Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2021 Sep 23:9:686664.
doi: 10.3389/fcell.2021.686664. eCollection 2021.

Systematic Characterization of Novel Immune Gene Signatures Predicts Prognostic Factors in Hepatocellular Carcinoma

Dafeng Xu  1 , Yu Wang  2 , Jincai Wu  1 , Yuliang Zhang  3 , Zhehao Liu  1 , Yonghai Chen  1 , Jinfang Zheng  1
Affiliations

Systematic Characterization of Novel Immune Gene Signatures Predicts Prognostic Factors in Hepatocellular Carcinoma

Dafeng Xu et al. Front Cell Dev Biol. .

Abstract

Background: The prognosis of patients with hepatocellular carcinoma (HCC) is negatively affected by the lack of effective prognostic indicators. The change of tumor immune microenvironment promotes the development of HCC. This study explored new markers and predicted the prognosis of HCC patients by systematically analyzing immune characteristic genes. Methods: Immune-related genes were obtained, and the differentially expressed immune genes (DEIGs) between tumor and para-cancer samples were identified and analyzed using gene expression profiles from TCGA, HCCDB, and GEO databases. An immune prognosis model was also constructed to evaluate the predictive performance in different cohorts. The high and low groups were divided based on the risk score of the model, and different algorithms were used to evaluate the tumor immune infiltration cell (TIIC). The expression and prognosis of core genes in pan-cancer cohorts were analyzed, and gene enrichment analysis was performed using clusterProfiler. Finally, the expression of the hub genes of the model was validated by clinical samples. Results: Based on the analysis of 730 immune-related genes, we identified 64 common DEIGs. These genes were enriched in the tumor immunologic related signaling pathways. The first 15 genes were selected using RankAggreg analysis, and all the genes showed a consistent expression trend across multi-cohorts. Based on lasso cox regression analysis, a 5-gene signature risk model (ATG10, IL18RAP, PRKCD, SLC11A1, and SPP1) was constructed. The signature has strong robustness and can stabilize different cohorts (TCGA-LIHC, HCCDB18, and GSE14520). Compared with other existing models, our model has better performance. CIBERSORT was used to assess the landscape maps of 22 types of immune cells in TCGA, GSE14520, and HCCDB18 cohorts, and found a consistent trend in the distribution of TIIC. In the high-risk score group, scores of Macrophages M1, Mast cell resting, and T cells CD8 were significantly lower than those of the low-risk score group. Different immune expression characteristics, lead to the different prognosis. Western blot demonstrated that ATG10, PRKCD, and SPP1 were highly expressed in cancer tissues, while IL18RAP and SLC11A1 expression in cancer tissues was lower. In addition, IL18RAP has a highly positive correlation with B cell, macrophage, Neutrophil, Dendritic cell, CD8 cell, and CD4 cell. The SPP1, PRKCD, and SLC11A1 genes have the strongest correlation with macrophages. The expression of ATG10, IL18RAP, PRKCD, SLC11A1, and SPP1 genes varies among different immune subtypes and between different T stages. Conclusion: The 5-immu-gene signature constructed in this study could be utilized as a new prognostic marker for patients with HCC.

Keywords: HCC; immune gene signatures; pan-cancer; prognosis; tumor immune infiltration cell.

PubMed Disclaimer

Conflict of interest statement

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Figures

FIGURE 1
FIGURE 1
(A,B) Volcano map and heat map of differentially expressed genes in GSE22058. (C,D) Volcano map and heat map of differentially expressed genes in GSE25097. (E,F) Volcano map and heat map of differentially expressed genes in GSE64041. (G,H) Volcano map and heat map of differentially expressed genes in GSE36376. (I) Bubble chart of differentially expressed genes in biological process. (J) Bubble chart of differentially expressed genes in cellular component. (K) Bubble chart of differentially expressed genes in molecular function. (L) Bubble chart of differentially expressed genes in KEGG pathway.
FIGURE 2
FIGURE 2
Box map of DEGs expression in tumor and para-cancer samples in different data sets. The differences in the expression of (A) HAMP, (B) C9, (C) CXCL14, (D) MARCO, (E) CXCL12, (F) HSD11B1, (G) C7, (H) C8A, and (I) MBL2 genes, respectively, in tumor and adjacent samples.
FIGURE 3
FIGURE 3
(A) For the changing trajectory of each independent variable, the horizontal axis represents the log value of the independent variable lambda, and the vertical axis represents the coefficient of the independent variable. (B) The confidence interval under each lambda. (C) RiskScore, survival time, survival status, and 5-gene expression in TCGA training set. (D) ROC curve and AUC of 5-gene signature classification. (E) The KM survival curve distribution of 5-gene signature in training set.
FIGURE 4
FIGURE 4
(A) RiskScore, survival time, survival status, and 5-gene expression in TCGA validation set. (B) ROC curve and AUC of 5-gene signature classification. (C) The KM survival curve distribution of 5-gene signature in validation set.
FIGURE 5
FIGURE 5
(A) RiskScore, survival time, survival status, and 5-gene expression in TCGA data sets. (B) ROC curve and AUC of 5-gene signature classification. (C) The KM survival curve distribution of 5-gene signature in TCGA data sets.
FIGURE 6
FIGURE 6
Prognostic analysis of clinical subgroups based on RiskScore. The horizontal axis represents survival time, and the vertical axis represents survival probability. Blue represents low expression group; red represents high expression group. (A–L) Based on Riskscore, survival curves were analyzed for Age ≤ 60, Age > 60, Female, Male, T1+T2, T3+T4, N0, M0, Stage I+II, Stage III+IV, Grade 1+2, and Grade 3+4 groups, respectively.
FIGURE 7
FIGURE 7
(A) Comparison of RiskScore among T Stage grouping samples. (B) Comparison of RiskScore among stage grouping samples. (C) Comparison of RiskScore among grade grouping samples. (D) Comparison of RiskScore between grouped samples with or without recurrence. * means P value < 0.05; ** means P value < 0.01; *** means P value < 0.005; **** means P value < 0.001; ns, no significant.
FIGURE 8
FIGURE 8
(A) Univariate Cox analysis. (B) Multivariate Cox analysis. (C) Nomogram model. (D): 1-, 3-, and 5- year correction curve of the model.
FIGURE 9
FIGURE 9
(A–C) Landscape of immune cell infiltration score in high-risk and low-risk groups of TCGA, HCCDB18, and GSE14520 data sets. (D) Comparison of CIBERSORT immune score between high-risk and low-risk groups in TCGA data set. (E) Comparison of estimate immune score between high-risk and low-risk groups in TCGA data set. (F) Heat map of correlation between CIBERSORT immune score and estimate immune score in high-risk and low-risk groups. *P < 0.05, **P < 0.01, ***P < 0.005, and ****P < 0.001; ns, no significant.
FIGURE 10
FIGURE 10
(A,B) The ROC curve of Hu’s risk model and the KM curve of high-risk and low-risk groups. (C,D) The ROC curve of Liu’s risk model and the KM curve of high-risk and low-risk groups. (E,F) The ROC curve of Ke’s risk model and the KM curve of high-risk and low-risk groups. (G,H) The ROC curve of Zheng’s risk model and the KM curve of high-risk and low-risk groups. (I) C-index comparison of different prognostic risk models.
FIGURE 11
FIGURE 11
Differential box diagram of 5-gene expression in pan-cancer. (A) ATG10, (B) IL18RAP, (C) SPP1, (D) PRKCD, (E) SLC11A1, and (F) the protein expression of the 5 genes in 4 pairs of LIHC and normal samples. *P < 0.05, **P < 0.01, ***P < 0.005, and ****P < 0.001.
FIGURE 12
FIGURE 12
Survival curve of five genes in pan-cancer. (A) ATG10, (B) IL18RAP, (C) SPP1, (D) PRKCD, and (E) SLC11A1.
FIGURE 13
FIGURE 13
The correlation between five genes and tumor microenvironment. (A) The correlation between genes and immune core. (B) The correlation between genes and StromalScore. (C–G) The correlation between ATG10, IL18RAP, SPP1, PRKCD, SLC11A1, and immune cells.
FIGURE 14
FIGURE 14
The potential mechanism of five genes in HCC. (A) Correlation between five genes and immune subtypes of pan-cancer. (B) Correlation between five genes and clinical stage. (C) Correlation between five genes and clinical grade. (D) Correlation between five genes and T stage. (E–I) GSEA enrichment analysis of ATG10, IL18RAP, SPP1, PRKCD, and SLC11A1. *P < 0.05, **P < 0.01, and ***P < 0.001.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

References

    1. Bauler T. J., Starr T., Nagy T. A., Sridhar S., Scott D., Winkler C. W., et al. (2017). Salmonella meningitis associated with monocyte infiltration in mice. Am. J. Pathol. 187 187–199. 10.1016/j.ajpath.2016.09.002 - DOI - PMC - PubMed
    1. Braliou G. G., Kontou P. I., Boleti H., Bagos P. G. (2019). Susceptibility to leishmaniasis is affected by host SLC11A1 gene polymorphisms: a systematic review and meta-analysis. Parasitol. Res. 118 2329–2342. 10.1007/s00436-019-06374-y - DOI - PubMed
    1. Chen R. X., Xia Y. H., Cui J. F., Xue T. C., Ye S. L. (2010). Osteopontin, a single marker for predicting the prognosis of patients with tumor-node-metastasis stage I hepatocellular carcinoma after surgical resection. J. Gastroenterol. Hepatol. 25 1435–1442. 10.1111/j.1440-1746.2010.06277.x - DOI - PubMed
    1. Cheung H., Chen N. J., Cao Z., Ono N., Ohashi P. S., Yeh W. C. (2005). Accessory protein-like is essential for IL-18-mediated signaling. J. Immunol. 174 5351–5357. 10.4049/jimmunol.174.9.5351 - DOI - PubMed
    1. Dai W., Wu J., Zhang S., Shi B., Xu X., Wang D., et al. (2017). Genes directly regulated by NF-kappaB in human hepatocellular carcinoma HepG2. Int. J. Biochem. Cell Biol. 89 157–170. 10.1016/j.biocel.2017.05.031 - DOI - PubMed