Identification and validation of a ferroptosis-related long non-coding RNA signature as a prognostic biomarker for hepatocellular carcinoma

Abstract

Background: Ferroptosis may play a central role in the development of hepatocellular carcinoma (HCC). Ferroptosis-related long non-coding RNAs (FRlncRNAs) show prognostic value in HCC through ferroptosis regulation and significant survival correlation. This study aimed to identify a prognostic FRlncRNA signature and explored its association with tumor immunity and oncogenic pathways.

Methods: The transcriptome sequencing information and matched clinical data of 365 HCC patients was obtained from The Cancer Genome Atlas (TCGA) database. Patients were randomly divided into training and testing groups. A prognostic FRlncRNA signature was established in the test set via a correlation analysis, univariate Cox regression analysis, and LLeast Absolute Shrinkage and Selection Operator (LASSO) regression analysis. We evaluated the signature's prognostic significance through clinical correlation analysis and developed a predictive nomogram for HCC outcomes. Immune function and checkpoint analyses were conducted to explore the association between the signature and HCC-related immunity. Additionally, a Gene Set Enrichment Analysis (GSEA) was conducted to detect the enriched pathways. Among genes identified, the oncogenic function of LINC01063 was studied both in vitro and in vivo.

Results: In the training set, we established a prognostic signature comprising seven FRlncRNAs and classified patients into low-risk (LR) and high-risk (HR) groups with different prognosis. Time-dependent receiver operating characteristic (ROC) analysis yielded area under the ROC curve (AUC) values of 0.745, 0.745, and 0.719 for 1-, 2-, and 3-year overall survival (OS). The prognostic impact of risk-groups was verified in the testing set. The HR patients exhibited greater infiltration of immune cells and elevated expression levels of immune checkpoint genes. Significant differences in the cytolytic activity and Type II interferon response between the LR and HR groups were found. Several signaling pathways were enriched in the HR group, indicating that the signature was closely associated with HCC development. Finally, among the seven FRlncRNAs in the signature, LINC01063 was validated as an oncogene. In vitro, the knockdown of LINC01063 inhibited cell proliferation, disrupted colony formation ability, and reduced the migration and invasion capacities of HCC cells, and in vivo, nude BALB/c mice injected with the LINC01063-knockdown HCC cells exhibited reduced tumor growth compared to the controls.

Conclusions: A signature of seven FRlncRNAs predicted outcome and correlated with immunity and activated oncogene pathways suggesting that the signature could be a predictor of efficacy of immunotherapy. LINC01063 was validated as a new oncogene in HCC. Our findings provide novel insights for prognostic assessment and precision therapy in HCC.

Keywords: Hepatocellular carcinoma (HCC); LINC01063 oncogene; ferroptosis; long non-coding RNA (lncRNA); prognostic signature.

Figure 1

Selection of FRlncRNAs using univariate Cox and LASSO regression analyses. (A) Forest plot showing the prognostic FRlncRNAs. (B,C) Identification of 16 prognostic FRlncRNAs in the HCC tissues. (D,E) LASSO-penalized Cox regression analysis. *, P<0.05; ***, P<0.001. FRlncRNAs, ferroptosis-related long non-coding RNAs; HCC, hepatocellular carcinoma; LASSO, least absolute shrinkage and selection operator.

Figure 2

Construction and validation of the prognostic FRlncRNA signature. (A) Survival curve of HCC patients based on the risk score in the training cohort. (B) Distribution of the risk scores in the training cohort. (C) Distribution of survival time in the training cohort. (D) Heatmap showing the expression levels of the seven prognostic FRlncRNAs between the HR and LR groups. (E,F) ROC curve analysis of the signature in the training cohort. (G) Survival curve of the HCC patients based on the risk score in the testing cohort. (H) Distribution of the risk scores in the testing cohort. (I) Distribution of survival time in the testing cohort. (J) The heatmap of the 7 FRlncRNAs between the two groups. (K,L) ROC curve analysis of the signature in the testing cohort. AUC, area under the curve; FRlncRNA, ferroptosis-related long non-coding RNA; HCC, hepatocellular carcinoma; HR, high-risk; LR, low-risk; ROC, receiver operating characteristic.

Figure 3

Correlations between different clinicopathological characteristics of the HCC patients and the risk score. (A-D) The forest plots of univariate (A) and multivariate (C) Cox regression analyses in the training cohorts, the forest plots of univariate (B) and multivariate (D) Cox regression analyses in the testing cohorts. (E-J) Clinical stratification analysis of OS in HCC patients based on risk scores, stratified by age (E), gender (F), grade (G), N stage (H), TNM stage (I), and T stage (J). HCC, hepatocellular carcinoma; N stage, nodal stage; OS, overall survival; T stage, tumor stage; TNM stage, tumor-node-metastasis stage.

Figure 4

Construction and assessment of a clinical prognostic nomogram. (A) Nomogram for predicting 1-, 3-, and 5-year OS probabilities. (B) Calibration plot for the nomogram. *, P<0.05. N, node; M, metastasis; OS, overall survival; T, tumor.

Figure 5

Immune-related characteristics in the LR and HR groups. (A) Heatmap of immune cell infiltration in the LR and HR groups. (B) Immune function differences based on the ssGSEA. (C) Expression levels of immune checkpoints between the LR and HR groups. *, P<0.05; **, P<0.01; ***, P<0.001; ns, not significant. APC, antigen presenting cell; CCR, chemokine receptor; HLA, human leukocyte antigen; HR, high-risk; LR, low-risk; MHC, major histocompatibility complex; ssGSEA, single-sample gene set enrichment analysis.

Figure 6

Enriched pathways in the HR group identified using the signature. (A-F) Significantly enriched KEGG pathways in the HR group. HR, high-risk; KEGG, Kyoto Encyclopedia of Genes and Genomes.

Figure 7

LINC01063 facilitates HCC progression. (A) LINC01063 expression was measured in LINC01063-knockdown (sh-LINC01063) HCC cells and their negative control (sh-NC) cells using qRT-PCR. (B,C) CCK-8 (B) and colony formation (C) assays were used to assess the cell proliferation ability of the cells in the indicated groups (original magnification, 1×, crystal violet-stained). (D,E) Transwell assay was used to verify the effect of knockdown of LINC01063 on the migration (D) and invasion (E) abilities of the HCC cells (original magnification, 20×, crystal violet-stained). (F) Xenograft assay was used to examine the effect of LINC01063 knockdown on HCC growth in vivo. *, P<0.05; **, P<0.01; ***, P<0.001. CCK-8, Cell Counting Kit-8; HCC, hepatocellular carcinoma; NC, negative control; OD, optical density; qRT-PCR, quantitative reverse transcription polymerase chain reaction.