Thiostrepton

XTP8 promotes hepatocellular carcinoma growth by forming a positive feedback loop with FOXM1 oncogene

Ming Han a, b, c, Hongping Lu b, c, Kai Han b, c, Xiaoxue Yuan b, c, Shunai Liu b, c, Yun Wang a, b, c, Jing Zhao a, b, c, Pu Liang b, c, Jun Cheng a, b, c,

Abstract

Hepatocellular carcinoma (HCC) is one of the most common cancer in the world and the main cause of cancer death. Chronic hepatitis B virus (HBV) infection is the major cause of HCC. HBx, as a transactivator, plays an important role in the occurrence and development process of HCC leading by HBV infection. XTP8, related to HBx, however, there are no studies on the function of XTP8 in HCC. In our research, we demonstrated that XTP8 was significantly up-regulated in HCC tissues compared with non-cancerous tissues in Oncomine, TCGA and GEO database. Moreover, Kaplan-Meier Plotter analysis indicated that patients with higher XTP8 expression had significantly lower overall survival. Our immunohistochemical results suggested that XTP8 protein expression in HCC tissues was dramatically higher compared with control normal tissues. In vivo xenograft experiments on nude mice, the overexpression of XTP8 promoted the tumorigenic ability of HepG2 cells. In HepG2 and Huh7 cells, XTP8 upregulated FOXM1 expression to promote cell proliferation and inhibited cell apoptosis. FOXM1 knockdown reduced promoter activity of XTP8 to downregulate XTP8 expression. Thiostrepton, an inhibitor of FOXM1, decreased XTP8 expression. Therefore, our study demonstrates that XTP8 is a valuable prognostic predictor for HCC and there is a novel positive regulatory feedback loop between XTP8 and FOXM1 promoting the development of HCC.

Keywords:
XTP8
Hepatocellular carcinoma
FOXM1

1. Introduction

Hepatocellular carcinoma (HCC) is one of most common cause of cancer mortality in the worldwide [1]. The infection of hepatitis B virus (HBV) has been shown to be a major risk factor for HCC, accounting for 55% of global cases [2]. HCC currently remains a major public health problem and one of the leading causes of cancer deaths in China. The occurrence and development of HCC is a complex process involving multiple factors, multiple genes and multiple stages. At present, the mechanism of HBV-induced HCC remains unclear, but previous studies have shown that the HBV viral genomic X gene (HBx) of HBV plays a crucial role in the pathogenesis of HCC caused by HBV [2,3]. Therefore, from the perspective of HBX, it is extremely necessary to explore new XTP8 protein (hepatitis B virus X Ag-Transactivated Protein 8, GenBank Accession: No. AF490257.1) was first isolated by our research team in 2003. XTP8 as the HBx trans-activated gene, is screened by suppression subtractive hybridization (SSH). The XTP8 gene is located on chromosome 5q12.1 and comprises 1590 bp CDS region that encodes a protein with 529 amino acid residues. It is also known as DEPDC1B or XTP1 [4]. Recent studies reported that XTP8 is associated with the development of multiple cancers including oral cancer [5,6], prostate cancer [7], non-small cell lung cancer [8], malignant melanoma [9], and soft tissue sarcoma [4]. However, the relationship between the XTP8 and HCC has not been reported so far. The concrete mechanism of XTP8 in the process of the development of HCC is still unknown. FoxM1, as an oncogene, plays a key role in the progression of HCC [10e12]. However, whether XTP8 promotes the HCC growth via interacting with FOXM1 remains unclear. In this study, we studied the role of XTP8 in HCC and explored the mechanism of HCC growth, to provide new evidences for the mechanism of HCC, especially in HBx-related HCC.

2. Materials and methods

2.1. Bioinformatics mining methods

TCGA [13] and Oncomine [14] databases were both selected to predict the expression level of XTP8 in HCC and normal liver tissues. In TCGA database, XTP8 mRNA expression information from 371 cases of HCC and 50 cases of normal liver tissues were downloaded from UCSC xena [15] (https://xenabrowser.net/). In addition, Oncomine database (https://www.oncomine.org) was also selected to search the differential expression levels between HCC and normal groups. Six mRNA expression datasets were downloaded from the GEO [16] (https://www.ncbi.nlm.nih.gov/geo/), with the accession number of GSE17548 GSE84005, and GSE22058, GSE25599, GSE50579 and GSE59259. Additionally, Kaplan-Meier Plotter database [17] (http://www.kmplot.com/) was used to draw the OS curves.

2.2. Immunohistochemistry

A tissue microarray (TMA) containing specimens from 40 HCC cases and 30 normal tissue samples was purchased from alenabio (BC03116a, China), which collected these tissues under the ethical standards, and with the donors completely informed and their consent requested. Use of the TMAs complied with relevant regulations of the Ethics Committee of Beijing Ditan Hospital, Capital Medical University. Immunohistochemistry staining for XTP8 detection was carried out with routine procedures using anti-XTP8 antibody (Invitrogen, USA) overnight at 4 C, and HRP-conjugated anti-rabbit antibody (Gene Tech, China) for 30 min. Slides were counterstained with hematoxylin and analyzed under the microscope (ZEISS, Germany).

2.3. Plasmids and siRNA oligonucleotides

The XTP8 CDS region (AF490257.1) was amplified and then inserted into the pcDNA3.1/myc-His() A vector and pCDH-CMVMCS-EF1-copGFP-T2A-Puro vector. The XTP8 promoter region (from 760 to þ160 bp) was amplified and then inserted into the pGL4.10 vector. The plasmids construction was routinely maintained in our previous studies [18,19]. The siRNA-XTP8 (siXTP8), siNC and siRNA-FOXM1 (siFOXM1) were synthesized by Genepharma Co. (Shanghai, China). The siRNA sequences are summarized in Table 1.

2.4. Cell culture, transfection and drug treatment

HepG2 and Huh7 cells were separately cultured in DMEM supplemented with 10% FBS and 1% penicillin and streptomycin (GIBCO, USA) at 37 C with 5% CO2. Cells were induced by thiostrepton for 24 h or transiently transfected with plasmid or siRNA using jetPRIME (Polyplus-transfection, France) according to the manufacturer’s protocol.

2.5. Cell proliferation assay

Cells were seeded in 96-well plates at a density of 10, 000 cells/ well with 100 mL culture medium before treatment. 10 mL CCK-8 solution (Dojindo, Japan) was added after treatment. After 30 min of incubation at 37 C, optical density values were obtained at 450 nm by using Varioskan Flash (Thermo Fisher Scientific, USA).

2.6. Caspase- 3/7 activity detection

Cells were seeded in 96-well plates at a density of 15, 000 cells/ well before treatment. Then 100 mL caspase-Glo 3/7 reagent (Promega, USA) were added to each well for 48 h after transfection or each well for 24 h after drug treatment. The cells were lysed at room temperature for 1 h. Luminescence was measured by using Veritas Microplate Luminometer (Turner Biosystems, USA).

2.7. RNA isolation and quantitative RT-PCR (qRT-PCR)

Total RNA from transfected HepG2 and Huh7 cells were separately using a Total RNA Kit (Omega, USA) and reverse transcribed into cDNA with the PrimeScript RT reagent kit (TaKaRa, Japan), according to the manufacturer’s instructions. Then cDNA was subjected to quantitative PCR (ABI, USA) amplification using specific primers. The primers sequences are listed in Table 2.

2.8. Western blot analysis

According to the routine procedure [19], adequately lysed cells were separated by 10% SDS-PAGE and transferred onto PVDF membranes. After blocking with 5% non-fat dry milk for 1 h, the membranes were incubated overnight at 4 C with primary antiXTP8 (Invitrogen, USA), anti-FOXM1 (CST, USA), anti-Bcl-2 (CST, USA), anti-Bax (CST, USA), and anti-b-actin (CST, USA) antibodies, respectively. Membranes were then washed for three times with TBS-Tween, and incubated with anti-rabbit or anti-mouse secondary antibodies (Zhongshan Jinqiao, China) for 1 h at room temperature, followed by three washes with TBS-Tween. Protein bands were detected using an enhanced chemiluminescence system (Millipore, USA) and analyzed with Bio1D software (Vilber, France).

2.9. Luciferase reporter assay

Lysis Buffer (Promega, USA). XTP8 promoter activities were measured on microplate luminometer, according to manufacturer’s Transfection assays were carried out using jetPRIME in 48-well culture dishes, with NC promoter reporter plasmids (pGL-XTP8 or pGL-NC), Renilla luciferase vector (pRL-TK) DNA and siNC (or siFOXM1). After 24 h of transfection, cells were lysed in Passive

2.10. Lentiviral infection and colony formation assay

Lentiviruses were produced in HEK293T packaging cell lines according to the routine procedure [19] after transfecting pCDHCMV-MCS-EF1-copGFP-T2A-Puro-(XTP8 or NC). In 6-cm dish, transduction was performed into HepG2 cells then selected in puromycin for 2 weeks. HepG2 cells with XTP8 overexpression or NC were seeded into 6-cm dish at a density of 2000 cells and cultured for 10 days until visible clones appeared. Cell colonies were stained using 0.1% crystal violet solution.

2.11. Nude mice xenograft model

To induce tumor formation in vivo, 4-week-old BALB/c male nude mice were purchased from Vital River Laboratory Animal Technology (Beijing, China) and 2 106 cells HepG2 cells (with XTP8 or NC) were subcutaneously inoculated into the underarm on the right side of nude mice (six in each group). Tumor size and the weight of tumors and nude mice were measured. Tumor volume was estimated according to the following formula: volume ¼ (a b2)/2. In the formula, a represents longest diameter while b represents shortest diameter. All experiments using mice were approved by the Institutional Animal Care and Use Committee of the Institute of Zoology (Chinese Academy of Sciences).

2.12. Statistical analysis

Each cell experiment was repeated at least three times. Data were analyzed by Student’s t tests, nonparametric tests or Chisquare test with the SPSS 17.0 software (IBM, USA) and GraphPad Prism 7 software (GraphPad Software Inc., USA). P < 0.05, P < 0.01 and P < 0.001 was considered statistically significant. 3. Results 3.1. XTP8 expression was highly upregulated in HCC To verify the XTP8 expression in human HCC tissues, data from several bioinformatics databases (Oncomine, TCGA and GEO) of normal group and HCC group were analyzed. In Oncomine database, the expression results of human XTP8 in two HCC research chips (Chen Liver and Wurmbach Liver) were obtained, and the relevant data were compared and analyzed (Fig. 1A and B). The results showed that XTP8 expression in HCC was highly expressed  compared with normal group (Fig. 1A, P ¼ 9.44E14; Fig. 1B, P ¼ 1.17 E6). In the TCGA HCC database (normal: 50 cases, HCC: 371 cases), XTP8 was significantly higher in HCC group than in normal group (P < 0.001) (Fig. 1C), the value of AUC was 0.8858 (Fig. 1D). In the paired 50 patients, the XTP8 was significantly expressed in the HCC group (P < 0.001) (Fig. 1E), and the value of AUC was 0.8636 (Fig.1F). By re-analyzing the mRNA database in TCGA HCC combined with clinical data, XTP8 expression also significantly increased in neoplasm histological grade (Fig. 1G), pathologic T (Fig. 1H) and pathologic stage (Fig. 1I), especially most significantly in neoplasm histological grade. In the GEO database, the expression results of XTP8 in different HCC research chips including GSE17548, GSE84005, GSE22058, GSE25599, GSE50579 and GSE59259 (Fig. 1JeO) were obtained. Compared with normal group, the expression level of XTP8 in the HCC group was significantly increased. To further clarify the relationship between XTP8 and overall survival of HCC patients, we performed online survival analysis using the Kaplan-Meier Plotter database. Compared with the high expression group of XTP8, the total survival time of patients with low expression was significantly prolonged (Fig. 1P). Among them, we found that the difference of survival time was also significant in female or male patients (Fig. 1Q, R). There were significant differences in each stage of HCC (Fig. 1SeU). To further investigate the association of XTP8 protein with HCC, in 70 cases including HCC tissue and normal tissues, we performed immunohistochemistry using tissue microarrays. According to the proportion of positive cells and staining intensity (Fig. 1V), the positive staining of XTP8 was analyzed. We scored 0, 1, 2 and 3, according to the depth of staining. Two groups were expressed, but staining intensity of XTP8 in HCC group was significantly stronger than that in normal group (Fig. 1WeY). Thus, the expression of XTP8 was positively correlated with the occurrence of HCC. The difference was statistically significant (P < 0.05, P < 0.01, P < 0.001). 3.2. XTP8 overexpression promoted HepG2 cells colony formation and tumor formation After packaging the lentivirus of XTP8 and NC, we screened the stable XTP8 and control HepG2 cells (XTP8 and NC) observed by the fluorescence microscope (Fig. 2A). The mRNA and protein differences between XTP8 and NC cells were detected by qRT-PCR (n ¼ 3) and Western blot respectively (Fig. 2B, C). The ability of colony formation was promoted by XTP8 overexpression (Fig. 2D). In order to further research the tumorigenic ability of XTP8 in vivo, we injected XTP8 and NC HepG2 cells into nude mice, respectively. The volume was calculated according to the formula. On the 11 d after inoculation, the tumor formation rate of the two groups was significantly different, and the volume of the tumor formed by XTP8 cells injection was significantly larger than that of NC. After 27 d of subcutaneous injection, the nude mice were sacrificed and a growth curve was drawn. The tumor volume was significantly increased after injection of XTP8 HepG2 cells compared with NC. The tumor weight increased significantly after injection of XTP8 HepG2 cells while mice weight increased not significantly. (Fig. 2EeJ, n ¼ 6). The difference was statistically significant (P < 0.05, P < 0.01, P < 0.001). 3.3. XTP8 promoted cell viability and inhibited cell apoptosis by regulating FOXM1 expression After transfecting XTP8 plasmid (XTP8), siXTP8 or negative controls (NC or siNC) in HepG2 and Huh7 cells, respectively, the XTP8 expression was detected by qRT-PCR (n ¼ 3) and Western blot (Fig. 3AeD). We used siXTP8-3 as siXTP8 to conduct follow-up experiments. Then the proliferation of HepG2 and Huh7 was detected at 24 h, 48 h and 72 h after transient transfection of XTP8 or siXTP8. At each time point, compared with the transfected NC group, the proliferation of XTP8 group was promoted, while the proliferation level of siXTP8 was inhibited (Fig. 3EeH, n ¼ 6).We further detected apoptosis of HepG2 and Huh7 after 48 h of transient overexpression of XTP8 or siXTP8 and their corresponding negative controls by detecting intracellular Caspase-3/7. After 48 h, the level of Caspase-3/7 in the XTP8 group was inhibited while the Caspase-3/7 activity was promoted after transfection of siXTP8 (Fig. 3IeL, n ¼ 6). In order to further clarify the gene function involved in the XTP8 gene, we conducted co-expression gene analysis in the TCGA database. XTP8 expression was strongly correlated to FOXM1(Figure 3M, P < 0.001). We found that FOXM1 was expressed higher in the HCC group that the normal group (Figure 3N, P < 0.001). In the matched data of 50 patients, FOXM1 was highly upregulated in the HCC group (Figure 3O, P < 0.001). Furthermore, we explored the expression of downstream genes by overexpressing and silencing XTP8 expression in HepG2 and Huh7 cells. We found that XTP8 silence downregulated FOXM1 mRNA (n ¼ 3) and protein expression levels, and Bcl-2 protein expression levels while it upregulated Bax protein expression levels (Fig. 3PeS). Accordingly, XTP8 overexpression upregulated FOXM1 mRNA (n ¼ 3) and protein expression levels, and Bcl-2 protein expression levels (Fig. 3TeW). These results further validated that XTP8 was a gene involved in the development of HCC by regulating FOXM1 expression. The difference was statistically significant (P < 0.05, P < 0.01, P < 0.001). 3.4. FOXM1inhibiton downregulated XTP8 expression to suppress cell viability and promote cell apoptosis To verify whether FOXM1 regulate XTP8, we transiently transfected three siRNA-FOXM1 (siFOXM1-1, -2, -3) and siNC in HepG2 cells and Huh7 cells, respectively, and used qRT-PCR to detect mRNA levels. The silence effect of siFOXM1-3 is the best (Fig. 4A, B), so we used siFOXM1-3 as siFOXM1 to conduct followup experiments. To explore the regulation of FOXM1 on XTP8 and its downstream, we silenced FOXM1 gene expression in HepG2 and Huh7 cells to explore the expression of XTP8. XTP8 expression was downregulated at mRNA (n ¼ 3) and protein expression levels (Fig. 4CeF). Our results further revealed that the proliferation activity of the cells was inhibited, and the activity of Caspase-3/7 was enhanced after FOXM1 silenced in HepG2 and Huh7 cells (Fig. 4HeJ, n ¼ 6). In order to explore whether FOXM1 was involved in the regulation of XTP8 promoter, we transfected the pGL-XTP8/ pGL-NC and the siFOXM1/siNC in HepG2 and Huh7 cells (Fig. 4K, L, n ¼ 3). The results showed that after silencing FOXM1, the XTP8 promoter activity decreased significantly. We detected the mRNA level in HepG2 and Huh7 cells by adding different concentrations of Thiostrepton (0, 1, 2 and 5 mmol/L). The mRNA levels of FOXM1 and XTP8 were down-regulated in both Thiosrepton at 2 and 5 mmol/L, and the inhibitory effect was optimal at 5 mmol/L (Figure 4M, N, P, Q, n ¼ 3). At the protein level, both FOXM1 and XTP8 were downregulated at 5 mmol/L Thiostrepton (Figure 4O, R). Cell viability was inhibited by adding different concentrations of Thiostrepton (0, 1, 2, and 5 mmol/L) in HepG2 and Huh7 cells, and decreased with concentration gradient (Fig. 4S, T, n ¼ 6). After 5 mmol/L Thiostrepton was used to induce cells apoptosis, caspase-3/7 activity significantly increased at the concentration of 5 mmol/L Thiostrepton. These results further suggested that Thiostrepton downregulated XTP8 expression to inhibit HCC growth (Figure 4U, V, n ¼ 6). The difference was statistically significant (P < 0.05, P < 0.01, P < 0.001). 4. Discussions HCC is a malignant tumor with high mortality [20]. In recent years, an increasing number of researches have shown that XTP8 gene plays an important role in tumor proliferation and metastasis [8,9]. However, the relationship between the XTP8 and HCC is still unknown. During the infection of HBV, HBx plays a vital role in the occurrence and development of HCC [2,21].Therefore, XTP8, as a HBx trans-activated gene could take part in progression of HCC. Currently effective molecular markers have been found in various cancers, but there is still a lack of effective molecular markers for the diagnosis and prognosis of HCC. Therefore, novel biomarkers for early diagnosis and prognosis of HCC are in demand [22]. According to the mass data from databases including the Oncomine database [14], TCGA database [13], the Gene Expression Omnibus (GEO) database [16] and the Kaplan-Meier plotter database [17], researchers have identified many new biomarkers and diagnosis genes of HCC. In this study, we found that the expression level of XTP8 gene is higher in HCC than in normal tissues based on above these databases. Besides these studies, we revealed that higher XTP8 gene expression predicts lower overall survival in HCC patients based on Kaplan-Meier Plotter database. This result suggested that expression of XTP8i is one of the prognostic factors for HCC patients. FOXM1 acts as a transcription factor involved in HCC [23]. Our results demonstrated that FOXM1 is associated with HCC and further validated previous results. The expression of XTP8 gene is highly correlated with FOXM1 gene, which suggested that the function of XTP8 gene can be similar to that of FOXM1. Relying on the database, we explored the correlation between XTP8 and HCC, and further found that XTP8 is associated with HCC in TMA, which enhances the credibility of relation between XTP8 and HCC. Besides that, we also conducted the classical colony formation and nude mice tumor formation experiments to verify this conclusion above. In our experiment, XTP8 can significantly promote colony formation and tumor formation for HepG2 cells, which further suggests that XTP8 promotes the development of HCC. At the same time, in HepG2 and Huh7 cells, we found that XTP8 can promote cell proliferation and inhibit cells apoptosis. In the downstream mechanism research, we further found that XTP8 regulates the expression levels of FOXM1, which plays an important role in the tumor metastasis and growth of HCC [12,24]. Overexpression of FOXM1 is associated with a variety of malignant features, indicating a poor prognosis in patients with HBV-HCC [10]. Due to the close relationship between FOXM1 and HCC, we explored whether FOXM1 as a transcription factor and oncogene to regulate XTP8. It was found that FOXM1 promotes cell proliferation and inhibit cell apoptosis. In our exploration, we further found that FOXM1 upregulates the promoter activity of XTP8 to increase XTP8 gene expression. But it is necessary to further conduct the chromatin immunoprecipitation assay (ChIP) experiment to verify the promoter research. Thiostrepton, as FOXM1 inhibitor, also downregulates XTP8 expression.
HBx, as a key gene of HCC occurrence and development, upregulates FOXM1 expression via the ERK/CREB pathway. FOXM1 inhibition significantly reduces HBx-enhanced HCC invasion and lung metastasis in vivo [10]. This study further suggests that XTP8 is involved in the regulation of HBx-mediated FOXM1 to promote HCC invasion, and further research in the future will provide some ideas for the study of HCC caused by HBx.
In summary, XTP8 is highly expressed in HCC tissues and is associated with survival prognosis of HCC. XTP8 gene can be used as an important biological indicator in the development of HCC to some extent. A positive feedback loop of XTP8 and FOXM1 promotes the process of development of HCC. This study preliminarily elucidates the initial mechanism of XTP8 regulation of HCC, and provide a new research perspective for HCC, especially HCC caused by HBx.

References

[1] R.L. Siegel, K.D. Miller, A. Jemal, Cancer statistics, 2015, Ca – Cancer J. Clin. 65 (2015) 5e29.
[2] M. Geng, X. Xin, L.Q. Bi, et al., Molecular mechanism of hepatitis B virus X protein function in hepatocarcinogenesis, World J. Gastroenterol. 21 (2015) 10732e10738.
[3] M.C. Kew, Epidemiology of chronic hepatitis B virus infection, hepatocellular carcinoma, and hepatitis B virus-induced hepatocellular carcinoma, Pathol. Biol. 58 (2010) 273e277.
[4] S. Pollino, M.S. Benassi, L. Pazzaglia, et al., Prognostic role of XTP1/DEPDC1B and SDP35/DEPDC1A in high grade soft-tissue sarcomas, Histol. Histopathol. 33 (2018) 597e608.
[5] P. Ahuja, K. Singh, In silico approach for SAR analysis of the predicted model of DEPDC1B: a novel target for oral cancer, Adv Bioinformatics 2016 (2016) 3136024.
[6] Y.F. Su, C.Y. Liang, C.Y. Huang, et al., A putative novel protein, DEPDC1B, is overexpressed in oral cancer patients, and enhanced anchorage-independent growth in oral cancer cells that is mediated by Rac1 and ERK, J. Biomed. Sci. 21 (2014) 67.
[7] S. Bai, T. Chen, T. Du, et al., High levels of DEPDC1B predict shorter biochemical recurrence-free survival of patients with prostate cancer, Oncol. Lett. 14 (2017) 6801e6808.
[8] Y. Yang, L. Liu, J. Cai, et al., DEPDC1B enhances migration and invasion of nonsmall cell lung cancer cells via activating Wnt/b-catenin signaling, Biochem. Biophys. Res. Commun. 450 (2014) 899e905.
[9] Y. Xu, W. Sun, B. Zheng, et al., DEPDC1B knockdown inhibits the development of malignant melanoma through suppressing cell proliferation and inducing cell apoptosis, Exp. Cell Res. 379 (2019) 48e54.
[10] L. Xia, W. Huang, D. Tian, et al., Upregulated FoxM1 expression induced by hepatitis B virus X protein promotes tumor metastasis and indicates poor prognosis in hepatitis B virus-related hepatocellular carcinoma, J. Hepatol. 57 (2012) 600e612.
[11] F.D. Meng, J.C. Wei, K. Qu, et al., FoxM1 overexpression promotes epithelialmesenchymal transition and metastasis of hepatocellular carcinoma, World J. Gastroenterol. 21 (2015) 196e213.
[12] B.N. Song, I.S. Chu, A gene expression signature of FOXM1 predicts the prognosis of hepatocellular carcinoma, Exp. Mol. Med. 50 (2018) e418.
[13] Z. Wang, M.A. Jensen, J.C. Zenklusen, A practical guide to the cancer genome atlas (TCGA), Methods Mol. Biol. 1418 (2016) 111e141.
[14] D.R. Rhodes, J. Yu, K. Shanker, et al., ONCOMINE: a cancer microarray database and integrated data-mining platform, Neoplasia 6 (2004) 1e6.
[15] M. Goldman, B. Craft, T. Swatloski, et al., The UCSC cancer genomics browser: update 2015, Nucleic Acids Res. 43 (2015) D812eD817.
[16] E. Clough, T. Barrett, The gene expression Omnibus database, Methods Mol. Biol. 1418 (2016) 93e110.
[17] A. Nagy, A. Lanczky, O. Menyhart, et al., Validation of miRNA prognostic power in hepatocellular carcinoma using expression data of independent datasets, Sci. Rep. 8 (2018) 9227.
[18] Z. Li, S. Feng, L. Zhou, et al., NS5ATP6 modulates intracellular triglyceride content through FGF21 and independently of SIRT1 and SREBP1, Biochem. Biophys. Res. Commun. 475 (2016) 133e139.
[19] C. Zhang, J. Wang, H. Zhang, et al., Hepatitis C virus core protein induces hepatic steatosis via Sirt1-dependent pathway, Liver Int. 38 (2018) 803e812. [20] J.M. Llovet, J. Zucman-Rossi, E. Pikarsky, et al., Hepatocellular carcinoma, Nat. Rev. Dis. Primers 2 (2016) 16018.
[21] S. Liu, S.S. Koh, C.G. Lee, Hepatitis B virus X protein and hepatocarcinogenesis, Int. J. Mol. Sci. 17 (2016) 940.
[22] A. Forner, J. Bruix, Biomarkers for early diagnosis of hepatocellular carcinoma, Lancet Oncol. 13 (2012) 750e751.
[23] N. Chai, H.H. Xie, J.P. Yin, et al., FOXM1 promotes proliferation in human hepatocellular carcinoma cells by transcriptional activation of CCNB1, Biochem. Biophys. Res. Commun. 500 (2018) 924e929.
[24] C. Liang, J. Zhao, H. Ge, et al., Clinicopathological and prognostic significance of FoxM1 in hepatocellular carcinoma patients: a meta-analysis, OncoTargets Ther. 11 (2018) 3561e3571.