Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Screening of the Key Genes for the Progression of Liver Cirrhosis to Hepatocellular Carcinoma Based on Bioinformatics

Screening of the Key Genes for the Progression of Liver Cirrhosis to Hepatocellular Carcinoma... Hindawi Journal of Oncology Volume 2022, Article ID 2515513, 17 pages https://doi.org/10.1155/2022/2515513 Research Article Screening of the Key Genes for the Progression of Liver Cirrhosis to Hepatocellular Carcinoma Based on Bioinformatics 1 2 1 1 1 2 Yuanbin Chen, Hongyan Qian, Xiao He , Jing Zhang, Song Xue, Yumeng Wu , 3 4 5 Jian Chen , Xuming Wu , and Suqing Zhang Medical School of Nantong University, Jiangsu, Nantong 226000, China Key Laboratory of Cancer Research Center Nantong, Affiliated Tumor Hospital of Nantong University, Jiangsu, Nantong 226361, China The Immunology Laboratory, Affiliated Tumor Hospital of Nantong University, Jiangsu, Nantong 226361, China Nantong Fourth People’s Hospital, Jiangsu Nantong 226001, China Department of Hepatobiliary and Pancreatic Surgery, Affiliated Tumor Hospital of Nantong University, Jiangsu, Nantong 226361, China Correspondence should be addressed to Jian Chen; cjsmx@sina.com, Xuming Wu; cxd050609@126.com, and Suqing Zhang; zsq7829@163.com Received 5 August 2022; Revised 22 August 2022; Accepted 30 August 2022; Published 26 September 2022 Academic Editor: Mingjun Zheng Copyright © 2022 Yuanbin Chen et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Hepatocellular carcinoma (HCC), which is among the most globally prevalent cancers, is strongly associated with liver cirrhosis. Using a bioinformatics approach, we have identified and investigated the hub genes responsible for the progression of cirrhosis into HCC. We analyzed the Gene Expression Omnibus (GEO) microarray datasets, GSE25097 and GSE17549, to identify differentially expressed genes (DEGs) in these two conditions and also performed protein-protein interaction (PPI) network analysis. STRING database and Cytoscape software were used to analyze the modules and locate hub genes following which the connections between hub genes and the transition from cirrhosis to HCC, progression of HCC, and prognosis of HCC were investigated. We used the Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis to detect the molecular mechanisms underlying the action of the primary hub genes. In all, 239 DEGs were obtained, with 94 of them showing evidence of upregulation and 145 showing evidence of downregulation in HCC tissues as compared to cirrhotic liver tissues. We identified six hub genes, namely, BUB1B, NUSAP1, TTK, HMMR, CCNA2, and KIF2C, which were upregulated and had a high diagnostic value for HCC. Besides, these six hub genes were positively related to immune cell infiltration. Since these genes may play a direct role in the progression of cirrhosis to HCC, they can be considered as potential novel molecular indicators for the onset and development of HCC. 1. Introduction raises the incidence rate of hepatic sclerosis to 84.6%, which in turn, raises the incidence of HCC to 49.9% [4]. To successfully prevent, diagnose, and treat HCC, it is Global rates of morbidity and death caused by hepatocellular crucial to understand how liver cirrhosis transforms into carcinoma (HCC), which is among the most common can- cers in the world and the second most lethal, are on the rise HCC. Many studies have showed that capillarization of liver sinusoidal endothelial cells, portal hypertension, immuno- [1]. Currently, both hepatitis B virus (HBV) and hepatitis C suppressive tumor microenvironment, etc., were important virus (HCV) have been identified as the most important factors promoting the development from liver cirrhosis to cause of HCC [2, 3]. HCC is most likely to occur in patients HCC [5, 6]. However, mitigating these factors did not with severe HBV infections, especially in those who suffer from posthepatitis cirrhosis. Posthepatitis cirrhosis also change the progression of the disease. And currently, there 2 Journal of Oncology CX CXCL6 CL6 MMP7 MMP7 PRO PROM M1 1 KC KCN NJ J1 16 6 SF SFR RP P5 5 60 CFTR CFTR CHS CHST4 T4 CLD CLDN10 N10 CLI CLIC6 C6 ANX ANXA3 A3 FGF23 FGF23 MAR MARC CO O FCN2 FCN2 GPM6A GPM6A CRHBP CRHBP CLEC1B CLEC1B FCN3 FCN3 IG IGJ J RS RSPO3 PO3 F FAM180A AM180A ES ESM M1 1 THBS4 THBS4 CYP7A1 CYP7A1 HO HOX XA A1 13 3 DU DUSP SP9 9 IG IGF2BP1 F2BP1 AFP AFP GP GPR158 R158 –1 C CO OX7B2 X7B2 S SL LC44A5 C44A5 PC PCO OL LC CE2 E2 0 FO FOX XN4 N4 FBX FBXO4 O43 3 AS ASPM PM NEK2 NEK2 NUF2 NUF2 –6 –3 0 36 SKA SKA1 1 KIF20A KIF20A –2 ZI ZIC2 C2 Log (Fold Change) I IGF2BP GF2BP3 3 Group group 1 group 2 (a) (b) C7 C7 C9 C9 CX CXCL14 CL14 T TMEM27 MEM27 CRHBP CRHBP FCN2 FCN2 CLEC4G CLEC4G FCN3 FCN3 OI OIT T3 3 SLC SLCO O1 1B B3 3 GREM2 GREM2 2 LP LPA A S SL LC22A1 C22A1 GB GBA3 A3 APO APOF F PGL PGLYRP2 YRP2 2 TENM1 TENM1 THRS THRSP P LIN LINC CO1093 O1093 CYP1A2 CYP1A2 AKRIB10 AKRIB10 SU SUL LT T1 1C C2 2 MA MAGEA6 GEA6 TOP TOP2 2A A ND NDC80 C80 RRM2 RRM2 CD CDK1 K1 PBK PBK C CCNB1 CNB1 –2 AS ASP PM M NEK2 NEK2 TT TTK K CENPF CENPF NUF2 NUF2 B BUB1B UB1B KIF20A KIF20A N NCAPG CAPG –2 0 2 DK DKK K1 1 S SP PINK1 INK1 –4 Log (Fold Change) ZI ZIC2 C2 Group group 1 group 2 (c) (d) Figure 1: Differentially expressed genes (DEGs) between HCC and cirrhotic liver tissues identified in GSE25097 and GSE17549 datasets. (a) A total of 627 genes were found to be elevated and 1716 genes depressed in HCC tissues as compared to cirrhotic liver tissues in the GSE25097 dataset. (b) The expression profiles of each of the top 20 DEGs identified from the GSE25097 dataset. (c) A total of 149 genes were found to be elevated and 285 genes depressed in HCC tissues as compared to cirrhotic liver tissues in the GSE17548 dataset. (d) The expression profiles of each of the top 20 DEGs in the GSE17548 dataset. are no ideal molecular markers that can help in distinguish- and prognosis, respectively. The Gene Ontology (GO) ing HCC from cirrhosis. To close this gap, the molecular and Kyoto Encyclopedia of Genes and Genomes (KEGG) detected the top different biological events and signaling aspects of HCC incidence, progress, and reasons for poor prognosis need to be further understood. pathways in elevated and depressed DEGs. The LASSO In this study, we analyzed two mRNA microarrays Cox regression model screened the highest predictive value from the GEO database to identify differentially expressed markers of HCC prognosis. The receiver operating charac- genes (DEGs) that vary in expression levels between HCC teristic (ROC) curve and immunoinfiltration analysis were tissues and cirrhotic liver tissues. Following this, protein- employed to analyze the role of these hub genes for HCC. protein interaction (PPI) network analyses and Kaplan- Based on these methods, we identified several genes that Meier curves investigated the connections between the could function as molecular markers to track the onset identified genes and those between identified hub genes and progression of HCC. –Log (P.adj) –Log (P.adj) 10 Journal of Oncology 3 GSE25097 GSE17548 GSE25097 GSE17548 533 94 55 1571 145 140 (a) (b) Figure 2: Identification of common genes from DEGs in the GSE25097 and GSE17548 datasets. (a) 94 genes were common upregulated, and (b) 145 genes were common downregulated. IGJ TIMD4 GPM6A UROC1 PCDH9 CLEC4G CD1D NR4A3 CD5L CLEC1B CCL21 COL4A4 NPY1R CNTN4 COLEC10 CXCL14 COL4A3 FCN2 PZP CCL19 DPT COLEC11 GPC3 HHIP ADAMTSL2 MCM8 IGF2BP3 ADAMTS13 CKAP2 ANK3 C7 FCN3 RAD51AP1NUSAP1 CFP CXCL12 CXCL2 SHBG CEP55 TSLP TRIM71 MBL2 ZWINT DUSP5 LIFR C9 ANLN KLF4 PTTG1 KDK CDC6 PLCXD3 EGR1 TRIP13 RRM2 C6 CCDC34 DCN KIF15 EPO PDGFRA AURKA KIF11 HGFAC KIF20A HGF IGF1 NUF2 UBE2C NEIL3 GYS2 BUB1B F9 EPHA2 CCNA2 CENPK NGFR C1orf112 CCNB2 CYR61 HMMR KMO SDS MCM2 SPC25 PSPH SOCS2 KIF2C HJURP ESR1 HPGD TTK SLC7A11 PRC1 CENPF GLS2 CDK1 RND3 CENPE ADH4 IDO2 CNDP1 KIF4A MAD2L1 CYP2C8 PTGS2 SHCBP1 PRKAR2B EZH2 AADAT IGFALS STIL MELK ASPA CYP1A2 FGFR2 PBK SLC22A1 CDKN2C DEPDC1B CCNE2 INMT ATAD2 NDC80 TOP2A CYP2E1 MCM10 HELLS BRIP1 TPX2 MND1 CAP2 DLGAP5 SGOL2 RBMS3 GNA14 NEK2 CCNB1 ASPM OIP5 NAT2 DTL CDC20 SLCO1B3 RBM24 FOXM1 CASC5 NCAPG OIT3 SPINK1 CYP8B1 CHRM3 ESM1 E2F8 CDKN3 FAM83D GPSM2 KIF14 CDCA5 RACGAP1 UBE2T SQLE PHLDA1 CYP39A1 MAGEA1 LHX2 FAM110C CD14 C8orf4 ANGPTL6 MT1X LCAT CETP MT1F APOF MT1G MAGEA12 ZIC2 FAM150B MARCO CLRN3 LECT2 MT1M (a) MELK NUSAP1 CENPF CCNB2 PBK KIF2C HMMR MAD2L1 CCNA2 TPX2 (b) Figure 3: Screen for hub genes. (a) PPI network. (b) Top 10 hub genes were identified by CytoHubba. 2. Materials and Methods GPL10687 platforms, respectively, was identified to screen for genes associated with liver cirrhosis and HCC [7]. The 2.1. Microarray Data. The GEO database, specifically, the MINiML files, which contained raw data, including those for GSE17548 and GSE25097 series based on GPL570 and all of the platforms, samples, and GSE records, were obtained, 4 Journal of Oncology GO GO p.adjust p.adjust tetrapyrrole binding tubulin binding heme binding microtubule binding 0.00015 steroid hydroxylase activity microtubule motor activity cyclin−dependent protein serine/threonine kinase mannose binding 0.02 regulator activity collagen−containing extracellular matrix spindle collagen trimer 0.00010 chromosomal region high−density lipoprotein particle chromosome, centromeric region pore complex condensed chromosome, centromeric region 0.01 protein kinase B signaling organelle fission 0.00005 protein activation cascade nuclear division complement activation chromosome segregation complement activation, lectin pathway mitotic nuclear division 0.02 0.04 0.06 0.08 0.10 0.12 0.1 0.2 0.3 0.4 GeneRatio GeneRatio Counts Counts (a) (b) KEGG p.adjust KEGG p.adjust 0.04 Cell cycle Amoebiasis 7.5e − 05 Human T−cell leukemia virus 1 infection NF−kappa B signaling pathway 0.03 5.0e − 05 Oocyte meiosis Chemical carcinogenesis 0.02 Progesterone−mediated oocyte maturation Tryptophan metabolism 2.5e − 05 0.01 p53 signaling pathway Histidine metabolism 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.04 0.06 0.08 0.10 GeneRatio GeneRatio Counts Counts 8 6 (c) (d) Figure 4: Functional enrichment analysis of hub genes in HCC. (a) GO analysis of DEGs in high expression samples. (b) GO analysis of DEGs in low expression samples. (c) KEGG analysis of DEGs in high expression samples. (d) KEGG analysis of DEGs in low expression samples. and the extracted data were log transformed for standardiza- gram. GEO2R, as a tool for interactive network, provides tion. Using preprocessCore, we normalized the data using users with the ability to compare two or more datasets that the median method. The annotated information included in are part of the GEO series to find DEGs [10]. The thresholds for statistical significance were set to log jfold changej >1 the platform was used to convert probes to gene symbols, which were then used in the normalization process. Probes and an adjusted p value of < 0.05. that matched more than one gene were excluded from these datasets. Several probes were utilized to detect the expression 2.3. Enrichment Analysis of DEGs. GO and KEGG databases value of each gene, and an average value from these was were utilized as references, and the “clusterProfiler” R pack- obtained. To eliminate the confounding effects of different age carried out an analysis of enrichment [11]. To correct for batches, we used the “removeBatchEffect” function of the multiple comparisons, the Benjamini–Hochberg approach “limma” package in R. Boxplots were used to analyze the was utilized, with a false discovery rate ðFDRÞ <0:05 indicat- cleaned datasets. A PCA plot was constructed to demonstrate ing statistical significance. the differences in the datasets before and after the removal of the batch effects [8, 9]. 2.4. Screening of Hub Genes. The STRING database (https:// www.string-db.org/) was utilized to get a PPI network, with 2.2. Identification of DEGs. DEGs between HCC tissues and a score of 0.4 or higher for minimum participation in inter- liver cirrhotic tissues were identified through GEO2R pro- actions [12]. The “Hubba” plug-in included in the Cytoscape Journal of Oncology 5 1.0 1.0 0.8 0.8 0.6 0.6 0.4 0.4 0.2 0.2 Overall Survival Overall Survival HR = 1.56 (1.10-2.21) HR = 1.74 (1.23-2.47) 0.0 P = 0.013 0.0 P = 0.002 0 30 60 90 120 030 60 90 120 Time (months) Time (months) BUB1B MELK Low Low High High (a) (b) 1.0 1.0 0.8 0.8 0.6 0.6 0.4 0.4 0.2 0.2 Overall Survival Overall Survival HR = 1.53 (1.08-2.16) HR = 1.52 (1.07-2.15) P = 0.016 P = 0.018 0.0 0.0 030 60 90 120 0 30 60 90 120 Time (months) Time (months) NUSAP1 MAD2L1 Low Low High High (c) (d) 1.0 1.0 0.8 0.8 0.6 0.6 0.4 0.4 0.2 0.2 Overall Survival Overall Survival HR = 1.70 (1.20-2.41) HR = 1.65 (1.16-2.33) 0.0 P = 0.003 0.0 P = 0.005 030 60 90 120 030 60 90 120 Time (months) Time (months) RRM2 TTK Low Low High High (e) (f) Figure 5: Continued. Survival probability Survival probability Survival probability Survival probability Survival probability Survival probability 6 Journal of Oncology 1.0 0.8 0.6 0.4 0.2 Overall Survival HR = 2.02 (1.42-2.87) 0.0 P < 0.001 0 30 60 90 120 Time (months) HMMR Low High (g) 1.0 0.8 0.6 0.4 0.2 Overall Survival HR = 1.65 (1.16-2.33) 0.0 P = 0.005 030 60 90 120 Time (months) CCNA2 Low High (h) 1.0 0.8 0.6 0.4 0.2 Overall Survival HR = 2.16 (1.51-3.08) P < 0.001 0.0 030 60 90 120 Time (months) KIF2C Low High (i) Figure 5: The prognostic potentials of the genes (a) BUB1B, (b) MELK, (c) MAD2L1, (d) NUSAP1, (e) RRM2, (f) TTK, (g) HMMR, (h) CCNA2, and (i) KIF2C were investigated. Patients diagnosed with HCC having higher levels of expression of these genes had lower overall survival statistics as compared to patients with lower levels of expression of these genes (logrank test, p <0:05). Based on the Cox pH model, HR was determined, and the 95% CI was shown as a dotted line. Survival probability Survival probability Survival probability Journal of Oncology 7 1.0 1.0 0.8 0.8 0.6 0.6 0.4 0.4 0.2 0.2 Progress Free Interval Progress Free Interval HR = 1.70 (1.26-2.27) HR = 1.74 (1.30-2.33) 0.0 P < 0.001 0.0 P < 0.001 0 30 60 90 120 030 60 90 120 Time (months) Time (months) BUB1B MELK Low Low High High (a) (b) 1.0 1.0 0.8 0.8 0.6 0.6 0.4 0.4 0.2 0.2 Progress Free Interval Progress Free Interval HR = 1.51 (1.13-2.02) HR = 1.71 (1.28-2.30) P = 0.006 0.0 0.0 P < 0.001 030 60 90 120 0 30 60 90 120 Time (months) Time (months) MAD2L1 CCNB2 Low Low High High (c) (d) 1.0 1.0 0.8 0.8 0.6 0.6 0.4 0.4 0.2 0.2 Progress Free Interval Progress Free Interval HR = 1.67 (1.24-2.23) HR = 1.64 (1.22-2.19) P = 0.001 0.0 0.0 P = 0.001 030 60 90 120 030 60 90 120 Time (months) Time (months) NUSAP1 RRM2 Low Low High High (e) (f) 1.0 1.0 0.8 0.8 0.6 0.6 0.4 0.4 0.2 0.2 Progress Free Interval Progress Free Interval HR = 1.69 (1.26-2.26) HR = 1.78 (1.33-2.39) 0.0 P < 0.001 0.0 P < 0.001 0 30 60 90 120 030 60 90 120 Time (months) Time (months) TTK HMMR Low Low High High (g) (h) Figure 6: Continued. Survival probability Survival probability Survival probability Survival probability Survival probability Survival probability Survival probability Survival probability 8 Journal of Oncology 1.0 1.0 0.8 0.8 0.6 0.6 0.4 0.4 0.2 0.2 Progress Free Interval Progress Free Interval HR = 1.69 (1.26-2.26) HR = 2.01 (1.50-2.70) 0.0 P < 0.001 0.0 P < 0.001 030 60 90 120 0 30 60 90 120 Time (months) Time (months) CCNA2 KIF2C Low Low High High (i) (j) Figure 6: The prognostic potentials of the genes (a) BUB1B, (b) MELK, (c) MAD2L1, (d) CCNB2, (e) NUSAP1, (f) RRM2, (g) TTK, (h) HMMR, (i) CCNA2, and (j) KIF2C were investigated. Patients diagnosed with HCC having higher levels of expression of these genes had lower cancer-free intervals as compared to patients with lower expression levels of these genes (logrank test, p <0:05). Based on the Cox pH model, HR was determined, and the 95% CI was shown as a dotted line. program was used to identify and choose the top 10 hub 3. Results nodes listed by degree [13]. 3.1. Screening for Differentially Expressed Genes. Two data- sets from the GEO database, namely, GSE25097 and 2.5. Survival Analysis. The raw RNA-sequencing data and GSE17548, were used to identify DEGs between cirrhotic liver accompanying clinical information were obtained from the and HCC tissues. We identified 2343 DEGs in GSE25097 Cancer Genome Atlas (TCGA) database. Log-rank tests dataset, of which 627 were elevated and 1716 were depressed obtained p values, hazard ratios (HR), and 95% confidence in HCC tissues when compared to cirrhotic liver tissues intervals (CI) for the two groups (cirrhotic liver tissues and (Figure 1(a)). In the GSE17548 dataset, we identified 434 HCC tissues). These results were then used to plot Kaplan- DEGs, of which 149 were upregulated and 285 were downreg- Meier (KM) survival analysis to assess distinctions in sur- ulated in HCC tissues when compared to cirrhotic liver tissues vival between the cirrhotic liver and HCC tissue groups (Figure 1(c)). The heatmap shows the expression levels of each [14, 15]. of the top 20 DEGs (Figures 1(b) and 1(d)). 3.2. Screening for Hub Genes. To further analyze the com- 2.6. Construction of Prognostic Signatures. To investigate the mon genes in the two datasets, Venn diagram was employed potential diagnostic utility of the hub genes identified, we to find 94 common upregulated genes and 145 common carried out least absolute shrinkage and selection opera- downregulated genes in HCC tissues as compared to cir- tor- (LASSO-) penalized Cox regression analysis [16]. rhotic liver tissues (Figures 2(a) and 2(b)). Using STRING The “glmnet” package in R package was used to develop and Cytoscape, we analyzed these 239 DEGs to identify a model for prognosis. A LASSO regression was carried those with interaction scores > 0:4. PPI network obtained a out with the assistance of a cross-validation of 10 folds, total of 183 nodes and 2193 edges (Figure 3(a)). CytoHubba with the penalty parameter (λ) adjusted to fulfil the was utilized to get the top 10 hub genes, namely, BUB1B, optimal value. Findings of the LASSO regression were MELK, MAD2L1, CCNB2, NUSAP1, RRM2, TTK, HMMR, used as the basis for calculating risk ratings. Patients CCNA2, and KIF2C (Figure 3(b)). diagnosed with HCC who participated in the TCGA study were grouped into low-risk and high-risk categories 3.3. GO and KEGG Enrichment Analyses of DEGs. To fur- based on the median risk score. A KM survival analysis ther examine the biological roles of the identified DEGs, was carried out to evaluate and contrast the variations we used the “clusterProfiler” package in R for GO and in overall survival (OS) that were observed in the two KEGG pathway enrichment analyses. The results of the groups [14, 17, 18]. GO analysis of upregulated DEGs indicated that this group contained genes related to biological processes 2.7. Immunoinfiltration Analysis. The immunogene module (including mitotic nuclear division, chromosome segrega- of the TIMER tool (https://cistrome.shinyapps.io/timer/) tion, nuclear division, and organelle fission), cellular was used to analyze correlations between the expression of components (including spindle, chromosomal region, hub genes and immunological infiltration (including infiltra- chromosome, centromeric region, and condensed chromo- tion levels of B cell, CD4 + T cell, CD8 + T cell, macrophage, some), and molecular functions (including tubulin bind- neutrophil, and dendritic cell) in HCC tissues from the ing, microtubule binding, microtubule motor activity, and TCGA [19, 20]. cyclin-dependent protein serine/threonine kinase regulator Survival probability Survival probability Journal of Oncology 9 6 ⁎⁎⁎ ⁎⁎⁎ 1 1 Normal Tumor Normal Tumor (a) (b) ⁎⁎⁎ 8 5 ⁎⁎⁎ 1 2 Normal Tumor Normal Tumor (c) (d) ⁎⁎⁎ ⁎⁎⁎ 8 5 Normal Tumor Normal Tumor (e) (f) ⁎⁎⁎ ⁎⁎⁎ 0 0 Normal Tumor Normal Tumor (g) (h) Figure 7: Continued. e expression of HMMR e expression of MAD2L1 e expression of BUB1B e expression of RRM2 Log (TPM+1) Log (TPM+1) Log (TPM+1) Log (TPM+1) 2 2 2 e expression of CCNA2 e expression of TTK e expression of NUSAP1 e expression of MELK Log (TPM+1) Log (TPM+1) Log (TPM+1) Log (TPM+1) 2 2 2 2 10 Journal of Oncology ⁎⁎⁎ Normal Tumor (i) Figure 7: Differential expression analysis of hub genes in TCGA dataset consisting of HCC tissue samples and normal tissue samples. The expression levels of (a) BUB1B, (b) MELK, (c) MAD2L1, (d) NUSAP1, (e) RRM2, (f) TTK, (g) HMMR, (h) CCNA2, and (i) KIF2C in HCC tissues (n = 374) were significantly higher than those in normal tissues (n =50). activity) (Figure 4(a)). Analysis of the downregulated HMMR, CCNA2, and KIF2C) were shown to have high pre- DEGs indicates that this group contains genes linked to dictive value for HCC prognosis. Patients diagnosed with biological processes (including complement activation, lec- HCC were split into two categories according to their risk scores. Figure 9(c) depicts the distributions of risk scores, tin pathway, complement activation, protein activation cascade, and protein kinase B signaling), cellular compo- survival statuses, and expression levels of these six genes in nents (including pore complex, high-density lipoprotein the patient population (Figure 9(c)). particle, collagen trimer, and collagen-containing extracel- In TCGA, the data on 374 HCC samples with detailed lular matrix), and molecular functions (including mannose clinicopathological information (Table 1) were evaluated for clinically relevant markers. These hub genes were mea- binding, steroid hydroxylase activity, heme binding, and tetrapyrrole binding) (Figure 4(b)). In addition, KEGG sured at mRNA levels in HCC tissues and normal tissues, analysis revealed that the elevated DEGs were intimately as well as the data was used to generate ROC curve. Our connected to the cell cycle, human T-cell leukemia virus results indicated that BUB1B, NUSAP1, TTK, HMMR, 1 infection, oocyte meiosis, progesterone-mediated oocyte CCNA2, and KIF2C were all upregulated in HCC at the mRNA levels. And the six hub genes had a high diagnostic maturation, and the p53 signaling pathway (Figure 4(c)). The downregulated DEGs were connected to amoebiasis, value, with AUCs of 0.961, 0.949, 0.971, 0.968, 0.970, and NF-kappa B signaling pathway, chemical carcinogenesis, 0.981, respectively (Figure 10). tryptophan metabolism, and histidine metabolism (Figure 4(d)). 3.6. Relationship between Hub Gene Expression and the Infiltration of Immune Cells. It has been shown that 3.4. Relationship between HCC Prognosis and Expression of tumor-associated fibroblasts in the stroma of the tumor Hub Genes. A univariate Cox regression analysis was carried microenvironment may affect a wide range of immune cells out to identify which hub genes were linked to HCC progno- that infiltrate the tumor. The effects of the hub genes identi- sis. We find that nine of the 10 identified hub genes (BUB1B, fied here on the recruitment of immune cells in the tumor MELK, MAD2L1, NUSAP1, RRM2, TTK, HMMR, CCNA2, microenvironment and hence on the prognosis of HCC are and KIF2C) showed prognostic significance (Figures 5 and as yet unknown. To investigate this, we analyzed the connec- 6). The expression profiles of these nine genes were then tions between BUB1B, NUSAP1, TTK, HMMR, CCNA2, evaluated in 374 HCC tissue samples and 50 normal liver tis- and KIF2C with immune infiltration in HCC and found that sue samples obtained from the TCGA database. Our findings the expression levels of them were positively associated with indicated that the expression levels of these nine hub genes the immune infiltration level of immune cells (Figure 11). in HCC tissues were significantly higher than those in nor- mal tissues (Figures 7 and 8). 4. Discussion 3.5. Construction of Prognostic Signatures of Hub Genes in HCC. The LASSO Cox regression model was utilized to Globally, HCC is the second deadliest and fifth most com- choose genes with the highest predictive value as potential monly occurring cancer [21]. The disease progression is quick markers of HCC prognosis. The value (λ =0:0088) was with malignancy at a high level, which combined with low detected because it was the lowest when compared to the incidences of early detection, usually points to a bad prognosis. median of the sum of the squared residuals (Figures 9(a) A high risk of developing HCC is associated with HBV or and 9(b)). Six possible predictors (BUB1B, NUSAP1, TTK, HCV infections, cirrhosis, and alcohol intake. Of these, e expression of KIF2C Log (TPM+1) 2 Journal of Oncology 11 ⁎⁎⁎ ⁎⁎⁎ 0 0 Normal Tumor Normal Tumor (a) (b) ⁎⁎⁎ ⁎⁎⁎ Normal Tumor Normal Tumor (c) (d) ⁎⁎⁎ ⁎⁎⁎ Normal Tumor Normal Tumor (e) (f) ⁎⁎⁎ ⁎⁎⁎ 0 0 Normal Tumor Normal Tumor (g) (h) Figure 8: Continued. e expression of MAD2L1 e expression of HMMR e expression of RRM2 e expression of BUB1B Log (TPM+1) Log (TPM+1) Log (TPM+1) Log (TPM+1) 2 2 2 e expression of CCNA2 e expression of TTK e expression of NUSAP1 e expression of MELK Log (TPM+1) Log (TPM+1) Log (TPM+1) Log (TPM+1) 2 2 2 2 12 Journal of Oncology ⁎⁎⁎ Normal Tumor (i) Figure 8: Differential expression analysis of hub genes in TCGA dataset consisting of HCC tissue samples and paired adjacent normal tissue samples. The expression levels of (a) BUB1B, (b) MELK, (c) MAD2L1, (d) NUSAP1, (e) RRM2, (f) TTK, (g) HMMR, (h) CCNA2, and (i) KIF2C in HCC tissues (n =50) were significantly higher than those in paired adjacent tissues (n =50). 9 9 9 8 8 8 8 8 8 7 6 6 6 5 4 4 4 4 4 4 3 3 2 2 2 2 2 1 0 986431 0.5 12.0 11.8 0.0 11.6 −0.5 11.4 −1.0 −7 −6 −5 −4 −3 −2 −7 −6 −5 −4 −3 −2 Log (𝜆) Log (𝜆) (a) (b) −1 Risk group Low High Status Dead Alive BUB1B 3 NUSAP1 2 TTK 1 HMMR CCNA2 −1 KIF2C −2 (c) Figure 9: Clinical significance of these hub genes in HCC patients’ data from TCGA. (a) Using the LASSO Cox regression model, the partial likelihood deviance versus log (λ) has been plotted. (b) Using the lambda parameter, chosen feature coefficients are shown. (c) Distribution of risk score, prognostic hub gene expression, and survival status of HCC patients. Partial Likelihood Deviance Survival time Risk score e expression of KIF2C Log (TPM+1) Coefficients Journal of Oncology 13 bioinformatics analyses using data from two gene chip data- Table 1: Baseline clinical information. sets (from cirrhotic liver and HCC tissues), we identified 239 Characteristic Levels Overall DEGs, of which 94 were elevated and 145 were depressed. n 374 Using Cytoscape, we were able to identify ten possible hub genes from these DEGs. The genes with the highest prognos- T1 183 (49.3%) tic potential were identified using the LASSO Cox regression T2 95 (25.6%) T stage, n (%) model. The hub genes that we have identified are intricately T3 80 (21.6%) connected to the incidence, progression, and prognosis of T4 13 (3.5%) HCC and therefore may be very useful in the early detection N0 254 (98.4%) and treatment of HCC. N stage, n (%) N1 4 (1.6%) We have identified BUB1B, NUSAP1, TTK, HMMR, M0 268 (98.5%) CCNA2, and KIF2C as potential predictive markers for M stage, n (%) HCC. Previous studies indicate that expression levels of M1 4 (1.5%) BUB1B, which is a spindle-assembly checkpoint gene Female 121 (32.4%) Gender, n (%) [26], were highly upregulated in multiple myeloma Male 253 (67.6%) patients and that these levels were strongly correlated with ≤60 177 (47.5%) unfavorable outcomes [27]. Another marker, NUSAP1, Age, n (%) >60 196 (52.5%) which is a microtubule-associated protein involved in ≤400 215 (76.8%) mitosis, is also known to participate in cell proliferation, AFP (ng/ml), n (%) >400 65 (23.2%) apoptosis, and repairing DNA damage in glioblastoma multiforme cells [28]. The protein kinase encoded by the No 208 (65.4%) Vascular invasion, n (%) TTK gene is necessary for mitotic checkpoints as well as Yes 110 (34.6%) the DNA damage response [29]. Elevated HMMR in Alive 244 (65.2%) mouse mammary epithelium enhances the rate of Brca1- OS event, n (%) Dead 130 (34.8%) mutant carcinogenesis as it is involved in modifying the A 219 (90.9%) phenotype of tumor cell and tumor microenvironment Child-Pugh grade, n (%) B 21 (8.7%) [30]. The CCNA2 gene also plays an important role in C 1 (0.4%) HCC, as the HBV genome integrates into one of the CCNA2 introns and forms an in-frame chimeric fusion Age, median (IQR) 61 (52, 69) with CCNA2 [31]. The KIF2C gene, belonging to the Kinesin family, has been shown to be significantly overex- pressed in several human malignancies [32]. Since mRNA is an essential component of all cells, including tumor cells, changes in mRNA levels of hub genes cirrhosis is the most significant risk factor, since 80–90% of HCC patients usually suffered from cirrhosis [22, 23]. can be used as molecular indicators for a variety of disorders, Patients diagnosed with HCC who undergo curative including cancer [13, 33, 34]. We find that the AUCs for therapy in the early stages of the disease have significantly BUB1B, NUSAP1, TTK, HMMR, CCNA2, and KIF2C were higher five-year survival rates [24]. However, the mechanism all >0.9, which indicates the expression levels of these genes for liver cirrhosis progresses into HCC is as yet unknown, can be used to differentiate between HCC tissues and normal though there are two theories about this process. One theory liver tissues. Besides, immune cells that have invaded a assumes that liver cirrhosis itself is a precancerous stage that tumor are called tumor-infiltrating cells. These cells are a leads to HCC due to internal hepatic interstitial changes and key part of the microenvironment of a tumor and are modulations in cell proliferation. The second theory postu- strongly related with carcinogenesis, progression, or metas- lates that cirrhosis affects hepatocyte proliferation by making tasis. In our results, we found the six hub genes were all pos- the cells more sensitive to carcinogenic factors in the itively associated with the immune infiltration level of external environment, which predisposes them to damage immune cells. All these results collectively suggested that that leads to the development of HCC [25]. Since the these hub genes could serve as diagnostic molecular markers rapid rate of cellular reproduction does not allow these for HCC. cells sufficient time for DNA repair, mutations accumulate Considering that the predictive signature was developed in newly produced cells, which pave the way to malignant and verified by the use of data from public databases, more transformation. experimental proof on top of the statistical evidence that We tried to address this gap in knowledge by analyzing we supplied will be required. the differences in gene expression profiles between normal It is concluded that BUB1B, NUSAP1, TTK, HMMR, liver tissues, cirrhotic liver tissues, and HCC tissues. We CCNA2, and KIF2C can be considered as potential novel screened several databases to get hub genes that may be molecular indicators for the onset and development of responsible for the progression of cirrhosis into HCC. We HCC, since they are linked to the transition from cirrhosis found that genes linked to mitotic nuclear division, chromo- to HCC. This study will prove important reference for trans- somal segregation, nuclear division, and organelle fission are lational medicine scientists, liver disease specialists, and bio- all intimately connected to this process. Through a series of informatics specialists. 14 Journal of Oncology 1.0 1.0 0.8 0.8 0.6 0.6 0.4 0.4 0.2 0.2 BUB1B NUSAP1 AUC: 0.961 AUC: 0.949 CI: 0.942-0.981 CI: 0.927-0.970 0.0 0.0 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 1 − specificity (FPR) 1 − specificity (FPR) (a) (b) 1.0 1.0 0.8 0.8 0.6 0.6 0.4 0.4 TTK HMMR 0.2 0.2 AUC: 0.971 AUC: 0.968 CI: 0.955-0.987 CI: 0.951-0.986 0.0 0.0 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 1 − specificity (FPR) 1 − specificity (FPR) (c) (d) 1.0 1.0 0.8 0.8 0.6 0.6 0.4 0.4 KIF2C 0.2 0.2 CCNA2 AUC: 0.981 AUC: 0.970 CI: 0.970-0.993 CI: 0.952-0.989 0.0 0.0 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 1 − specificity (FPR) 1 − specificity (FPR) (e) (f) Figure 10: Receiver operating characteristic analysis (ROC) of (a) BUB1B, (b) NUSAP1, (c) TTK, (d) HMMR, (e) CCNA2, and (f) KIF2C in HCC patients’ data (n = 424). Sensitivity (TPR) Sensitivity (TPR) Sensitivity (TPR) Sensitivity (TPR) Sensitivity (TPR) Sensitivity (TPR) Journal of Oncology 15 Purity B Cell CD8+T Cell CD4+T Cell Macrophage Neutrophil Dendritic Cell 5 cor = 0.15 partial.cor = 0.5 partial.cor = 0.326 partial.cor = 0.373 partial.cor = 0.481 partial.cor = 0.434 partial.cor = 0.476 p = 5.13e–03 p = 3.39e–23 p = 6.70e–10 p = 9.12e–13 p = 3.67e–21 p = 2.90e–17 p = 1.32e–20 –1 0.25 0.50 0.75 1.00 0.1 0.2 0.3 0.4 0.2 0.4 0.6 0.0 0.1 0.2 0.3 0.4 0.0 0.1 0.2 0.3 0.05 0.10 0.15 0.20 0.25 0.50 0.75 1.00 Infiltration Level (a) Purity B Cell CD8+T Cell CD4+T Cell Macrophage Neutrophil Dendritic Cell cor = 0.15 partial.cor = 0.489 partial.cor = 0.341 partial.cor = 0.356 partial.cor = 0.461 partial.cor = 0.415 partial.cor = 0.487 p = 5.13e–03 p = 4.14e–22 p = 8.74e–11 p = 9.80e–12 p = 2.48e–19 p = 7.84e–16 p = 1.28e–21 0.25 0.50 0.75 1.00 0.1 0.2 0.3 0.4 0.2 0.4 0.6 0.0 0.1 0.2 0.3 0.4 0.0 0.1 0.2 0.3 0.05 0.10 0.15 0.20 0.25 0.50 0.75 1.00 Infiltration Level (b) Purity B Cell CD8+T Cell CD4+T Cell Macrophage Neutrophil Dendritic Cell partial.cor = 0.464 partial.cor = 0.313 partial.cor = 0.308 partial.cor = 0.422 partial.cor = 0.355 partial.cor = 0.424 cor = 0.148 p = 5.92e–03 p = 9.19e–20 p = 3.27e–09 p = 5.53e–09 p = 3.48e–16 p = 1.16e–11 p = 2.92e–16 0.25 0.50 0.75 1.00 0.1 0.2 0.3 0.4 0.2 0.4 0.6 0.0 0.1 0.2 0.3 0.4 0.0 0.1 0.2 0.3 0.05 0.10 0.15 0.20 0.25 0.50 0.75 1.00 Infiltration Level (c) Purity B Cell CD8+T Cell CD4+T Cell Macrophage Neutrophil Dendritic Cell 5 cor = 0.166 partial.cor = 0.399 partial.cor = 0.271 partial.cor = 0.267 partial.cor = 0.351 partial.cor = 0.368 partial.cor = 0.406 p = 1.99e–03 p = 1.47e–14 p = 3.69e–07 p = 4.91e–17 p = 2.54e–11 p = 1.75e–12 p = 6.84e–15 0.25 0.50 0.75 1.00 0.1 0.2 0.3 0.4 0.2 0.4 0.6 0.0 0.1 0.2 0.3 0.4 0.0 0.1 0.2 0.3 0.05 0.10 0.15 0.20 0.25 0.50 0.75 1.00 Infiltration Level (d) Purity B Cell CD8+T Cell CD4+T Cell Macrophage Neutrophil Dendritic Cell cor = 0.152 partial.cor = 0.476 partial.cor = 0.354 partial.cor = 0.318 partial.cor = 0.402 partial.cor = 0.366 partial.cor = 0.489 p = 7.31e–21 p = 1.63e–11 p = 1.60e–09 p = 1.21e–14 p = 2.41e–12 p = 8.02e–22 p = 4.56e–03 0.25 0.50 0.75 1.00 0.1 0.2 0.3 0.4 0.2 0.4 0.6 0.0 0.1 0.2 0.3 0.4 0.0 0.1 0.2 0.3 0.05 0.10 0.15 0.20 0.25 0.50 0.75 1.00 Infiltration Level (e) Purity B Cell CD8+T Cell CD4+T Cell Macrophage Neutrophil Dendritic Cell cor = 0.123 partial.cor = 0.482 partial.cor = 0.363 partial.cor = 0.275 partial.cor = 0.449 partial.cor = 0.362 partial.cor = 0.47 p = 2.09e–21 p = 4.31e–12 p = 2.12e–07 p = 2.73e–18 p = 4.21e–12 p = 4.66e–29 6 p = 2.19e–02 0.25 0.50 0.75 1.00 0.1 0.2 0.3 0.4 0.2 0.4 0.6 0.0 0.1 0.2 0.3 0.4 0.0 0.1 0.2 0.3 0.05 0.10 0.15 0.20 0.25 0.50 0.75 1.00 Infiltration Level (f) Figure 11: Relationships between immune infiltration in HCC tissues and expression levels of (a) BUB1B, (b) NUSAP1, (c) TTK, (d) HMMR, (e) CCNA2, and (f) KIF2C from the TIMER database. CCNA2 Expression Level (log2 TPM) HMMR Expression Level (log2 TPM) KIF2C Expression Level (log2 TPM) TTK Expression Level (log2 TPM) NUSAP1 Expression Level (log2 TPM) BUB1B Expression Level (log2 TPM) LIHC LIHC LIHC LIHC LIHC LIHC 16 Journal of Oncology [10] S. Davis and P. S. Meltzer, “GEOquery: a bridge between the Data Availability Gene Expression Omnibus (GEO) and BioConductor,” Bioin- The datasets analyzed during the current study are available formatics, vol. 23, no. 14, pp. 1846-1847, 2007. in TCGA (https://portal.gdc.cancer.gov/) and GEO reposi- [11] G. Yu, L. G. Wang, Y. Han, and Q. Y. He, “clusterProfiler: an R tory (https://www.ncbi.nlm.nih.gov/geo/). package for comparing biological themes among gene clus- ters,” OMICS, vol. 16, no. 5, pp. 284–287, 2012. [12] D. Szklarczyk, A. Franceschini, S. Wyder et al., “STRING v10: Conflicts of Interest protein-protein interaction networks, integrated over the tree of life,” Nucleic Acids Research, vol. 43, no. D1, pp. D447– The authors declare no competing interests. D452, 2015. [13] C. Zhang, W. Zhang, H. Cui et al., “Role of hub genes in the occurrence and development of testicular cancer based on bio- Authors’ Contributions informatics,” International Journal of General Medicine, vol. 15, pp. 645–660, 2022. Yuanbin Chen and Hongyan Qian are equal first authors. [14] Z. Zhang, E. Lin, H. Zhuang et al., “Construction of a novel gene-based model for prognosis prediction of clear cell renal cell Acknowledgments carcinoma,” Cancer Cell International, vol. 20, no. 1, p. 27, 2020. [15] W. Lin, S. Wu, X. Chen et al., “Characterization of hypoxia sig- This study was supported by the Nantong Science and Tech- nature to evaluate the tumor immune microenvironment and nology Bureau (grant no. JC2020025), the Health Commit- predict prognosis in glioma groups,” Frontiers in Oncology, tee of Nantong (grant no. MB2020021), and the Clinical vol. 10, p. 796, 2020. Medicine Special Program from Nantong University (grant [16] R. Tibshirani, “The lasso method for variable selection in the Cox no. 2019JY017). This work was also supported by Scientific model,” Statistics in Medicine,vol.16,no.4,pp. 385–395, 1997. Research Project of Health Commission of Jiangsu Province 6 [17] F. Xu, X. Huang, Y. Li, Y. Chen, and L. Lin, “m A-related (grant no. M2020009) lncRNAs are potential biomarkers for predicting prognoses and immune responses in patients with LUAD,” Molecular Therapy-Nucleic Acids, vol. 24, pp. 780–791, 2021. References [18] Y. Ji and Y. Xue, “Identification and clinical validation of 4- lncRNA signature for predicting survival in head and neck [1] W. Shi, L. Feng, S. Dong et al., “FBXL6 governs c-MYC to pro- squamous cell carcinoma,” Oncotargets and Therapy, vol. 13, mote hepatocellular carcinoma through ubiquitination and pp. 8395–8411, 2020. stabilization of HSP90AA1,” Cell Communication and Signal- [19] J. Wang, J. Sun, L. N. Liu et al., “Siglec-15 as an immune sup- ing: CCS, vol. 18, no. 1, p. 100, 2020. pressor and potential target for normalization cancer immu- [2] H. Nordenstedt, D. L. White, and H. B. El-Serag, “The chang- notherapy,” Nature Medicine, vol. 25, no. 4, pp. 656–666, 2019. ing pattern of epidemiology in hepatocellular carcinoma,” [20] R. Ravi, K. A. Noonan, V. Pham et al., “Bifunctional immune Digestive and Liver Disease, vol. 42, Supplement 3, pp. S206– checkpoint-targeted antibody-ligand traps that simultaneously S214, 2010. disable TGFβ enhance the efficacy of cancer immunotherapy,” [3] L. Kulik and H. B. El-Serag, “Epidemiology and management Nature Communications, vol. 9, no. 1, p. 741, 2018. of hepatocellular carcinoma,” Gastroenterology, vol. 156, [21] N. Wen, Y. Cai, F. Li et al., “The clinical management of hepa- no. 2, article e471, pp. 477–491.e1, 2019. tocellular carcinoma worldwide: a concise review and compar- [4] G. V. Papatheodoridis, P. Lampertico, S. Manolakopoulos, and ison of current guidelines: 2022 update,” Bioscience Trends, A. Lok, “Incidence of hepatocellular carcinoma in chronic vol. 16, no. 1, pp. 20–30, 2022. hepatitis B patients receiving nucleos(t)ide therapy: a system- atic review,” Journal of Hepatology, vol. 53, no. 2, pp. 348– [22] M. Walker, H. B. El-Serag, Y. Sada et al., “Cirrhosis is under- recognised in patients subsequently diagnosed with hepatocel- 356, 2010. lular cancer,” Alimentary Pharmacology & Therapeutics, [5] W. L. Tsai and R. T. Chung, “Viral hepatocarcinogenesis,” vol. 43, no. 5, pp. 621–630, 2016. Oncogene, vol. 29, no. 16, pp. 2309–2324, 2010. [23] L. A. Beste, S. L. Leipertz, P. K. Green, J. A. Dominitz, D. Ross, [6] H. Li, “Intercellular crosstalk of liver sinusoidal endothelial and G. N. Ioannou, “Trends in burden of cirrhosis and hepato- cells in liver fibrosis, cirrhosis and hepatocellular carcinoma,” cellular carcinoma by underlying liver disease in US veterans, Digestive and Liver Disease, vol. 54, no. 5, pp. 598–613, 2022. 2001-2013,” Gastroenterology, vol. 149, no. 6, pp. 1471– [7] G. Yildiz, A. Arslan-Ergul, S. Bagislar et al., “Genome-wide 1482.e5, 2015. transcriptional reorganization associated with senescence-to- [24] H. B. El-Serag, “Hepatocellular carcinoma,” The New England immortality switch during human hepatocellular carcinogene- Journal of Medicine, vol. 365, no. 12, pp. 1118–1127, 2011. sis,” PLoS One, vol. 8, no. 5, article e64016, 2013. [25] S. A. Sharma, M. Kowgier, B. E. Hansen et al., “Toronto HCC [8] X. Zhang, W. Zhang, Y. Jiang, K. Liu, L. Ran, and F. Song, “Identification of functional lncRNAs in gastric cancer by inte- risk index: a validated scoring system to predict 10-year risk of HCC in patients with cirrhosis,” Journal of hepatology, vol. 68, grative analysis of GEO and TCGA data,” Journal of Cellular Biochemistry, vol. 120, no. 10, pp. 17898–17911, 2019. no. 1, pp. 92–99, 2018. [9] J. Sun, J. Huang, J. Lan et al., “Overexpression of CENPF cor- [26] T. Rio Frio, J. Lavoie, N. Hamel et al., “Homozygous BUB1B relates with poor prognosis and tumor bone metastasis in mutation and susceptibility to gastrointestinal neoplasia,” breast cancer,” Cancer Cell International, vol. 19, no. 1, The New England Journal of Medicine, vol. 363, no. 27, p. 264, 2019. pp. 2628–2637, 2010. Journal of Oncology 17 [27] X. Tang, M. Guo, P. Ding et al., “BUB1B and circBUB1B_ 544aa aggravate multiple myeloma malignancy through evok- ing chromosomal instability,” Signal Transduction and Tar- geted Therapy, vol. 6, no. 1, p. 361, 2021. [28] Y. Zhao, J. He, Y. Li, S. Lv, and H. Cui, “NUSAP1 potentiates chemoresistance in glioblastoma through its SAP domain to stabilize ATR,” Signal Transduction and Targeted Therapy, vol. 5, no. 1, p. 44, 2020. [29] F. Chen, E. P. Szymanski, K. N. Olivier et al., “Whole-exome sequencing identifies the 6q12-q16 linkage region and a candi- date gene, TTK, for pulmonary nontuberculous mycobacterial disease,” American Journal of Respiratory and Critical Care Medicine, vol. 196, no. 12, pp. 1599–1604, 2017. [30] F. Mateo, Z. He, L. Mei et al., “Modification of BRCA1- associated breast cancer risk by HMMR overexpression,” Nature Communications, vol. 13, no. 1, p. 1895, 2022. [31] Y. T. Chiu, J. K. Wong, S. W. Choi et al., “Novel pre-mRNA splicing of intronically integrated HBV generates oncogenic chimera in hepatocellular carcinoma,” Journal of Hepatology, vol. 64, no. 6, pp. 1256–1264, 2016. [32] S. Wei, M. Dai, C. Zhang et al., “KIF2C: a novel link between Wnt/β-catenin and mTORC1 signaling in the pathogenesis of hepatocellular carcinoma,” Protein & Cell, vol. 12, no. 10, pp. 788–809, 2021. [33] Y. Zheng, Y. Luo, X. Chen et al., “The role of mRNA in the development, diagnosis, treatment and prognosis of neural tumors,” Molecular Cancer, vol. 20, no. 1, p. 49, 2021. [34] E. Wieczorek and E. Reszka, “mRNA, microRNA and lncRNA as novel bladder tumor markers,” Clinica Chimica Acta, vol. 477, pp. 141–153, 2018. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Oncology Hindawi Publishing Corporation

Screening of the Key Genes for the Progression of Liver Cirrhosis to Hepatocellular Carcinoma Based on Bioinformatics

Screening of the Key Genes for the Progression of Liver Cirrhosis to Hepatocellular Carcinoma Based on Bioinformatics

Abstract

Hepatocellular carcinoma (HCC), which is among the most globally prevalent cancers, is strongly associated with liver cirrhosis. Using a bioinformatics approach, we have identified and investigated the hub genes responsible for the progression of cirrhosis into HCC. We analyzed the Gene Expression Omnibus (GEO) microarray datasets, GSE25097 and GSE17549, to identify differentially expressed genes (DEGs) in these two conditions and also performed protein-protein interaction (PPI) network analysis. STRING database and Cytoscape software were used to analyze the modules and locate hub genes following which the connections between hub genes and the transition from cirrhosis to HCC, progression of HCC, and prognosis of HCC were investigated. We used the Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis to detect the molecular mechanisms underlying the action of the primary hub genes. In all, 239 DEGs were obtained, with 94 of them showing evidence of upregulation and 145 showing evidence of downregulation in HCC tissues as compared to cirrhotic liver tissues. We identified six hub genes, namely, BUB1B, NUSAP1, TTK, HMMR, CCNA2, and KIF2C, which were upregulated and had a high diagnostic value for HCC. Besides, these six hub genes were positively related to immune cell infiltration. Since these genes may play a direct role in the progression of cirrhosis to HCC, they can be considered as potential novel molecular indicators for the onset and development of HCC.

Loading next page...
 
/lp/hindawi-publishing-corporation/screening-of-the-key-genes-for-the-progression-of-liver-cirrhosis-to-1fPGgXboAV
Publisher
Hindawi Publishing Corporation
ISSN
1687-8450
eISSN
1687-8469
DOI
10.1155/2022/2515513
Publisher site
See Article on Publisher Site

Abstract

Hindawi Journal of Oncology Volume 2022, Article ID 2515513, 17 pages https://doi.org/10.1155/2022/2515513 Research Article Screening of the Key Genes for the Progression of Liver Cirrhosis to Hepatocellular Carcinoma Based on Bioinformatics 1 2 1 1 1 2 Yuanbin Chen, Hongyan Qian, Xiao He , Jing Zhang, Song Xue, Yumeng Wu , 3 4 5 Jian Chen , Xuming Wu , and Suqing Zhang Medical School of Nantong University, Jiangsu, Nantong 226000, China Key Laboratory of Cancer Research Center Nantong, Affiliated Tumor Hospital of Nantong University, Jiangsu, Nantong 226361, China The Immunology Laboratory, Affiliated Tumor Hospital of Nantong University, Jiangsu, Nantong 226361, China Nantong Fourth People’s Hospital, Jiangsu Nantong 226001, China Department of Hepatobiliary and Pancreatic Surgery, Affiliated Tumor Hospital of Nantong University, Jiangsu, Nantong 226361, China Correspondence should be addressed to Jian Chen; cjsmx@sina.com, Xuming Wu; cxd050609@126.com, and Suqing Zhang; zsq7829@163.com Received 5 August 2022; Revised 22 August 2022; Accepted 30 August 2022; Published 26 September 2022 Academic Editor: Mingjun Zheng Copyright © 2022 Yuanbin Chen et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Hepatocellular carcinoma (HCC), which is among the most globally prevalent cancers, is strongly associated with liver cirrhosis. Using a bioinformatics approach, we have identified and investigated the hub genes responsible for the progression of cirrhosis into HCC. We analyzed the Gene Expression Omnibus (GEO) microarray datasets, GSE25097 and GSE17549, to identify differentially expressed genes (DEGs) in these two conditions and also performed protein-protein interaction (PPI) network analysis. STRING database and Cytoscape software were used to analyze the modules and locate hub genes following which the connections between hub genes and the transition from cirrhosis to HCC, progression of HCC, and prognosis of HCC were investigated. We used the Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis to detect the molecular mechanisms underlying the action of the primary hub genes. In all, 239 DEGs were obtained, with 94 of them showing evidence of upregulation and 145 showing evidence of downregulation in HCC tissues as compared to cirrhotic liver tissues. We identified six hub genes, namely, BUB1B, NUSAP1, TTK, HMMR, CCNA2, and KIF2C, which were upregulated and had a high diagnostic value for HCC. Besides, these six hub genes were positively related to immune cell infiltration. Since these genes may play a direct role in the progression of cirrhosis to HCC, they can be considered as potential novel molecular indicators for the onset and development of HCC. 1. Introduction raises the incidence rate of hepatic sclerosis to 84.6%, which in turn, raises the incidence of HCC to 49.9% [4]. To successfully prevent, diagnose, and treat HCC, it is Global rates of morbidity and death caused by hepatocellular crucial to understand how liver cirrhosis transforms into carcinoma (HCC), which is among the most common can- cers in the world and the second most lethal, are on the rise HCC. Many studies have showed that capillarization of liver sinusoidal endothelial cells, portal hypertension, immuno- [1]. Currently, both hepatitis B virus (HBV) and hepatitis C suppressive tumor microenvironment, etc., were important virus (HCV) have been identified as the most important factors promoting the development from liver cirrhosis to cause of HCC [2, 3]. HCC is most likely to occur in patients HCC [5, 6]. However, mitigating these factors did not with severe HBV infections, especially in those who suffer from posthepatitis cirrhosis. Posthepatitis cirrhosis also change the progression of the disease. And currently, there 2 Journal of Oncology CX CXCL6 CL6 MMP7 MMP7 PRO PROM M1 1 KC KCN NJ J1 16 6 SF SFR RP P5 5 60 CFTR CFTR CHS CHST4 T4 CLD CLDN10 N10 CLI CLIC6 C6 ANX ANXA3 A3 FGF23 FGF23 MAR MARC CO O FCN2 FCN2 GPM6A GPM6A CRHBP CRHBP CLEC1B CLEC1B FCN3 FCN3 IG IGJ J RS RSPO3 PO3 F FAM180A AM180A ES ESM M1 1 THBS4 THBS4 CYP7A1 CYP7A1 HO HOX XA A1 13 3 DU DUSP SP9 9 IG IGF2BP1 F2BP1 AFP AFP GP GPR158 R158 –1 C CO OX7B2 X7B2 S SL LC44A5 C44A5 PC PCO OL LC CE2 E2 0 FO FOX XN4 N4 FBX FBXO4 O43 3 AS ASPM PM NEK2 NEK2 NUF2 NUF2 –6 –3 0 36 SKA SKA1 1 KIF20A KIF20A –2 ZI ZIC2 C2 Log (Fold Change) I IGF2BP GF2BP3 3 Group group 1 group 2 (a) (b) C7 C7 C9 C9 CX CXCL14 CL14 T TMEM27 MEM27 CRHBP CRHBP FCN2 FCN2 CLEC4G CLEC4G FCN3 FCN3 OI OIT T3 3 SLC SLCO O1 1B B3 3 GREM2 GREM2 2 LP LPA A S SL LC22A1 C22A1 GB GBA3 A3 APO APOF F PGL PGLYRP2 YRP2 2 TENM1 TENM1 THRS THRSP P LIN LINC CO1093 O1093 CYP1A2 CYP1A2 AKRIB10 AKRIB10 SU SUL LT T1 1C C2 2 MA MAGEA6 GEA6 TOP TOP2 2A A ND NDC80 C80 RRM2 RRM2 CD CDK1 K1 PBK PBK C CCNB1 CNB1 –2 AS ASP PM M NEK2 NEK2 TT TTK K CENPF CENPF NUF2 NUF2 B BUB1B UB1B KIF20A KIF20A N NCAPG CAPG –2 0 2 DK DKK K1 1 S SP PINK1 INK1 –4 Log (Fold Change) ZI ZIC2 C2 Group group 1 group 2 (c) (d) Figure 1: Differentially expressed genes (DEGs) between HCC and cirrhotic liver tissues identified in GSE25097 and GSE17549 datasets. (a) A total of 627 genes were found to be elevated and 1716 genes depressed in HCC tissues as compared to cirrhotic liver tissues in the GSE25097 dataset. (b) The expression profiles of each of the top 20 DEGs identified from the GSE25097 dataset. (c) A total of 149 genes were found to be elevated and 285 genes depressed in HCC tissues as compared to cirrhotic liver tissues in the GSE17548 dataset. (d) The expression profiles of each of the top 20 DEGs in the GSE17548 dataset. are no ideal molecular markers that can help in distinguish- and prognosis, respectively. The Gene Ontology (GO) ing HCC from cirrhosis. To close this gap, the molecular and Kyoto Encyclopedia of Genes and Genomes (KEGG) detected the top different biological events and signaling aspects of HCC incidence, progress, and reasons for poor prognosis need to be further understood. pathways in elevated and depressed DEGs. The LASSO In this study, we analyzed two mRNA microarrays Cox regression model screened the highest predictive value from the GEO database to identify differentially expressed markers of HCC prognosis. The receiver operating charac- genes (DEGs) that vary in expression levels between HCC teristic (ROC) curve and immunoinfiltration analysis were tissues and cirrhotic liver tissues. Following this, protein- employed to analyze the role of these hub genes for HCC. protein interaction (PPI) network analyses and Kaplan- Based on these methods, we identified several genes that Meier curves investigated the connections between the could function as molecular markers to track the onset identified genes and those between identified hub genes and progression of HCC. –Log (P.adj) –Log (P.adj) 10 Journal of Oncology 3 GSE25097 GSE17548 GSE25097 GSE17548 533 94 55 1571 145 140 (a) (b) Figure 2: Identification of common genes from DEGs in the GSE25097 and GSE17548 datasets. (a) 94 genes were common upregulated, and (b) 145 genes were common downregulated. IGJ TIMD4 GPM6A UROC1 PCDH9 CLEC4G CD1D NR4A3 CD5L CLEC1B CCL21 COL4A4 NPY1R CNTN4 COLEC10 CXCL14 COL4A3 FCN2 PZP CCL19 DPT COLEC11 GPC3 HHIP ADAMTSL2 MCM8 IGF2BP3 ADAMTS13 CKAP2 ANK3 C7 FCN3 RAD51AP1NUSAP1 CFP CXCL12 CXCL2 SHBG CEP55 TSLP TRIM71 MBL2 ZWINT DUSP5 LIFR C9 ANLN KLF4 PTTG1 KDK CDC6 PLCXD3 EGR1 TRIP13 RRM2 C6 CCDC34 DCN KIF15 EPO PDGFRA AURKA KIF11 HGFAC KIF20A HGF IGF1 NUF2 UBE2C NEIL3 GYS2 BUB1B F9 EPHA2 CCNA2 CENPK NGFR C1orf112 CCNB2 CYR61 HMMR KMO SDS MCM2 SPC25 PSPH SOCS2 KIF2C HJURP ESR1 HPGD TTK SLC7A11 PRC1 CENPF GLS2 CDK1 RND3 CENPE ADH4 IDO2 CNDP1 KIF4A MAD2L1 CYP2C8 PTGS2 SHCBP1 PRKAR2B EZH2 AADAT IGFALS STIL MELK ASPA CYP1A2 FGFR2 PBK SLC22A1 CDKN2C DEPDC1B CCNE2 INMT ATAD2 NDC80 TOP2A CYP2E1 MCM10 HELLS BRIP1 TPX2 MND1 CAP2 DLGAP5 SGOL2 RBMS3 GNA14 NEK2 CCNB1 ASPM OIP5 NAT2 DTL CDC20 SLCO1B3 RBM24 FOXM1 CASC5 NCAPG OIT3 SPINK1 CYP8B1 CHRM3 ESM1 E2F8 CDKN3 FAM83D GPSM2 KIF14 CDCA5 RACGAP1 UBE2T SQLE PHLDA1 CYP39A1 MAGEA1 LHX2 FAM110C CD14 C8orf4 ANGPTL6 MT1X LCAT CETP MT1F APOF MT1G MAGEA12 ZIC2 FAM150B MARCO CLRN3 LECT2 MT1M (a) MELK NUSAP1 CENPF CCNB2 PBK KIF2C HMMR MAD2L1 CCNA2 TPX2 (b) Figure 3: Screen for hub genes. (a) PPI network. (b) Top 10 hub genes were identified by CytoHubba. 2. Materials and Methods GPL10687 platforms, respectively, was identified to screen for genes associated with liver cirrhosis and HCC [7]. The 2.1. Microarray Data. The GEO database, specifically, the MINiML files, which contained raw data, including those for GSE17548 and GSE25097 series based on GPL570 and all of the platforms, samples, and GSE records, were obtained, 4 Journal of Oncology GO GO p.adjust p.adjust tetrapyrrole binding tubulin binding heme binding microtubule binding 0.00015 steroid hydroxylase activity microtubule motor activity cyclin−dependent protein serine/threonine kinase mannose binding 0.02 regulator activity collagen−containing extracellular matrix spindle collagen trimer 0.00010 chromosomal region high−density lipoprotein particle chromosome, centromeric region pore complex condensed chromosome, centromeric region 0.01 protein kinase B signaling organelle fission 0.00005 protein activation cascade nuclear division complement activation chromosome segregation complement activation, lectin pathway mitotic nuclear division 0.02 0.04 0.06 0.08 0.10 0.12 0.1 0.2 0.3 0.4 GeneRatio GeneRatio Counts Counts (a) (b) KEGG p.adjust KEGG p.adjust 0.04 Cell cycle Amoebiasis 7.5e − 05 Human T−cell leukemia virus 1 infection NF−kappa B signaling pathway 0.03 5.0e − 05 Oocyte meiosis Chemical carcinogenesis 0.02 Progesterone−mediated oocyte maturation Tryptophan metabolism 2.5e − 05 0.01 p53 signaling pathway Histidine metabolism 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.04 0.06 0.08 0.10 GeneRatio GeneRatio Counts Counts 8 6 (c) (d) Figure 4: Functional enrichment analysis of hub genes in HCC. (a) GO analysis of DEGs in high expression samples. (b) GO analysis of DEGs in low expression samples. (c) KEGG analysis of DEGs in high expression samples. (d) KEGG analysis of DEGs in low expression samples. and the extracted data were log transformed for standardiza- gram. GEO2R, as a tool for interactive network, provides tion. Using preprocessCore, we normalized the data using users with the ability to compare two or more datasets that the median method. The annotated information included in are part of the GEO series to find DEGs [10]. The thresholds for statistical significance were set to log jfold changej >1 the platform was used to convert probes to gene symbols, which were then used in the normalization process. Probes and an adjusted p value of < 0.05. that matched more than one gene were excluded from these datasets. Several probes were utilized to detect the expression 2.3. Enrichment Analysis of DEGs. GO and KEGG databases value of each gene, and an average value from these was were utilized as references, and the “clusterProfiler” R pack- obtained. To eliminate the confounding effects of different age carried out an analysis of enrichment [11]. To correct for batches, we used the “removeBatchEffect” function of the multiple comparisons, the Benjamini–Hochberg approach “limma” package in R. Boxplots were used to analyze the was utilized, with a false discovery rate ðFDRÞ <0:05 indicat- cleaned datasets. A PCA plot was constructed to demonstrate ing statistical significance. the differences in the datasets before and after the removal of the batch effects [8, 9]. 2.4. Screening of Hub Genes. The STRING database (https:// www.string-db.org/) was utilized to get a PPI network, with 2.2. Identification of DEGs. DEGs between HCC tissues and a score of 0.4 or higher for minimum participation in inter- liver cirrhotic tissues were identified through GEO2R pro- actions [12]. The “Hubba” plug-in included in the Cytoscape Journal of Oncology 5 1.0 1.0 0.8 0.8 0.6 0.6 0.4 0.4 0.2 0.2 Overall Survival Overall Survival HR = 1.56 (1.10-2.21) HR = 1.74 (1.23-2.47) 0.0 P = 0.013 0.0 P = 0.002 0 30 60 90 120 030 60 90 120 Time (months) Time (months) BUB1B MELK Low Low High High (a) (b) 1.0 1.0 0.8 0.8 0.6 0.6 0.4 0.4 0.2 0.2 Overall Survival Overall Survival HR = 1.53 (1.08-2.16) HR = 1.52 (1.07-2.15) P = 0.016 P = 0.018 0.0 0.0 030 60 90 120 0 30 60 90 120 Time (months) Time (months) NUSAP1 MAD2L1 Low Low High High (c) (d) 1.0 1.0 0.8 0.8 0.6 0.6 0.4 0.4 0.2 0.2 Overall Survival Overall Survival HR = 1.70 (1.20-2.41) HR = 1.65 (1.16-2.33) 0.0 P = 0.003 0.0 P = 0.005 030 60 90 120 030 60 90 120 Time (months) Time (months) RRM2 TTK Low Low High High (e) (f) Figure 5: Continued. Survival probability Survival probability Survival probability Survival probability Survival probability Survival probability 6 Journal of Oncology 1.0 0.8 0.6 0.4 0.2 Overall Survival HR = 2.02 (1.42-2.87) 0.0 P < 0.001 0 30 60 90 120 Time (months) HMMR Low High (g) 1.0 0.8 0.6 0.4 0.2 Overall Survival HR = 1.65 (1.16-2.33) 0.0 P = 0.005 030 60 90 120 Time (months) CCNA2 Low High (h) 1.0 0.8 0.6 0.4 0.2 Overall Survival HR = 2.16 (1.51-3.08) P < 0.001 0.0 030 60 90 120 Time (months) KIF2C Low High (i) Figure 5: The prognostic potentials of the genes (a) BUB1B, (b) MELK, (c) MAD2L1, (d) NUSAP1, (e) RRM2, (f) TTK, (g) HMMR, (h) CCNA2, and (i) KIF2C were investigated. Patients diagnosed with HCC having higher levels of expression of these genes had lower overall survival statistics as compared to patients with lower levels of expression of these genes (logrank test, p <0:05). Based on the Cox pH model, HR was determined, and the 95% CI was shown as a dotted line. Survival probability Survival probability Survival probability Journal of Oncology 7 1.0 1.0 0.8 0.8 0.6 0.6 0.4 0.4 0.2 0.2 Progress Free Interval Progress Free Interval HR = 1.70 (1.26-2.27) HR = 1.74 (1.30-2.33) 0.0 P < 0.001 0.0 P < 0.001 0 30 60 90 120 030 60 90 120 Time (months) Time (months) BUB1B MELK Low Low High High (a) (b) 1.0 1.0 0.8 0.8 0.6 0.6 0.4 0.4 0.2 0.2 Progress Free Interval Progress Free Interval HR = 1.51 (1.13-2.02) HR = 1.71 (1.28-2.30) P = 0.006 0.0 0.0 P < 0.001 030 60 90 120 0 30 60 90 120 Time (months) Time (months) MAD2L1 CCNB2 Low Low High High (c) (d) 1.0 1.0 0.8 0.8 0.6 0.6 0.4 0.4 0.2 0.2 Progress Free Interval Progress Free Interval HR = 1.67 (1.24-2.23) HR = 1.64 (1.22-2.19) P = 0.001 0.0 0.0 P = 0.001 030 60 90 120 030 60 90 120 Time (months) Time (months) NUSAP1 RRM2 Low Low High High (e) (f) 1.0 1.0 0.8 0.8 0.6 0.6 0.4 0.4 0.2 0.2 Progress Free Interval Progress Free Interval HR = 1.69 (1.26-2.26) HR = 1.78 (1.33-2.39) 0.0 P < 0.001 0.0 P < 0.001 0 30 60 90 120 030 60 90 120 Time (months) Time (months) TTK HMMR Low Low High High (g) (h) Figure 6: Continued. Survival probability Survival probability Survival probability Survival probability Survival probability Survival probability Survival probability Survival probability 8 Journal of Oncology 1.0 1.0 0.8 0.8 0.6 0.6 0.4 0.4 0.2 0.2 Progress Free Interval Progress Free Interval HR = 1.69 (1.26-2.26) HR = 2.01 (1.50-2.70) 0.0 P < 0.001 0.0 P < 0.001 030 60 90 120 0 30 60 90 120 Time (months) Time (months) CCNA2 KIF2C Low Low High High (i) (j) Figure 6: The prognostic potentials of the genes (a) BUB1B, (b) MELK, (c) MAD2L1, (d) CCNB2, (e) NUSAP1, (f) RRM2, (g) TTK, (h) HMMR, (i) CCNA2, and (j) KIF2C were investigated. Patients diagnosed with HCC having higher levels of expression of these genes had lower cancer-free intervals as compared to patients with lower expression levels of these genes (logrank test, p <0:05). Based on the Cox pH model, HR was determined, and the 95% CI was shown as a dotted line. program was used to identify and choose the top 10 hub 3. Results nodes listed by degree [13]. 3.1. Screening for Differentially Expressed Genes. Two data- sets from the GEO database, namely, GSE25097 and 2.5. Survival Analysis. The raw RNA-sequencing data and GSE17548, were used to identify DEGs between cirrhotic liver accompanying clinical information were obtained from the and HCC tissues. We identified 2343 DEGs in GSE25097 Cancer Genome Atlas (TCGA) database. Log-rank tests dataset, of which 627 were elevated and 1716 were depressed obtained p values, hazard ratios (HR), and 95% confidence in HCC tissues when compared to cirrhotic liver tissues intervals (CI) for the two groups (cirrhotic liver tissues and (Figure 1(a)). In the GSE17548 dataset, we identified 434 HCC tissues). These results were then used to plot Kaplan- DEGs, of which 149 were upregulated and 285 were downreg- Meier (KM) survival analysis to assess distinctions in sur- ulated in HCC tissues when compared to cirrhotic liver tissues vival between the cirrhotic liver and HCC tissue groups (Figure 1(c)). The heatmap shows the expression levels of each [14, 15]. of the top 20 DEGs (Figures 1(b) and 1(d)). 3.2. Screening for Hub Genes. To further analyze the com- 2.6. Construction of Prognostic Signatures. To investigate the mon genes in the two datasets, Venn diagram was employed potential diagnostic utility of the hub genes identified, we to find 94 common upregulated genes and 145 common carried out least absolute shrinkage and selection opera- downregulated genes in HCC tissues as compared to cir- tor- (LASSO-) penalized Cox regression analysis [16]. rhotic liver tissues (Figures 2(a) and 2(b)). Using STRING The “glmnet” package in R package was used to develop and Cytoscape, we analyzed these 239 DEGs to identify a model for prognosis. A LASSO regression was carried those with interaction scores > 0:4. PPI network obtained a out with the assistance of a cross-validation of 10 folds, total of 183 nodes and 2193 edges (Figure 3(a)). CytoHubba with the penalty parameter (λ) adjusted to fulfil the was utilized to get the top 10 hub genes, namely, BUB1B, optimal value. Findings of the LASSO regression were MELK, MAD2L1, CCNB2, NUSAP1, RRM2, TTK, HMMR, used as the basis for calculating risk ratings. Patients CCNA2, and KIF2C (Figure 3(b)). diagnosed with HCC who participated in the TCGA study were grouped into low-risk and high-risk categories 3.3. GO and KEGG Enrichment Analyses of DEGs. To fur- based on the median risk score. A KM survival analysis ther examine the biological roles of the identified DEGs, was carried out to evaluate and contrast the variations we used the “clusterProfiler” package in R for GO and in overall survival (OS) that were observed in the two KEGG pathway enrichment analyses. The results of the groups [14, 17, 18]. GO analysis of upregulated DEGs indicated that this group contained genes related to biological processes 2.7. Immunoinfiltration Analysis. The immunogene module (including mitotic nuclear division, chromosome segrega- of the TIMER tool (https://cistrome.shinyapps.io/timer/) tion, nuclear division, and organelle fission), cellular was used to analyze correlations between the expression of components (including spindle, chromosomal region, hub genes and immunological infiltration (including infiltra- chromosome, centromeric region, and condensed chromo- tion levels of B cell, CD4 + T cell, CD8 + T cell, macrophage, some), and molecular functions (including tubulin bind- neutrophil, and dendritic cell) in HCC tissues from the ing, microtubule binding, microtubule motor activity, and TCGA [19, 20]. cyclin-dependent protein serine/threonine kinase regulator Survival probability Survival probability Journal of Oncology 9 6 ⁎⁎⁎ ⁎⁎⁎ 1 1 Normal Tumor Normal Tumor (a) (b) ⁎⁎⁎ 8 5 ⁎⁎⁎ 1 2 Normal Tumor Normal Tumor (c) (d) ⁎⁎⁎ ⁎⁎⁎ 8 5 Normal Tumor Normal Tumor (e) (f) ⁎⁎⁎ ⁎⁎⁎ 0 0 Normal Tumor Normal Tumor (g) (h) Figure 7: Continued. e expression of HMMR e expression of MAD2L1 e expression of BUB1B e expression of RRM2 Log (TPM+1) Log (TPM+1) Log (TPM+1) Log (TPM+1) 2 2 2 e expression of CCNA2 e expression of TTK e expression of NUSAP1 e expression of MELK Log (TPM+1) Log (TPM+1) Log (TPM+1) Log (TPM+1) 2 2 2 2 10 Journal of Oncology ⁎⁎⁎ Normal Tumor (i) Figure 7: Differential expression analysis of hub genes in TCGA dataset consisting of HCC tissue samples and normal tissue samples. The expression levels of (a) BUB1B, (b) MELK, (c) MAD2L1, (d) NUSAP1, (e) RRM2, (f) TTK, (g) HMMR, (h) CCNA2, and (i) KIF2C in HCC tissues (n = 374) were significantly higher than those in normal tissues (n =50). activity) (Figure 4(a)). Analysis of the downregulated HMMR, CCNA2, and KIF2C) were shown to have high pre- DEGs indicates that this group contains genes linked to dictive value for HCC prognosis. Patients diagnosed with biological processes (including complement activation, lec- HCC were split into two categories according to their risk scores. Figure 9(c) depicts the distributions of risk scores, tin pathway, complement activation, protein activation cascade, and protein kinase B signaling), cellular compo- survival statuses, and expression levels of these six genes in nents (including pore complex, high-density lipoprotein the patient population (Figure 9(c)). particle, collagen trimer, and collagen-containing extracel- In TCGA, the data on 374 HCC samples with detailed lular matrix), and molecular functions (including mannose clinicopathological information (Table 1) were evaluated for clinically relevant markers. These hub genes were mea- binding, steroid hydroxylase activity, heme binding, and tetrapyrrole binding) (Figure 4(b)). In addition, KEGG sured at mRNA levels in HCC tissues and normal tissues, analysis revealed that the elevated DEGs were intimately as well as the data was used to generate ROC curve. Our connected to the cell cycle, human T-cell leukemia virus results indicated that BUB1B, NUSAP1, TTK, HMMR, 1 infection, oocyte meiosis, progesterone-mediated oocyte CCNA2, and KIF2C were all upregulated in HCC at the mRNA levels. And the six hub genes had a high diagnostic maturation, and the p53 signaling pathway (Figure 4(c)). The downregulated DEGs were connected to amoebiasis, value, with AUCs of 0.961, 0.949, 0.971, 0.968, 0.970, and NF-kappa B signaling pathway, chemical carcinogenesis, 0.981, respectively (Figure 10). tryptophan metabolism, and histidine metabolism (Figure 4(d)). 3.6. Relationship between Hub Gene Expression and the Infiltration of Immune Cells. It has been shown that 3.4. Relationship between HCC Prognosis and Expression of tumor-associated fibroblasts in the stroma of the tumor Hub Genes. A univariate Cox regression analysis was carried microenvironment may affect a wide range of immune cells out to identify which hub genes were linked to HCC progno- that infiltrate the tumor. The effects of the hub genes identi- sis. We find that nine of the 10 identified hub genes (BUB1B, fied here on the recruitment of immune cells in the tumor MELK, MAD2L1, NUSAP1, RRM2, TTK, HMMR, CCNA2, microenvironment and hence on the prognosis of HCC are and KIF2C) showed prognostic significance (Figures 5 and as yet unknown. To investigate this, we analyzed the connec- 6). The expression profiles of these nine genes were then tions between BUB1B, NUSAP1, TTK, HMMR, CCNA2, evaluated in 374 HCC tissue samples and 50 normal liver tis- and KIF2C with immune infiltration in HCC and found that sue samples obtained from the TCGA database. Our findings the expression levels of them were positively associated with indicated that the expression levels of these nine hub genes the immune infiltration level of immune cells (Figure 11). in HCC tissues were significantly higher than those in nor- mal tissues (Figures 7 and 8). 4. Discussion 3.5. Construction of Prognostic Signatures of Hub Genes in HCC. The LASSO Cox regression model was utilized to Globally, HCC is the second deadliest and fifth most com- choose genes with the highest predictive value as potential monly occurring cancer [21]. The disease progression is quick markers of HCC prognosis. The value (λ =0:0088) was with malignancy at a high level, which combined with low detected because it was the lowest when compared to the incidences of early detection, usually points to a bad prognosis. median of the sum of the squared residuals (Figures 9(a) A high risk of developing HCC is associated with HBV or and 9(b)). Six possible predictors (BUB1B, NUSAP1, TTK, HCV infections, cirrhosis, and alcohol intake. Of these, e expression of KIF2C Log (TPM+1) 2 Journal of Oncology 11 ⁎⁎⁎ ⁎⁎⁎ 0 0 Normal Tumor Normal Tumor (a) (b) ⁎⁎⁎ ⁎⁎⁎ Normal Tumor Normal Tumor (c) (d) ⁎⁎⁎ ⁎⁎⁎ Normal Tumor Normal Tumor (e) (f) ⁎⁎⁎ ⁎⁎⁎ 0 0 Normal Tumor Normal Tumor (g) (h) Figure 8: Continued. e expression of MAD2L1 e expression of HMMR e expression of RRM2 e expression of BUB1B Log (TPM+1) Log (TPM+1) Log (TPM+1) Log (TPM+1) 2 2 2 e expression of CCNA2 e expression of TTK e expression of NUSAP1 e expression of MELK Log (TPM+1) Log (TPM+1) Log (TPM+1) Log (TPM+1) 2 2 2 2 12 Journal of Oncology ⁎⁎⁎ Normal Tumor (i) Figure 8: Differential expression analysis of hub genes in TCGA dataset consisting of HCC tissue samples and paired adjacent normal tissue samples. The expression levels of (a) BUB1B, (b) MELK, (c) MAD2L1, (d) NUSAP1, (e) RRM2, (f) TTK, (g) HMMR, (h) CCNA2, and (i) KIF2C in HCC tissues (n =50) were significantly higher than those in paired adjacent tissues (n =50). 9 9 9 8 8 8 8 8 8 7 6 6 6 5 4 4 4 4 4 4 3 3 2 2 2 2 2 1 0 986431 0.5 12.0 11.8 0.0 11.6 −0.5 11.4 −1.0 −7 −6 −5 −4 −3 −2 −7 −6 −5 −4 −3 −2 Log (𝜆) Log (𝜆) (a) (b) −1 Risk group Low High Status Dead Alive BUB1B 3 NUSAP1 2 TTK 1 HMMR CCNA2 −1 KIF2C −2 (c) Figure 9: Clinical significance of these hub genes in HCC patients’ data from TCGA. (a) Using the LASSO Cox regression model, the partial likelihood deviance versus log (λ) has been plotted. (b) Using the lambda parameter, chosen feature coefficients are shown. (c) Distribution of risk score, prognostic hub gene expression, and survival status of HCC patients. Partial Likelihood Deviance Survival time Risk score e expression of KIF2C Log (TPM+1) Coefficients Journal of Oncology 13 bioinformatics analyses using data from two gene chip data- Table 1: Baseline clinical information. sets (from cirrhotic liver and HCC tissues), we identified 239 Characteristic Levels Overall DEGs, of which 94 were elevated and 145 were depressed. n 374 Using Cytoscape, we were able to identify ten possible hub genes from these DEGs. The genes with the highest prognos- T1 183 (49.3%) tic potential were identified using the LASSO Cox regression T2 95 (25.6%) T stage, n (%) model. The hub genes that we have identified are intricately T3 80 (21.6%) connected to the incidence, progression, and prognosis of T4 13 (3.5%) HCC and therefore may be very useful in the early detection N0 254 (98.4%) and treatment of HCC. N stage, n (%) N1 4 (1.6%) We have identified BUB1B, NUSAP1, TTK, HMMR, M0 268 (98.5%) CCNA2, and KIF2C as potential predictive markers for M stage, n (%) HCC. Previous studies indicate that expression levels of M1 4 (1.5%) BUB1B, which is a spindle-assembly checkpoint gene Female 121 (32.4%) Gender, n (%) [26], were highly upregulated in multiple myeloma Male 253 (67.6%) patients and that these levels were strongly correlated with ≤60 177 (47.5%) unfavorable outcomes [27]. Another marker, NUSAP1, Age, n (%) >60 196 (52.5%) which is a microtubule-associated protein involved in ≤400 215 (76.8%) mitosis, is also known to participate in cell proliferation, AFP (ng/ml), n (%) >400 65 (23.2%) apoptosis, and repairing DNA damage in glioblastoma multiforme cells [28]. The protein kinase encoded by the No 208 (65.4%) Vascular invasion, n (%) TTK gene is necessary for mitotic checkpoints as well as Yes 110 (34.6%) the DNA damage response [29]. Elevated HMMR in Alive 244 (65.2%) mouse mammary epithelium enhances the rate of Brca1- OS event, n (%) Dead 130 (34.8%) mutant carcinogenesis as it is involved in modifying the A 219 (90.9%) phenotype of tumor cell and tumor microenvironment Child-Pugh grade, n (%) B 21 (8.7%) [30]. The CCNA2 gene also plays an important role in C 1 (0.4%) HCC, as the HBV genome integrates into one of the CCNA2 introns and forms an in-frame chimeric fusion Age, median (IQR) 61 (52, 69) with CCNA2 [31]. The KIF2C gene, belonging to the Kinesin family, has been shown to be significantly overex- pressed in several human malignancies [32]. Since mRNA is an essential component of all cells, including tumor cells, changes in mRNA levels of hub genes cirrhosis is the most significant risk factor, since 80–90% of HCC patients usually suffered from cirrhosis [22, 23]. can be used as molecular indicators for a variety of disorders, Patients diagnosed with HCC who undergo curative including cancer [13, 33, 34]. We find that the AUCs for therapy in the early stages of the disease have significantly BUB1B, NUSAP1, TTK, HMMR, CCNA2, and KIF2C were higher five-year survival rates [24]. However, the mechanism all >0.9, which indicates the expression levels of these genes for liver cirrhosis progresses into HCC is as yet unknown, can be used to differentiate between HCC tissues and normal though there are two theories about this process. One theory liver tissues. Besides, immune cells that have invaded a assumes that liver cirrhosis itself is a precancerous stage that tumor are called tumor-infiltrating cells. These cells are a leads to HCC due to internal hepatic interstitial changes and key part of the microenvironment of a tumor and are modulations in cell proliferation. The second theory postu- strongly related with carcinogenesis, progression, or metas- lates that cirrhosis affects hepatocyte proliferation by making tasis. In our results, we found the six hub genes were all pos- the cells more sensitive to carcinogenic factors in the itively associated with the immune infiltration level of external environment, which predisposes them to damage immune cells. All these results collectively suggested that that leads to the development of HCC [25]. Since the these hub genes could serve as diagnostic molecular markers rapid rate of cellular reproduction does not allow these for HCC. cells sufficient time for DNA repair, mutations accumulate Considering that the predictive signature was developed in newly produced cells, which pave the way to malignant and verified by the use of data from public databases, more transformation. experimental proof on top of the statistical evidence that We tried to address this gap in knowledge by analyzing we supplied will be required. the differences in gene expression profiles between normal It is concluded that BUB1B, NUSAP1, TTK, HMMR, liver tissues, cirrhotic liver tissues, and HCC tissues. We CCNA2, and KIF2C can be considered as potential novel screened several databases to get hub genes that may be molecular indicators for the onset and development of responsible for the progression of cirrhosis into HCC. We HCC, since they are linked to the transition from cirrhosis found that genes linked to mitotic nuclear division, chromo- to HCC. This study will prove important reference for trans- somal segregation, nuclear division, and organelle fission are lational medicine scientists, liver disease specialists, and bio- all intimately connected to this process. Through a series of informatics specialists. 14 Journal of Oncology 1.0 1.0 0.8 0.8 0.6 0.6 0.4 0.4 0.2 0.2 BUB1B NUSAP1 AUC: 0.961 AUC: 0.949 CI: 0.942-0.981 CI: 0.927-0.970 0.0 0.0 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 1 − specificity (FPR) 1 − specificity (FPR) (a) (b) 1.0 1.0 0.8 0.8 0.6 0.6 0.4 0.4 TTK HMMR 0.2 0.2 AUC: 0.971 AUC: 0.968 CI: 0.955-0.987 CI: 0.951-0.986 0.0 0.0 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 1 − specificity (FPR) 1 − specificity (FPR) (c) (d) 1.0 1.0 0.8 0.8 0.6 0.6 0.4 0.4 KIF2C 0.2 0.2 CCNA2 AUC: 0.981 AUC: 0.970 CI: 0.970-0.993 CI: 0.952-0.989 0.0 0.0 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 1 − specificity (FPR) 1 − specificity (FPR) (e) (f) Figure 10: Receiver operating characteristic analysis (ROC) of (a) BUB1B, (b) NUSAP1, (c) TTK, (d) HMMR, (e) CCNA2, and (f) KIF2C in HCC patients’ data (n = 424). Sensitivity (TPR) Sensitivity (TPR) Sensitivity (TPR) Sensitivity (TPR) Sensitivity (TPR) Sensitivity (TPR) Journal of Oncology 15 Purity B Cell CD8+T Cell CD4+T Cell Macrophage Neutrophil Dendritic Cell 5 cor = 0.15 partial.cor = 0.5 partial.cor = 0.326 partial.cor = 0.373 partial.cor = 0.481 partial.cor = 0.434 partial.cor = 0.476 p = 5.13e–03 p = 3.39e–23 p = 6.70e–10 p = 9.12e–13 p = 3.67e–21 p = 2.90e–17 p = 1.32e–20 –1 0.25 0.50 0.75 1.00 0.1 0.2 0.3 0.4 0.2 0.4 0.6 0.0 0.1 0.2 0.3 0.4 0.0 0.1 0.2 0.3 0.05 0.10 0.15 0.20 0.25 0.50 0.75 1.00 Infiltration Level (a) Purity B Cell CD8+T Cell CD4+T Cell Macrophage Neutrophil Dendritic Cell cor = 0.15 partial.cor = 0.489 partial.cor = 0.341 partial.cor = 0.356 partial.cor = 0.461 partial.cor = 0.415 partial.cor = 0.487 p = 5.13e–03 p = 4.14e–22 p = 8.74e–11 p = 9.80e–12 p = 2.48e–19 p = 7.84e–16 p = 1.28e–21 0.25 0.50 0.75 1.00 0.1 0.2 0.3 0.4 0.2 0.4 0.6 0.0 0.1 0.2 0.3 0.4 0.0 0.1 0.2 0.3 0.05 0.10 0.15 0.20 0.25 0.50 0.75 1.00 Infiltration Level (b) Purity B Cell CD8+T Cell CD4+T Cell Macrophage Neutrophil Dendritic Cell partial.cor = 0.464 partial.cor = 0.313 partial.cor = 0.308 partial.cor = 0.422 partial.cor = 0.355 partial.cor = 0.424 cor = 0.148 p = 5.92e–03 p = 9.19e–20 p = 3.27e–09 p = 5.53e–09 p = 3.48e–16 p = 1.16e–11 p = 2.92e–16 0.25 0.50 0.75 1.00 0.1 0.2 0.3 0.4 0.2 0.4 0.6 0.0 0.1 0.2 0.3 0.4 0.0 0.1 0.2 0.3 0.05 0.10 0.15 0.20 0.25 0.50 0.75 1.00 Infiltration Level (c) Purity B Cell CD8+T Cell CD4+T Cell Macrophage Neutrophil Dendritic Cell 5 cor = 0.166 partial.cor = 0.399 partial.cor = 0.271 partial.cor = 0.267 partial.cor = 0.351 partial.cor = 0.368 partial.cor = 0.406 p = 1.99e–03 p = 1.47e–14 p = 3.69e–07 p = 4.91e–17 p = 2.54e–11 p = 1.75e–12 p = 6.84e–15 0.25 0.50 0.75 1.00 0.1 0.2 0.3 0.4 0.2 0.4 0.6 0.0 0.1 0.2 0.3 0.4 0.0 0.1 0.2 0.3 0.05 0.10 0.15 0.20 0.25 0.50 0.75 1.00 Infiltration Level (d) Purity B Cell CD8+T Cell CD4+T Cell Macrophage Neutrophil Dendritic Cell cor = 0.152 partial.cor = 0.476 partial.cor = 0.354 partial.cor = 0.318 partial.cor = 0.402 partial.cor = 0.366 partial.cor = 0.489 p = 7.31e–21 p = 1.63e–11 p = 1.60e–09 p = 1.21e–14 p = 2.41e–12 p = 8.02e–22 p = 4.56e–03 0.25 0.50 0.75 1.00 0.1 0.2 0.3 0.4 0.2 0.4 0.6 0.0 0.1 0.2 0.3 0.4 0.0 0.1 0.2 0.3 0.05 0.10 0.15 0.20 0.25 0.50 0.75 1.00 Infiltration Level (e) Purity B Cell CD8+T Cell CD4+T Cell Macrophage Neutrophil Dendritic Cell cor = 0.123 partial.cor = 0.482 partial.cor = 0.363 partial.cor = 0.275 partial.cor = 0.449 partial.cor = 0.362 partial.cor = 0.47 p = 2.09e–21 p = 4.31e–12 p = 2.12e–07 p = 2.73e–18 p = 4.21e–12 p = 4.66e–29 6 p = 2.19e–02 0.25 0.50 0.75 1.00 0.1 0.2 0.3 0.4 0.2 0.4 0.6 0.0 0.1 0.2 0.3 0.4 0.0 0.1 0.2 0.3 0.05 0.10 0.15 0.20 0.25 0.50 0.75 1.00 Infiltration Level (f) Figure 11: Relationships between immune infiltration in HCC tissues and expression levels of (a) BUB1B, (b) NUSAP1, (c) TTK, (d) HMMR, (e) CCNA2, and (f) KIF2C from the TIMER database. CCNA2 Expression Level (log2 TPM) HMMR Expression Level (log2 TPM) KIF2C Expression Level (log2 TPM) TTK Expression Level (log2 TPM) NUSAP1 Expression Level (log2 TPM) BUB1B Expression Level (log2 TPM) LIHC LIHC LIHC LIHC LIHC LIHC 16 Journal of Oncology [10] S. Davis and P. S. Meltzer, “GEOquery: a bridge between the Data Availability Gene Expression Omnibus (GEO) and BioConductor,” Bioin- The datasets analyzed during the current study are available formatics, vol. 23, no. 14, pp. 1846-1847, 2007. in TCGA (https://portal.gdc.cancer.gov/) and GEO reposi- [11] G. Yu, L. G. Wang, Y. Han, and Q. Y. He, “clusterProfiler: an R tory (https://www.ncbi.nlm.nih.gov/geo/). package for comparing biological themes among gene clus- ters,” OMICS, vol. 16, no. 5, pp. 284–287, 2012. [12] D. Szklarczyk, A. Franceschini, S. Wyder et al., “STRING v10: Conflicts of Interest protein-protein interaction networks, integrated over the tree of life,” Nucleic Acids Research, vol. 43, no. D1, pp. D447– The authors declare no competing interests. D452, 2015. [13] C. Zhang, W. Zhang, H. Cui et al., “Role of hub genes in the occurrence and development of testicular cancer based on bio- Authors’ Contributions informatics,” International Journal of General Medicine, vol. 15, pp. 645–660, 2022. Yuanbin Chen and Hongyan Qian are equal first authors. [14] Z. Zhang, E. Lin, H. Zhuang et al., “Construction of a novel gene-based model for prognosis prediction of clear cell renal cell Acknowledgments carcinoma,” Cancer Cell International, vol. 20, no. 1, p. 27, 2020. [15] W. Lin, S. Wu, X. Chen et al., “Characterization of hypoxia sig- This study was supported by the Nantong Science and Tech- nature to evaluate the tumor immune microenvironment and nology Bureau (grant no. JC2020025), the Health Commit- predict prognosis in glioma groups,” Frontiers in Oncology, tee of Nantong (grant no. MB2020021), and the Clinical vol. 10, p. 796, 2020. Medicine Special Program from Nantong University (grant [16] R. Tibshirani, “The lasso method for variable selection in the Cox no. 2019JY017). This work was also supported by Scientific model,” Statistics in Medicine,vol.16,no.4,pp. 385–395, 1997. Research Project of Health Commission of Jiangsu Province 6 [17] F. Xu, X. Huang, Y. Li, Y. Chen, and L. Lin, “m A-related (grant no. M2020009) lncRNAs are potential biomarkers for predicting prognoses and immune responses in patients with LUAD,” Molecular Therapy-Nucleic Acids, vol. 24, pp. 780–791, 2021. References [18] Y. Ji and Y. Xue, “Identification and clinical validation of 4- lncRNA signature for predicting survival in head and neck [1] W. Shi, L. Feng, S. Dong et al., “FBXL6 governs c-MYC to pro- squamous cell carcinoma,” Oncotargets and Therapy, vol. 13, mote hepatocellular carcinoma through ubiquitination and pp. 8395–8411, 2020. stabilization of HSP90AA1,” Cell Communication and Signal- [19] J. Wang, J. Sun, L. N. Liu et al., “Siglec-15 as an immune sup- ing: CCS, vol. 18, no. 1, p. 100, 2020. pressor and potential target for normalization cancer immu- [2] H. Nordenstedt, D. L. White, and H. B. El-Serag, “The chang- notherapy,” Nature Medicine, vol. 25, no. 4, pp. 656–666, 2019. ing pattern of epidemiology in hepatocellular carcinoma,” [20] R. Ravi, K. A. Noonan, V. Pham et al., “Bifunctional immune Digestive and Liver Disease, vol. 42, Supplement 3, pp. S206– checkpoint-targeted antibody-ligand traps that simultaneously S214, 2010. disable TGFβ enhance the efficacy of cancer immunotherapy,” [3] L. Kulik and H. B. El-Serag, “Epidemiology and management Nature Communications, vol. 9, no. 1, p. 741, 2018. of hepatocellular carcinoma,” Gastroenterology, vol. 156, [21] N. Wen, Y. Cai, F. Li et al., “The clinical management of hepa- no. 2, article e471, pp. 477–491.e1, 2019. tocellular carcinoma worldwide: a concise review and compar- [4] G. V. Papatheodoridis, P. Lampertico, S. Manolakopoulos, and ison of current guidelines: 2022 update,” Bioscience Trends, A. Lok, “Incidence of hepatocellular carcinoma in chronic vol. 16, no. 1, pp. 20–30, 2022. hepatitis B patients receiving nucleos(t)ide therapy: a system- atic review,” Journal of Hepatology, vol. 53, no. 2, pp. 348– [22] M. Walker, H. B. El-Serag, Y. Sada et al., “Cirrhosis is under- recognised in patients subsequently diagnosed with hepatocel- 356, 2010. lular cancer,” Alimentary Pharmacology & Therapeutics, [5] W. L. Tsai and R. T. Chung, “Viral hepatocarcinogenesis,” vol. 43, no. 5, pp. 621–630, 2016. Oncogene, vol. 29, no. 16, pp. 2309–2324, 2010. [23] L. A. Beste, S. L. Leipertz, P. K. Green, J. A. Dominitz, D. Ross, [6] H. Li, “Intercellular crosstalk of liver sinusoidal endothelial and G. N. Ioannou, “Trends in burden of cirrhosis and hepato- cells in liver fibrosis, cirrhosis and hepatocellular carcinoma,” cellular carcinoma by underlying liver disease in US veterans, Digestive and Liver Disease, vol. 54, no. 5, pp. 598–613, 2022. 2001-2013,” Gastroenterology, vol. 149, no. 6, pp. 1471– [7] G. Yildiz, A. Arslan-Ergul, S. Bagislar et al., “Genome-wide 1482.e5, 2015. transcriptional reorganization associated with senescence-to- [24] H. B. El-Serag, “Hepatocellular carcinoma,” The New England immortality switch during human hepatocellular carcinogene- Journal of Medicine, vol. 365, no. 12, pp. 1118–1127, 2011. sis,” PLoS One, vol. 8, no. 5, article e64016, 2013. [25] S. A. Sharma, M. Kowgier, B. E. Hansen et al., “Toronto HCC [8] X. Zhang, W. Zhang, Y. Jiang, K. Liu, L. Ran, and F. Song, “Identification of functional lncRNAs in gastric cancer by inte- risk index: a validated scoring system to predict 10-year risk of HCC in patients with cirrhosis,” Journal of hepatology, vol. 68, grative analysis of GEO and TCGA data,” Journal of Cellular Biochemistry, vol. 120, no. 10, pp. 17898–17911, 2019. no. 1, pp. 92–99, 2018. [9] J. Sun, J. Huang, J. Lan et al., “Overexpression of CENPF cor- [26] T. Rio Frio, J. Lavoie, N. Hamel et al., “Homozygous BUB1B relates with poor prognosis and tumor bone metastasis in mutation and susceptibility to gastrointestinal neoplasia,” breast cancer,” Cancer Cell International, vol. 19, no. 1, The New England Journal of Medicine, vol. 363, no. 27, p. 264, 2019. pp. 2628–2637, 2010. Journal of Oncology 17 [27] X. Tang, M. Guo, P. Ding et al., “BUB1B and circBUB1B_ 544aa aggravate multiple myeloma malignancy through evok- ing chromosomal instability,” Signal Transduction and Tar- geted Therapy, vol. 6, no. 1, p. 361, 2021. [28] Y. Zhao, J. He, Y. Li, S. Lv, and H. Cui, “NUSAP1 potentiates chemoresistance in glioblastoma through its SAP domain to stabilize ATR,” Signal Transduction and Targeted Therapy, vol. 5, no. 1, p. 44, 2020. [29] F. Chen, E. P. Szymanski, K. N. Olivier et al., “Whole-exome sequencing identifies the 6q12-q16 linkage region and a candi- date gene, TTK, for pulmonary nontuberculous mycobacterial disease,” American Journal of Respiratory and Critical Care Medicine, vol. 196, no. 12, pp. 1599–1604, 2017. [30] F. Mateo, Z. He, L. Mei et al., “Modification of BRCA1- associated breast cancer risk by HMMR overexpression,” Nature Communications, vol. 13, no. 1, p. 1895, 2022. [31] Y. T. Chiu, J. K. Wong, S. W. Choi et al., “Novel pre-mRNA splicing of intronically integrated HBV generates oncogenic chimera in hepatocellular carcinoma,” Journal of Hepatology, vol. 64, no. 6, pp. 1256–1264, 2016. [32] S. Wei, M. Dai, C. Zhang et al., “KIF2C: a novel link between Wnt/β-catenin and mTORC1 signaling in the pathogenesis of hepatocellular carcinoma,” Protein & Cell, vol. 12, no. 10, pp. 788–809, 2021. [33] Y. Zheng, Y. Luo, X. Chen et al., “The role of mRNA in the development, diagnosis, treatment and prognosis of neural tumors,” Molecular Cancer, vol. 20, no. 1, p. 49, 2021. [34] E. Wieczorek and E. Reszka, “mRNA, microRNA and lncRNA as novel bladder tumor markers,” Clinica Chimica Acta, vol. 477, pp. 141–153, 2018.

Journal

Journal of OncologyHindawi Publishing Corporation

Published: Sep 26, 2022

There are no references for this article.