Lipid Metabolism-related lncRNA Model Identifies AC026412.3 as a Driver of Fatty Acid β-oxidation in Hepatocellular Carcinoma

Figures

• 

  Fig. 1. Prognostic model and risk score based on LRLs in HCC. (A) Flowchart illustrating the complete bioinformatic workflow of this study. This includes the acquisition and preprocessing of TCGA-LIHC transcriptomic and clinical data, identification of lipid metabolism-related genes from MSigDB, Pearson correlation–based screening of LRLs, differential expression analysis, construction of the prognostic model using univariate Cox, LASSO, and multivariate Cox regression, followed by validation in training and testing cohorts, and downstream functional characterization, including enrichment analyses, immune infiltration evaluation, and experimental validation of key lncRNAs. (B) Volcano plot showing differentially expressed LRLs. (C) Heatmap displaying the expression patterns of the top 50 differentially expressed LRLs among the 653 identified LRLs in HCC samples. (D) Forest plot showing the hazard ratios and confidence intervals of 180 prognosis-related LRLs identified by univariate Cox regression analysis. (E) Heatmap illustrating the expression profiles of all 180 prognosis-related LRLs across HCC samples and their associations with clinical characteristics. (F) LASSO regression curve…

• 

  Fig. 2. Construction of the LRL-based prognostic model in HCC. (A, F, K) Distribution of risk scores in the training, testing, and overall cohorts, respectively, ranked from low to high. (B, G, L) Survival status and survival time of patients in the training, testing, and overall cohorts according to their risk scores. (C, H, M) Heatmaps showing the expression patterns of the prognostic LRLs between the high- and low-risk groups in the training, testing, and overall cohorts. (D, I, N) Kaplan–Meier survival curves comparing overall survival between the high- and low-risk groups in the training, testing, and overall cohorts. (E, J, O) Time-dependent ROC curves evaluating the predictive accuracy of the prognostic model for 1-, 3-, and 5-year overall survival in the training, testing, and overall cohorts. LRL, lipid metabolism-related lncRNA; HCC, hepatocellular carcinoma; ROC, receiver operating characteristic; AUC, area under the curve.

• 

  Fig. 3. Association of the LRL-based prognostic model with clinical features in HCC. (A–B) Univariate and multivariate Cox regression analyses of age, sex, grade, stage, and risk score in relation to overall survival. (C) Multivariate ROC curves comparing the predictive accuracy of the risk score and clinical characteristics (age, sex, grade, stage). (D–G) Kaplan–Meier survival curves stratified by clinical subgroups (age <65/≥65 (D), sex (E), grade G1–2/G3–4 (F), stage I–II/III–IV (G)), showing differences in overall survival between high- and low-risk groups. LRL, lipid metabolism-related lncRNA; HCC, hepatocellular carcinoma; ROC, receiver operating characteristic; AUC, area under the curve.

• 

  Fig. 4. Functional analysis of DEGs and mutation profiles between high- and low-risk groups. (A) Volcano plot showing DEGs between high- and low-risk groups. (B) Heatmap illustrating the expression patterns of DEGs in HCC samples. (C) KEGG pathway enrichment analysis of DEGs performed using clusterProfiler and presented as a bubble plot. Significantly enriched pathways were identified based on adjusted P values. (D) GO enrichment analysis of DEGs performed using clusterProfiler to systematically characterize functional alterations associated with risk stratification and presented as a bubble plot. (E) Circos plot displaying the distribution and regulation status of DEGs in GO pathways. (F) GSEA showing significantly enriched pathways in the high-risk group. (G) GSEA showing significantly enriched pathways in the low-risk group. (H, I) Waterfall plots illustrating somatic mutation profiles generated using maftools, visualizing mutational landscapes in high- and low-risk groups and highlighting differences in mutation frequency and the distribution of key driver genes. (J) Kaplan–Meier survival curves comparing overall survival between high-TMB and low-TMB groups. (K) Kaplan–Meier survival…

• 

  Fig. 5. Identification of AC026412.3 as a key lipid metabolism–related lncRNA in HCC. (A) qRT-PCR analysis of lncRNA expression levels in HuH-7 cells transfected with shNC, shAC026412.3, shNRAV, or shAL031985.3 plasmids (n = 3 per group). (B) Colony formation assays showing the clonogenic growth ability of HuH-7 cells after transfection with shNC, shAC026412.3, shNRAV, or shAL031985.3 plasmids (n = 3 per group). (C) BODIPY staining showing lipid droplet accumulation in HuH-7 cells after transfection with shNC, shAC026412.3, shNRAV, or shAL031985.3 plasmids (n = 3 per group). Scale bar = 25 µm. (D) qRT-PCR analysis of AC026412.3 expression in the normal human liver cell line THLE-2 and HCC cell lines HuH-7, HepG2, and Li-7 (n = 3 per group). (E) qRT-PCR analysis of AC026412.3 expression in HCC tissues and matched adjacent non-tumorous tissues from 18 patients with HCC (n = 18 pairs). (F) AC026412.3 expression in HCC tissues (n = 371) and normal liver samples (n = 160) from public datasets. (G) Kaplan–Meier survival curve showing the association between AC026412.3 expression and overall survival in HCC patients (n = 350). (H) Bar plot showing KEGG pathway enrichment of…

• 

  Fig. 6. Effects of AC026412.3 on lipid accumulation, cell behavior, and tumor growth in HCC. (A) qRT-PCR analysis of AC026412.3 expression in HuH-7 cells transfected with vector or AC026412.3 overexpression plasmids (n = 3 per group). (B) BODIPY staining showing lipid droplet accumulation in HuH-7 cells transfected with vector or AC026412.3 overexpression plasmids (n = 3 per group). Scale bar = 25 µm. (C–D) Colony formation and Transwell assays showing the proliferation and invasion abilities of HuH-7 cells transfected with vector or AC026412.3 overexpression plasmids (n = 3 per group). (E) Representative images of xenograft tumors derived from HuH-7 cells transfected with vector or AC026412.3 overexpression plasmids (n = 5 per group). (F–G) Quantification of tumor weight and tumor volume in xenograft models derived from HuH-7 cells transfected with vector or AC026412.3 overexpression plasmids (n = 5 per group). (H–I) Measurement of TC and TG levels in xenograft tumor tissues derived from HuH-7 cells transfected with vector or AC026412.3 overexpression plasmids (n = 3 per group, randomly selected from five tumors in each group). (J–K) Colony formation and Transwell assays showing the…

• 

  Fig. 7. Transcriptomic analysis and expression validation of candidate genes associated with AC026412.3. (A) PCA showing the clustering of transcriptomic profiles between HuH-7 cells overexpressing AC026412.3 and control cells (n = 3 per group). (B) Volcano plot illustrating differentially expressed genes following AC026412.3 overexpression (n = 3 per group). (C) Pearson correlation analysis identifying genes significantly correlated with AC026412.3 expression (correlation coefficient > 0.5, P < 0.001). (D) Heatmap displaying the top 25 upregulated and top 25 downregulated genes following AC026412.3 overexpression (n = 3 per group). (E) Spearman correlation analysis using the GEPIA platform showing the relationship between AC026412.3 expression and candidate lipid metabolism-related genes (ABCA1, AQP7, ANGPTL4, SLC2A5, and SLC22A5) in HCC samples. (F) Pearson correlation analysis between AC026412.3 and SLC22A5 mRNA expression in 18 paired HCC samples (n = 18). (G) qRT-PCR analysis of SLC22A5 mRNA expression in HuH-7 cells transfected with shNC or shAC026412.3 plasmids (n = 3 per group). (H) qRT-PCR analysis of SLC22A5 mRNA expression in HuH-7 cells transfected with vector or AC026412.3…

• 

  Fig. 8. Analysis of fatty acid β-oxidation and energy metabolism associated with AC026412.3 in HCC. (A) KEGG pathway enrichment analysis of differentially expressed genes following AC026412.3 overexpression (n = 3 per group). (B) Heatmap displaying differentially expressed lipid metabolism-related genes in HuH-7 cells after AC026412.3 overexpression (n = 3 per group). (C) Pearson correlation analysis between AC026412.3 expression and the mRNA levels of CPT1A, CPT2, and CACT in clinical HCC samples (n = 18). (D) qRT-PCR analysis of CPT1A, CPT2, and CACT mRNA expression in HuH-7 cells with AC026412.3 overexpression or knockdown (n = 3 per group). (E) Western blot analysis of CPT1A, CPT2, and CACT protein levels in HuH-7 cells with AC026412.3 overexpression or knockdown (three independent experiments). (F) Measurement of fatty acid β-oxidation activity in HuH-7 cells with AC026412.3 overexpression or knockdown (n = 3 per group). (G) Measurement of intracellular ATP levels in HuH-7 cells with AC026412.3 overexpression or knockdown (n = 3 per group). Data are presented as mean ± SD. n.s., not significant; *P < 0.05, **P < 0.01, ***P < 0.001. HCC, hepatocellular carcinoma; KEGG, Kyoto…