Identification of common hub genes in gastrointestinal cancers and their role in prognosis and Immune cell infiltration in hepatocellular carcinoma

doi:10.21203/rs.3.rs-3092429/v1

Download PDF

Research Article

Identification of common hub genes in gastrointestinal cancers and their role in prognosis and Immune cell infiltration in hepatocellular carcinoma

https://doi.org/10.21203/rs.3.rs-3092429/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background

Common hub genes (CHGs) in gastrointestinal cancers (GICs) and their prognostic potential and immune infiltration in hepatocellular carcinoma (HCC) are limited.

Methods

Differential analysis of genes in the six GICs was performed, and CHGs were identified using the survival and protein-protein interaction (PPI) network analyses. A polygenic prognostic model was constructed based on the CHGs using the LASSO-Cox regression algorithm and validated in two cohorts. CHGs-related multivariate Cox regression and immune infiltration analyses were performed.

Results

Ten CHGs (TOP2A, CCNB1, CDK1, MAD2L1, RRM2, CCNA2, BUB1B, CDC6, AURKA, and BUB1) were obtained. In The Cancer Genome Atlas (TCGA) cohort, a 3-gene (CCNB1, MAD2L1, and BUB1) prognostic model was built. A significant difference was showed in the overall survival of patients between the two groups. Moreover, both type I and II interferon response pathways decreased in high-risk HCC patients. These results were also validated in the ICGC and internal cohorts.

Conclusions

10 CHGs in six GICs was revealed. In HCC, a prognostic model based on the 3 CHGs in GICs was constructed. Both type I and II interferon response pathways markedly decreased in the high-risk group of HCC patients. This might provide novel perspectives for the targeting and immunotherapy of GICs.

Hepatocellular carcinoma

GIC

GIC common hub gene signature

Immune infiltration

Gastrointestinal cancer (GIC) is one of the most lethal malignant tumors. About 1 million gastric cancer, 840,000 liver cancer, 570,000 esophageal cancer, 460,000 pancreatic cancer, and 1.8 million colorectal cancer cases were reported in 2018 worldwide. [1]. These cancers are mostly diagnosed late, which leads to poor prognosis [2]. The current study suggested that the diagnosis and treatment of GIC could be improved by identifying its predictive markers. The results of the current study showed similarities between prognostic markers and molecular therapeutic targets of various GIC types. For example, the human epidermal growth factor receptor 2 (HER2) is overexpressed to varying degrees in almost all histological types of GIC, including esophageal, gastric, pancreatic, and colon cancer. The majority of GIC can be treated with therapies targeting the HER2 mutation [3]. Therefore, HER2 can be considered a promising drug target for GIC. Numerous studies reported that microsatellite instability (MSI) was imperative in promoting the progression of colorectal and gastric cancer and could be used for the early detection of precancerous lesions. Immune checkpoint inhibitors (ICIs) can be used to treat MSI-related GICs [4, 5]. In esophageal cancer, tissue inhibitor metalloproteinase 1 (TIMP1) is highly expressed and correlated with tumor size and other clinicopathologic features. The patients with abnormal overexpression of TIMP1 exhibited lower overall survival in 76% of patients. In gastric and colon cancer, the high expression levels of TIMP1 were associated with a shorter time of disease progression, suggesting that it was a poor prognostic factor for the recurrence of gastric and colon cancer. Studies revealed that miR-6745/TIMP1 axis participated in the occurrence of gastric cancer; therefore, it could be used as a therapeutic target to prevent the progression of cancer [6]. Studies also showed that it was possible to find common GIC markers and use them as common therapeutic targets to develop a broad-spectrum anticancer drug for the simultaneous treatment of multiple GICs.

Liver cancer is among the top five causes of cancer-related deaths and among the top 10 cancers with the highest cases worldwide [7]. Mostly, HCC patients are diagnosed at a later stage due to the lack of previous symptoms, which leads to an overall poor prognosis [8, 9]. Therefore, the identification of effective biomarkers or molecular targets for the early diagnosis of HCC is urgently needed to improve its prognosis. Recently, immunotherapy using tumor markers, such as ICIs, revealed superior effectiveness against HCC in clinical trials [10]. Previous studies have shown that the tumor immune microenvironment (TIME) is involved in the immune response to tumors and immune escape [11]. The cytokines, including type I interferon (IFN) and IFN-γ, are produced by immune cells and are a key factor in effective anti-tumor immune response [12]. Studies have shown that HCC can escape the IFN-γ-related immune response by the loss of its receptors, induction of immunosuppressive cytokines, such as IL-10, and checkpoint ligands, such as PDL1 [13]. Type I IFN is beneficial to the immune response; however, it also triggers negative signals, such as the overexpression of IL-10 in HCC, which downregulates the T-cell stimulation ability of APC, promotes the upregulation of PDL1 in monocytes, and impairs the recruitment of tumor-specific human T cells [12], thereby, affecting the immunosuppressive state of the tumor. However, it is still unclear whether the core genes shared by GIC affect the TIME of HCC.

In the current study, the common hub genes (CHGs) of GIC were screened using public transcriptomic data. A signature model, based on 3 CHGs of the TCGA-HCC cohort, was built. This study aimed to find the characteristics of TIME and use internal and external cohorts for validation. The flow chart of the current study is shown in Fig. 1. The results revealed the CHGs of GIC as well as their potential relevance to prognosis and TIME in HCC patients, which might offer a new perspective on the therapy of both GIC and HCC.

2.1 Data Collection and Processing

The differentially expressed genes (DEGs) in various cancers, including colon, liver, esophageal, gastric, pancreatic, and rectal cancer, were obtained from the Gene Expression Profiling Interactive Analysis 2 (GEPIA2) database (http://gepia2.cancer-pku.cn/) using the R package "limma". The significance threshold of P-value and |log2(FC)| was adjusted at 0.01 and 1, respectively. Among 424 HCC patients, 361 patients had primary HCC according to pathological diagnosis. Clinical data were obtained from the Cancer Genome Atlas (TCGA) database (https://portal.gdc.cancer.gov/). Moreover, 50 non-tumor samples were also obtained from the GDC portal (accessed on January 14, 2022, https://portal.gdc.cancer.gov). Furthermore, the sequencing data of 231 HCC tumor samples obtained from the International Cancer Genome Consortium (ICGC) database (accessible on June 1, 2022) and 116 tumor samples obtained from the database of Guangxi Medical University Cancer Hospital, Nanning, China, were included in this study.

2.2 Functional Enrichment Analysis

The Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses were conducted using the Database for Annotation, Visualization, and Integrated Discovery (DAVID) database (https://david.ncifcrf.gov/) [14, 15]. The obtained DEGs were entered into the DAVID database, and the enrichment analyses were performed utilizing the "Homo sapiens" species as background to obtain the enriched GO functional annotations and KEGG pathway analysis outcomes, which were then displayed.

2.3 Protein-Protein Interaction Network Analysis

Based on obtained DEGs, a protein-protein interaction (PPI) network was constructed using the Search Tool for the Retrieval of Interacting Genes (STRING) database (https://cn.string-db.org/) [16]. The PPI network was then displayed and examined for modules using the Cytoscape program and CytoHubba plugin. The correlations among nodes and edges were evaluated to identify core genes using the Degree algorithm.

2.4 Expression and Survival Analysis of Hub Genes

The expression patterns of the core genes in cancer tissues, including colon, liver, esophageal, gastric, pancreatic, and rectal cancer, and non-cancer tissues were examined using the GEPIA database [17]. Based on the optimal cutoff value in cancer tissues, the expression levels of genes were classified into high and low categories. Moreover, the correlations between the expression levels of core genes and the patient’s prognosis were examined by generating the survival curves using overall survival and disease-free survival as endpoints. Furthermore, 116 RNA-seq liver cancer data obtained from our center were used to confirm the expression levels of core genes and prognosis [18].

2.5 Methylation Analysis

The epigenetic regulation of gene expression was evaluated using the University of ALabama at Birmingham cancer data analysis Portal (UALCAN) (http://ualcan.path.uab.edu/) based on the methylation of promoter regions in cancers using the TCGA database. The promoter methylation levels of core genes between healthy and liver cancer tissues were analyzed and compared using the UALCAN database [19, 20].

2.6 Construction and Verification of CHGs Prediction Model

The least absolute shrinkage and selection operator (LASSO) regression analysis was performed using the R package "glmnet" to eliminate overfitting. Moreover, four CHGs were screened out from 10 CHGs to construct the prognostic model in TCGA. The risk score of single samples was calculated using Eq. (1).

Risk score = ∑I = 1nCoef(Xi)∗ Exp(Xi) (1)

where Coef(Xi) and Exp(Xi) were risk coefficients and gene expression levels, respectively.

The samples of patients were categorized into two groups, including high-risk (n = 180) and low-risk (n = 181) groups based on the median risk scores.

The overall survival of patients was compared among subgroups using the Kaplan–Meier (K–M) survival curves. The receiver operating characteristic (ROC) curves over time were drawn using the R package "timeROC" according to risk scores. The accuracy and prediction abilities of the model were confirmed by calculating the area under the curve (AUC). Principal component analysis (PCA) and t-distributed stochastic neighbor embedding (t-SNE) were performed using the R packages "stats" and "Rtsne", respectively. The independent prognostic value was verified using univariate and multivariate Cox regression analyses, and the overall survival values of the two groups were predicted. Moreover, the reliability of the model’s risk scores was validated using the data of the ICGC and internal cohorts.

2.7 Analysis of Functional Enrichment and Immune Cell Infiltration

The expression levels of shared DEGs between two groups in the three cohorts were confirmed using the R package “limma” (fold-change > 1.2 and P < 0.05). The GO and KEGG analyses were performed using the R package "clusterprofiler". Moreover, immune-related scores of the two risk groups were calculated and validated using the single-sample Gene-Set Enrichment Analysis (ssGSEA) with a single sample using the R package “GSVA” [21].

2.8 Statistic Analyses

The expression levels of genes in tumor and non-tumor tissues were compared using a student’s t-test. The overall survival of patients in the two groups was compared using K–M survival curve analysis and log-rank test. Moreover, the difference in clinical information between the groups in the three cohorts was compared using a Chi-square test. The P-values were adjusted using Benjamini–Hochberg to confirm the weight of the difference, and the Mann-Whitney U-test was used to compare the ssGSEA scores of immune cells and pathways between the risk groups. The R software package (version 4.1.2) was used for all the statistical analyses. A two-sided P-value of < 0.05 was considered statistically significant.

3.1 Enrichment of Common DEGs in GIC

The DEGs in colorectal cancer were determined using the GEPIA database. A total of 2,706 genes showed upregulation and 2,682 genes showed downregulation. In liver cancer, 1,486 genes were upregulated, while 783 genes were downregulated. In esophageal cancer, 2,643 genes were upregulated while 1,435 genes were downregulated. In gastric cancer, 3,751 genes showed upregulation while 928 genes showed downregulation. In pancreatic cancer, 8,743 genes were upregulated and 483 genes were downregulated. Moreover, 2,995 upregulated genes and 2,949 downregulated genes were found in rectal cancer (Fig. 2A).

The intersection of the DEGS in these six cancers was taken to obtain the common DEGs in GIC, which included 240 upregulated and 25 downregulated genes (Fig. 2B, Supplementary Table 1).

Table 1

Survival analysis and P-values of ten hub genes
	TOP2A		CDK1	MAD2L1	RRM2	CCNA2	BUB1B	CDC6	AURKA	BUB1
GIC OS	0.0064	0.002	0.00029	0.00041	0.0072	0.0038	0.0017	0.00053	0.0051	0.0033
GIC DFS	0.0026	2.90E-06	2.00E-05	2.40E-09	1.40E-06	7.20E-05	4.00E-04	3.30E-08	1.20E-06	0.00028
LIHC OS	7.60E-05	1.80E-06	1.40E-05	0.0038	0.00035	0.0014	9.20E-05	0.00046	0.0021	8.40E-05
LIHC DFS	4.00E-05	6.00E-04	2.30E-06	0.0057	0.00032	0.0022	5.10E-05	0.00035	0.0014	0.00014

GIC, Gastrointestinal cancer; OS, Overall survival; DFS, Disease-free survival; LIHC, Liver hepatocellular carcinoma.

3.2 Enrichment Analysis and PPI of Common DEGs

The GO functional annotations and KEGG enrichment analysis were performed on the 265 common DEGs. The results revealed that the common DEGs were involved in the following GO annotations: biological processes (BP), including nuclear division, organelle fission, and chromosome segregation; cellular components (CC), including chromosome regions and dense bodies; and molecular functions (MF), including ATPase activity (Fig. 2C). The KEGG pathway analysis showed that the downregulated DEGs were more substantially involved in the protein digestion and absorption pathways, while the upregulated DEGs were involved in the cell cycle system (Fig. 2D).A PPI network of the 265 common DEGs was constructed, and the top 10 genes, including TOP2A, CCNB1, CDK1, MAD2L1, RRM2, CCNA2, BUB1B, CDC6, AURKA, and BUB1, were selected as CHGs (Fig. 2E).

3.3 Expression Levels and Survival Prognosis of CHGs

The expression levels of 10 CHGs and survival outcomes were examined using the GEPIA2 database. The expression levels of the 10 CHGs were considerably higher in tumor tissues as compared to those in non-tumor tissues in all six cancers (Fig. 3A, Supplementary Table 2). The survival analysis showed that the pan-cancer patients with higher expression levels of the 10 CHGs had shorter overall survival and disease-free survival (Fig. 3B and 3C), and the high expression levels of the 10 CHGs in each of the six cancers were significantly correlated with shorter overall survival and disease-free survival (Fig. 3D and 3E, Supplementary Fig. 1, Supplementary Table 3). The results of the expression levels and prognosis analysis of these 10 CHGs were validated using the RNA sequencing analysis of tumor tissues obtained from 116 HCC patients at Guangxi Medical University Cancer Hospital, Nanning, China. (Fig. 3F-3H, Supplementary Fig. 2, Supplementary Fig. 3).

3.4 Methylation Analysis of the CHGs

The methylation analysis of the 10 CHGs was performed using the UALCAN database. The results showed that seven genes, including BUB1, BUB1B, CCNA2, CCNB1, CDK1, RRM2, and TOP2A, exhibited a significant correlation between their promoter methylation levels and HCC (P-values were 5.58E-04 for BUB1, 1.35E-10 for BUB1B, 1.78E-14 for CCNA2, 1.96E-05 for CCNB1, 4.26E-06 for CDK1, 1.11E-16 for RRM2, and 3.21E-07 for TOP2A). The BUB1B gene showed higher methylation levels in its promoter region in cancer as compared to those in the adjacent normal tissue, while the other six genes (60%) exhibited lower methylation levels in their promoter regions in cancer as compared to those in the adjacent normal tissue. The aberrant methylation levels in regulatory areas accounted for a major portion of the differential gene expression in cancer (Supplementary Fig. 4).

3.5 Construction and Validation of Prognostic Models

The univariate analysis of the 10 CHGs revealed that they were related to the prognosis of HCC (Fig. 4C). LASSO-Cox regression analysis of the 10 CHGs showed three significant prognosis-related genes, including Cyclin B1 (CCNB1), Mitotic Arrest Deficient-like 1 (MAD2L1), and Budding Uninhibited by Benomyl1 (BUB1), which were further screened, as shown in Figs. 4A and 4B. A prognostic model of three genes was constructed, and the risk score was calculated using Eq. (2).

Prognostic models (Risk score = e ^{(0.229* CCNB1 expression + 0.012* MAD2L1 expression + 0.016* BUB1 expression)}) (2)

The patients were divided into a high-risk group (n = 180) and a low-risk group (n = 181) based on the selected median value of the risk scores (Fig. 4D). The PCA and t-SNE plots revealed differences in the orientations of the two groups, as shown in Figs. 4E and 4F, which suggested that the classification based on the median value of the risk score was more precise. K-M survival curve analysis revealed that the overall survival of the low-risk group was longer than that of the high-risk group (P < 0.001) (Fig. 6D), which suggested that the patients in the high-risk group had a significantly worse outcome than that of the low-risk group’s patients (Fig. 4H). ROC curve analysis revealed that the AUC values of the 1-, 2-, and 3-year survival rates were 0.752, 0.700, and 0.670, respectively, indicating that the survival risk scores had an excellent predictive ability for the prognosis of patients (Fig. 4I).

Moreover, the validity and robustness of the model were identified in our internal and ICGC RNA-seq data cohorts (Fig. 5).

3.6 Prognostic Potential of the Three-Gene Signature Model

Cox regression analysis was performed for multiple clinical characteristics. The univariate and multivariate regression analyses revealed that in the TCGA cohort, risk scores were related to overall survival. Thus, the risk scores were an excellent independent prognostic predictor factor (Fig. 6A). The findings were verified in the ICGC and internal cohorts (Figs. 6B and 6C).

3.7 Functional Enrichment Analysis in Three Cohorts

A total of 56 common DEGs between the high-risk and low-risk groups were obtained from all cohorts (Fig. 7A). The correlations between functions and biological pathways of DEGs and risk scores were explored by performing GO enrichment analysis (Fig. 7B). The results showed that the DEGs were significantly enriched in the following GO annotations: BPs, including maintenance of stem cell population and cell number; CC, including heterochromatin and collagen trimer; and MF, including DNA catalytic activity, carbon-oxygen lyase activity, and structural constituent of extracellular matrix. The KEGG analysis showed that these DEGs were significantly enriched in neuroactive ligand-receptor interactions (Fig. 7C).The scores of different immune pathways between the two groups were obtained using the ssGSEA algorithm to compare the differences in immune pathways. The comparison of the scores indicated that the co-stimulation scores of macrophages, dendritic cells (DC), immature DCs (iDCs), and Treg cells increased significantly in the high-risk group. Type I and II interferon response, neutrophils, and cytolytic activity were significantly reduced (P < 0.05, Fig. 7D and 7G). This reduction of type I and II interferon response in the high-risk group’s patients was validated in the ICGC and internal cohorts (P < 0.05, Fig. 7H and 7I).

The current study revealed that 10 CHGs, including TOP2A, CCNB1, CDK1, MAD2L1, RRM2, CCNA2, BUB1B, CDC6, AURKA, and BUB1, were associated with the prognosis of GIC. A prognostic model based on three common GIC CHGs, including CCNB1, MAD2L1, and BUB1, was established using LASSO-cox regression analysis to detect HCC prognosis. Moreover, the immune cell infiltration analysis showed a significant reduction in type I and type II IFN response pathways-related genes in high-risk HCC patients. These results were verified in the internal and ICGC cohorts.

Studies have shown that various prognostic-related hub genes, including integrin alpha subunit 5 (ITGA5), and microRNA, which are associated with the survival and clinicopathological features of GIC, might be potential biomarkers for the diagnosis and treatment of GIC [22, 23]. Studies investigating the role of the CHGs, which were identified in the current study in these GICs, are limited. The overexpression of TOP2A, a prognostic marker related to low survival rate, could induce the progression and recurrence of colorectal and liver cancers mainly through the regulation and replication of DNA topological state [24]. Wong et al. showed that TOP2A was overexpressed in HCC and was associated with histological grade, microvascular invasion, tumor progression, shortened survival, and resistance to chemotherapy [25]. CCNB1 could promote the phosphorylation of AKT and PI3K in liver cancer and play a crucial role in the abnormal cell cycle by promoting the ubiquitination of P53 as well as reducing its expression [26]. It has been reported that ellagic acid (EA), an important polyphenol compound, plays an anticancer role in various types of cancers, including colorectal, liver, and pancreatic cancers, by affecting different hub genes, such as CDK1, CCNA2, and CCNB1 [27]. MAD2L1, a pre-oncogene, with upregulated expression levels in gastric cancer, maintains the separation of chromosome and spindle during mitosis and acts as a checkpoint [28]. Wang et al. showed that the hepatitis B virus (HBV) could activate the expression of ribonucleoside reductase subunits M2 (RRM2) as well as the activity of the RR enzyme, thereby replicating viral DNA in host hepatocytes[29]. Moreover, RRM2 is also an important gene for Gemcitabine resistance in pancreatic cancer; therefore, RRM2 might be an important therapeutic target for HBsV-related HCC and pancreatic cancer [30].

Currently, studies on prognostic models related to HCC, such as copper death and cell senescence, have been successively conducted [31, 32]. The current study selected three genes among the 10 CHG genes to successfully construct a prognostic model of HCC. Recent studies suggested that the expression levels of CHGs, including CCNB1, MAD2L1, and BUB1, were greatly different in tumor and non-tumor tissues, and were greatly associated with the adverse survival rates of HCC patients [33]. These findings were consistent with the findings of the current study. Few studies have shown that CyclinB1 (CCNB1) is a key gene in the development of liver cancer [34]. Studies have shown that ncRNA-mediated upregulation of MAD2L1 is a predictor of adverse outcomes and tumor invasion in HCC [35]. MiR-200c-5p could inhibit the expression of the MAD2L1 gene, which led to inhibiting the migration, invasion, and proliferation of HCC cells [36]. In HCC, BUB1 is upregulated and is directly regulated by miR-490-5p, which leads to inhibiting the viability, proliferation, and migration of tumor cells and inducing cell cycle arrest and apoptosis [37]. Studies showed that these genes could significantly affect the progression and prognosis of HCC, thereby suggesting the prediction potential of the CHGs-based prognostic model.

Moreover, the functional enrichment analysis revealed that the common DEGs in the six GICs were mostly concentrated on the pathways related to cell proliferation and cell division. Studies have shown that the occurrence and development of both GIC and HCC are related to cell cycle disorders, and cyclin-dependent kinase 4/6 inhibitors are considered promising therapeutic strategies for GIC [38]. Therefore, CHGs might affect the cell cycle and promote the occurrence and progression of GIC by interfering with DNA or chromosome replication. Studies suggested that the changes in genomic and epigenomic DNA are present in almost all GICs [4]. Abnormal DNA methylation is a biomarker for the prognosis and diagnosis of GICs [39]. These findings were also consistent with those of the current study.

The treatment of HCC has been greatly changed by the use of ICIs [40, 41]. TIME performs an imperative part in the efficacy of ICIs [42]. IFN, a key factor of immune response in the TIME, is crucial for tumor immune surveillance, and the combination of type I and II IFNs could be regarded as a hopeful new anti-tumor therapy [43, 44]. In the current study, the immune functions of type I and II IFN response pathways were limited in the TIME of HCC. Studies have shown that ferroptosis and pyroptosis are significantly associated with the TIME of liver cancer. The low-risk group patients in the ferroptosis and pyroptosis models had a much better prognosis in predicting overall survival and anti-tumor immune function as compared to those in the high-risk group. In both internal and external cohorts of the ferroptosis and pyroptosis high-risk groups, the type I and II IFN response pathways were significantly lower than those in the low-risk group [45, 46]. This further validated the results of the current study and might be one of the specific mechanisms, by which, CHGs could affect the TIME of HCC. Moreover, a study reported that in tumor tissues, higher expression of IL-17 significantly inhibited IFN-γ-induced apoptosis in HCC cells and promoted HCC [47]. In NK cells in patients with chronic HBV infection, the secretion capacity of IFN-γ was reduced, which was involved in the progression of HCC. Moreover, the NK cell functions in HCC patients, including IFN-γ production, were downregulated as compared to those in healthy people [48]. IFIT3 could enhance the anti-tumor effects of IFN-α by promoting IFN-α response both in vitro and in vivo as well as could promote IFN-α response and therapeutic effects by enhancing IFN-α signaling in HCC [49]. These studies suggested that CHGs might inhibit the type I and II IFN response pathway and reduce the expression level of IFN, thereby, reducing its effects on HCC immunogenicity, immune infiltration, and adaptive immune attack as well as promoting the occurrence and progression of HCC. This also suggested that a decrease in the expression levels of IFN could disrupt the homeostasis of the cell cycle and cell death, mediated by the interaction of tumors with innate and adaptive immunity. The tumor cells that survive immune attacks are less sensitive to IFN; therefore, they are less immunogenic and are not detected easily by the adaptive immune system. Moreover, aged tumor cells can persist during this stage. Moreover, the cells that do not respond to IFN can acquire stem cell capabilities, such as self-renewal, to maintain the survival of this cell population, thereby promoting tumor survival [50]. In HCC, cytotoxic clearance of disseminated tumor cells is facilitated by the production of perforin by activated NK cells or by interactions between intrahepatic lymphocytes and Kupffer cells[51, 52]. CHGs might evade the role of immune effector cells, including liver-resident and liver-infiltrating NK cells, in the immune surveillance of HCC.

The current study built a reliable predictive prognostic model based on CHGs in HCC count using LASSO Cox regression analysis by searching for prognosis-related CHGs in GIC. Meanwhile, the model was verified using external and internal cohorts, which significantly improved the reliability of the data. However, there were certain limitations to the current study, such as the lack of experimental data to verify the correlations between CHGs and immune activity. Secondly, numerous prognostic genes and immune checkpoints of HCC were not considered, and only 19 immune cells were included, which might not be comprehensive enough; therefore, further research is needed.

In conclusion, the current study revealed 10 CHGs, including TOP2A, CCNB1, CDK1, MAD2L1, RRM2, CCNA2, BUB1B, CDC6, AURKA, and BUB1, in six GICs. A prognostic model was constructed based on three GIC CHGs, including CCNB1, MAD2L1, and BUB1 for HCC. Moreover, the results showed that the type I and II IFN response pathways were greatly reduced in the high-risk group of HCC patients. These findings might provide a novel perspective for targeted therapy of GIC as well as new ideas for immunotherapy of HCC.

Author Contributions:

Designed the research, analyzed the data, and revised the manuscript (BDX, CLY). Wrote the manuscript and participated in data preparation, experiments, analysis of data and figure preparation (BSC, SJL, ZZW, JWH, YKL, MZ, RS, HPY). Read and approved the manuscript for publication (all authors). All authors have read and agreed to the published version of the manuscript.

Acknowledgements

This research was supported by the National Natural Science Foundation of China (82260573, 81960450), High-level Innovation Team and Outstanding Scholar Program in Guangxi Colleges and Universities, Key Laboratory of Early Prevention and Treatment for Regional High Frequency Tumor, Ministry of Education/Guangxi, Independent Research Project (GKE-ZZ202215), Guangxi Key Research and Development Program (AA18221001, AB18050020). The authors would like to thank all the reviewers who participated in the review and MJEditor (www.mjeditor.com) for its linguistic assistance during the preparation of this manuscript.

Institutional Review Board Statement:

All procedures involving patient samples were approved by the Ethics Committee of Guangxi Medical University Affiliated Tumor Hospital, and written informed consent was obtained from all patients.

Informed Consent Statement:

Informed consent was obtained from all subjects involved in the study.

Data Availability Statement:

All datasets analyzed during the current study are available from the corresponding author upon reasonable request.

Conflicts of Interest:

The authors have no conflicts of interest related to this publication.

Arnold, M.; Abnet, C.C.; Neale, R.E.; Vignat, J.; Giovannucci, E.L.; McGlynn, K.A.; Bray, F. Global Burden of 5 Major Types of Gastrointestinal Cancer. Gastroenterology 2020, 159, 335-349 e315, doi:10.1053/j.gastro.2020.02.068.
Cao, W.; Chen, H.D.; Yu, Y.W.; Li, N.; Chen, W.Q. Changing profiles of cancer burden worldwide and in China: a secondary analysis of the global cancer statistics 2020. Chin Med J (Engl) 2021, 134, 783-791, doi:10.1097/CM9.0000000000001474.
El Dika, I.; Ilson, D.H. Current and future therapies for targeting HER2 mutations in gastrointestinal cancer. Expert Rev Anticancer Ther 2018, 18, 1085-1092, doi:10.1080/14737140.2018.1510324.
Grady, W.M.; Yu, M.; Markowitz, S.D. Epigenetic Alterations in the Gastrointestinal Tract: Current and Emerging Use for Biomarkers of Cancer. Gastroenterology 2021, 160, 690-709, doi:10.1053/j.gastro.2020.09.058.
Puliga, E.; Corso, S.; Pietrantonio, F.; Giordano, S. Microsatellite instability in Gastric Cancer: Between lights and shadows. Cancer Treat Rev 2021, 95, 102175, doi:10.1016/j.ctrv.2021.102175.
Macedo, F.C.; Cunha, N.; Pereira, T.C.; Soares, R.F.; Monteiro, A.R.; Bonito, N.; Valido, F.; Sousa, G. A prospective cohort study of TIMP1 as prognostic biomarker in gastric and colon cancer. Chin Clin Oncol 2022, 11, 43, doi:10.21037/cco-22-69.
Villanueva, A. Hepatocellular Carcinoma. N Engl J Med 2019, 380, 1450-1462, doi:10.1056/NEJMra1713263.
El-Serag, H.B.; Marrero, J.A.; Rudolph, L.; Reddy, K.R. Diagnosis and treatment of hepatocellular carcinoma. Gastroenterology 2008, 134, 1752-1763, doi:10.1053/j.gastro.2008.02.090.
Schmidt, S.; Follmann, M.; Malek, N.; Manns, M.P.; Greten, T.F. Critical appraisal of clinical practice guidelines for diagnosis and treatment of hepatocellular carcinoma. J Gastroenterol Hepatol 2011, 26, 1779-1786, doi:10.1111/j.1440-1746.2011.06891.x.
Yang, C.; Zhang, H.; Zhang, L.; Zhu, A.X.; Bernards, R.; Qin, W.; Wang, C. Evolving therapeutic landscape of advanced hepatocellular carcinoma. Nat Rev Gastroenterol Hepatol 2022, doi:10.1038/s41575-022-00704-9.
Zhang, Q.; Yu, X.; Zheng, Q.; He, Y.; Guo, W. A Molecular Subtype Model for Liver HBV-Related Hepatocellular Carcinoma Patients Based on Immune-Related Genes. Front Oncol 2020, 10, 560229, doi:10.3389/fonc.2020.560229.
Sangro, B.; Sarobe, P.; Hervas-Stubbs, S.; Melero, I. Advances in immunotherapy for hepatocellular carcinoma. Nat Rev Gastroenterol Hepatol 2021, 18, 525-543, doi:10.1038/s41575-021-00438-0.
Nagao, M.; Nakajima, Y.; Kanehiro, H.; Hisanaga, M.; Aomatsu, Y.; Ko, S.; Tatekawa, Y.; Ikeda, N.; Kanokogi, H.; Urizono, Y.; et al. The impact of interferon gamma receptor expression on the mechanism of escape from host immune surveillance in hepatocellular carcinoma. Hepatology 2000, 32, 491-500, doi:10.1053/jhep.2000.16470.
Sherman, B.T.; Hao, M.; Qiu, J.; Jiao, X.; Baseler, M.W.; Lane, H.C.; Imamichi, T.; Chang, W. DAVID: a web server for functional enrichment analysis and functional annotation of gene lists (2021 update). Nucleic Acids Res 2022, doi:10.1093/nar/gkac194.
Huang da, W.; Sherman, B.T.; Lempicki, R.A. Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat Protoc 2009, 4, 44-57, doi:10.1038/nprot.2008.211.
Szklarczyk, D.; Gable, A.L.; Lyon, D.; Junge, A.; Wyder, S.; Huerta-Cepas, J.; Simonovic, M.; Doncheva, N.T.; Morris, J.H.; Bork, P.; et al. STRING v11: protein-protein association networks with increased coverage, supporting functional discovery in genome-wide experimental datasets. Nucleic Acids Res 2019, 47, D607-D613, doi:10.1093/nar/gky1131.
Tang, Z.; Kang, B.; Li, C.; Chen, T.; Zhang, Z. GEPIA2: an enhanced web server for large-scale expression profiling and interactive analysis. Nucleic Acids Res 2019, 47, W556-W560, doi:10.1093/nar/gkz430.
Yang, C.L.; Qiu, X.; Lin, J.Y.; Chen, X.Y.; Zhang, Y.M.; Hu, X.Y.; Zhong, J.H.; Tang, S.; Li, X.Y.; Xiang, B.D.; et al. Potential Role and Clinical Value of PPP2CA in Hepatocellular Carcinoma. J Clin Transl Hepatol 2021, 9, 661-671, doi:10.14218/JCTH.2020.00168.
Chandrashekar, D.S.; Bashel, B.; Balasubramanya, S.A.H.; Creighton, C.J.; Ponce-Rodriguez, I.; Chakravarthi, B.; Varambally, S. UALCAN: A Portal for Facilitating Tumor Subgroup Gene Expression and Survival Analyses. Neoplasia 2017, 19, 649-658, doi:10.1016/j.neo.2017.05.002.
Chandrashekar, D.S.; Karthikeyan, S.K.; Korla, P.K.; Patel, H.; Shovon, A.R.; Athar, M.; Netto, G.J.; Qin, Z.S.; Kumar, S.; Manne, U.; et al. UALCAN: An update to the integrated cancer data analysis platform. Neoplasia 2022, 25, 18-27, doi:10.1016/j.neo.2022.01.001.
Rooney, M.S.; Shukla, S.A.; Wu, C.J.; Getz, G.; Hacohen, N. Molecular and genetic properties of tumors associated with local immune cytolytic activity. Cell 2015, 160, 48-61, doi:10.1016/j.cell.2014.12.033.
Macha, M.A.; Seshacharyulu, P.; Krishn, S.R.; Pai, P.; Rachagani, S.; Jain, M.; Batra, S.K. MicroRNAs (miRNAs) as biomarker(s) for prognosis and diagnosis of gastrointestinal (GI) cancers. Curr Pharm Des 2014, 20, 5287-5297, doi:10.2174/1381612820666140128213117.
Zhu, H.; Wang, G.; Zhu, H.; Xu, A. ITGA5 is a prognostic biomarker and correlated with immune infiltration in gastrointestinal tumors. BMC Cancer 2021, 21, 269, doi:10.1186/s12885-021-07996-1.
Carvalho, R.F.; do Canto, L.M.; Cury, S.S.; Frostrup Hansen, T.; Jensen, L.H.; Rogatto, S.R. Drug Repositioning Based on the Reversal of Gene Expression Signatures Identifies TOP2A as a Therapeutic Target for Rectal Cancer. Cancers (Basel) 2021, 13, doi:10.3390/cancers13215492.
Wong, N.; Yeo, W.; Wong, W.L.; Wong, N.L.; Chan, K.Y.; Mo, F.K.; Koh, J.; Chan, S.L.; Chan, A.T.; Lai, P.B.; et al. TOP2A overexpression in hepatocellular carcinoma correlates with early age onset, shorter patients survival and chemoresistance. Int J Cancer 2009, 124, 644-652, doi:10.1002/ijc.23968.
Xia, P.; Zhang, H.; Xu, K.; Jiang, X.; Gao, M.; Wang, G.; Liu, Y.; Yao, Y.; Chen, X.; Ma, W.; et al. MYC-targeted WDR4 promotes proliferation, metastasis, and sorafenib resistance by inducing CCNB1 translation in hepatocellular carcinoma. Cell Death & Disease 2021, 12, doi:10.1038/s41419-021-03973-5.
Cheshomi, H.; Bahrami, A.R.; Matin, M.M. Ellagic acid and human cancers: a systems pharmacology and docking study to identify principal hub genes and main mechanisms of action. Mol Divers 2021, 25, 333-349, doi:10.1007/s11030-020-10101-6.
Ding, X.; Fu, Q.; Chen, W.; Chen, L.; Zeng, Q.; Zhang, S.; He, L. Targeting of MAD2L1 by miR-515-5p involves the regulation of cell cycle arrest and apoptosis of colorectal cancer cells. Cell Biol Int 2022, 46, 840-848, doi:10.1002/cbin.11774.
<pterostilbene inhibits hepatocellula source am j cancer res so 2021 11 6 2975 2989.pdf>.
Fisher, S.B.; Patel, S.H.; Bagci, P.; Kooby, D.A.; El-Rayes, B.F.; Staley, C.A.; Adsay, N.V.; Maithel, S.K. An analysis of human equilibrative nucleoside transporter-1, ribonucleoside reductase subunit M1, ribonucleoside reductase subunit M2, and excision repair cross-complementing gene-1 expression in patients with resected pancreas adenocarcinoma. Cancer 2013, 119, 445-453, doi:10.1002/cncr.27619.
Gao, B.; Wang, Y.; Lu, S. Cellular senescence affects energy metabolism, immune infiltration and immunotherapeutic response in hepatocellular carcinoma. Sci Rep 2023, 13, 1137, doi:10.1038/s41598-023-28436-z.
Wang, Y.; Zhang, Y.; Wang, L.; Zhang, N.; Xu, W.; Zhou, J.; Zhao, Y.; Zhu, W.; Zhang, T.; Wang, L. Development and experimental verification of a prognosis model for cuproptosis-related subtypes in HCC. Hepatol Int 2022, 16, 1435-1447, doi:10.1007/s12072-022-10381-0.
Yang, W.X.; Pan, Y.Y.; You, C.G. CDK1, CCNB1, CDC20, BUB1, MAD2L1, MCM3, BUB1B, MCM2, and RFC4 May Be Potential Therapeutic Targets for Hepatocellular Carcinoma Using Integrated Bioinformatic Analysis. Biomed Res Int 2019, 2019, 1245072, doi:10.1155/2019/1245072.
Zou, Y.; Ruan, S.; Jin, L.; Chen, Z.; Han, H.; Zhang, Y.; Jian, Z.; Lin, Y.; Shi, N.; Jin, H. CDK1, CCNB1, and CCNB2 are Prognostic Biomarkers and Correlated with Immune Infiltration in Hepatocellular Carcinoma. Med Sci Monit 2020, 26, e925289, doi:10.12659/MSM.925289.
Liu, S.; Miao, M.; Kang, L. Upregulation of MAD2L1 mediated by ncRNA axis is associated with poor prognosis and tumor immune infiltration in hepatocellular carcinoma: A review. Medicine (Baltimore) 2023, 102, e32625, doi:10.1097/MD.0000000000032625.
Li, Y.; Bai, W.; Zhang, J. MiR-200c-5p suppresses proliferation and metastasis of human hepatocellular carcinoma (HCC) via suppressing MAD2L1. Biomed Pharmacother 2017, 92, 1038-1044, doi:10.1016/j.biopha.2017.05.092.
Xu, B.; Xu, T.; Liu, H.; Min, Q.; Wang, S.; Song, Q. MiR-490-5p Suppresses Cell Proliferation and Invasion by Targeting BUB1 in Hepatocellular Carcinoma Cells. Pharmacology 2017, 100, 269-282, doi:10.1159/000477667.
Zeng, F.; Zhou, Y.; Khowtanapanich, T.; Saengboonmee, C. Cyclin-Dependent Kinase 4/6 Inhibitors: A Potential Breakthrough Therapy for Malignancies of Gastrointestinal Tract. In Vivo 2022, 36, 1580-1590, doi:10.21873/invivo.12868.
Wong, C.C.; Li, W.; Chan, B.; Yu, J. Epigenomic biomarkers for prognostication and diagnosis of gastrointestinal cancers. Semin Cancer Biol 2019, 55, 90-105, doi:10.1016/j.semcancer.2018.04.002.
Chiew Woon, L.; Joycelyn Jie Xin, L.; Su Pin, C. Nivolumab for the treatment of hepatocellular carcinoma. Expert Opin Biol Ther 2020, 20, 687-693, doi:10.1080/14712598.2020.1749593.
Ren, D.; Hua, Y.; Yu, B.; Ye, X.; He, Z.; Li, C.; Wang, J.; Mo, Y.; Wei, X.; Chen, Y.; et al. Predictive biomarkers and mechanisms underlying resistance to PD1/PD-L1 blockade cancer immunotherapy. Mol Cancer 2020, 19, 19, doi:10.1186/s12943-020-1144-6.
Oura, K.; Morishita, A.; Tani, J.; Masaki, T. Tumor Immune Microenvironment and Immunosuppressive Therapy in Hepatocellular Carcinoma: A Review. Int J Mol Sci 2021, 22, doi:10.3390/ijms22115801.
Gocher, A.M.; Workman, C.J.; Vignali, D.A.A. Interferon-gamma: teammate or opponent in the tumour microenvironment? Nat Rev Immunol 2022, 22, 158-172, doi:10.1038/s41577-021-00566-3.
Koustas, E.; Trifylli, E.M.; Sarantis, P.; Papadopoulos, N.; Karapedi, E.; Aloizos, G.; Damaskos, C.; Garmpis, N.; Garmpi, A.; Papavassiliou, K.A.; et al. Immunotherapy as a Therapeutic Strategy for Gastrointestinal Cancer-Current Treatment Options and Future Perspectives. Int J Mol Sci 2022, 23, doi:10.3390/ijms23126664.
Fang, C.; Liu, S.; Feng, K.; Huang, C.; Zhang, Y.; Wang, J.; Lin, H.; Wang, J.; Zhong, C. Ferroptosis-related lncRNA signature predicts the prognosis and immune microenvironment of hepatocellular carcinoma. Sci Rep 2022, 12, 6642, doi:10.1038/s41598-022-10508-1.
Huo, J.; Cai, J.; Guan, G.; Liu, H.; Wu, L. A Ferroptosis and Pyroptosis Molecular Subtype-Related Signature Applicable for Prognosis and Immune Microenvironment Estimation in Hepatocellular Carcinoma. Front Cell Dev Biol 2021, 9, 761839, doi:10.3389/fcell.2021.761839.
Li, J.; Zeng, M.; Yan, K.; Yang, Y.; Li, H.; Xu, X. IL-17 promotes hepatocellular carcinoma through inhibiting apoptosis induced by IFN-gamma. Biochem Biophys Res Commun 2020, 522, 525-531, doi:10.1016/j.bbrc.2019.11.134.
Chen, Y.; Tian, Z. HBV-Induced Immune Imbalance in the Development of HCC. Front Immunol 2019, 10, 2048, doi:10.3389/fimmu.2019.02048.
Yang, Y.; Zhou, Y.; Hou, J.; Bai, C.; Li, Z.; Fan, J.; Ng, I.O.L.; Zhou, W.; Sun, H.; Dong, Q.; et al. Hepatic IFIT3 predicts interferon-alpha therapeutic response in patients of hepatocellular carcinoma. Hepatology 2017, 66, 152-166, doi:10.1002/hep.29156.
von Locquenghien, M.; Rozalen, C.; Celia-Terrassa, T. Interferons in cancer immunoediting: sculpting metastasis and immunotherapy response. J Clin Invest 2021, 131, doi:10.1172/JCI143296.
Bayon, L.G.; Izquierdo, M.A.; Sirovich, I.; van Rooijen, N.; Beelen, R.H.; Meijer, S. Role of Kupffer cells in arresting circulating tumor cells and controlling metastatic growth in the liver. Hepatology 1996, 23, 1224-1231, doi:10.1002/hep.510230542.
Brodbeck, T.; Nehmann, N.; Bethge, A.; Wedemann, G.; Schumacher, U. Perforin-dependent direct cytotoxicity in natural killer cells induces considerable knockdown of spontaneous lung metastases and computer modelling-proven tumor cell dormancy in a HT29 human colon cancer xenograft mouse model. Mol Cancer 2014, 13, 244, doi:10.1186/1476-4598-13-244.

No competing interests reported.

SupplementaryMaterial.docx

Download PDF

Version 1

posted

You are reading this latest preprint version

Identification of common hub genes in gastrointestinal cancers and their role in prognosis and Immune cell infiltration in hepatocellular carcinoma

Status:

Version 1

Abstract

Background

Methods

Results

Conclusions

Figures

1. Introduction

2 Material and Methods

2.1 Data Collection and Processing

2.2 Functional Enrichment Analysis

2.3 Protein-Protein Interaction Network Analysis

2.4 Expression and Survival Analysis of Hub Genes

2.5 Methylation Analysis

2.6 Construction and Verification of CHGs Prediction Model

2.7 Analysis of Functional Enrichment and Immune Cell Infiltration

2.8 Statistic Analyses

3 Results

3.1 Enrichment of Common DEGs in GIC

3.2 Enrichment Analysis and PPI of Common DEGs

3.3 Expression Levels and Survival Prognosis of CHGs

3.4 Methylation Analysis of the CHGs

3.5 Construction and Validation of Prognostic Models

3.6 Prognostic Potential of the Three-Gene Signature Model

3.7 Functional Enrichment Analysis in Three Cohorts

4 Discussion

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1