Screening and function analysis of hub genes and pathways in hepatocellular carcinoma via bioinformatics approaches

Abstract

BACKGROUND:

Liver carcinoma is a major cause of cancer-related death worldwide. Up to date, the mechanisms of liver cancerigenesis and development have not been fully understood. Multi-genes and pathways were involved in the tumorigenesis of liver cancer.

OBJECTIVE:

The aim of the present study was to screen key genes and pathways in liver cancerigenesis and development by using bioinformatics methods.

METHODS:

A dataset GSE64041 were retrieved from GEO database and the differentially expressed genes (DEGs) were screened out. Then the DEG functions were annotated by gene ontology (GO) and pathway enrichment analysis, respectively. The hub genes were further selected by protein-protein interaction (PPI) analysis. Afterwards, the mRNA and protein expressions as well as the prognostic values of the hub genes were assessed.

RESULTS:

As a result, 208 up-regulated and 82 down-regulated genes were screened out. These DEGs were mainly enriched in cell cycle and metabolism-related pathways. Through PPI analysis, TOP2A, PRDM10, CDK1, AURKA, BUB1, PLK1, CDKN3, NCAPG, BUB1B and CCNA2 were selected as hub genes, which were all over-expressed in liver cancers relative to those in normal tissues, respectively. Among them, PLK1 and CCNA2 were suggested to be prognostic factors for liver carcinoma.

CONCLUSION:

In conclusion, the present study identified several hub genes, and cell cycle and metabolism-related pathways that may play critical roles in the tumorigenesis of liver cancer. Future validation laboratory experiments are required to confirm the results.

Keywords

Liver carcinoma hub genes cell cycle prognosis bioinformatics

1. Introduction

Liver cancer, usually known as hepatocellular carcinoma (HCC), accounts for 70–85% primary malignant tumor in the liver, which is a major cause of cancer death and its incidence is increasing [1]. Previous reports have indicated that several epidemiological factors such as aflatoxins exposure [2], chronic virus hepatitis [3], and genetic factors [4] might have an association with liver cancer risk. Nevertheless, the underlying mechanisms of liver cancer genesis and development are still unclear.

The genesis of liver cancer is a complex biological process involving multiple genes and steps. Evidence suggested that dysfunctional angiogenesis, chronic inflammation, interaction of proinflammatory cytokines, endocrine hormones, adipokines and cell metabolism alteration might be involved in cancer development and progression [5]. Thus, the treatment strategy for liver cancer has been comprehensive therapy such as surgery, chemotherapy, and even biotherapy with differing advantages and disadvantages [6]. Nevertheless, most patients with liver cancer are diagnosed at intermediate or advanced disease stages, and in this case curative approaches are often not feasible. The current standard treatment for patients with advanced liver cancer is the multikinase inhibitor Sorafenib [7]. Though Sorafenib can significantly increase overall survival in these patients, seven large, randomized phase III clinical trials evaluating other molecular therapies in the first-line and second-line settings have failed to improve on the results observed with this drug [8]. Therefore, effective treatment options are needed for advanced liver cancer patients.

The exact mechanisms of liver cancerigenesis are still unknown. A number of studies have revealed possible roles of some genes and pathways such as PIN1 [9], Wnt/beta-actin [10], and Nrf2/Keap1 [11] in the development of liver cancer. However, these reports only concentrated on any certain molecule, gene or pathway, ignoring that the cancerigenesis and development process involves aberrant expression of a variety of genes and pathways, among which some hub genes/proteins might diffusively interact with other genes/proteins and thus play a key role in the malignant transformation process of cells [12]. Moreover, these hub genes/proteins may act as prognostic biomarkers or treatment targets for cancers [13]. Therefore, it is urgent to explore novel biomarkers for liver carcinoma with a powerful genome-wide technology.

Microarray is one of the widely used high-through- put tool for performing global gene expression profiles. In the present study, we aimed to find possible hub genes/proteins and screen prognostic biomarkers for liver carcinoma by using bioinformatics methods. The differentially expressed genes (DEGs) between liver cancer tissues and paired non-cancerous tissues were screened out by analyzing the microarray-based big data. Then, the functions and roles of selected candidate genes in liver cancer were further evaluated. A protein-protein interaction (PPI) network was constructed to screen the hub nodes (hub genes/proteins) that have been found to play important roles in many networks, because highly connected hub genes are expected to play critical roles in biological processes [14]. Afterwards, both the mRNA and protein expressions as well as the prognostic values of the hub genes were assessed.

2. Materials and methods

2.1 Data source

To obtain microarray-based big data, datasets were retrieved in the Gene Expression Omnibus (GEO) database (http://www.ncbi.nlm.nih.gov/geo/).

Datasets met the following criteria were considered. First, the study type must be Expression profiling by array; second, the studies should only concern human sapiens; third, the study must include both liver cancer tissues and paired non-cancerous tissues. As a consequence, a gene expression profile (GSE64041) comprising 60 liver carcinoma tissue samples and 60 paired non-cancerous liver specimens, which was deposited by Makowska et al. [15], met the criteria, and thus it was selected for further analysis.

2.2 Screening of DEGs

To gain the DEGs between liver cancer and non-cancerous tissue, the data of GSE64041 were submitted to GEO2R that is based on limma R packages from the Bioconductor project [16] for comparison. The data were downloaded for further screening. Genes that met the cut-off criteria of $P<$ 0.05 and a $|$ log fold-change $|$ of $>$ 1.0 were screened out as DEGs.

2.3 Functional enrichment analysis of DEGS

To learn the possible functions of the DEGs, the functional enrichment analyses of the DEGs, mainly involving gene ontology (GO) function analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis were carried out by using Gather tool [17]. In the GO analysis, genes were classified into hierarchical categories and the gene network was presented according to biological processes [18]. A $P$ -value of less than 0.05 was considered as statistically significant. In the KEGG pathways analysis, the DEGs were annotated [19] and the threshold of significance was based on P values. Also, a $P$ -value of less than 0.05 was considered as statistically significant.

Table 1
The most significant up-regulated and down-regulated DEGs in GSE64041 (Top ten, liver cancer versus non-cancerous tissue)

Probe set	Gene symbol	t		B		Fold change		$p$ value
Down-regulated
8053341	REG3A	$-$ 6.	60397	11.	72321	$-$ 2.	62488	1.13E-09
8114964	SPINK1	$-$ 6.	85654	12.	94729	$-$ 2.	35063	3.20E-10
8084630	NMRAL1P1	$-$ 8.	75511	22.	65842	$-$ 2.	34195	1.48E-14
8175234	GPC3	$-$ 7.	01685	13.	73453	$-$ 1.	94576	1.42E-10
8102800	SLC7A11	$-$ 8.	25822	20.	04891	$-$ 1.	93613	2.16E-13
8167254	SSX1	$-$ 6.	34142	10.	47336	$-$ 1.	87677	4.10E-09
7927694	PHYHIPL	$-$ 8.	21563	19.	82698	$-$ 1.	65853	2.71E-13
8136336	AKR1B10	$-$ 4.	41847	2.	233265	$-$ 1.	61133	2.18E-05
7923086	ASPM	$-$ 9.	65124	27.	43665	$-$ 1.	5879	1.10E-16
8117054	CAP2	$-$ 13.	5616	48.	38471	$-$ 1.	57412	4.95E-26
Up-regulated
8106418	CRHBP	12.	09655	40.	61003	2.	700262	1.48E-22
7914075	FCN3	11.	1056	35.	27825	2.	617092	3.53E-20
7928330	OIT3	13.	13233	46.	12448	2.	521529	5.08E-25
8110183	CDHR2	11.	54748	37.	6603	2.	423885	3.06E-21
7958056	STAB2	13.	84654	49.	87529	2.	376714	1.07E-26
8148029	COLEC10	10.	54784	32.	26708	2.	240457	7.76E-19
7961102	CLEC1B	13.	28557	46.	93317	2.	231518	2.21E-25
7964183	GLS2	9.	127202	24.	63333	2.	214384	1.95E-15
8113369	SLCO4C1	10.	81622	33.	71613	2.	19836	1.75E-19
7968678	FREM2	10.	75994	33.	41227	2.	165268	2.40E-19

2.4 Protein-protein interaction network construction by string and cytoscape

To screen the possible hub genes/proteins that might play a central role in the development of liver cancer, the DEGs were submitted to STRING and Cytoscape tools to predict the interaction relationship among them. In the STRING tool, a combined score of not less than 0.40 (median confidence score) was considered significant. The hub proteins were selected based on its relations with other proteins, which were sorted by the degree value in the network.

2.5 Assessment of the mRNA expression of the hub genes

To evaluate the comparison of the mRNA expression levels of the selected hub genes in liver cancer vs normal tissues, the relevant data were assessed in Oncomine database that is a platform containing about 18,000 cancer gene expression microarrays for researchers [20]. The individual gene expression level of the hub genes was analyzed in this database. The rank for a gene is the median rank for that gene across each of the analyses. For the comparisons, 1.5 fold change, P-value $=$ 0.05, and top 10% gene rank were used as threshold.

2.6 Assessment of the protein expression of the hub genes

To further learn the protein expression of the hub genes in liver cancer and normal liver tissues, they were respectively evaluated in the Human Protein Atlas (HPA) database [21]. The proteins were assessed by immunohistochemistry in tissue microarrays. The staining/expression of the proteins were classified according to the depth of stain as not detected, low, medium and high.

2.7 Assessment of the prognostic value of the hub genes

To evaluate the prognostic value of the selected hub genes in liver carcinoma, the top 10 hub genes were submitted to ProggeneV2 database [22] for analysis. A liver cancer cohort in The Cancer Genome Atlas (TCGA) was used to assess the effect of the genes/proteins on the overall survival time of the patients. The probability of survival and its significance was assessed by using Kaplan-Meier method and log-rank test, respectively. Moreover, hazard ratio (HR) value was also calculated. Two-tailed P value of less than 0.05 was considered as statistically significant.

Figure 1.

Go analysis (a) and KEGG enrichment analysis (b).

3. Results

3.1 Identification of DEGs from the selected gene expression profile

The dataset GSE64041 was retrieved from GEO database, which was a gene expression profile conducted on Affymetrix Human Gene 1.0 ST Array platform. GSE64041 included 60 cancer specimens and 60 paired non-cancer specimens. After the comparison, a total of 208 up-regulated and 82 down-regulated genes were screened out according to the criteria (Table 1, Supplementary Table 1).

3.2 Functional annotation and pathway enrichment of DEGS

To perform the function annotation of the DEGs, GO analysis was conducted. The results showed that the DEGs were enriched in 68 GO items, most of which were enriched in items regarding cell cycle and cell metabolism, such as GO:0000278 [6]: mitotic cell cycle, GO:0000280 [7]: nuclear division, GO:00454 49 [6]: regulation of transcription, GO:0007088 [7]: regulation of mitosis, and GO:0019222 [4]: regulation of metabolism. The top 20 GO items were listed in Fig. 1a.

Figure 2.

Protein-Protein Interaction networks of the 10 hub genes (proteins) including TOP2A, PRDM10, CDK1, AURKA, BUB1, PLK1, CDKN3, NCAPG, BUB1B and CCNA2.

Table 2

The hub genes in the protein-protein interaction network (top 10)

Node_name	Degree	Closeness		Betweenness		Stress	Clustering coefficient
TOP2A	52	105.	9167	8525.	548	35138	0.34163
PRDM10	49	109.	2833	10596.	17	59932	0.11565
CDK1	39	95.	78333	830.	7177	8318	0.62348
AURKA	37	95.	7	1079.	107	9478	0.66967
BUB1	36	92.	48333	683.	9743	5464	0.70952
PLK1	33	92.	15	617.	1131	5926	0.71402
CDKN3	33	92.	7	470.	8779	6442	0.78788
NCAPG	33	84.	33333	28.	92226	624	0.82386
BUB1B	33	84.	33333	28.	92226	624	0.82386
CCNA2	33	84.	33333	28.	92226	624	0.82386

Figure 3.

Pooled analyses on the mRNA expression of the ten hub genes in liver carcinoma vs normal tissues.

The results of KEGG pathway analysis showed that the DEGs were enriched in 6 pathways, including gamma-Hexachlorocyclohexane degradation, Tryptophan metabolism, Fatty acid metabolism, Bile acid biosynthesis, Cell cycle, and Stilbene, coumarine and lignin biosynthesis, in accordance with the results of the GO analysis (Fig. 1b).

3.3 Protein-protein interaction network construction

In order to screen the hub genes/proteins, the DEGs were submitted to STRING website for analysis. The DEGs with a confidence of 0.40 were involved for establishing PPI networks. In the network, the top 10 node proteins, including TOP2A, PRDM10, CDK1, AURKA, BUB1, PLK1, CDKN3, NCAPG, BUB1B and CCNA2, showed a strong association with other proteins (more than 5), suggesting that they have higher hub degrees (Table 2 and Fig. 2). Therefore, the hub genes/proteins may play a central role in the genesis and progression of liver carcinoma.

3.4 Over-expression of mRNA of the hub genes in liver carcinoma

The mRNA expression levels of TOP2A, PRDM10, CDK1, AURKA, BUB1, PLK1, CDKN3, NCAPG, BUB1B and CCNA2 were analyzed in Oncomine database, respectively. The clinical specimens of liver cancer vs normal tissue were compared, of which the data met the inclusion criteria (P value of less than 0.05) were involved. There are five datasets involved in this analysis, as shown in Fig. 3. For each gene, a comparative combined analysis was conducted and an average P value was presented. The P values for the ten hub genes were all less than 0.05 and the columns were red or white, but not blue, indicating that the mRNAs of the ten hub genes were significantly up-regulated in liver cancer tissues, compared with those in normal tissues.

Figure 4.

Assessment of the protein expression levels of TOP2A, PRDM10, CDK1, AURKA, PLK1, NCAPG, and CCNA2 in liver carcinoma and normal tissues.

3.5 Over-expression of the hub proteins in liver carcinoma

To evaluate the protein expression of the hub genes in liver carcinoma, relevant data were retrieved from the HPA database. As a consequence, expression of 7 out of 10 hub proteins both in liver cancer and normal liver tissue were available in the database. There was a lack of relevant data regarding the remaining 3 proteins, namely, BUB1, BUB1B and CDKN3, in the database. As shown in Fig. 4, ‘medium’ to ‘high staining’ of most hub proteins were shown in liver cancer, while ‘not detected’ to ‘low staining’ in normal tissues, indicating that these hub proteins were over-expressed in liver carcinoma, compared with those in normal tissues.

3.6 The prognostic value of the hub genes

To evaluate the prognostic values of the hub genes, the top 10 hub genes were assessed in a TCGA liver cancer cohort, respectively. The gene expressions were divided into two groups (high expression and low expression) according to the median expression level. The survival curves were drawn to assess the relationship between the hub gene expression levels and the prognosis of liver cancer, respectively. As a result, among the ten hub genes, expressions of PLK1 and CCNA2 might affect the overall survival time of liver cancer patients. The log-rank test showed that the patients with high expression of PLK1 [HR: 1.15 (1.01–1.31), $P=$ 0.040] and CCNA2 [HR: 1.16 (1.01–1.32), $P=$ 0.037], respectively, had a shorter survival time, compared with the patients with low expression, suggesting that high expression of PLK1 or CCNA2, respectively, might be a prognostic factor for liver carcinoma (Fig. 5, Supplementary Fig. 1).

Figure 5.

The overall survival Kaplan-Meier curves in liver cancer patients with high and low expression of PLK1 (a) and CCNA2 (b). The P value of the log-rank test for PLK1 was 0.040 ( $<$ 0.05), and for CCNA2 was 0.037 ( $<$ 0.05).

4. Discussion

In the present study, a gene expression profile GSE64041 containing 60 liver cancer specimens and paired non-cancerous specimens were retrieved and analyzed. DEGs were obtained and their functions were annotated by GO and KEGG pathway analysis. Most of the DEGs were enriched in pathways regarding cell cycle and cell metabolism. Several hub genes were screened out, including TOP2A, PRDM10, CDK1, AURKA, BUB1, PLK1, CDKN3, NCAPG, BUB1B and CCNA2, of which PLK1 and CCNA2 might have a prognostic value for liver carcinoma.

The genesis and development of liver carcinoma is a complicated biological process that includes multi-steps and genes. Previous studies have addressed the molecular mechanisms of liver carcinoma via bioinformatics methods. In 2015, Liu et al. [23] screened DEGs of liver cancer vs normal controls by pooling GSE6222 and GSE41804, and found several key genes. However, GSE6222 contained only 6 liver cancer samples and 2 healthy control samples, while GSE41804 included 10 liver cancer samples and 10 non-cancerous samples. The sample sizes of these two datasets were limited. Besides, the samples in GSE41804 were related to both HCV infection and Il28B genetic variation, indicating that the samples of GSE41804 could not be representative of a great proportion of liver carcinoma. In the present study, 60 liver cancer specimens and 60 paired non-cancer specimens were involved. The larger sample size and the paired controls might increase power to address the mechanisms of liver cancer genesis. Recently, another study conducted by He et al. [24] compared the liver cancer tissues and cirrhosis, and screened out the DEGs. Then, the functions of DEGs were further annotated. However, though cirrhosis is regarded as a pre-cancerous stage of liver cancer, not all liver cancer cases were developed from cirrhosis. Hence, the results of the study could only be appropriate for the situation that liver cancer cases were originated from cirrhosis. The data of GSE64041 were uploaded by Makowska et al. [15]. Their study focused on the molecular classification of liver cancer and the subclass-specific gene expression patterns. The hub genes were not screened. However, in the present study, we screened the hub genes as biomarkers for liver cancer by mining the same data. The study goals between Makowska et al. [15] and the present study were different.

The genesis and progression of liver carcinoma is a complex process, in which multiple genes and pathways may play important roles. Using bioinformatics methods, we analyzed the microarray-based high throughput data. As a result, a number of DEGs were screened out. GO analysis showed that these DEGs were mainly enriched in items regarding cell cycle and metabolism. In the KEGG pathway analysis, the DEGs were mostly enriched in six pathways that happened to be related to cell cycle and cell metabolisms. The results of the pathway analysis were in accordance with those of the GO analysis, indicating that aberrant functions of cell cycle and cell metabolism might play a key role during the genesis and development of liver carcinoma.

To screen any hub genes that might play critical roles among the DEGs, we constructed the PPI network. Consequently, the network was established by using STRING and Cytoscape tools. The hub genes were listed in the order of the degree that illustrates the relations of any certain gene with other genes. The top 10 node proteins, including TOP2A, PRDM10, CDK1, AURKA, BUB1, PLK1, CDKN3, NCAPG, BUB1B and CCNA2, were selected. Using Oncomine evaluation, the mRNA expression of the ten genes was higher in liver carcinoma than that in normal liver tissues, respectively. Then, the data of the HPA database showed that the protein expressions of these genes were in line with their mRNA expressions. This result indicates that the top 10 hub genes/proteins were over-expressed in liver cancer, and thus might play a critical role in its genesis and progression.

TOP2A (topoisomerase II $\alpha$ ) encodes a nuclear enzyme that controls DNA topological structure and cell cycle progression [25]. Over-expression of TOP2A was detected in liver cancer tissues, compared with those in normal tissues [26]. PRDM10 is a member of PRDM protein family that has been shown to control corneal endothelial cell differentiation and proliferation [27]. Moreover, it has been suggested to affect prostate cancer cell viability [28]. CDK1 (cyclin-dependent kinase 1) promotes cell cycle gene expression and is necessary for faithful cell division [29]. Targeting CDK1 can inhibit cellular proliferation of liver cancer cells [30]. As a negative regulator of P53, AURKA (Aurora kinase A) promotes tumor growth and cell survival [31]. Together with Myc, AURKA regulates expression of each other at the transcriptional level, and contributes to the genesis of liver carcinoma [32]. BUB1 is a gene of the spindle assembly checkpoint that controls proper chromosome segregation during mitosis [33]. Thus, it plays important roles in cell proliferation and progression of cancers [34]. CDKN3 (Cyclin-dependent kinase inhibitor 3) is essential for normal mitosis and G1/S transition [35], and its over-expression in human cancer usually indicates poor survival for patients [36]. Therefore, it also acts as a target in cancer treatment research [37]. NCAPG (non-SMC condensin I complex subunit G) organizes the coiling topology of individual chromatids, and contributes to restructuring chromatin into rod-shaped mitotic chromosomes and ensuring the segregation of sister chromatid during cell division. Hence, it takes part in epithelial-mesenchymal transition, and thus plays a role in the progression of liver carcinoma [38]. As a checkpoint related gene, BUB1B over-expression might increase risk of prostate cancer [39]. Taken together, most of the above mentioned hub genes/proteins may have a relationship with cell cycle, indicating that aberrant cell cycle functions might be a key cause or an essential mechanism underlying liver carcinoma genesis and development.

To evaluate the prognostic values of the hub genes, a TCGA cohort was used. Among these ten hub genes, only two genes (PLK1 and CCNA2) showed their prognostic value. PLK1 (Polo-like kinase 1) has been reported to be an oncogene in liver cancer [40]. It is a cell cycle protein that plays various roles in promoting cell cycle progression, particularly regulating the mitotic spindle formation checkpoint at the M-phase [41]. Besides, it is a proviral host-factor for hepatitis B virus replication [42]. Thus, PLK1 has been regarded as a target for liver cancer research [43]. CCNA2 (Cyclin A2) is expressed at elevated levels from S phase until early mitosis [44], and might promote liver cancer formation [45]. Hence, suppression of CCNA2 by miR-22 shows the therapeutic merit in liver cancer [46]. Nevertheless, their prognostic value in liver carcinoma has rarely been reported in the literature. In the present study, the data showed that the patients with low PLK1/CCNA2 expression level had a longer overall survival time relative to those of patients with high PLK1/CCNA2 expression level, respectively, suggesting that PLK1 and CCNA2 may act prognostic factors for liver carcinoma. Moreover, they may also be used as potential therapeutic targets for liver cancer.

In conclusion, the present study provides preliminary research on the mechanisms underlying liver carcinoma genesis and development. Microarray-based data were retrieved and DEGs were screened out by bioinformatics methods. Then, the functions of the DEGs were annotated by GO analysis and pathway enrichment analysis. They were mainly enriched in cell cycle and cell metabolism-related pathways. Afterwards, ten hub genes out of the DEGs were screened out and analyzed. Both the mRNA and protein expression of the hub genes were over-expressed in liver carcinoma, compared with normal liver tissues. Among the hub genes, PLK1 and CCNA2 have been indicated to be prognostic factors for liver carcinoma. Overall, the results of the present study gave a valuable indication for liver cancer research. However, future validation experiments are warranted to confirm the results.

Footnotes

Conflict of interest

The authors declare that they have no conflict of interest.

Supplementary data

The supplementary files are available to download from https://dx-doi-org.web.bisu.edu.cn/10.3233/CBM-171160.

References

Erstad

D.J.

and Tanabe

K.K.

, Hepatocellular carcinoma: early-stage management challenges, J Hepatocell Carcinoma 4 (2017), 81–92.

Moudgil

Redhu

Dhanda

and Singh

, A review of molecular mechanisms in the development of hepatocellular carcinoma by aflatoxin and hepatitis B and C viruses, J Environ Pathol Toxicol Oncol 32 (2013), 165–175.

Stepien

Hughes

D.J.

Hybsier

Bamia

Tjonneland

Overvad

Affret

His

Boutron-Ruault

M.C.

Katzke

Kuhn

Aleksandrova

Trichopoulou

Lagiou

Orfanos

Palli

Sieri

Tumino

Ricceri

Panico

Bueno-de-Mesquita

H.B.

Peeters

P.H.

Weiderpass

Lasheras

Bonet Bonet

Molina-Portillo

Dorronsoro

Huerta

J.M.

Barricarte

Ohlsson

Sjoberg

Werner

Shungin

Wareham

Khaw

K.T.

Travis

R.C.

Freisling

Cross

A.J.

Schomburg

and Jenab

, Circulating copper and zinc levels and risk of hepatobiliary cancers in Europeans, Br J Cancer 116 (2017), 688–696.

Tang

Liu

Yuan

Xie

Liu

and Wen

, Genetic factors associated with risk of metabolic syndrome and hepatocellular carcinoma, Oncotarget 8 (2017), 35403–35411.

Tahergorabi

Khazaei

Moodi

and Chamani

, From obesity to cancer: a review on proposed mechanisms, Cell Biochem Funct 34 (2016), 533–545.

Schlachterman

Craft

W.W.

, Jr. Hilgenfeldt

Mitra

and Cabrera

, Current and future treatments for hepatocellular carcinoma, World J Gastroenterol 21 (2015), 8478–8491.

Song

M.J.

and Bae

S.H.

, Newer treatments for advanced hepatocellular carcinoma, Korean J Intern Med 29 (2014), 149–155.

Llovet

J.M.

Villanueva

Lachenmayer

and Finn

R.S.

, Advances in targeted therapies for hepatocellular carcinoma in the genomic era, Nat Rev Clin Oncol 12 (2015), 408–424.

Cheng

C.W.

Leong

K.W.

and Tse

, Understanding the role of PIN1 in hepatocellular carcinoma, World J Gastroenterol 22 (2016), 9921–9932.

10.

Chen

Rajasekaran

and Hui

K.M.

, Atypical regulators of Wnt/beta-catenin signaling as potential therapeutic targets in Hepatocellular Carcinoma, Exp Biol Med (Maywood) 242 (2017), 1142–1149.

11.

Jeddi

Soozangar

Sadeghi

M.R.

Somi

M.H.

and Samadi

, Contradictory roles of Nrf2/Keap1 signaling pathway in cancer prevention/promotion and chemoresistance, DNA Repair (Amst) 54 (2017), 13–21.

12.

Jin

Wang

Huang

G.Z.

Han

L.T.

Tang

Z.Y.

Fan

D.G.

and Zhang

S.Z.

, Identifying hub genes and dysregulated pathways in hepatocellular carcinoma, Eur Rev Med Pharmacol Sci 19 (2015), 592–601.

13.

Yang

Han

Yuan

Hei

and Liang

, Gene co-expression network analysis reveals common system-level properties of prognostic genes across cancer types, Nat Commun 5 (2014), 3231.

14.

Langfelder

Mischel

P.S.

and Horvath

, When is hub gene selection better than standard meta-analysis? PLoS One 8 (2013), e61505.

15.

Makowska

Boldanova

Adametz

Quagliata

Vogt

J.E.

Dill

M.T.

Matter

M.S.

Roth

Terracciano

and Heim

M.H.

, Gene expression analysis of biopsy samples reveals critical limitations of transcriptome-based molecular classifications of hepatocellular carcinoma, J Pathol Clin Res 2 (2016), 80–92.

16.

Barrett

Wilhite

S.E.

Ledoux

Evangelista

Kim

I.F.

Tomashevsky

Marshall

K.A.

Phillippy

K.H.

Sherman

P.M.

Holko

Yefanov

Lee

Zhang

Robertson

C.L.

Serova

Davis

and Soboleva

, NCBI GEO: archive for functional genomics data sets-update, Nucleic Acids Res 41 (2013), D991–995.

17.

Chang

J.T.

and Nevins

J.R.

, GATHER: a systems approach to interpreting genomic signatures, Bioinformatics 22 (2006), 2926–2933.

18.

Tian

Wang

Guo

Liu

and Teng

, An improved method for functional similarity analysis of genes based on Gene Ontology, BMC Syst Biol 10 (2016), 119.

19.

Kanehisa

Sato

Kawashima

Furumichi

and Tanabe

, KEGG as a reference resource for gene and protein annotation, Nucleic Acids Res 44 (2016), D457–462.

20.

Rhodes

D.R.

Kalyana-Sundaram

Mahavisno

Varambally

Briggs

B.B.

Barrette

T.R.

Anstet

M.J.

Kincead-Beal

Kulkarni

Varambally

Ghosh

and Chinnaiyan

A.M.

, Oncomine 3.0: genes, pathways, and networks in a collection of 18,000 cancer gene expression profiles, Neoplasia 9 (2007), 166–180.

21.

Ponten

Jirstrom

and Uhlen

, The Human Protein Atlas – a tool for pathology, J Pathol 216 (2008), 387–393.

22.

Goswami

C.P.

and Nakshatri

, PROGgeneV2: enhancements on the existing database, BMC Cancer 14 (2014), 970.

23.

Liu

Chang

Wang

and Lu

, Identification of hepatocellular carcinoma-associated hub genes and pathways by integrated microarray analysis, Tumori 101 (2015), 206–214.

24.

Yin

Gong

Xiao

Shi

Ding

and He

, Bioinformatics analysis of key genes and pathways for hepatocellular carcinoma transformed from cirrhosis, Medicine (Baltimore) 96 (2017), e6938.

25.

de Resende

M.F.

Vieira

Chinen

L.T.

Chiappelli

da Fonseca

F.P.

Guimaraes

G.C.

Soares

F.A.

Neves

Pagotty

Pellionisz

P.A.

Barkhordarian

Brant

and Rocha

R.M.

, Prognostication of prostate cancer based on TOP2A protein and gene assessment: TOP2A in prostate cancer, J Transl Med 11 (2013), 36.

26.

Panvichian

Tantiwetrueangdet

Angkathunyakul

and Leelaudomlipi

, TOP2A amplification and overexpression in hepatocellular carcinoma tissues, Biomed Res Int 2015 (2015), 381602.

27.

Rolev

O’Donovan

D.G.

Georgiou

Rajan

M.S.

and Chittka

, Identification of Prdm genes in human corneal endothelium, Exp Eye Res 159 (2017), 114–122.

28.

Zhu

Song

Chen

Wang

Zhao

Wang

and Li

, PRDM16 is associated with evasion of apoptosis by prostatic cancer cells according to RNA interference screening, Mol Med Rep 14 (2016), 3357–3361.

29.

Landry

B.D.

Mapa

C.E.

Arsenault

H.E.

Poti

K.E.

and Benanti

J.A.

, Regulation of a transcription factor network by Cdk1 coordinates late cell cycle gene expression, EMBO J 33 (2014), 1044–1060.

30.

Zhang

Huang

Ran

Xiong

Zhong

Fan

Wang

and Ye

, miR-582-5p inhibits proliferation of hepatocellular carcinoma by targeting CDK1 and AKT3, Tumour Biol 36 (2015), 8309–8316.

31.

Sehdev

Katsha

Arras

Peng

Soutto

Ecsedy

Zaika

Belkhiri

and El-Rifai

, HDM2 regulation by AURKA promotes cell survival in gastric cancer, Clin Cancer Res 20 (2014), 76–86.

32.

Han

Tian

Zhang

Feng

and Li

, Aurora kinase A mediates c-Myc’s oncogenic effects in hepatocellular carcinoma, Mol Carcinog 54 (2015), 1467–1479.

33.

de Voer

R.M.

Geurts van Kessel

Weren

R.D.

Ligtenberg

M.J.

Smeets

Vreede

Kamping

E.J.

Verwiel

E.T.

Hahn

M.M.

Ariaans

Spruijt

van Essen

Houge

Schackert

H.K.

Sheng

J.Q.

Venselaar

van Ravenswaaij-Arts

C.M.

van Krieken

J.H.

Hoogerbrugge

and Kuiper

R.P.

, Germline mutations in the spindle assembly checkpoint genes BUB1 and BUB3 are risk factors for colorectal cancer, Gastroenterology 145 (2013), 544–547.

34.

Takagi

Miki

Shibahara

Nakamura

Ebata

Watanabe

Ishida

Sasano

and Suzuki

, BUB1 immunolocalization in breast carcinoma: its nuclear localization as a potent prognostic factor of the patients, Horm Cancer 4 (2013), 92–102.

35.

Nalepa

Barnholtz-Sloan

Enzor

Dey

Gehlhausen

J.R.

Lehmann

A.S.

Park

S.J.

Yang

Chen

Guan

Chen

Renbarger

Yang

F.C.

Parada

L.F.

and Clapp

, The tumor suppressor CDKN3 controls mitosis, J Cell Biol 201 (2013), 997–1012.

36.

Fan

Chen

Huang

Shen

Welsh

E.A.

Teer

J.K.

Cai

Cress

W.D.

and Wu

, Overexpression of major CDKN3 transcripts is associated with poor survival in lung adenocarcinoma, Br J Cancer 113 (2015), 1735–1743.

37.

Berumen

Espinosa

A.M.

and Medina

, Targeting CDKN3 in cervical cancer, Expert Opin Ther Targets 18 (2014), 1149–1162.

38.

Liu

Liang

Liu

Huang

Yin

Zhou

Feng

Zou

and Wu

, Overexpression of nonSMC condensin I complex subunit G serves as a promising prognostic marker and therapeutic target for hepatocellular carcinoma, Int J Mol Med 40 (2017), 731–738.

39.

Chen

Cai

Z.D.

Wang

Liu

Z.Z.

Lin

Z.Y.

Y.D.

Liang

Y.X.

Han

Z.D.

Liu

J.C.

and Zhong

W.D.

, Overexpression of BUB1B contributes to progression of prostate cancer and predicts poor outcome in patients with prostate cancer, Onco Targets Ther 9 (2016), 2211–2220.

40.

Wang

Chang

Wang

Qiang

and Chen

, Polo-like Kinase 1-targeting Chitosan Nanoparticles Suppress the Progression of Hepatocellular Carcinoma, Anticancer Agents Med Chem 17 (2017), 948–954.

41.

Bouhlal

Ouled-Haddou

Debuysscher

Singh

A.R.

Ossart

Reignier

Hocini

Fouquet

Al Baghami

Eugenio

M.S.

Nguyen-Khac

Regimbeau

J.M.

and Marcq

, RB/PLK1-dependent induced pathway by SLAMF3 expression inhibits mitosis and control hepatocarcinoma cell proliferation, Oncotarget 7 (2016), 9832–9843.

42.

Diab

A.M.

Foca

Fusil

Lahlali

Jalaguier

Amirache

N’Guyen

Isorce

Cosset

F.L.

Zoulim

Andrisani

O.M.

and Durantel

, Polo-like-kinase 1 is a proviral host-factor for hepatitis B virus replication, Hepatology (2017).

43.

Chauhan

Zubair

Nadeem

Ansari

S.A.

Ansari

M.Y.

and Mohammad

, Escheriosome-mediated cytosolic delivery of PLK1-specific siRNA: potential in treatment of liver cancer in BALB/c mice, Nanomedicine (Lond) 9 (2014), 407–420.

44.

Kanakkanthara

Jeganathan

K.B.

Limzerwala

J.F.

Baker

D.J.

Hamada

Nam

H.J.

van Deursen

W.H.

Hamada

Naylor

R.M.

Becker

N.A.

Davies

B.A.

van Ree

J.H.

Mer

Shapiro

V.S.

Maher

L.J.

, 3rd Katzmann

D.J.

and van Deursen

J.M.

, Cyclin A2 is an RNA binding protein that controls Mre11 mRNA translation, Science 353 (2016), 1549–1552.

45.

Hung

Y.H.

Huang

H.L.

Chen

W.C.

Yen

M.C.

Cho

C.Y.

Weng

T.Y.

Wang

C.Y.

Chen

Y.L.

Chen

L.T.

and Lai

M.D.

, Argininosuccinate lyase interacts with cyclin A2 in cytoplasm and modulates growth of liver tumor cells, Oncol Rep 37 (2017), 969–978.

46.

Yang

Liu

H.X.

and Wan

Y.J.

, MiR-22-silenced cyclin A expression in colon and liver cancer cells is regulated by bile acid receptor, J Biol Chem 290 (2015), 6507–6515.

Screening and function analysis of hub genes and pathways in hepatocellular carcinoma via bioinformatics approaches

Abstract

BACKGROUND:

OBJECTIVE:

METHODS:

RESULTS:

CONCLUSION:

Keywords

1. Introduction

2. Materials and methods

2.1 Data source

2.2 Screening of DEGs

2.3 Functional enrichment analysis of DEGS

Table 1 The most significant up-regulated and down-regulated DEGs in GSE64041 (Top ten, liver cancer versus non-cancerous tissue)

2.5 Assessment of the mRNA expression of the hub genes

2.6 Assessment of the protein expression of the hub genes

2.7 Assessment of the prognostic value of the hub genes

3.1 Identification of DEGs from the selected gene expression profile

3.2 Functional annotation and pathway enrichment of DEGS

3.4 Over-expression of mRNA of the hub genes in liver carcinoma

3.6 The prognostic value of the hub genes

Footnotes

Conflict of interest

Supplementary data

References

Table 1
The most significant up-regulated and down-regulated DEGs in GSE64041 (Top ten, liver cancer versus non-cancerous tissue)