HSP 90 interacts with HMGCR and promotes the progression of hepatocellular carcinoma

Heat shock protein 90 (HSP90) has been reported to promote the growth and inhibit apoptosis of hepatocellular carcinoma (HCC) cells. However, the underlying mechanisms are not fully understood. Immunostaining of the tissue array demonstrated that HSP90 was upregulated in HCC clinical samples and was associated with clinical features. HSP90 interacted with 3-hydroxy-3-methylglutaryl-CoA reductase (HMGCR), the rate-limiting enzyme of mevalonate pathway, in the immunoprecipitation assay and regulated its protein expression level by inhibiting protein degradation. In addition, lovastatin, an inhibitor of HMGCR, impaired the oncogenic functions of HSP90 in the cell growth, migration and colony formation assays. Taken together, this study demonstrated that HSP90 promoted the progression of HCC by positively regulating the mevalonate pathway and indicated that HSP90 may be a promising therapeutic target.


Introduction
Hepatocellular carcinoma (HCC) is one of the most common malignancies worldwide (1). Although great progress has been achieved in the diagnosis and treatment of HCC, the overall survival of HCC patients remains poor (2). Alteration of metabolism is a hallmark of cancer cells (3). Thus, targeting metabolic pathways may be a promising strategy for the treatment of HCC.
The mevalonate (MVA) pathway is responsible for cholesterol synthesis in mammals (4). 3-Hydroxy-3-methylglutaryl-CoA reductase (HMGCR) is a rate-limiting enzyme for the MVA pathway (5). Numerous studies have demonstrated the oncogenic roles of the MVA pathway in tumorigenesis (6,7). It has been reported that enzymes involved in the MVA pathway were upregulated in breast cancer and HMGCR promoted the transformation of benign breast epithelium (2). In gastric cancer cells, HMGCR has been reported to activate the hedgehog signaling pathway and promotes the growth and migration of cancer cells (8). In HCC, HMGCR has been demonstrated to regulate the activation of yes-associated protein and the chemoresistance of cancer cells (2). The protein expression level of HMGCR is tightly controlled at multiple levels. Sterol regulatory element-binding protein 2 is a key modulator of HMGCR via regulation of its transcription (9). On the other hand, ubiquitin fusion degradation (UFD) and autocrine motility factor receptor (GP78) promoted the degradation of the HMGCR protein (10). In addition, several microRNAs have been reported to target HMGCR (11). Statins, which are HMGCR inhibitors, have attracted attention for their role in lowing cholesterol and anti-cancer activity (2). Therefore, increased understanding of the regulation of HMGCR may improve the treatment of HCC.
Heat-shock protein 90 (HSP90) is a molecular chaperon and serves an important role in the protein folding, stability and degradation of proteins referred to as 'client proteins' (12). Numerous oncogenes (including epidermal growth factor receptor, human epidermal growth factor receptor and protein kinase B) are client proteins (13)(14)(15). Due to environmental stress (hypoxia, low pH and lack of nutrients), cancer cells are more dependent on the function of HSP90, which may explain the significant upregulation of HSP90 in cancer tissues (16,17). Several studies have reported the function of HSP90 in HCC by regulating cell proliferation and apoptosis, and the inhibitor for HSP90, 17-DMAG, has shown anti-cancer activity (18,19). However, the detail underlying mechanism of the function of HSP90 in HCC remain largely unknown. This study examined the expression pattern of HSP90 in HCC and investigated the underlying molecular mechanism.  (20). Primers for HSP90 were as follows: Plasmid construction. The coding sequence of HMGCR was inserted into pcDNA3.1 (Addgene, Inc., Cambridge, MA, USA) containing myc tag by HindIII and EcoRV, the coding sequence of HSP90 was inserted into P23 (Addgene, Inc.) containing Flag tag and green fluorescent protein by XbaI and EcoRI, and the coding sequence of UFD was inserted into pCMV-HA (Addgene, Inc.) using PstI and BamHI. All restriction enzymes were obtained from Takara Biomedical Technology (Beijing) Co., Ltd. An empty vector was used as the control. The expression vectors were transfected into 7404 and Huh-7 cells using Lipofectamine ® 2000 reagent (Thermo Fisher Scientific, Inc.) for 8 h at 37˚C, following the manufacturer's protocol. A total of 3x10 5 cells were seeded in 6-well plates and cells were transfected with 1 µg/µl plasmids using Lipofectamine ® 2000. Cells were selected using fluorescence-activated cell sorting by detecting expression of green fluorescent protein (independently expressed).

Cell
Knocking down the expression of HMGCR in HCC cells. The lentivirus (multiplicity of infection=1) used to knock down the expression of HMGCR was purchased from Shanghai GeneChem Co., Ltd. (Shanghai, China). The sequences for HMGCR-short hairpin (sh)RNA were as follows: Forward, 5'-AAG TCA TAG TGG GGA CAG TGA-3'; reverse, 5'-AAA CTC TGA TGG AAA CTC ATG-3'. The sequence for control shRNA was as follows: 5'-AGG GGA GGG GAA AAA TTT T-3'. Cells were selected with 1 µg/ml puromycin for 1 week. Following recovery for 7 days, subsequent assays were performed.
Western blotting. Proteins were extracted from HCC cells using radioimmunoprecipitation assay lysis buffer (Cell Signaling Technology, Inc., Danvers, MA, USA) containing protease and phosphatase inhibitors. Cell lysates were centrifuged (12,000 x g for 20 min at 4˚C) and the concentration was quantified using a Bradford assay. A total of 20 µg proteins were separated by 10% SDS-PAGE and transferred to polyvinylidene difluoride membranes (EMD Millipore, Billerica, MA, USA). Membranes were blocked with 5% milk at room temperature for 1 h, and probed with primary and secondary antibodies. The immunoreactive protein bands were visualized using an Enhanced Chemiluminescence kit (Pierce; Thermo Fisher Scientific, Inc.    0.003% H 2 O 2 and 0.03% 3,3'-diaminobenzidine (DAB) in 0.05 mol/l Tris-HCl (pH 7.6). Immunohistochemistry for each sample was repeated three times. Slides were then developed with DAB and counterstained with hematoxylin at room temperature for 2 min. The slides were evaluated using a light microscope.
Staining was assessed using the H-score method (21), according to the following equation: H-score=Σ(i x Pi). The staining intensity i) of the tumour cells was graded from 0-3, and Pi was the percentage of stained cells at each staining intensity. Tissues with a final HSP90 H-score <45 were classified as 'HSP90 low expression' and those with a final score ≥45 were classified as 'HSP90 high expression'.
Immunoprecipitation assay. The immunoprecipitation assay was conducted as previously described (4). Cells were washed with ice-cold PBS and lysed in Tris-buffered saline (pH 7.4) containing 50 mM Tris, 150 mM NaCl, 1% NP-40, 1 mM EDTA, 1 mM Na 3 VO 4 , 10 mM NaF, 2.5 mg/ml aprotinin and leupeptin, 1 mM β-glycerophosphate and 4-(2-aminoethyl) benzenesulfonyl fluoride hydrochloride and 10 mM iodoacetate. Lysates were incubated on ice for 15 min prior to removal of cellular debris and nuclei via centrifugation at 10,000 x g for 20 min at 4˚C. Cell lysates were incubated overnight with the HMGCR antibody (1:1,000; cat. no. ab174830; Abcam) at 4˚C. IgG (1:1,000; cat. no. 7074; Cell Signaling Technology, Inc.) was used as control. Protein A Sepharose beads (GE Healthcare Life Sciences) in a 1:1 mixture with 50 mM Tris buffer (pH 7.0) were added to the cell lysates and incubated for an additional 4 h at 4˚C. The immunoprecipitates were washed four times in Tris-buffered saline and boiled for 5 min in 40 µl Laemmli buffer containing 0.02% blue bromophenol and 2% β-mercaptoethanol.
Cell migration assay. Cell migration was evaluated using a Boyden chamber. The lower chamber was loaded with 0.152 ml medium containing 10% FBS. After 8 h, cells that migrated to the lower surface of filters were detected with traditional hematoxylin and eosin staining for 5 min at room temperature. The migrated cells were counted under an inverted microscope and four fields of view were taken for each group. The experiments were repeated three times.
Overexpression of RasV12 in normal hepatic cells. The RasV12 plasmid (cat. no. 46746) was purchased from Addgene, Inc. and its coding sequence was inserted into a P23 lentiviral vector (cat. no. 100633; Addgene, Inc.) containing Flag tag and green fluorescent protein using BamHI and EcoRI. The lentivirus was produced using 293T cells as previously described (6), and used to infect L02 at a multiplicity of infection of 1. Green fluorescent protein-positive cells were sorted using fluorescence-activated cell sorting and collected. Following a two-week recovery period, subsequent experiments were performed.
In vivo metastasis assay. Control cells (7404/P23) and HSP90 overexpressed cells (7404/Flag-HSP90) were labeled with the luciferase gene. A total of 1x10 6 cells were injected into nude mice through the tail vein. A total of 12 mice (male, five weeks old, ~20 g) were obtained from the Shanghai Laboratory Animal Center (Shanghai, China) and each group contained six mice. The mice were housed under standard conditions of 25˚C with 27% humidity, free access to food and water, and a 12 h light/dark cycle. The mice were treated with lovastatin (50 mg/kg) every three days. Control mice were treated with ethanol. After 6 weeks, the mice were injected intraperitoneally with 150 mg/kg luciferin (Sigma-Aldrich; Merck KGaA) and after 5 min the metastasis foci were examined using Living Image 2.0 software (PerkinElmer, Inc., Waltham, MA, USA). The mice were sacrificed on the sixth week following imaging.
Statistics analysis. Each experiment was repeated at least three times. The data are presented as the means ± standard deviation. Statistical analysis was performed using SPSS 15.0 (SPSS, Inc., Chicago, IL, USA). Statistical comparisons between two groups were performed using Student's t-test. For comparison among multiple groups, a two-way analysis of variance (ANOVA) was performed. Scheffe analysis was performed following ANOVA. For survival analysis, Kaplan-Meier curves were constructed and the P-value was determined using logrank test. For the association between the HSP90 expression and clinical features, the χ 2 test was performed . P<0.05 was considered to indicate a statistically significant difference. The correlation between HSP90 and HMGCR expression was calculated using the Pearson's correlation coefficient.

P90 is upregulated in HCC and is associated with clinical features.
To examine the expression pattern of HSP90, this study used a HCC tissue microarray that included 297 patients with HCC. Immunohistochemistry staining was performed to evaluate the expression of HSP90 in HCC tissues. High HSP90 expression (score >50) was found in ~80% (236/297) clinical samples, and the expression of HSP90 was associated with tumor size and tumor nodes metastasis (TNM) stage (Table I).
In addition, the expression of HSP90 in a normal liver cell line (L02) and HCC cell lines (Hep3B, MHCC97, 7404, QGY and Huh-7) was investigated. Lower expression levels of HSP90 mRNA was observed in normal liver cells (L02), and higher HSP90 mRNA expression levels were observed in HCC cell lines (Fig. 1A). In addition, it was found that oncogenic RasV12 expression was associated with enhanced expression of HSP90, which may explain the upregulation of HSP90 in HCC tissues (Fig. 1B). Furthermore, the expression of HSP90 was positively associated with the protein expression level of HMGCR (Fig. 1C), suggesting that HSP90 may modulate the expression of HMGCR.

HSP90 elevates the protein expression level of HMGCR.
Subsequently, the present study examined the modulation of HMGCR by HSP90. Overexpression of HSP90 in 7404 (P53 mutant) and Huh-7 (P53 mutant) cells significantly upregulated the expression of HMGCR (Fig. 2A). However, knock-down of HSP90 expression by two independent small interfering RNA sequences significantly decreased the protein level of HMGCR (Fig. 2B). In addition, the expression of HSP90 did not affect the mRNA expression levels of HMGCR (Fig. 2C), suggesting that HSP90 regulates the protein levels of HMGCR at the post-translation level. Consistent with this hypothesis, it was found that knock-down HSP90 promoted the ubiquitination of HMGCR (Fig. 2D), suggesting that HSP90 upregulates the protein levels of HMGCR by inhibiting its degradation.
HSP90 interacts with HMGCR. To explore the underlying mechanism by which HSP90 may inhibit the degradation of HMGCR, the interaction between HSP90 and HMGCR was examined. Immunofluorescent staining demonstrated the co-localization of HSP90 and HMGCR (Fig. 3A). Consistent with these observations, exogenously expressed HSP90 (Flag-HSP90) and HMGCR (myc-HMGCR) formed a complex in 7404 cells (Fig. 3B). In addition, the fusion protein GST-HSP90 formed a complex with endogenously expressed HMGCR (Fig. 3C). Furthermore, endogenously expressed HMGCR interacted with HSP90 in the immunoprecipitation assay using anti-HSP90 antibody (Fig. 3D). These data demonstrated an interaction between HMGCR and HSP90. The study subsequently investigated the biological significance of the interaction between HSP90 and HMGCR. Overexpression of HSP90 attenuated the degradation of HMGCR induced by the expression of UFD (Fig. 3E). Taken together, these results suggested that HSP90 interacted with HMGCR and inhibited its degradation. Lovastatin, the inhibitor of HMGCR, inhibits the growth and migration of HCC cells. To study whether HMGCR mediated the biological functions of HSP90 in HCC cells, 10 µM lovastatin, an inhibitor of HMGCR, was used to rescue the functions of HSP90. HSP90 promoted the growth (Fig. 4A), migration (Fig. 4B) and anchorage-independent growth of HCC cells (Fig. 4C), which was attenuated by the treatment of lovastatin. These results demonstrated that HSP90 regulated cell growth and migration by elevating the activity of HMGCR.
Lovastatin inhibits the metastasis of HCC cells in vivo. As presented in Table I, the expression of HSP90 was positively associated with the tumor TNM stage. Subsequently, this study examined whether HSP90 promoted the metastasis of HCC cells and whether its promotion on metastasis of HCC cells was dependent on the activity of HMGCR. As demonstrated in Fig. 5A, forced expression of HSP90 promoted the metastasis of 7404 cells, which were was injected into nude mice through the tail vein. Furthermore, the administration of lovastatin effectively blocked the metastasis of 7404 cells in vivo, which was demonstrated by the fluorescence intensity, photon counts and prolonged the survival of the mice (Fig. 5A-C). Taken together, these data suggested that HSP90 promoted the metastasis of HCC cells, which was dependent on the activity of HMGCR.

Discussion
Although several studies have demonstrated the oncogenic roles of HSP90 in the progression of HCC by promoting the proliferation and inhibiting apoptosis of cancer cells, the detailed underlying mechanism of the oncogenic roles of HSP90 in HCC remains poorly understood (22,23). The findings of the current study demonstrated that the expression of HSP90 was significantly associated with clinical features. In addition, the expression of HSP90 was induced upon the activation of Ras signaling. Furthermore, HSP90 inhibited the degradation of HMGCR, the rate-limiting enzyme of the mevalonate pathway, which suggested a role for HSP90 in cholesterol metabolism. Notably, lovastatin, an inhibitor of HMGCR, reduced the malignant behavior of cancer cells via overexpression of HSP90 both in vitro and in vivo, suggesting that the clinical use of statins may be beneficial for HCC treatment. These data suggested that HSP90 promoted the progression of HCC by activating the mevalonate pathway.
An important finding of this study was the regulation of HMGCR degradation by HSP90. A previous study demonstrated that UFD and the E3 ligase GP78 cooperated to promote the ubiquitination and degradation of HMGCR (10). The present study demonstrated an interaction between HSP90 and HMGCR, and HSP90 inhibited the ubiquitination of HMGCR induced by UFD. Based on these observations, it is speculated that HSP90 may compete with UFD for binding with HMGCR.
Another notable finding was the induction of HSP90 by RasV12. Aberrant activation of RasV12 signaling is observed in ~100% of HCC clinical samples (24), which may provide an explanation for the upregulation of HSP90 in HCC samples. In addition, HSP90 has been reported to positively regulate β-catenin/T-cell factor signaling (25), suggesting that HSP90 may be an important protein for various signaling pathways.
In the current study, lovastatin, an inhibitor of HMGCR, exhibited anti-cancer activity in HCC cells. Lovastatin impaired the growth, migration and metastasis of HCC cells driven by HSP90. Combination of 17-dimethylaminoethylamino-17-demethoxygeldanamycin (an inhibitor of HSP90) and lovastatin (the inhibitor for HMGCR) may be useful for HCC therapy.
In conclusion, the present study may have elucidated a mechanism through which HSP90 promotes the progression of HCC, and HSP90 may be a promising therapeutic target. Further investigation of the therapeutic effects of HSP90 and HMGCR inhibitors for HCC will provide novel insights.

Funding
This study was supported by the Nature Science Foundation of Inner Mongolia (grant no. 2015MS08144).

Availability of data and materials
All data generated or analyzed during this study are included in this published article.

Authors' contributions
LD and RG designed the present study. LX, CZ, HL and ZC performed the experiments.

Ethics approval and consent to participate
This study was approved by the ethics committee of Inner Mongolia People's Hospital, and written informed consent was obtained from patients.

Patient consent for publication
Written informed consent was obtained from patients.