Identification of a Prognostic Model Based on Immune-Related Genes of Lung Squamous Cell Carcinoma

Abstract

Immune-related genes (IRGs) play considerable roles in tumor immune microenvironment (IME). This research aimed to discover the differentially expressed immune-related genes (DEIRGs) based on the Cox predictive model to predict survival for lung squamous cell carcinoma (LUSC) through bioinformatics analysis. First of all, the differentially expressed genes (DEGs) were acquired based on The Cancer Genome Atlas (TCGA) using the limma R package, the DEIRGs were obtained from the ImmPort database, whereas the differentially expressed transcription factors (DETFs) were acquired from the Cistrome database. Thereafter, a TFs-mediated IRGs network was constructed to identify the candidate mechanisms for those DEIRGs in LUSC at molecular level. Moreover, Gene Ontology (GO), together with Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis, was conducted for exploring those functional enrichments for DEIRGs. Besides, univariate as well as multivariate Cox regression analysis was conducted for establishing a prediction model for DEIRGs biomarkers. In addition, the relationship between the prognostic model and immunocytes was further explored through immunocyte correlation analysis. In total, 3,599 DEGs, 223 DEIRGs, and 46 DETFs were obtained from LUSC tissues and adjacent non-carcinoma tissues. According to multivariate Cox regression analysis, 10 DEIRGs (including CALCB, GCGR, HTR3A, AMH, VGF, SEMA3B, NRTN, ENG, ACVRL1, and NR4A1) were retrieved to establish a prognostic model for LUSC. Immunocyte infiltration analysis showed that dendritic cells and neutrophils were positively correlated with IRGs, which possibly exerted an important part within the IME of LUSC. Our study identifies a prognostic model based on IRGs, which is then used to predict LUSC prognosis and analyze immunocyte infiltration. This may provide a novel insight for exploring the potential IRGs in the IME of LUSC.

Keywords: a Cox prediction model; immune-related genes (IRGs); lung squamous cell carcinoma; prognostic biomarkers; transcription factors (TFs) mediated IRGs network.

Figure 1

Identification of differential expression genes, IRGs and TFs in LUSC vs. normal tissues. (A–C) The volcano plot of differential expression genes, IRGs and TFs in LUSC vs. normal tissues. (D–F) The hierarchical clustering heat maps of differential expression genes, IRGs and TFs in LUSC vs. normal tissues.

Figure 2

The flow diagram of the whole study.

Figure 3

Functional enrichment analysis of differential expression IRGs in LUSC. (A) The outer circle shows the expression (logFC) of differential expression IRGs in each enriched GO terms: red dots which were on each Go terms indicated the up-regulation differential expression IRGs. Blue dots indicated the down-regulation differential expression IRGs. The inner-circle shows the prominence of GO terms (log10-adjusted P-values). (B) The circle represents the relationship between statistically top 30 differential expression IRGs and their GO terms. (C) The top five most significant GO terms and their annotations. (D) The top 10 pathways which were enriched in differential expression IRGs were showed in the dot plot. (E) The top 12 pathways which were enriched in differential expression IRGs were showed in the barplot. (F) the significantly statistically different 21 pathways were used Cytoscape software for constructing a pathway-IRG network with differential expression IRGs. The green rectangles indicate the pathways, the red circles indicate the up-regulation differential expression IRGs, the blue circles indicate the down-regulation differential expression IRGs.

Figure 4

OS-related DEIRGs and TFs-IRGs regulatory network. (A) The forest map of OS-related DEIRGs in LUSC. Red and green dots indicate the high and low-risk, respectively. (B) Regulatory network between prognosis-related DEIRGs and DETFs in LUSC. The red and blue circles indicate the high and low-risk DEIRGs, respectively. The green diamonds indicate DETFs. Solid and dashed lines in the network showed that there is a positive and negative correlation between prognosis-related DEIRGs and DETFs.

Figure 5

Prognosis value of 10 differential expression IRGs in LUSC patients. (A) Kaplan-Meier curve analysis for OS (overall survival) in LUSC patients using the 10 differential expression IRGs signature. (B) ROC curve analysis of the prognostic 10 differential expressions IRGs signature. (C) The risk score analysis of prognostic 10 differential expressions IRGs signature in LUSC high-risk group and low-risk group. (D) The survival status analysis of prognostic 10 differential expressions IRGs signature in LUSC high-risk group and low-risk group. (E) A risk heat-map constructed from 10 differential expression IRGs from 431 LUSC patients.

Figure 6

Univariate and multivariate independent prognostic analysis in LUSC. (A) Univariate independent prognostic analysis forest map of the prognostic immune-related genes model and LUSC clinicopathological characteristics. (B) Multivariate independent prognostic analysis forest map of prognostic immune-related genes model and LUSC clinicopathological characteristics. The red dots in the forest map shows that the clinical characteristic is a high-risk factor. The green dots in the forest map shows that the clinical characteristic is a low-risk factor.

Figure 7

Relationships between the clinical-pathological characteristics and the expressions of differential expression IRGs in LUSC. (A–D) Differences in the expression of DEIRGs between the pathological TNM stages I-II/III-IV in LUSC. (E–H) Differences in the expression of DEIRGs between the pathological T1-T2/T3-T4 stages in LUSC. (I–L) Differences in the expression of DEIRGs between the pathological M0/M1 stages in LUSC. (M) Differences in the expression of DEIRGs between the pathological N0/N1-N3 stages in LUSC.

Figure 8

Relationships between prognostic value and degree of infiltration of six types of immune cells. (A) B cells; (B) CD4 T cells; (C) CD8 T cells; (D) Dendritic cells; (E) Macrophages; (F) Neutrophils.