Comprehensive analysis of KLF2 as a prognostic biomarker associated with fibrosis and immune infiltration in advanced hepatocellular carcinoma

Figures

• 

  Fig. 1. The expression distribution and prognostic survival value of KLF2 in tumor tissues and normal tissues of TCGA-LIHC and ICGC-LIRI. A The distribution of KLF2 expression across different types of tumor and normal tissues. B The level of KLF2 expression was significantly lower in tumor tissue in ICGC-LIRI (G1: normal liver tissue; G2: primary tumor tissue). C–F Kaplan–Meier survival analysis showed patients with higher KLF2 expression had a significantly better overall survival (OS), progression free survival (PFS), disease free survival (DFS) and disease specific survival (DSS) from TCGA-LIHC. G Kaplan–Meier survival analysis of KLF2 from ICGC-LIRI dataset. *P < 0.05, **P < 0.01, ***P < 0.001

• 

  Fig. 2. An overview of KLF2 genetic alterations landscape in HCC. A The presentation of KLF2 mutation landscape in HCC. B Putative copy number alterations of KLF2 in HCC. C Visualization of KLF2 genetic landscape in HCC. D The correlation between KLF2 methylation and the mRNA expression level. E The heatmap of the methylation level of KLF2 in HCC and normal samples. F The Kaplan–Meier survival curves of different islands of KLF2 in HCC. G The expression distribution of m6A-related genes in HCC and normal tissues. H The correlation between KLF2 expression and m6A-related genes in HCC. *P < 0.05, **P < 0.01, ***P < 0.001

• 

  Fig. 3. A single-cell level expression analysis of KLF2 in liver tissue. A UMAP plot and cell fraction plot show the cell line type clustering and cell line ratio of liver cancer based on the single-cell sequencing analysis from the SCP542. B, C UMAP plot and Cell Fraction plot show the cell type clustering and cell ratio in the normal liver tissue based on the single-cell sequencing analysis from the GSE 115469 and the GSE130073. D TSNE plot shows KLF2 expression in different cell lines of liver cancer. E, F Box bars show the KLF2 expression in different cell types at single cell level in normal liver tissue

• 

  Fig. 4. Analysis of KLF2 targets (KLFTs) and KLF2 expression in different cohorts. A, B GSEA-GO analysis and GSEA-KEGG analysis [24] on the 98 targets of the transcription factor KLF2. C The correlations between KLF2 and relevant pathway score. D KLF2 expression is analyzed in different tissues from GSE25097. G1: normal liver tissue; G2: liver cirrhosis tissue; G3: HCC tissue. E KLF2 expression is analyzed in different tissues from GSE6764. G1: liver cirrhosis tissue; G2: early HCC tissue; G3: advanced HCC tissue. F The expression distribution of KLFTs-17 in the normal and HCC samples from TCGA. G The correlations between KLFTs-17 and CAFs marked genes. The expression levels of PPP1R12A, SP1 and SMARCAD1 are analyzed in the GSE25097 (H) and GSE6764 (I) (G1: liver cirrhosis tissue; G2: HCC tissue). *P < 0.05, **P < 0.01, ***P < 0.001

• 

  Fig. 5. Subtype classification based on KLFTs-17 s. A Cumulative distribution function (CDF) based on consensus clustering. B Relative change in the area under the CDF curve (CDF delta area). C Heat map of consistent clustering results when k = 2, rows and columns represent samples, and different colors represent different subtype groups. D Heat map of KLF2-target genes expression between the two subgroups. E The expression distribution of m6A-related marker genes between the two subgroups. F Kaplan–Meier survival analysis of the two subgroups. G The distribution of clinical characteristics in samples of different subgroups, in which the horizontal axis represents different groups, the vertical axis represents the percentage of clinical information contained in corresponding grouped samples, and different colors represent different clinical information; The above table represents the distribution of a clinical feature in two groups (*P < 0.05). H The heatmap of the differential gene expression between the two subgroups. I The volcano plot was constructed using the fold change values and P adjust. Red dots indicate upregulated genes; blue dots indicate downregulated genes; grey dots…

• 

  Fig. 6. Confirmation of the prognostic factor SPP1. A Multivariate Cox regression among CAFs-related genes followed by multivariate Cox regression in the subgroup C1. B Nomogram showing the prediction of the 1-year, 2-year, and 3-year overall survival of HCC cancer patients. C Calibration curve for the overall survival nomogram model in the discovery group. D The distribution of SPP1 expression across different types of tumor and normal tissues. E GSEA-Hallmark analysis on the Spp1-related genes in HCC. F, G Correlation of SPP1 expression with the clinical characteristics of Grade and T stage in TCGA_LIHC. H–K Kaplan–Meier survival analysis show patients from TCGA_LIHC with high level of SPP1 expression had a significantly worse OS, PFS, DSS, and DFS. L Kaplan–Meier survival analysis of KLF2 from ICGC_LIRI dataset. M The ROC curve for SPP1 diagnosis

• 

  Fig. 7. Multidimensional analysis between KLF2 expression and immune microenvironment in HCC. A KLF2 expression positively correlated with stromal score, immune score, and ESTIMATE score in HCC. B KLF2 expression is significantly positively related to infiltrating levels of CD4+ T cells, CD8+ T cells, B cells, dendritic cells, macrophages, and neutrophils in HCC. C Heat map of the correlation between the expression of KLF2 and immune-checkpoint-related genes. The different colors represent the trend of gene expression in different samples. D Correlation analysis between KLF2 gene expression and TMB. E, F Scatter plots show KLF2 expression is associated with PDCD1 and CD274. G, H Comparison of response to ICB treatments in KLF2 high- and low-expression groups. I, J Prognostic KM curves of high- and low-expression groups of KLF2. *P < 0.05, **P < 0.01, ***P < 0.001

• 

  Fig. 8. Comprehensive analysis of marker gene CD3D of CXCR6 CD8+ T cell. A UMAP plot and Cell Fraction plot show the cell type clustering and cell ratio of hepatic immune cells based on the single-cell sequencing analysis of GSE125188. B A network of the correlation between the cell type and drug response is also shown on the right (datasets are marked in red, cell types are marked in yellow, and drugs are marked in blue). C The heatmap shows association of KLF2 expression and marker genes of CXCR6 CD8+ T cell. D, E GSEA enrichment plot of CXCR6 CD8+-T-cell marker genes from GSE125188 scRNA dataset and differential genes expressed in CXCR6 CD8+-T-cell induced by the drug Isoflupredone. F Matrix plot of signature genes in GSE125188 dataset referring to CXCR6 CD8+ T cell and Isoflupredone. G, H Scatter plots show the correlation of CD3D with PDCD1 and CD274 (PDL1). I GSEA-Hallmark analysis on the CD3D-related genes in HCC. J The expression distribution of CD3D in the different grades of LIHC. K, L Comparison of response to ICB treatments in KLF2 high- and low-expression groups in IMvigor210 cohort 2018 and Kim cohort 2019. M, N Prognostic KM curves of high- and low-expression groups of…