Theranostics and Precision Medicine for the Management of Hepatocellular Carcinoma, Volume 2: Diagnosis, Therapeutic Targets, and Molecular Mechanisms

Theranostics and Precision Medicine for the Management of Hepatocellular Carcinoma, Volume Two: Diagnosis, Therapeutic Targets and Molecular Mechanisms for Hepatocellular Carcinoma Progression provides comprehensive information about ongoing research and clinical data surrounding liver cancer. The book presents detailed descriptions about diagnostics and therapeutic options for easy understanding, with a focus on precision medicine approaches to improve treatment outcomes. The volume discusses topics such as computational approaches for identification of biomarkers, enzymes and pathways of HCC, circulating and epigenetic biomarkers, drug resistance, metabolic pathways, and small molecule-target therapies. In addition, it discusses immunotherapies, immune check point inhibitors and nanotechnology-based therapies.

This book is a valuable resource for cancer researchers, oncologists, graduate students, hepathologists and members of biomedical research who need to understand more about liver cancer to apply in their research work or clinical setting.

• Provides detailed information on traditional and novel diagnostic tools for hepatocellular carcinoma
• Discusses promising targeted therapies, both available and in development, explaining the best option to use for specific cases
• Brings recent findings in immunotherapies, immune checkpoint inhibitors and nanotechnology-based therapeutic approaches for treatment of HCC

• Language: English
• Publisher: Academic Press
• Release date: Apr 2, 2022
• ISBN: 9780323993654

Book preview

Preface

Ganji Purnachandra Nagaraju and Ramakrishna Vadde

Hepatocellular carcinoma (HCC) is a form of primary liver cancer that has emerged as one of the most frequently diagnosed human malignancies worldwide. It predominantly targets individuals with underlying conditions, such as cirrhosis, hepatitis B or hepatitis C infections, and alcohol-related disease. Despite the widespread occurrences of HCC, its underlying mechanisms that lead to tumor progression are still unclear. In contrast to various other cancers, systemic therapies for HCC, including chemotherapy and radiotherapy, are not effective. Currently, the only viable options for managing advanced HCC are surgical resection and transplantation. Therefore it is crucial to investigate pathways and factors that lead to HCC tumor suppression with the goal of developing novel therapeutic options for treating HCC.

In the current series we have three different volumes. Volume 1 discusses the biology, pathophysiology, and progression of liver cancer. Volume 2 discusses multiple signaling and molecular mechanisms that are associated with HCC progression and metastasis, including renowned and novel biomarkers, as well as diagnostic approaches for advanced HCC. Volume 3 discusses several therapeutic targets for HCC and therapeutic molecules, such as phytochemicals and small molecules, as well as clinical assessment, management, and precision medicine.

Volume 2 includes a detailed discussion about HCC diagnosis. It also elaborates on the computational analysis of HCC prognosis–associated genes and various other computational approaches to identifying HCC biomarkers, enzymes, and molecular pathways; the effect of RNA sequencing for understanding HCC; and the role of next-generation sequencing in liver malignancies. Further, the tumor microenvironment is elucidated along with related pathways, particularly metabolic pathway–based targeted therapy to manage HCC with a computational approach. The roles of ion channels, tyrosine kinases, and transcription factors in hepatic cancer are explained in depth. Additionally, regulatory activities of multiple transcriptional factors in liver cancer are discussed, along with the role of specificity protein 1.

A comprehensive discussion of conventional and novel biomarkers for diagnosis and prognosis of advanced liver cancer is included in this volume. The role of molecular signaling and associated drug resistance as well as challenges of multidrug resistance and future prospects are examined. Finally, up-to-date treatment strategies, including the potential therapeutic role of ROS-mediated pathways, dysregulated cell-signaling pathways, and LKB1/STK11-mediated signaling transduction in HCC, are scrutinized.

The aim of this series is not only to illustrate the experimental biochemistry and clinical significance of HCC but also to stress the functional and pathophysiological roles of different signaling pathways. Knowledge of these roles represents an essential step in understanding the tumor markers in such a way that their application will result in medical and economic advances. This series will provide novel ideas to researchers and scholars as well as innovative future perspectives in the field of research and clinical applications.

Chapter 1

Hepatocellular carcinoma diagnosis

Gayathri Chalikonda¹, Sekani Allen¹, Ramakrishna Vadde² and Ganji Purnachandra Nagaraju³,    ¹University of Nevada, Reno, NV, United States,    ²Department of Bioinformatics and Biotechnology, Yogi Vemana University, Kadapa, India,    ³School of Medicine, Division of Hematology and Oncology, University of Alabama, Birmingham, AL, United States

Abstract

Hepatocellular carcinoma (HCC) is a common malignancy that develops as a result of chronic liver disease and often cirrhosis. Although imaging modalities are the mainstay of screening and detection, their limitations include difficulties in identifying HCC before progression to an advanced stage and differentiating HCC from benign cirrhotic nodules. Alpha-fetoprotein is a carcinoembryonic glycoprotein that is commonly used to screen for HCC. However, its low sensitivity of approximately 60% allows several cancers to go undetected. Several studies have looked into promising biomarkers to improve early detection with a lower false-negative rate.

Keywords

Hepatocellular carcinoma; biomarkers; α-fetoprotein; protein induced by vitamin K deficiency; PIVKA-II; des-gamma-carboxyprothrombin; phenylalanyl-tryptophan; glycocholate; Golgi protein 73; phosphatidylinositol proteoglycan; heat shock protein; Dickkopf-1; osteopontin; microRNA

Abbreviations

HCC Hepatocellular carcinoma

AFP Alpha-fetoprotein

PIVKA-II Protein induced by vitamin K deficiency

DCP Des-gamma-carboxyprothrombin

DKK1 Dickkopf-1

Phe-Trp Phenylalanyl-tryptophan

GCA Glycocholate

GP73 Golgi protein 73

GPC-3 Phosphatidylinositol proteoglycan

HSP Heat shock protein

DKK1 Dickkopf-1

OPN Osteopontin

miRNA MicroRNA

miR-122 MicroRNA-122

Introduction

Hepatocellular carcinoma (HCC) is a malignancy that develops as a result of chronic liver disease and often cirrhosis, with etiologies including prolonged alcohol use, untreated hepatitis, or nonalcohol-associated steatohepatitis [1]. HCC is a common cancer that is often discovered in its later stages, owing to nonexistent or vague symptoms, including upper abdominal pain, weight loss, and early satiety. Also, HCC may present with paraneoplastic syndromes, including hypoglycemia, excessive ILGF-2 production, excessive erythropoietin production, or diarrhea. Inadequate surveillance for high-risk patients contributes to diagnosis at advanced stages as well. Therefore early and accurate diagnosis of HCC is imperative.

Current guidelines encourage regular surveillance of high-risk individuals. If the initial lesion that is noted is benign and less than 1 cm in diameter, the primary care physician is encouraged to reimage the tumor every 6 months. If the lesion presents as a malignant lesion 1 cm in diameter or greater, then closer surveillance or more advanced imaging modalities are recommended. Although ultrasonography, computed tomography, and magnetic resonance imaging are critical for HCC diagnosis, both expense and operator-dependent image quality limit the use of imaging modalities. Also, with these different methods it is difficult to differentiate HCC from benign cirrhotic nodules. If diagnosed in its early stages, HCC treatment options include surgical resection and transplantation. However advanced-stage HCC treatment is generally limited to sorafenib (which is a first-line drug), regorafenib (which is a second-line drug), and lenvatinib [1]. Therefore further the development of early detection modalities is critical for improving prognosis.

In recent years, several research studies have focused on identifying serum tumor biomarkers for early identification of HCC. Briefly, tumor biomarkers are molecular markers related to the growth, proliferation, and/or pathogenesis of cells [2,3]. When looking for the ideal tumor marker in HCC, we are looking for (1) tumor biomarkers to distinguish tumor/cancer types, (2) biomarkers to predict tumor development, and metastasis, and (3) biomarkers to evaluate treatment efficacy [2]. In this chapter we will comment on several tumor biomarkers that have been used and researched in diagnosing, tracking, and predicting treatment outcomes.

Alpha-fetoprotein (AFP), introduced in the 1960s, is a commonly used tumor marker today [3]. AFP is a carcinoembryonic serum glycoprotein; its main function is to transfer several large molecules in the embryonic and perinatal period. AFP expression is higher in highly metastatic and invasive cancer cells; because of this, clinicians have used AFP to screen for HCC since 1970. Although the specificity of AFP is 72%–90%, its sensitivity is approximately 60% [4,5]. For example, AFP levels are normal in nearly 30% of cases with advanced HCC patients. Additionally, AFP levels may be elevated in benign hepatitis and cirrhosis. Owing to the high false-negative rates of HCC, the use of AFP alone is controversial. With the addition of other tumor biomarkers we may be able to improve HCC diagnosis accuracy [2]. On the other hand, AFP-L3 is more specific for HCC than its other glycoforms; however, while it has 92% specificity, its sensitivity is only 37% [6]. Despite AFP’s extensive use in the clinical setting, it has many limitations.

Protein induced by vitamin K deficiency (PIVKA-II), or des-gamma-carboxyprothrombin (DCP), is an incorrect prothrombin structure resulting from defective posttranslational carboxylation of its precursor protein. This new protein is able to promote malignant proliferation in HCC. Therefore this marker has high sensitivity as well as specificity, allowing differentiation of HCC from hepatitis and cirrhosis. This HCC marker has a high diagnostic rate in Japan, Europe, and the United States, as levels are directly related to cancer size, staging, Child-Pugh score, and recurrence. When the protein level is elevated above 32.09 mAU/mL, it has approximately 51% sensitivity and 85% specificity for identifying HCC [7]. When DCP and ACP were used together, the sensitivity increased to 87%, while the specificity decreased to 69% [6]. Therefore this biomarker proves to be a useful tool in aiding HCC screening. For instance, following tumor resection, a decrease in PIVKA-II below a threshold value would suggest a successful operation. However, if this level rises again, it would suggest local recurrence rather than HCC metastasis [2]. Therefore PIVKA-II monitoring can help in evaluating development, metastasis, or even recurrence of HCC.

The biomarker panel of glycocholate (GCA) and phenylalanyl-tryptophan (Phe-Trp), two serum metabolites, has been shown to identify HCC in high-risk populations with increased accuracy when compared to AFP. Owing to its high selectivity, with 86%–95% diagnostic accuracy in identifying AFP-negative cancers, this panel is a potential method for predicting HCC development in susceptible patients prior to symptom presentation [3]. Thus Phe-Trp and GCA are another set of promising HCC biomarkers.

Golgi protein 73 (GP73) is a glycoprotein on Golgi cell membranes. GP73 is rarely expressed in hepatocyte cells except for hepatocyte cells that are adjacent to the bile duct surfaces. It is a housekeeping gene that is useful for differentiating between healthy and cells that are infected with virus from HCC, as it is elevated only in malignant cells [6]. In patients who had hepatitis B versus HCC and patients who had hepatitis B versus healthy patients, GP73 had a specificity of 92%–97.4% and a sensitivity of 74.6%–76.9% for detecting HCC. Additionally, it has been found that GP73 is not associated with age, sex, or AFP levels but with tumor burden, invasion, and differentiation [8].

Phosphatidylinositol proteoglycan (GPC-3) is a glycoprotein that plays a significant role in proliferation, differentiation, regulation, and migration of tumor cells. GPC-3 is present in most HCC cell lines and absent in normal lung, liver, and kidney tissue. Therefore GPC-3 has 76.9% sensitivity and 96.8% specificity. Additionally, in HCC cases with tumors less than 3 cm in diameter, GPC-3 can be elevated up to 76%, while its expression in cirrhotic nodules is much lower. Therefore GPC-3 can help in differentiating benign from malignant tumors [9].

Heat shock protein (HSP) is a molecular chaperone protein that is active during cellular stress. HSP70 expression is more correlated with early HCC than other lesions secondary to hypoxia. In malignant cells this protein promotes growth and inhibits aging. Additionally, HSP70 is important for interaction with the extracellular matrix during tumor metastasis. Also, HSP90 may be helpful for early HCC detection, especially in regard to small HCC [2]. HSP90 is positively correlated with metastatic ability. Overall, HSPs help direct attention toward cancerous cells.

Dickkopf-1 (DKK1) is a WNT/B-catenin signaling pathway antagonist. DKK1 is expressed in cancer cells such as multiple myeloma, HCC, and prostate cancer [10]. DKK1 is a serum biomarker that is specific for all cancers. In studies performed, DKK1 levels were significantly elevated in HCC patients; this test had a specificity of 87.2% and a sensitivity of 73.8% [11]. A panel with both AFP and DKK1 may be useful for AFP-negative HCC detection. Additionally, higher preoperative DKK1 levels are associated with lower survival rate and time [2]. Therefore DKK1 appears to be a promising biomarker as well.

Osteopontin (OPN) is a protein phosphatase that is found in Kupffer cells, ductal epithelial cells, and stellate cells but is notably absent in hepatocytes. Also, OPN expression is much higher in HCC patients than in healthy patients or patients with other liver diseases; sensitivity reached 87% and specificity reached 82% [2,12]. OPN is a potential diagnostic tumor biomarker that can be combined with tumor biomarkers for HCC detection, although further studies are required to confirm OPN’s role in hepatic fibrosis and carcinogenesis.

MicroRNAs (miRNAs) are 17- to 25-nucleotide noncoding RNA segments that control posttranscriptional gene expression, affecting cellular processes including carcinogenesis. While several sequences have been linked to HCC, microRNA-122 (miR-122) is one that plays a significant part in the process of hepatocarcinogenesis [7,13]. In the recent quantitative metaanalysis by Xiao-Fei Zhao et al. on miR-122 and its use in diagnosis of HCC, when miR-122 was used for HCC versus hepatitis infection, a sensitivity of 79% and a specificity of 82% were demonstrated; used for HCC versus liver cirrhosis or dysplasia, MiR-122 demonstrated 65% sensitivity and 75% specificity. Overall, the study showed that measuring miR-122 allows for HCC identification [13]. In other studies, a panel with seven different miRNAs was used to detect AFP-negative HCC [5]. In another study, a different panel of seven microRNAs was shown to diagnose hepatitis B related HCC highly accurately [6]. As various microRNA sequences have been associated with HCC, establishing a panel for early diagnosis of this cancer may help improve prognosis.

Treatment options for advanced-stage HCC are limited, so novel early detection techniques are necessary for early treatment and improvement of outcomes [1,14].

Conflict of interest

None to disclose

Funding

None to declare

References

1. Sim HW, Knox J. Hepatocellular carcinoma in the era of immunotherapy. Curr Probl Cancer. 2018;42:40–48.

2. Zong J, Fan Z, Zhang Y. Serum tumor markers for early diagnosis of primary hepatocellular carcinoma. J Hepatocell Carcinoma. 2020;7:413–422.

3. Luo P, Yin P, Hua R, et al. A large-scale, multicenter serum metabolite biomarker identification study for the early detection of hepatocellular carcinoma. Hepatology. 2018;67:662–675.

4. Marrero JA, Lok AS. Newer markers for hepatocellular carcinoma. Gastroenterology. 2004;127:S113–S119.

5. Luo P, Wu S, Yu Y, et al. Current status and perspective biomarkers in AFP negative HCC: towards screening for and diagnosing hepatocellular carcinoma at an earlier stage. Pathol Oncol Res. 2020;26:599–603.

6. Tsuchiya N, Sawada Y, Endo I, Saito K, Uemura Y, Nakatsura T. Biomarkers for the early diagnosis of hepatocellular carcinoma. World J Gastroenterol. 2015;21:10573–10583.

7. Wang X, Zhang W, Liu Y, et al. Diagnostic value of prothrombin induced by the absence of vitamin K or antagonist-II (PIVKA-II) for early stage HBV related hepatocellular carcinoma. Infect Agent Cancer. 2017;12:47.

8. Kladney RD, Bulla GA, Guo L, et al. GP73, a novel Golgi-localized protein upregulated by viral infection. Gene. 2000;249:53–65.

9. Ibrahim TR, Abdel-Raouf SM. Immunohistochemical study of glypican-3 and HepPar-1 in differentiating hepatocellular carcinoma from metastatic carcinomas in FNA of the liver. Pathol Oncol Res. 2015;21:379–387.

10. Chae W-J, Ehrlich AK, Chan PY, et al. The Wnt antagonist Dickkopf-1 promotes pathological type 2 cell-mediated inflammation. Immunity. 2016;44:246–258.

11. Jeng J-E, Chuang L-Y, Chuang W-L, Tsai J-F. Serum Dickkopf-1 as a biomarker for the diagnosis of hepatocellular carcinoma. Chin Clin Oncol. 2012;1:4.

12. O'Regan A, Berman JS. Osteopontin: a key cytokine in cell-mediated and granulomatous inflammation. Int J Exp Pathol. 2000;81:373–390.

13. Zhao XF, Li N, Lin DD, Sun LB. Circulating MicroRNA-122 for the diagnosis of hepatocellular carcinoma: a meta-analysis. Biomed Res Int. 2020;2020:5353695.

14. Hristova VA, Chan DW. Cancer biomarker discovery and translation: proteomics and beyond. Expert Rev Proteom. 2019;16:93–103.

Chapter 2

Computational analysis of prognosis-related genes in liver cancer

Vigneshwar Suriya Prakash Sinnarasan, Dahrii Paul, Mathavan Muthaiyan, Dinakara Rao Ampasala and Amouda Venkatesan,    Centre for Bioinformatics, Pondicherry University, Kalapet, India

Abstract

Improved understanding of liver cancer (LC) molecular mechanisms is required to determine therapeutic targets and personalized medicine. The current study analyzed the gene expression data of LC using various computational biology approaches to find the hub genes. Transcriptome profiles were downloaded from the Cancer Genome Atlas (TCGA) and Sequence Read Archive (SRA) databases. The overlapping differentially expressed genes (DEGs) among three datasets were identified by using R software. The functional enrichment analysis was performed by using Enrichr, and the constructed protein–protein interaction network identified the hub genes using Cytoscape. Overall survival was analyzed by plotting Kaplan-Meier survival curves. A total of 412 overlapping DEGs were identified from TCGA and the SRA datasets. Also, 10 hub genes, AURKB, BUB1B, NCAPG, CCNB2, PDZ Binding Kinase, TOP2A, BUB1, KIF20A, CCNB1, and cell division cycle associated 8, were identified from overlapping genes. The prognosis value of hub genes indicated that downregulation is closely associated with patient survival, and the prognosis is poor. In conclusion, all identified hub genes were found to be associated with prognosis; hence they may be treated as potential therapeutic targets in treating LC.

Keywords

Liver cancer; next-generation sequencing (NGS); bioinformatics analysis

Abbreviations

AURKB aurora kinase subfamily of serine/threonine kinases

BUB1 mitotic checkpoint serine/threonine kinase

BUB1B mitotic checkpoint serine/threonine-protein kinase BUB1 beta

CCNB1 cyclin B1

CCNB2 cyclin B2

CDCA8 cell division cycle associated 8

DEGs differently expressed genes

GEPIA2 Gene expression profiling interactive analysis

GO gene ontology

HTS high-throughput sequencing

KEGG Kyoto Encyclopedia of genes and genomes

KIF20A kinesin family member 20A

LC liver cancer

LIHC liver hepatocellular carcinoma

MCC maximal clique centrality

NCAPG non-SMC condensin I complex subunit G

NCBI National Center for Biotechnology Information

NGS next-generation sequencing

PBK PDZ Binding Kinase

PPI protein–protein interaction

SRA Sequence Read Archive

STRING Search Tool for the Retrieval of Interacting Genes

T9 threonine 9

TCGA the Cancer Genome Atlas

TOP2A topoisomerase II alpha

Introduction

Liver cancer (LC) is a type of cancer that originates in the liver and results in high morbidity and mortality rates. More than 800,000 cases are reported every year around the world [1]. According to the 2018 global cancer statistics, there were an estimated 841,080 cases of LC and 781,631 deaths from LC, which accounted for 4.7% cases and 8.2% deaths of all cancer [2]. LC is the fourth leading cause of cancer death around the world [2]. Hepatocellular carcinoma (HCC) and intrahepatic cholangiocarcinoma (ICC) are the most common forms of LC. About 75%–85% of cases of LC are HCC, and 10%–15% are ICC [3]. Other rare types may cumulatively make up approximately 15% of cases of LC [4]. The major risk factors for LC are hepatitis virus, foods adulterated with aflatoxin, diabetes, cirrhosis, inherited metabolic diseases, alcohol, tobacco, and nonalcoholic fatty liver disease [1,4]. At present, different diagnostic approaches are used: biopsy, imaging, and the molecular method. Surgery, radiotherapy, chemotherapy, immunotherapy, targeted drug therapy, and embolization therapy are common treatment modalities [1]. The overall 5-year survivability of LC patients is around 18%, but depending on treatment and disease stage, the survival rates may vary [1]. If LC is not diagnosed at an early stage, the survival rate may drop to 11% (regional stage) and 2% (distant stage) [1]. Although diagnosis and treatment have significantly improved, patients in the advanced metastatic stage outcomes still have poor prognoses. There is a possibility of misclassification or misdiagnosis of LC as other cancers originating from LC [3]. Therefore more effort is required to understand the molecular mechanism underlying LC progression, potential drug targets, and diagnostic biomarkers.

Many high-throughput sequencing (HTS) techniques are used for the comprehensive analysis of cancer. HTS techniques have enabled genetic-based landscaping in many cancer studies. A study of the differentially expressed gene (DEG) profile of HCC has enabled its classification [5,6]. Because of its morphologically, etiologically, and genomewise highly heterogeneous composition, LC presents as a complex tumor Therefore the investigation of DEGs play a crucial role in the assessment of tumor progression, potential drug targets, and diagnostic markers.

The Cancer Genome Atlas (TCGA) is a collaborative effort led by the National Cancer Institute and the National Human Genome Research Institute to map different types of cancer. The database contains the following data: level 1 (raw data), level 2 (normalized data), level 3 (aggregated data), and level 4 (region of interest data) of 33 cancer types, including genomic, epigenomic, transcriptomic, and proteomic data [7]. In this study, taking advantage of the TCGA database, which allows access to transcriptome profiling for various cancer conditions and RNA-seq data from the Sequence Read Archive (SRA) database, DEGs associated with cancerous LC are identified. From functional enrichment and the protein–protein interaction (PPI) network, the essential genes and pathways identified are closely related to LC. Further, the hub genes are associated with LC screened by DEG analysis, network, and cluster construction using bioinformatics tools (Fig. 2.1).

Figure 2.1 The workflow.

Materials and methods

Data collection

The gene expression data of liver HCC (LIHC) were downloaded from TCGA (https://www.cancer.gov/tcga) using TCGA biolinks [8]. The data contain 371 LIHC samples and 50 normal controls. LC transcriptome datasets (PRJNA668068 and PRJNA563853) were downloaded from the SRA database (https://www.ncbi.nlm.nih.gov/sra) of the National Center for Biotechnology Information (NCBI) [9]. The PRJNA668068 dataset contains 6 samples (3 normal and 3 tumor cases), and the PRJNA563853 dataset contains 12 samples (6 normal and 6 tumor cases).

Identification of differentially expressed genes

Analysis of the TCGA-LIHC gene expression profiles and SRA data was carried out by using R software (v3.6.1). Initially, the fastqcr (v0.1.2) [10] was used to check the quality of the transcriptome data to remove the low-quality reads as well as adapters by trimFastq in seqTools (v3.6). The preprocessed sequences were aligned with the human reference genome (hg38) to map the reads to find the expressed transcripts by using Rsubread (v2.0) [11]. From the generated BAM file, the count matrix was constructed by using Genomic Alignments (v1.8.4). Finally, DEseq2 was used to find the DEGs between LC tumor and normal samples. The threshold values for TCGA-LIHC were FDR<0.05, a |log twofold change| ≥ ±2 and for SRA downloaded transcriptome data |log twofold change| ≥ ±2, P<.05, to identify DEGs. The PRJNA668068 and PRJNA563853 datasets were used to find the overlapping DEGs from TCGA and transcriptome. A Venn diagram was constructed by using the online portal Venny [12].

Functional enrichment analysis of differentially expressed genes

The online web server [13] Enrichr for gene enrichment analysis accepts a set of genes as input to explore the functions. gene ontology (GO) describes the functional annotation of the DEGs identified by the Enrichr [11]. GO annotation reveals the biological process, cellular component, and molecular function of interesting genes. The pathway enrichment of DEGs was performed by using the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway database [14].

Network and module analysis

The identified DEGs were used to construct the PPI network using the online resource STRING (Search Tool for the Retrieval of Interacting Genes) [15]. Later, the PPI network from STRING was visualized by using Cytoscape (Version 3.8.1). The PPI network was analyzed by using the Cytohubba plugin to find the hub genes based on maximal clique centrality (MCC) [16].

Survival analysis

The overall survival analysis of key genes was performed by plotting Kaplan-Meier (KM) survival curves using Gene Expression Profiling Interactive Analysis (GEPIA2) [17]. GEPIA2, an online web server, was used to analyze the survival of large-scale gene expression profiles based on TCGA and GTEx databases. The prognosis of the key genes of LC was estimated by GEPIA2.

Results

Identification of differentially expressed genes

TCGA-LIHC data and the transcriptome data (PRJNA668068 and PRJNA563853 datasets) were analyzed to identify DEGs in normal and tumor samples. There were 2032 DEGs (286 upregulated and 1746 downregulated) in TCGA-LIHC, 2358 DEGs (1462 upregulated and 896 downregulated) in PRJNA668068, and 2987 DEGs (1344 upregulated and 1643 downregulated) in PRJNA563853. From three datasets, 412 overlapping DEGs (104 upregulated and 308 downregulated) are shown in Fig. 2.2 and tabulated in Table 2.1.

Figure 2.2 Venn diagram of DEGs for TCGA-LIHC, PRJNA668068, and PRJNA563853. DEGs, Differentially expressed genes.

Table 2.1

Functional analysis

To understand the biological functions of 412 DEGs, GO analysis was performed using Enrichr. Mainly, DEGs were enriched in the mitotic spindle organization, microtubule motor activity, microtubule cytoskeleton organization involved in mitosis, binding regulating cell growth, key signaling events, and the movement of the fundamental cellular process, as shown in Tables 2.2 and 2.3. The KEGG pathway analysis reveals that DEGs were mainly enriched in the cell cycle, oocyte meiosis, mineral absorption, retinol metabolism, progesterone-mediated oocyte maturation, and p53 signaling pathway (Fig. 2.3).

Table 2.2

Table 2.3

Figure 2.3 Top 10 functional pathways associated with DEGs. DEGs, Differentially expressed genes.

Protein–protein interaction network and hub genes

The PPI network analysis of key genes was performed by using STRING and visualized in Cytoscape software, resulting in 207 nodes and 2977 interactions (Fig. 2.4) with interacting proteins with key genes. Based on the interacting proteins, hub genes responsible for LC were identified by the Cytohubba plugin. It was found that AURKB, BUB1B, NCAPG, CCNB2, PBK, TOP2A, BUB1, KIF20A, CCNB1, and CDCA8 are highly interacting genes and are responsible for LC based on MCC values (Fig.