Sage Journals: Discover world-class research

Abstract

BACKGROUND:

Hepatocellular carcinoma (HCC) is a highly heterogeneous disease, which makes prognostic prediction challenging.We aimed to investigate association of TNFRSF4 expression with the immune infiltration and gene mutation in HCC.

METHODS:

In this study, the expression profiles and corresponding clinical data of HCC patients were downloaded from the Cancer Genome Atlas (TCGA) database.

Kaplan-Meier and Cox regression were used to evaluate the clinical value of TNFRSF4. ESTIMATE and CIBERSORT algorithms were applied to investigate the infiltration ratio of 22 immune cells. The WGCNA and LASSO COX algorithms were performed, establishing a prognostic risk model that was then validated by HCC samples from GEO. Finally, the effects on gene mutation occurring in HCC patients of TNFRSF4 expression and risk score were appraised.

RESULTS:

In HCC tissues, it was found the TNFRSF4 expression profile was significantly different with age, gender, tumor grade, disease stage, prominently affecting the survival outcome and prognosis of patients. Univariate and multivariate COX regression analysis suggested that TNFRSF4 was an independent prognostic marker. Samples of high/low expression of TNFRSF4 were screened for differential genes, and then the WGCNA and LASSO COX constructed a 13-gene signature, excellently dividing samples into hign/low risk groups. Compared with the low-risk group, the overall survival (OS) of high-risk group was markedly lower, with $P<$ 0.0001. By ROC curve analysis, the predictive ability of the 13-gene signature was further confirmed. Both the high/low TNFRSF4 expression and the high/low risk score were demonstrated to exert effects on the frequency of gene mutation in HCC.

CONCLUSIONS:

As an independent prognostic marker of HCC, TNFRSF4 was found simultaneously to affect the immune infiltration of cells and the frequency of gene mutations.

Keywords

TNFRSF4 hepatocellular carcinoma prognosis immune infiltration mutation

Abbreviations

AUC, area under curve;

GEO, gene expression omnibus;

GM, gene mutation;

HCC, hepatocellular carcinoma;

IRGs, immune-related genes;

KM, Kaplan-Meier;

LASSO COX, least absolute shrinkage and selection operator;

ROC, receiver operating characteristic;

RS, risk score;

TCGA, cancer genome atlas;

WGCNA, weighted gene co-expression network analy- sis.

1. Introduction

HCC is the most common type of liver cancer globally, is expected to significantly increase over the next 10 years [1, 2].

Although there are many ways to treat HCC, less than 9% of HCC patients can survive over five years [3]. The high recurrence and metastasis lead to the poor overall survival [4]. Therefore, the determination of reliable prognostic markers is crucial to improve the prognosis of HCC patients.

The tumor immune microenvironment is the environment where tumor cells interact with the immune system. In addition, the mutual interactions between tumor cells and the surrounding tumor microenvironment will eventually promote tumor development and metastasis [5]. Immunotherapy is the most significant way to treat cancer [6], wherein PD-1/PD-L1 pathway is regared as the best target for immunotherapy. Although many studies have reported the impact of PD-1 pathway on prognosis, fewer patients benefit from PD-1/PD-L1 immunotherapy [7, 8]. Thus, the immune microenvironment urgently awaits to be better understood to improve prognosis prediction.

TNFRSF4 (tumor necrosis factor receptor 4), a costimulatory receptor of the TNF family, plays a significant role in the immune microenvironment [9]. TNFRSF4 expression could be reduced by regulatory T (Treg) cell, thereby contributing to developments of various malignant tumors [10]. However, the clinical and prognostic role of TNFRSF4 in HCC development is perplexing. Therefore, the study aims to investigate the prognostic performance of TNFRSF4 in HCC.

Figure 1.

General flow chart.

By our previous researches, several prognostic biomarkers and potential inflammation-modulating therapeutic targets were identified through experimental and bioinformatics method [11, 12, 13]. In this study, univariate and multivariate COX were utilized to confirm the potentiality of the TNFRSF4 for improving the survival prediction for HCC. CIBERSORT algorithm was used to score immune cells including 22 kinds. Prognostic immune signatures were constructed based on WGCNA and LASSO COX algorithms. Finally, the effects of TNFRSF4 expression and risk score were evaluated on gene mutation frequency.

2. Materials and methods

TCGA provided patient gene expression and clinical characteristics data (https://tcga-data.nci.nih.gov/tcga/). The microarray dataset (GSE76427 from GEO) (https:// www.ncbi.nlm.nih.gov/geo/) was used to validate the results from the TCGA dataset. GEPIA2 (https://gepia2. cancer-pku.cn/) and ENCORI (https://starbase.sysu. edu.cn/) online databases were applied to carry out single gene screening. The flow diagram was indicated by Fig. 1.

2.1 Prognostic analysis

Univariate Cox analysis was implemented to appraise the association of the overall survival outcome with TNFRSF4 expression level as well as clinical characteristics as age, gender, stage (T, N, M). Following, Multivariate Cox analysis were performed, indicating TNFRSF4 was an independent prognostic marker of HCC (referring to Appendix 1, 2).

2.2 Immune cell infiltration

For all the 363 HCC samples (referring to Appendix 3), ESTIMATE algorithm was first used to calculate immune score, and then Wilcoxon test adopted to analyze differential expression of TNFRSF4, and the end, the CIBERSORT software [14] implemented to calculate the expression ratio of TNFRSF4 in 22 immune cells.

2.3 Differentially expressed genes (DEG)

The DESeq2 method was taken to yield differential expression of TNFRSF4 using the statistical test. Fold-change (FC) of gene expression was calculated, taking a threshold of logFC $>$ 1 and $P$ value $<$ 0.05 for DEG selection.

2.4 WGCNA analysis

The WGNCA (R package) was performed to construct the co-expression network [15]. In the network, paired Pearson coefficient was employed to assess the co-expression correlation among all genes in the adjacency matrix. Genes with a high degree of full correlation were clustered into the same module. Then, the topological overlap of internal modules and adjacent modules was served as the functional modules. Finally, the correlation between gene expression modules and traits were constructed.

$\displaystyle\alpha_{ij}=S_{ij}^{\beta},\quad S_{ij}=\left|{\textit{cor}\left(% {x_{i},x_{j}}\right)}\right|$ (1)

In Eq. (1), $x_{i}$ and $x_{j}$ indicate the expression vector of genes $i$ and $j$ , respectively. cor, denoting the Pearson coefficient of these two vectors, is converted into the interaction coefficient represented by $S_{ij}$ . Finally, we used a formula to convert the connection coefficient to the weighting coefficient $W_{ij}$ :

$\displaystyle W_{ij}=\frac{l_{ij}+\alpha_{ij}}{\textit{min}\{k_{i},k_{j}\}+1-% \alpha_{ij}}$ (2) $\displaystyle l_{ij}=\sum\limits_{u}\alpha_{iu}\alpha_{uj},\quad k_{i}=\sum% \limits_{u}\alpha_{iu}$ (3)

Finally, we conduct module mining to identify the relationship between gene expression modules and immune markers.

2.5 LASSO COX analysis

Under the univariate COX regression and LASSO regression analysis, 13 genes were incorporated into the model. The risk score (RS) of each sample was calculated using Eq. (4) below:

$\displaystyle\textit{Risk score}(\textit{RS})=\sum\limits_{i=1}^{N}(\textit{% Exp}_{i}\ast C_{i})$ (4)

Almonst, $N$ is the number of prognostic lumps that permeate immune cells, $\textit{Exp}_{i}$ is the expression profile of bump permeate immune cells, with the $C_{i}$ representative of the regression coefficient determined by LASSO analysis.

2.6 Mutation landscape analysis

The ‘maftools’ package in R [16] was executed to observe the effects of TNFRSF4 expression and RS grouping on gene mutations occuring in HCC samples (in this step, excluding samples occurring no mutations, 353 samples were left).

Figure 2.

Screening of single gene TNFRSF4.

Figure 3.

Clinical analysis of TNFRSF4. (A) Violin chart shows the relative expression of TNFRSF4 in normal and tumor samples; (B) KM curve of TCGA cohort high ( $n=$ 183) and low TNFRSF4 expression group ( $n=$ 182). (C) 1, 3 and 5 year ROC curve; (D–F) Box plot of TNFRSF4 expression in different disease stage and tumor grade.

All the statistical analysis in this manuscript were performed by R software (https://www.r-project.org/), with the creteria of $p<$ 0.05 for a $T$ -test of being statistical significance. The median of sample was used as the critical value dividing high/low groups.

3. Results

3.1 TNFRSF4 identification in HCC and its prognosis evaluation

318 genes were resulted from the intersection of HCC whole genome, immune-related genes, and the mutation genes, then subject to GEPIA 2 and ENCORI databases to analyze the survival outcome and differential expression, yielding 10 genes with the $P<$ 0.05. In consideration of previously reported 9 genes in HCC, the remained gene TNFRSF4 amongst was identified as a prognosis marker (Fig. 2).

The original microarray data was downloaded from TCGA, and the TNFRSF4 expression in HCC samples were analyzed. In Fig. 3A, it is indicated TNFRSF4 expression in HCC samples was up-regulated compared with the case of non-tumor samples. Kaplan-Meier analysis showed that the overall survival outcome of high TNFRSF4 expression group was poor ( $P<$ 0.0001, Fig. 3B). The ROC curve shows that the TNFRSF4 expression was of good predictive performance, with the 1, 3 and 5-year AUCs separately of 0.66, 0.60, and 0.64 (seeing Fig. 3C). Furthermore, the GEO data GSE76427 ( $n=$ 116) was used to verify the prognostic significance of gene TNFRSF4. Patients with high TNFRSF4 expression showed a low survival outcome with $P<$ 0.001, which was consistent with the results from TCGA. Moreover, the TNFRSF4 expression in HCC patients was greatly different in diverse stages (Fig. 3D–F).

Figure 4.

TNFRSF4 prognostic value. Univariate and multivariate regression analysis of different clinicopathological variables in TCGA (A–B) and GEO (C–D).

Subsequently, Univariate Cox regression analysis was implemented to evaluate the prognostic value of TNFRSF4 expression, age, gender, tumor grade and disease stage (Fig. 4A). Then subject to multivariate Cox regression, it was found TNFRSF4 expression was an independent prognostic biomarker (Fig. 4B). The conclusion was further confirmed by the GEO dataset (Fig. 4C and D), demonstrating the TNFRSF4 could be an independent biomarker for HCC.

3.2 Analysis of immune microenvironment of high/low TNFRSF4 expression group

Figure 5.

Comparisons of 22 important immune fractions between high/low TNFRSF4 groups. (A) Differences in the immune invasion of 22 cases with high and low TNFRSF4 expression; (B) In TCGA-LIHC, Violin chart showing the difference in the distribution of 22 immune cells in the high TNFRSF4 expression (red) and the low TNFRSF4 expression (blue). (C) In GEO-LIHC, Violin chart showing the difference in the distribution of 22 immune cells in the high TNFRSF4 expression (red) and the low TNFRSF4 expression (blue).

We used the CIBERSORT algorithm to investigate the immune infiltration level involved in 22 resistant cell types for HCC patients. There was a significant difference in the immune invasion of 22 cases with high and low TNFRSF4 expression (Fig. 5A). Also, we assess the correlation of 22 anti-tumor cell groups. We compare the effects of TNFRSF4 expression in the distribution of immune infiltration in both the TCGA and GEO datasets after removing samples with $P>$ 0.05. In the TCGA dataset, it is displayed by Fig. 5B that the T cells CD8, hair follicle helper T cells, T regulatory cells, macrophages M0, and mast cells in the T were significantly different in the high and low NFRSF4 expression. For the GEO dataset, it is indicated that the plasma cells, T CD4 memory activated cells, T cells CD8, macrophages M0, macrophages M2, and activated dendritic cells have striking differences in the high and low TNFRSF4 expression (Fig. 5C). Intersected with the results above, immune cells T cells CD8 and macrophages M0 were overlapped, thus selected for subsequent analysis.

3.3 The immune-assocaited modul screened by WGCNA

Figure 6.

Screening the highly relevant modules based on WGCNA. (A) The volcano map of differential genes in TNFRSF4 expression group; (B) Unscaled fitting index of soft threshold power ( $R^{2}=$ 0.85). (C) Dendrogram of differential gene clusters based on different indicators. (D) Association diagram between modules. (E) Heatmap drew by gene co-expression similarity matrix. (F) Heat map of the correlation between gene modules and immune markers.

In Fig. 6A, there were 1475 differential genes (referring to appendix4) of high and low TNFRSF4 expression, including 1384 up-regulated genes (colored in red) and 91 down-regulated genes (colored in blue). Then, these 1475 differential genes were subject to WGCNA analysis to explore the relation of the TNFRSF4 expression with immune markers. Genes as CX3CR1, KLRG1 and GNLY representing T CD8 cell markers, and CYP27A1, CD68 and CXLC8 denoting macrophage M0 markers, were selected out as traits for co-expression analysis [17, 18, 19, 20, 21, 22, 23]. Average linkage clustering identified 5 gene co-expression modules, with each module separately containing 42, 81, 277, 476, and 579 genes colored by gray, yellow, brown, blue, and cyan (Fig. 6C). Correlation analysis between the modules and immune markers revealed that TNFRSF4 expression level was greatly related to the turquoise module (Fig. 6F).

3.4 Construction of the 13-gene signature based on risk score

Figure 7.

Identification of immune-related genes. (A–B) LASSO analysis determines the number of factors. (C) Analysis of prognostic risk scoring model. (D) Kaplan-Meier survival curve in the TCGA cohort. (E) 1, 3 and 5 year ROC curve.

Included in the turquoise module, 579 IRGs were related to the survival outcome, and then subject to LASSO analysis, ultimately 13 IRGs were determined as the 13-gene signature, referring to ART5, C15orf48, CCR3, CCR7, CD8A, LPCAT1, MMP10, S100A9, SLC16A3, SLC1A5, SMOX, SPP1, UBASH3A (Fig. 7A and B). In Fig. 7C, HCC patients were divided into high/low risk group according to the risk scoring model. Subsequently, Kaplan-Meier analysis was implemented in Fig. 7D, showing that the overall survival rate of the high-risk group was lower than that of the low-risk group. In the time-dependent ROC (tROC) curves, it was demonstrated that the classifier possessed a strong predictive ability, with the AUCs of 1, 3, and 5 years of 0.785, 0.778, and 0.775, respectively (Fig. 7E).

3.5 Mutation landscapes

Figure 8.

Gene mutation waterfall chart. (A–B) The waterfall chart of gene mutation between high-risk and low-risk groups of signature. (C–D) Cascade of gene mutation between high and low TNFRSF4 expression groups.

In this step, mutation landscapes were depicted for 353 samples, including 174 samples of high TNFRSF4 expression and high risk score (referring to Fig. 8A and C), and 179 samples of low TNFRSF4 expression and low risk score (referring to Fig. 8B and D). By exploring the correlation of TNFRSF4 expression and gene mutations for the TCGA dataset, it was found that the gene mutation frequency in TNFRSF4 high expression group was higher than that of the low. Moreover, the situation was the same in high and low-risk groups. Amonst, the gene mutation frequency of TP53, TIN and CTNNB1 were the highest, indicating that the gene mutation frequency was influenced by high/low TNFRSF4 expression and highlow risk score. Conclusively, TNFRSF4 could be served as a biomarker for HCC classification.

4. Discussion

TNFRSF4 uses T cells to conduct costimulatory signals and enhance cell proliferation, survival, and migration [24]. The prognostic role of the TNFRSF4 remains to be explored in liver cancer. The study directed an in-depth analysis of the expression, mutation, and prognostic value of TNFRSF4 in HCC. It was believed the TNFRSF4 expression in HCC was higher than that in normal liver tissue, consistent with the conclusion of previous studies [25]. Upregulated in clinical HCC samples, TNFRSF4 may serve as a potential prognostic biomarker.

TNFRSF4, a T cell costimulatory molecule, is associated with activation of inflammatory immunity in HCC [26]. The activation of TNFRSF4 promoted the inflation of activated T cells and memory T cells to exacerbate autoimmune diseases [27] . Besides, TNFRSF4 imaging probe predicted early local immunotherapy for HCC [28]. Piconese et al. found that TNFRSF4 can induce the proliferation of regulatory T cells in HCC [29]. However, few efforts were made on the prognostic gene signature associated with TNFRSF4.The study dedicated to exploring the relationship of TNFRSF4 expression with immune infiltration and frequency of gene mutation.

The effects of TNFRSF4 on the immune process were reported to be significant [30, 31], and the interaction between them could reduce the immune suppression of regulatory T cells, thus enhancing the immune responses to definite antigens [32, 33]. Moreover, we also found the TNFRSF4 expression were in correlation with different immune markers and immune infiltration. Therefore, it was inferred the TNFRSF4 expression may have a potential impact on tumor immunity.

In this manuscript, it was found the TNFRSF4 expression was greatly different in the immune infiltration level of T CD8 cells and macrophages M0. By previous investigations, the average density of T CD8 cells in liver cancer was indicated to be significantly lower than that in normal liver tissue, thereby served as a benchmark for evaluating tumor immune checkpoints [34, 35]. T CD8 cells differentiate into an exhaustion state in HCC to form a strong barrier against powerful anti-tumor immunity [36]. Macrophage M0 was believed to be a poor prognostic factor of HCC tumor immune microenvironment [37]. Overexpression of immune-related genes C15orf48 [38, 39], CCR3 [40, 41], CCR7 [42], MMP10 [43], S100A9 [44], SLC1A5 [45] and SMOX [46] are involved in the progression of many malignancies play a key role. The expression levels of CD8A [47] and SLC16A3 [48] were positively correlated with high immune infiltration in multiple tumor types. LPCAT1 [49] and SPP1 [49] may promote HCC progression by accelerating cell growth, migration and metastasis. However, the biological roles of ART5 and UBASH3A in HCC remain unclear. Integratively, the relationship between different immune cell markers and the TNFRSF4 expression level suggested the TNFRSF4 was important to regulating the immune micro-environment of HCC.

Following, HCC patients were divided into high and low-risk groups to further explore the prognostic model, displaying TP53 was greatly up-regulated in both the high TNFRSF4 expression group and high-risk group. TP53, as a tumor suppressor, its mutation accounted for a relatively large proportion of HCC [51]. For the low TNFRSF4 expression group and the low-risk group, CTNNB1 were significantly up-regulated. CTNNB1 mutations can cause a high enrichment of DNA methylation at the CpG sites in the HCC [52]. Accordingly, TNFRSF4 expression was believed to not only affect the immune microenviroment, but also the gene mutation in liver cancer.

5. Conclusions

Conclusively, TNFRSF4 expression was suggested to be an independent indicator for clinical prognosis of HCC, and high expression of TNFRSF4 is associated with immune infiltration and frequency of gene mutation. However, the potentiality of TNFRSF4 in HCC still awaits to be further experimentally validated.

Ethical statement

Our study did not require an ethical board approval because it did not contain human or animal trials.

Consent for publication

Not applicable.

Availability of data and materials

All data generated or analyzed during this study are included in this published article.

Authors contributions

Conception: Huan Hu.

Interpretation or analysis of data: Huan Ding.

Preparation of the manuscript: Di Wang.

Revision for important intellectual content: Qingjia Chi.

Supervision: Feifei Tian, Han Zhao.

Footnotes

Acknowledgments

We appreciate the generosity of TCGA and GEO database for sharing hepatocellular carcinomadata.

Conflict of interest

The authors have declared that no competing interests exist.

References

Bray

Ferlay

Soerjomataram

Siegel

R.L.

Torre

L.A.

and Jemal

, Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries, Ca-a Cancer Journal for Clinicians 68(6) (2018), 394–424.

Valery

P.C.

Laversanne

Clark

P.J.

Petrick

J.L.

McGlynn

K.A.

and Bray

, Projections of primary liver cancer to 2030 in 30 countries worldwide, Hepatology 67(2) (2018), 600–611.

Karan

Umair

Ashton

Natasha

Jason

S.S.

Jeremy

C.J.

and Kabir

, Factors affecting time to treatment in hepatocellular cancer, Journal of Clinical Oncology 38(15_suppl) (2020).

Sim

H.W.

and Knox

, Hepatocellular carcinoma in the era of immunotherapy, Current Problems in Cancer 42(1) (2018), 40–48.

Lei

J.K.

W.X.

R.G.

Yang

and Tan

H.B.

, Immune cells within the tumor microenvironment: Biological functions and roles in cancer immunotherapy, Cancer Letters 470 (2020), 126–133.

Abbott

and Ustoyev

, Cancer and the Immune System: The History and Background of Immunotherapy, Seminars in Oncology Nursing 35(5) (2019), 150923.

Calderaro

Rousseau

Amaddeo

Mercey

Charpy

Costentin

Luciani

Zafrani

E.S.

Laurent

Azoulay

Lafdil

and Pawlotsky

J.M.

, Programmed death ligand 1 expression in hepatocellular carcinoma: Relationship With clinical and pathological features, Hepatology 64(6) (2016), 2038–2046.

Sideras

Biermann

Verheij

Takkenberg

B.R.

Mancham

Hansen

B.E.

Schutz

H.M.

de Man

R.A.

Sprengers

Buschow

S.I.

Verseput

M.C.

Boor

P.P.

Pan

van Gulik

T.M.

Terkivatan

Ijzermans

J.N.

Beuers

U.H.

Sleijfer

Bruno

M.J.

and Kwekkeboom

, PD-L1, Galectin-9 and CD8(+) tumor-infiltrating lymphocytes are associated with survival in hepatocellular carcinoma, Oncoimmunology 6(2) (2017), e1273309.

Xie

Sun

and Wang

, The OX40/OX40L Axis Regulates T Follicular Helper Cell Differentiation: Implications for Autoimmune Diseases, Frontiers in Immunology 12 (2021), 670637.

10.

Smith

C.A.

Farrah

and Goodwin

R.G.

, The TNF receptor superfamily of cellular and viral proteins: activation, costimulation, and death, Cell 76(6) (1994), 959–962.

11.

Tian

Zhou

Kang

Luo

Fan

Yan

and Liang

, The small-molecule inhibitor selectivity between IKKalpha and IKKbeta kinases in NF-kappaB signaling pathway, Journal of Receptors and Signal Transduction 35(4) (2015), 307–318.

12.

Tian

L.X.

Tang

Wang

Zhang

Liu

Chen

Zhu

J.Y.

and Liang

H.P.

, Knockout of cytochrome P450 1A1 enhances lipopolysaccharide-induced acute lung injury in mice by targeting NF-kappaB activation, Febs Open Bio 10(11) (2020), 2316–2328.

13.

Ding

Tian

and Liang

, A dual immune signature of CD8+ T cells and MMP9 improves the survival of patients with hepatocellular carcinoma, Bioscience Reports 41(3) (2021).

14.

Chen

Khodadoust

M.S.

Liu

C.L.

Newman

A.M.

and Alizadeh

A.A.

, Profiling Tumor Infiltrating Immune Cells with CIBERSORT, Methods Mol Biol 1711 (2018), 243–259.

15.

Langfelder

and Horvath

, WGCNA: an R package for weighted correlation network analysis, Bmc Bioinformatics 9 (2008), 559.

16.

Mayakonda

Lin

D.C.

Assenov

Plass

and Koeffler

H.P.

, Maftools: efficient and comprehensive analysis of somatic variants in cancer, Genome Research 28(11) (2018), 1747–1756.

17.

Gerlach

Moseman

E.A.

Loughhead

S.M.

Alvarez

Zwijnenburg

A.J.

Waanders

Garg

de la Torre

J.C.

and von Andrian

U.H.

, The Chemokine Receptor CX3CR1 Defines Three Antigen-Experienced CD8 T Cell Subsets with Distinct Roles in Immune Surveillance and Homeostasis, Immunity 45(6) (2016), 1270–1284.

18.

Mach

Gao

Lemonnier

Lecardonnel

Oswald

I.P.

Estelle

and Rogel-Gaillard

, The peripheral blood transcriptome reflects variations in immunity traits in swine: towards the identification of biomarkers, Bmc Genomics 14 (2013), 894.

19.

Park

G.H.

Kim

K.Y.

Cheong

J.Y.

Cho

S.W.

and Kwack

, Association of GNLY genetic polymorphisms with chronic liver disease in a Korean population, Dna and Cell Biology 31(9) (2012), 1492–1498.

20.

Herndler-Brandstetter

Ishigame

Shinnakasu

Plajer

Stecher

Zhao

Lietzenmayer

Kroehling

Takumi

Kometani

Inoue

Kluger

Kaech

S.M.

Kurosaki

Okada

and Flavell

R.A.

, KLRG1(+) Effector CD8(+) T Cells Lose KLRG1, Differentiate into All Memory T Cell Lineages, and Convey Enhanced Protective Immunity, Immunity 48(4) (2018), 716–729.

21.

Gyoneva

Hosur

Gosselin

Zhang

Ouyang

Cotleur

A.C.

Peterson

Allaire

Challa

Cullen

Roberts

Miao

Reynolds

T.L.

Glass

C.K.

Burkly

and Ransohoff

R.M.

, Cx3cr1-deficient microglia exhibit a premature aging transcriptome, Life Science Alliance 2(6) (2019).

22.

Meng

Liu

Guan

Fan

Zhou

Zhang

and Liang

, The establishment of immune infiltration based novel recurrence predicting nomogram in prostate cancer, Cancer Medicine 8(11) (2019), 5202–5213.

23.

Stenger

Hanson

D.A.

Teitelbaum

Dewan

Niazi

K.R.

Froelich

C.J.

Ganz

Thoma-Uszynski

Melian

Bogdan

Porcelli

S.A.

Bloom

B.R.

Krensky

A.M.

and Modlin

R.L.

, An antimicrobial activity of cytolytic T cells mediated by granulysin, Science 282(5386) (1998), 121–125.

24.

Nuebling

Schumacher

C.E.

Hofmann

Hagelstein

Schmiedel

B.J.

Maurer

Federmann

Rothfelder

Roerden

Dorfel

Schneider

Jung

and Salih

H.R.

, The Immune Checkpoint Modulator OX40 and Its Ligand OX40L in NK-Cell Immunosurveillance and Acute Myeloid Leukemia, Cancer Immunology Research 6(2) (2018), 209–221.

25.

Gaspar

Pravin

Rodrigues

Uhlenbroich

Everett

K.L.

Wollerton

Morrow

Tuna

and Brewis

, CD137/OX40 Bispecific Antibody Induces Potent Antitumor Activity that Is Dependent on Target Coengagement, Cancer Immunology Research 8(6) (2020), 781–793.

26.

Xie

Liao

Luo

Liao

Gong

Deng

Yuan

and Zeng

, OX40 expression in hepatocellular carcinoma is associated with a distinct immune microenvironment, specific mutation signature, and poor prognosis, Oncoimmunology 7(4) (2018), e1404214.

27.

Huang

Pan

Liu

Lin

Zhan

Zhang

and Zhan

, Immune Function and Mechanism of Costimulating Molecules PD-1 and OX40 in Rheumatoid Arthritis, Journal of Interferon and Cytokine Research 40(11) (2020), 530–539.

28.

Chen

Cao

Gao

Jin

and Shan

, Early Therapeutic Vaccination Prediction of Hepatocellular Carcinoma via Imaging OX40-Mediated Tumor Infiltrating Lymphocytes, Molecular Pharmaceutics 16(10) (2019), 4252–4259.

29.

Piconese

Timperi

Pacella

Schinzari

Tripodo

Rossi

Guglielmo

Mennini

Grazi

G.L.

Di Filippo

Brozzetti

Fazzi

Antonelli

Lozzi

M.A.

Sanchez

and Barnaba

, Human OX40 tunes the function of regulatory T cells in tumor and nontumor areas of hepatitis C virus-infected liver tissue, Hepatology 60(5) (2014), 1494–1507.

30.

Rothfelder

Hagelstein

Roerden

Blumenstock

Hofmann

Nuebling

Jung

Salih

H.R.

and Dorfel

, Expression of the Immune Checkpoint Modulator OX40 in Acute Lymphoblastic Leukemia Is Associated with BCR-ABL Positivity, Neoplasia 20(11) (2018), 1150–1160.

31.

Gorris

Halilovic

Rabold

van Duffelen

Wickramasinghe

I.N.

Verweij

Wortel

Textor

J.C.

de Vries

and Figdor

C.G.

, Eight-Color Multiplex Immunohistochemistry for Simultaneous Detection of Multiple Immune Checkpoint Molecules within the Tumor Microenvironment, Journal of Immunology 200(1) (2018), 347–354.

32.

Deng

Zhao

Zhang

Jia

Wang

Zhou

and He

, OX40 (CD134) and OX40 ligand, important immune checkpoints in cancer, Oncotargets and Therapy 12 (2019), 7347–7353.

33.

Lin

Zhang

and Zhang

, Therapeutic strategies for the costimulatory molecule OX40 in T-cell-mediated immunity, Acta Pharmaceutica Sinica B 10(3) (2020), 414–433.

34.

Yarchoan

Xing

Luan

Sharma

R.B.

Popovic

Pawlik

T.M.

Kim

A.K.

Zhu

Jaffee

E.M.

Taube

J.M.

and Anders

R.A.

, Characterization of the Immune Microenvironment in Hepatocellular Carcinoma, Clinical Cancer Research 23(23) (2017), 7333–7339.

35.

Yang

Y.M.

Kim

S.Y.

and Seki

, Inflammation and Liver Cancer: Molecular Mechanisms and Therapeutic Targets, Seminars in Liver Disease 39(1) (2019), 26–42.

36.

Zheng

Goswami

Meng

Zhang

Cao

Zhu

Zhang

Duan

Chen

Gao

and Zhang

, PD1(Hi) CD8(+) T cells correlate with exhausted signature and poor clinical outcome in hepatocellular carcinoma, Journal for Immunotherapy of Cancer 7(1) (2019), 331.

37.

Farha

Jairath

N.K.

Lawrence

T.S.

and El

N.I.

, Characterization of the Tumor Immune Microenvironment Identifies M0 Macrophage-Enriched Cluster as a Poor Prognostic Factor in Hepatocellular Carcinoma, Jco Clinical Cancer Informatics 4 (2020), 1002–1013.

38.

Zhou

Wang

Luo

Ding

Zhang

Wang

and Liu

, A novel gene, NMES1, downregulated in human esophageal squamous cell carcinoma, International Journal of Cancer 101(4) (2002), 311–316.

39.

Sova

Feng

Geiss

Wood

Strauss

Rudolf

Lieber

and Kiviat

, Discovery of novel methylation biomarkers in cervical carcinoma by global demethylation and microarray analysis, Cancer Epidemiology Biomarkers & Prevention 15(1) (2006), 114–123.

40.

Johrer

Zelle-Rieser

Perathoner

Moser

Hager

Ramoner

Gander

Holtl

Bartsch

Greil

and Thurnher

, Up-regulation of functional chemokine receptor CCR3 in human renal cell carcinoma, Clinical Cancer Research 11(7) (2005), 2459–2465.

41.

Sharma

Singh

Sharma

and Saxena

, Gene Expression Profiling of Chemokines and Their Receptors in Low and High Grade Astrocytoma, Asian Pac J Cancer Prev 18(5) (2017), 1307–1313.

42.

Muller

Homey

Soto

Catron

Buchanan

M.E.

McClanahan

Murphy

Yuan

Wagner

S.N.

Barrera

J.L.

Mohar

Verastegui

and Zlotnik

, Involvement of chemokine receptors in breast cancer metastasis, Nature 410(6824) (2001), 50–56.

43.

Klupp

Neumann

Kahlert

Diers

Halama

Franz

Schmidt

Koch

Weitz

Schneider

and Ulrich

, Serum MMP7, MMP10 and MMP12 level as negative prognostic markers in colon cancer patients, Bmc Cancer 16 (2016), 494.

44.

Jiang

Wang

Zhang

and Cao

, Predictive value of S100A9 for lymph node metastasis in cervical cancer, Zhong Nan Da Xue Xue Bao Yi Xue Ban 45(6) (2020), 701–708.

45.

Yoo

H.C.

Park

S.J.

Nam

Kang

Kim

Yeo

J.H.

Kim

J.K.

Heo

Lee

H.S.

Lee

M.Y.

Lee

C.W.

Kang

J.S.

Kim

Y.H.

Lee

Choi

Hwang

G.S.

Bang

and Han

J.M.

, A Variant of SLC1A5 Is a Mitochondrial Glutamine Transporter for Metabolic Reprogramming in Cancer Cells, Cell Metabolism 31(2) (2020), 267–283.

46.

Huang

Wang

Zhang

H.J.

Zhou

Y.L.

and Shi

J.H.

, SMOX expression predicts the prognosis of non-small cell lung cancer, Annals of Translational Medicine 9(13) (2021), 1048.

47.

Zhu

Xiao

Wang

and Hu

, Pan-Cancer Molecular Characterization of m(6)A Regulators and Immunogenomic Perspective on the Tumor Microenvironment, Frontiers in Oncology 10 (2020), 618374.

48.

Xue

Liu

Xie

and Luo

, Prognostic Value of SLC16A3(MCT4) in Lung Adenocarcinoma and Its Clinical Significance, International Journal of General Medicine 14 (2021), 8413–8425.

49.

R.Q.

J.D.

X.F.

Dang

Y.W.

Yang

L.J.

Huang

Z.G.

Liu

L.M.

Liao

L.F.

Yang

and Chen

, LPCAT1 overexpression promotes the progression of hepatocellular carcinoma, Cancer Cell International 21(1) (2021), 442.

50.

Long

Zhang

Wan

Lin

Bai

Xiong

and Zhao

, A four-gene-based prognostic model predicts overall survival in patients with hepatocellular carcinoma, Journal of Cellular and Molecular Medicine 22(12) (2018), 5928–5938.

51.

Comprehensive and Integrative Genomic Characterization of Hepatocellular Carcinoma, Cell 169(7) (2017), 1327–1341.

52.

Song

Ravasz

Dong

Wang

and Ewing

R.M.

, A Protein Interaction between beta-Catenin and Dnmt1 Regulates Wnt Signaling and DNA Methylation in Colorectal Cancer Cells, Molecular Cancer Research 13(6) (2015), 969–981.

Elevated expression of TNFRSF4 impacts immune cell infiltration and gene mutation in hepatocellular carcinoma

Abstract

BACKGROUND:

METHODS:

RESULTS:

CONCLUSIONS:

Keywords

Abbreviations

1. Introduction

2.1 Prognostic analysis

2.2 Immune cell infiltration

2.3 Differentially expressed genes (DEG)

2.4 WGCNA analysis

3.1 TNFRSF4 identification in HCC and its prognosis evaluation

5. Conclusions

Ethical statement

Consent for publication

Availability of data and materials

Authors contributions

Footnotes

Acknowledgments

Conflict of interest

References