Screening for Lipid-Metabolism-Related Genes and Identifying the Diagnostic Potential of ANGPTL6 for HBV-Related Early-Stage Hepatocellular Carcinoma

Figures

• 

  Figure 1. Flow chart of this study. Abbreviations: GEO, Gene Expression Omnibus; T, tumorous tissues of HCC patients; N, adjacent nontumorous tissues of HCC patients; DEGs, differentially expressed genes; HPA, Human Protein Atlas; DEPs, differential expression proteins; Nanoscale Liquid Chromatography–Tandem Mass Spectrometry, Nano-LC-MS/MS; TCGA, The Cancer Genome Atlas; GTEx, Genotype-Tissue Expression; ANGPTL6, angiopoietin-like protein 6; AFP, alpha-fetoprotein; ELISA, enzyme-linked immunosorbent assay; HCC, hepatocellular carcinoma; HCs, healthy controls; CHB, chronic hepatitis B; ROC; receiver operating characteristic.

• 

  Figure 2. KEGG pathway and biological process enrichment analyses of 1312 DEGs from the GSE102079 dataset. (A) KEGG pathway and biological process enrichment analyses of 578 upregulated DEGs. (B) KEGG pathway and biological process enrichment analyses of 734 downregulated DEGs. In these bubble charts, the y-axis labels present enriched the top 3 terms of KEGG pathway, UniProt KW biological process, and GO biological process analyses. The x-axis labels present gene counts, which represent the number of DEGs enriched in a KEGG pathway term or a biological process term. The bubble sizes represent gene counts, and color intensity represents -log10 (p values) of the significantly enriched terms. Abbreviations: KEGG, Kyoto Encyclopedia of Genes and Genomes; KW, Keywords; GO, Gene Ontology.

• 

  Figure 3. Venn diagrams of overlapping genes. (A) Thirty overlapping genes between 1312 DEGs (from the GSE102079 dataset) and 144 hepatic metabolism-related genes (from a gene expression cluster of the HPA website) represent hepatic metabolism-related DEGs in a Venn diagram. (B) Forty-two overlapping genes between 1312 DEGs (from the GSE102079 dataset) and 80 hepatic HLM-related genes (from a gene expression cluster of the HPA website) represent hepatic HLM-related DEGs in a Venn diagram.

• 

  Figure 4. The heatmap of 10 DEPs between HCC tissues and paired adjacent nontumorous tissues. (A) The heatmap of 10 DEPs. (B) The annotated information of the DEG-encoded proteins. Abbreviations: ADH4, alcohol dehydrogenase 4 (class II), pi polypeptide; ANGPTL6, angiopoietin-like 6; CLEC4G, C-type lectin domain family 4 member G; CLEC4M, C-type lectin domain family 4 member M; CYP2C9, cytochrome P450 family 2 subfamily C member 9; GYS2, glycogen synthase 2; MBL2, mannose-binding lectin 2; PHYHD1, phytanoyl-CoA dioxygenase domain containing 1; SLC27A5, solute carrier family 27 member 5; TTC36, tetratricopeptide repeat domain 36; FC, fold change.

• 

  Figure 5. Correlations of methylation levels of (A) PHYHD1, (B) TTC36, and (C) SLC27A5 with corresponding mRNA levels (analyzed by the cBioPortal platform). The y axis shows the mRNA expression z-scores of the target gene relative to all samples. The x axis is the methylation level of the target gene (obtained from the HumanMethylation450 BeadChip platform). A Spearman correlation coefficient < −0.7 was considered to indicate a strong negative correlation. A p-value < 0.05 was considered statistically significant. Abbreviations: HM450, HumanMethylation450 BeadChip; RSEM, RNA-Seq by Expectation-Maximization.

• 

  Figure 6. SurvivalMeth data showed a clinical correlation between TTC36 methylation levels. (A) Methylation levels of 2 CpGs (cg24222440 and cg16806210) of TTC36 in HCC cancer tissues with different pathological stages. (B) A total of 375 HCC patient samples were divided into high- and low-risk groups at the optimal cutoff (prognostic index) of 2.808. The prognostic index was calculated according to the DNA methylation matrix and the coefficient (obtained from the proportional hazards regression model) of TTC36, and a high level of prognostic index implies a high risk for the patient. The “red curve” indicates the high-risk group. The “green curve” indicates the low-risk group. (C) The methylation levels of 3 CpGs (cg01128850, cg16806210, and cg24222440) of TTC36 in the high- and low-risk groups. *** p < 0.001. (D) The survival curve of the Kaplan–Meier plot showed that HCC patients in the high-risk group had a lower survival probability [HR = 2.05 (1.11–3.80), p < 0.01].

• 

  Figure 7. Correlation between SLC27A5 and infiltration levels of immune cells in HCC. Correlation analysis of SLC27A5 expression and infiltration levels of immune cells in HCC using the TIMER database. p < 0.05 and partial correlations < −0.3 or > 0.3 were considered to be statistically significant correlations. Abbreviations: cor, correlation. TPM, transcripts per million; TIMER, Tumor Immune Estimation Resource.

• 

  Figure 8. The prognostic value of ADH4 expression for HCC patients. Kaplan-Meier survival curves were plotted using the online GEPIA platform. (A) Correlation of low ADH4 expression levels with poor DFS in HCC patients. The DFS curve for patients with high ADH4 expression levels is shown by the red solid line. The DFS curve for patients with low ADH4 expression levels is shown by the blue solid line; (B) Correlation of low ADH4 expression levels with poor OS of HCC patients. The OS curve for patients with high ADH4 expression levels is shown by the red solid line. The OS curve for patients with low ADH4 expression levels is shown by the blue solid line. The 95% confidence intervals are shown by the dashed line. p < 0.05 in the log-rank test was considered statistically significant. Abbreviations: DFS, disease-free survival; OS, overall survival; n, number; GEPIA, Gene Expression Profiling Interactive Analysis.

• 

  Figure 9. Potential diagnostic values of ANGPTL6 and AFP levels for early-stage and all-stage HCC in the GSE63898 dataset. (A) ROC curves for ANGPTL6 to differentiate HCC with early-stage (BCLC 0~A staging) and nontumorous cirrhotic tissues. (B) ROC curves for AFP to differentiate early-stage HCC and nontumorous cirrhotic tissues. (C) ROC curves for ANGPTL6 to differentiate all-stage HCC and nontumorous cirrhotic tissues. (D) ROC curves for AFP to differentiate all-stage HCC and nontumorous cirrhotic tissues. Abbreviations: AFP, alpha-fetoprotein; AUC, area under the ROC curve; CI, confidence interval.

• 

  Figure 10. Potential diagnostic values of ANGPTL6 and AFP levels for early-stage and all-stage HCC in the GSE76427 dataset. (A) ROC curves for ANGPTL6 to differentiate HCC with early-stage (BCLC 0~A staging) and paired adjacent nontumorous tissues. (B) ROC curves for AFP to differentiate early-stage HCC and paired adjacent nontumorous tissues. (C) ROC curves for ANGPTL6 to differentiate all-stage HCC and paired adjacent nontumorous tissues. (D) ROC curves for AFP to differentiate all-stage HCC and paired adjacent nontumorous tissues.

• 

  Figure 11. The expression levels of serum ANGPTL6 and AFP between early-stage HCC patients and HCs. (A) High levels of serum ANGPTL6 in early-stage HCC patients (n = 27) compared with those in HCs (n = 17) (p = 0.035). (B) High levels of serum AFP in early-stage HCC patients (n = 27) compared with those in HCs (n = 17) (p < 0.001). (C) The correlation of serum ANGPTL6 and serum AFP was analyzed using Spearman’s rank correlation analysis. * p < 0.05. *** p ≤ 0.001. Abbreviations: num, number.

• 

  Figure 12. The potential value of serum ANGPTL6 for the diagnosis of early HCC patients with normal serum AFP levels (AFP ≤ 7 ng/mL). (A) Serum ANGPTL6 levels were higher in early HCC patients with normal serum AFP levels (n = 32) than in CHB patients with normal serum AFP levels (n = 30) (p = 0.013). Serum ANGPTL6 levels were higher in early HCC patients with normal serum AFP levels (n = 32) than in HCs (n = 32) (p = 0.001). (B) A ROC curve of the serum ANGPTL6 levels was used to distinguish between early HCC patients with normal serum AFP levels and the noncancer group (CHB patients and HCs). * p < 0.05. *** p ≤ 0.001.