits antitumor activity against MET-addicted gastric cancers
.
Cancer Res
2016
;
76
:
3895
903
.
39.
Komada
M
,
Hatsuzawa
K
,
Shibamoto
S
,
Ito
F
,
Nakayama
K
,
Kitamura
N
. 
Proteolytic processing of the hepatocyte growth factor/scatter factor receptor by furin
.
FEBS Lett
1993
;
328
:
25
9
.
40.
Graveel
CR
,
Tolbert
D
,
Vande Woude
GF
. 
MET: a critical player in tumorigenesis and therapeutic target
.
Cold Spring Harb Perspect Biol
2013
;
5
:
a009209
.
41.
Baltschukat
S
,
Engstler
BS
,
Huang
A
,
Hao
HX
,
Tam
A
,
Wang
HQ
, et al
Capmatinib (INC280) is active against models of non-small cell lung cancer and other cancer types with defined mechanisms of MET activation
.
Clin Cancer Res
2019
;
25
:
3164
75
.
42.
Kou
J
,
Musich
PR
,
Staal
B
,
Kang
L
,
Qin
Y
,
Yao
ZQ
, et al
Differential responses of MET activations to MET kinase inhibitor and neutralizing antibody
.
J Transl Med
2018
;
16
:
253
.
43.
Wang
M
,
Casey
PJ
. 
Protein prenylation: unique fats make their mark on biology
.
Nat Rev Mol Cell Biol
2016
;
17
:
110
22
.
44.
Shen
N
,
Shao
Y
,
Lai
SS
,
Qiao
L
,
Yang
RL
,
Xue
B
, et al
GGPPS, a new EGR-1 target gene, reactivates ERK 1/2 signaling through increasing Ras prenylation
.
Am J Pathol
2011
;
179
:
2740
50
.