Skip to main content
Download PDF

Liver cancer-specific prognostic model developed using endoplasmic reticulum stress-related LncRNAs and LINC01011 as a potential therapeutic target

BMC Medical Genomics volume 18, Article number: 71 (2025) Cite this article

Abstract

Liver cancer is a serious malignancy worldwide, and long noncoding RNAs (lncRNAs) have been implicated in its prognosis.It remains unclear how lncRNAs related to endoplasmic reticulum stress (ERS) influence liver cancer prognosis. Here, we analyzed RNA and clinical data from the Cancer Genome Atlas and sourced ERS-related genes from the Molecular Signatures Database. Co-expression analysis identified ERS-related lncRNAs, and Cox regression analysis as well as least absolute shrinkage and selection operator regression highlighted three lncRNAs for a prognostic model. Based on median risk scores, we classified patients into two risk groups. The high-risk group displayed poor prognosis, and this finding was validated in the test set. According to consistency clustering, the patients were assigned to two clusters, and tumor microenvironment scores were computed. Patients with a high mutation burden had worse outcomes. Furthermore, immune infiltration analysis indicated more immune cells and mutations in checkpoint molecules among high-risk individuals. Drug sensitivity varied between the risk groups. LINC01011 was selected for functional assays. Colony formation assay and CCK-8 assay revealed that silencing LINC01011 suppressed liver cancer cell proliferation. Transwell and scratch assays indicated that silencing LINC01011 inhibited liver cancer cell migration. Western blotting assay revealed that inhibiting LINC01011 induced apoptosis and simultaneously inhibited epithelial-mesenchymal transition. These findings confirm the validity of the prognostic model and indicate that LINC01011 could serve as a potential research target.

Peer Review reports

Introduction

Liver cancer is a common gastrointestinal malignancy with a substantial health burden on humans [1]. It ranks third and sixth in terms of worldwide cancer occurrence and major cause of cancer-related deaths, respectively [2]. Key risk factors associated with liver cancer include chronic alcohol consumption, chronic hepatitis B/C virus infections, persistent liver diseases, and ingestion of aflatoxin-contaminated food [3]. Despite current research, the specific pathways underlying liver cancer progression remain elusive, resulting in poor prognosis of liver cancer patients. Therefore, to design effective therapeutic strategies against liver cancer, it is crucial to develop robust prognostic models and identify viable biological targets.

The endoplasmic reticulum (ER) functions as an indispensable eukaryotic cell component and regulates the synthesis, folding, and secretion of proteins. Certain physiological and biochemical factors can impair the ability of ER to correctly fold and synthesize proteins. Misfolded protein accumulation induces ERS, thereby triggering the unfolded protein response (UPR) [4]. UPR is an evolutionarily conserved adaptive response involving a series of signaling pathways and regulatory processes to ensure the proper folding of unfolded proteins or promote the degradation of misfolded ones, ultimately restoring ER homeostasis [5]. In certain types of tumors, persistent changes in the tumor microenvironment (TME) can trigger ERS. Additionally, ERS regulates tumor progression by various mechanisms such as angiogenesis, tumor cell proliferation, and metastasis [6]. Consequently, genes related to ERS are vital in determining the prognosis and treatment strategies for liver cancer.

Long noncoding RNAs (lncRNAs) are over 200-nucleotide-long nonprotein-coding RNAs with a pivotal role in human cancers and frequently exhibit elevated expression levels [7]. Based on recent evidence, lncRNAs appears to be closely related to ERS and are involved in modulating pathways associated with cancer [8]. For instance, lncRNA LUCRC influences the ERS response, thereby affecting the growth and tumor formation in colorectal cancer [9]. Similarly, lncRNA H19 enhances resveratrol’s inhibitory effect on gastric cancer cell migration and proliferation by regulating ERS [10]. Hence, we believe that a prognostic model developed using ERS-related lncRNAs could have profound implications for liver cancer prognosis, and these lncRNAs may serve as promising research targets for designing appropriate therapeutic approaches for liver cancer.

Methods

Data acquisition

The Cancer Genome Atlas (TCGA) (http://cancergenome.nih.gov/) was accessed for collecting clinical samples (n = 377), RNA data (tumor sample = 372, normal sample = 50), and mutation data (n = 372). Next, “endoplasmic reticulum stress” was entered as a keyword in the Molecular Signatures Database [11] (MSigDB, http://www.gsea-msigdb.org/gsea/msigdb/index.jsp), and a set of 295 genes associated with ERS was obtained.

Screening of ERS-related LncRNAs with differential expression

A co-expression network analysis was accomplished with “limma” package contained in R platform, and 581 ERS-associated lncRNAs were identified using Pearson’s correlation coefficient > 0.6 and p < 0.001 as the criteria. Subsequently, the following statistical thresholds were applied: false discovery rate (FDR) < 0.05 and|log2 fold change (FC)| > 1; this step revealed 355 lncRNAs with remarkable variations in expression levels between liver cancer tissues and adjacent normal tissues.

Prognostic model development and validation

We constructed and validated a prognostic prediction model using a systematic analytical approach. Initially, the 355 differentially expressed ERS-related lncRNAs (DElncRNAs) were subjected to three analyses for stepwise selection: (1) univariate Cox regression, (2) least absolute shrinkage and selection operator (LASSO) regression, and (3) multivariate Cox regression. Three key lncRNAs were chosen as model construction parameters. For the selected lncRNAs from all samples, the relative expression levels were determined; a risk scoring system was then developed, and the 370 patient samples were assigned to high-risk and low-risk groups by using the median risk score as the cutoff. To validate the efficacy of the model, the samples were randomly allocated to a training set and a test set (n = 185 for each set). The training and test sets were used for model construction and model validation, respectively. During model evaluation, R packages, including “survival,” “caret,” “glmnet,” “survminet,” and “timeROC,” were integrated for prognostic function analysis. Additionally, the packages “regplot,” “survival,” and “rms” were utilized for constructing nomograms for predicting the survival likelihood of liver cancer patients at 1, 3, and 5 years.

Enrichment analysis

Differentially expressed genes (DEGs) were distinguished between the risk groups according to|log2 FC| > 1 and FDR < 0.05. A total of 4,231 differentially DEGs were identified, including 4,088 upregulated genes and 144 downregulated genes. Next, by employing the “clusterProfiler” package in R software, these DEGs were subjected to Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses. Furthermore, gene set enrichment analysis (GSEA) was undertaken for determining substantially enriched key pathways in both risk groups.

Mutation data analysis

The mutation data were annotated and displayed with the “maftools” package. By using the “survival” package, the relationship between mutation levels and survival time was analyzed; furthermore, a comprehensive survival analysis was conducted regarding the risk levels of liver cancer patients in relation to their tumor mutation burden (TMB).

Consensus clustering analysis of ERS-related LncRNAs

By using three ERS-related lncRNAs, the sample data of liver cancer patients was subjected to clustering analysis with the “ConsensusClusterPlus” package. The optimal number of clusters was determined as 2, and the liver cancer patients were assigned to two subgroups.

Tumor immune microenvironment (TIME) and drug sensitivity analysis

We applied the “estimate” package in R for evaluating the TIME and determined ESTIMATE, immune, and stromal scores for both risk cohorts. Additionally, ssGSEA was used to quantify 16 distinct immune cell populations, and subsequent comparative analyses were conducted between the groups. The “ggboxplot” package was used to illustrate these scoring disparities through boxplot visualization. The levels of expression of 44 immune checkpoint proteins were estimated across various patient subgroups.

To assess drug sensitivity, the “oncoPredict” package was applied to analyze liver cancer patient samples, and boxplots were generated to highlight differences in drug responsiveness among the different subgroups.

Cell culture and transfection

HepG2, MHCC97H, Hep3B, Huh7, and HCC LM3 liver cancer cell lines were procured from the Shanghai Cell Bank of Chinese Academy of Sciences (catalog numbers: SCSP-510, SCSP-5092, SCSP-5045, SCSP-526, and SCSP-5093, respectively). Mycoplasma detection test and short tandem repeat analysis were performed for the cell lines. DMEM (Wisent, Nanjing, China) containing 10% fetal bovine serum (FBS, Wisent) and 1% penicillin-streptomycin (Beyotime, Shanghai, China) was utilized for culturing the cells under a 5% CO2 environment at 37 °C.

LINC01011 expression was silenced using siRNAs (GenePharm, Shanghai, China). After the cell confluence was 60–70%, we replaced the medium with a serum-free medium. Lipofectamine 3000 (Invitrogen, USA) was utilized for cell transfection. Supplementary Table 1 shows the four LINC01011 siRNA sequences.

RNA extraction and quantitative reverse transcription polymerase chain reaction

RNA was extracted with TRIzol reagent by using cell samples (density: 1–5 × 107 cells/mL). Following cell lysis, cDNA was synthesized using HiScript II Q RT SuperMix for qPCR (R222-01) as recommended by the manufacturer. ChamQ Universal SYBR qPCR Master Mix (Q111-02) was employed for detecting qPCR products. The reaction system was prepared in accordance with the standard operating procedure. The reaction parameters were (1) 95 °C for 30 s and (2) 95 °C for 10 s and 60 °C for 30 s for 40 cycles. Supplementary Table 2 shows primer sequences.

CCK-8 assay

Liver cancer cells (3000–5000 cells/well) were added to a 96-well plate and grown for 24 h. After achieving cell adherence, the treatment was started. Next, culture medium (90 µL) and CCK-8 assay solution (10 µL; BestBio, Shanghai, China) were added to the wells. Subsequently, the plate was kept in the dark. Absorbance for data analysis was estimated at 450 nm.

Cell colony formation assay

Actively proliferating liver cancer cells were acquired, digested, and resuspended. Cell suspension was added to a 6-well plate, ensuring that each well contained 800–1000 cells. Cultivation was performed for 14 days. Subsequently, the resultant colonies were fixed in situ, stained, photographed, and enumerated.

Transwell assay

To evaluate cell migration, Hep3B and HepG2 cells were resuspended in a medium with 5% FBS at 5 × 104 cells/well cell density after 24 h of culture in a serum-free medium. These cells were then placed in the upper chamber of Transwell inserts (Corning, USA), and a medium supplemented with 20% FBS was added to the lower chamber. For invasion assays, the upper chamber was precoated with a diluted Matrigel solution (Corning, USA). For migration and invasion assays, the cells were cultured for 24 and 48 h, respectively. After fixing, the cells were stained with crystal violet and visualized and enumerated by microscopy.

Wound healing assay

Two types of cancer cells were added to a 6-well plate at 5 × 105 cells/well cell density. After cell confluence was approximately 90%, a 200-µL pipette tip was utilized for creating a scratch in the cell monolayer. Subsequently, a serum-free medium was used to treat the cells. At 0, 24, and 48 h after creating the scratch, microscopy images were acquired for assessing alterations in the wound area.

Extraction of proteins and Western blotting (WB) assay

RIPA lysis buffer was used for extracting proteins. Briefly, 1000 µL RIPA solution, 10 µL PMSF, and 1 µL phosphatase inhibitor were mixed for protein extraction. The BCA method was employed for estimating extracted protein concentration. Following protein quantification, the samples were subjected to electrophoretic separation on SDS-PAGE gels. The separated protein bands were shifted to methanol-activated PVDF membranes. Next, the membranes were blocked and incubated with primary antibodies overnight at 4 °C; subsequently, secondary antibodies were added, and the membranes were incubated for additional 2 h at room temperature. Finally, proteins were identified with an ECL chemiluminescent substrate kit (Biosharp, catalog number: BL520B). The relevant antibodies are listed in Supplementary Table 3.

Data analysis

Bioinformatics and statistical data analysis were conducted with R software (v 4.3.3) and GraphPad Prism (v 9.5.0), respectively. Image analysis was performed with ImageJ software. Differential analysis was performed using the “limma” package. Kaplan-Meier(K-M) survival analysis combined with the Log-Rank test was used to evaluate survival rates. Cox regression analysis was conducted to assess survival factors. The LASSO method was applied for variable selection. Receiver operating characteristic (ROC) curves were used to visualize model accuracy. Student’s t-test (two-tailed) and ANOVA were implemented for assessing statistical significance. A p-value of < 0.05 was considered statistically significant.

Results

Screening and identification of ERS-related LncRNAs and prognostic model construction

To obtain ERS-related lncRNAs, we initially screened out 295 ERS-associated genes from the MSigDB. We then performed co-expression analysis with the lncRNA data from the TCGA transcriptomic dataset and identified 581 ERS-related lncRNAs(Supplementary Table 4). To determine DElncRNAs in liver cancer tissues and normal tissues, FDR < 0.05 and|log2 FC| > 1 were utilized as the cutoff criteria, and 355 DElncRNAs were obtained. The top 50 lncRNAs with significant differential expression were visualized in a heatmap (Fig. 1A). After removing the missing data from the downloaded clinical samples, we integrated the data with the DElncRNAs. We carried out LASSO regression analysis as well as univariate and multivariate Cox regression analyses and identified 13 lncRNAs (Fig. 1B), 8 lncRNAs (AC026356.1, AC026412.3, AC016747.1, AC026356.1, AC026412.3, AC016747.1, AL031985.3, MIR4435-2HG, WAC-AS1, AL355574.1, and LINC01011), and 3 lncRNAs (AC026356.1, AC026412.3, and LINC01011), respectively. The prognostic model was developed with the 3 lncRNAs.

$$\eqalign{ Risk\,score{\rm{ }} & = {\rm{ }}AC026356.1\,expression\,level{\rm{ }} \times {\rm{ }}0.5223 \cr & + {\rm{ }}AC026412.3\,expression\,level{\rm{ }} \times {\rm{ }}0.7261{\rm{ }} \cr & + {\rm{ }}LINC01011\,expression\,level{\rm{ }} \times {\rm{ }}0.6208 \cr} $$

According to median risk scores, the samples were allocated to low-and high-risk groups.

Fig. 1
figure 1

Screening of lncRNAs related to ERS. (A) Heatmap of differentially expressed lncRNAs associated with ERS. (B) Univariate Cox regression and (C, D) LASSO regression analyses for screening lncRNAs related to ERS

Full size image

Performance evaluation of the prognostic model

To explore the relationship of risk scores with survival in risk groups, we performed survival analysis with the K-M method and plotted the corresponding survival curves (Fig. 2A). High-risk liver cancer patients showed poor prognosis. To assess whether the prognostic model displayed accuracy over time, the receiver operating characteristic (ROC) curve was plotted. The area under the curve (AUC) values were 0.716, 0.638, and 0.652 at 1, 3, and 5 years of survival, respectively (Fig. 2B); this finding showed adequate performance of the prognostic model for accurate prediction. To predict survival probabilities of liver cancer patients at 1, 3, and 5 years, a nomogram constructed using patients’ gender, age, tumor grade, risk score, tumor stage, and TNM stage (Fig. 2C). Additionally, we utilized Cox regression analysis for confirming the independent prognostic value of the risk score for liver cancer (Figs. 2D-E).

Fig. 2
figure 2

Performance of the developed prognostic model. (A) K-M survival curves. (B) ROC curves indicated specificity and sensitivity to predict patient OS. (C) A nomogram for liver cancer patients was constructed using sex, grade, age, tumor stage, and risk score for predicting the survival likelihood of patients at 1, 3, and 5 years. (D, E) Univariate and multivariate Cox regression analyses

Full size image

To validate the effectiveness of this model, we plotted heatmaps of the three lncRNAs, the ranking distribution of risk scores, and scatterplots for patient survival for training and test sets (Fig. 3A-C and D-F, respectively). Higher risk scores correlated with increased mortality rates in patients with liver cancer, which validated the risk model as a relevant prognostic resource.

Fig. 3
figure 3

Performance of ERS-related lncRNAs for patient prognosis. (A) lncRNA expression profile. (B) Risk score. (C and D-F) Patient survival status in training and test sets, respectively

Full size image

Enrichment analysis

To examine the biological functions of ERS-associated genes, those with significant differential expression in high- and low-risk groups were identified and assessed using GO, KEGG, and GSEA. GO enrichment analysis showed the enrichment of biological functions associated with the positive regulation of cell activation and leukocyte activation (Fig. 4A). KEGG analysis revealed that cytokine-cytokine receptor interaction as well as the PI3K-Akt signaling pathway were remarkably enriched (Fig. 4B). Additionally, GSEA analysis of the KEGG results indicated that ECM receptor interaction and cytokine-cytokine receptor interaction pathways were notably enriched in the high-risk group (Figs. 4C-D). These observations imply that these genes were strongly associated with cancer-related biological processes.

Fig. 4
figure 4

Risk model functional enrichment analyses. (A, B) GO enrichment and KEGG pathway enrichment analyses and (C, D) GSEA in the risk groups

Full size image

Tumor mutation data analysis

After integrating the risk score data, mutation data were annotated and visualized with the “maftools” package. Significant mutations were present in both risk groups, including TTN, TP53, CTNNB1, MUC16, PCLO, OBSCN, LRP1B, RYR2, CSMD3, ALB, APOB, XIRP2, ARID1A, CACNA1E, ADGRV1, HMCN1, USH2A, ABCA1B, AXIN1, and HMCN1 (Figs. 5A-B).

Our analysis explored the link between TMB and patient survival and revealed that individuals with a higher TMB tended to have reduced survival rates (Fig. 5C). Additionally, when we integrated risk scores with mutation burden data, liver cancer patients with both elevated mutation levels and high-risk scores had the most poor survival outcomes (Fig. 5D).

Fig. 5
figure 5

Tumor mutation analysis. (A, B) Waterfall plot of gene mutation state in liver cancer samples. (C) K-M survival curve analysis in two mutational groups. (D) K-M survival curve analysis for risk and mutation condition

Full size image

Differences in drug sensitivity among liver cancer patients

To analyze variations in drug responsiveness across distinct patient cohorts, we utilized the “oncoPredict” tool to determine sensitivity in individuals with liver cancer. High-risk patients exhibited higher responsiveness to 5-fluorouracil, AGI-5198, AZD4547, BPD-00008900, BDP-00009066, and BMS-536,924 (Figs. 6A-F). The low-risk category showed increased sensitivity to AT13148, axitinib, AZD1332, cisplatin, GSK2578215A, and GSK269962A (Figs. 6G-L).

Fig. 6
figure 6

Comparison of the chemosensitivity of the two groups. (A) 5-Fluorouracil sensitivity. (B) AGI-5198 sensitivity. (C) AZD4547 sensitivity. (D) BPD-00008900 sensitivity. (E) BDP-009066 sensitivity. (F) BMS-536,924 sensitivity. (G) AT13148 sensitivity. (H) Axitinib sensitivity. (I) AZD1332 sensitivity. (J) Cisplatin sensitivity. (K) GSK2578215A sensitivity. (L) GSK269962A sensitivity

Full size image

TME analysis and immune infiltration analysis of the risk groups, and consensus clustering analysis of patients with liver cancer

To investigate the variances in TMEs across the two risk categories, we analyzed samples from patients in both groups. High-risk category individuals displayed increased ESTIMATE scores, immune scores and stromal scores (Figs. 7A-C). This finding suggests that high-risk patients have a TME marked by reduced tumor purity. Immune cell infiltration analysis showed a greater prevalence of immature dendritic cells, activated dendritic cells (aDCs), plasmacytoid dendritic cells(pDCs), CD8+T cells, macrophages, T follicular helper cells, regulatory T cells (Tregs), Th1 and Th2 cells, and tumor-infiltrating lymphocytes in high-risk patients; low-risk patients exhibited high abundance of mast cells (Fig. 7D). Additionally, as shown by immune checkpoint analysis, high expression of various molecules, including CD40LG, CD86, LGALS9, HHLA2, and NRP1, was observed in high-risk patients (Fig. 7E). This suggests that immunotherapy might be more potent in high-risk patients.

According to the risk signature, we utilized a consensus clustering analysis, which grouped patient samples into two clusters (Fig. 7F). Analysis of the overall survival (OS) rates in these clusters showed that cluster 1 patients displayed a worse prognosis (Fig. 7G). We utilized a Sankey diagram to illustrate the relationship between risk grouping and clustering. The C1 cluster was entirely classified as the high-risk group, whereas the C2 cluster included all patients from the low-risk group and a subset of patients from the high-risk group(Figure 7H).We then evaluated differences in the TME between the clusters by performing a scoring analysis. Cluster 1 patients showed higher ESTIMATE, immune, and stromal scores than cluster 2 patients (Figs. 7I-K).

Fig. 7
figure 7

TME features in two risk groups and two subgroups. (A-C) ESTIMATE, immune, and stromal scores, respectively. (D) Immune cell infiltration. (E) Immune checkpoints in the groups. (F) Consensus matrix K = 2 clusters. (G) K-M survival curves. (H) Sankey diagram illustrates the relationship between consensus clustering and risk grouping. (I) ESTIMATE scores. (J) Immune score. (K) Stromal scores between the clusters

Full size image

Silencing LINC01011 suppresses proliferative, migratory, and invasive capabilities of liver cancer cells

The prognostic model included 3 lncRNAs, of which AC026356.1 and AC026412.3 have previously been shown to have significant prognostic relevance in liver cancer. Therefore, we chose LINC01011 for further experimental investigation. The GEPIA database search showed that LINC01011 is expressed at higher levels in liver cancer (Fig. 8A). To assess knockdown efficiency, we used siRNAs to knockdown LINC01011 expression in HepG2 and Hep3B cells and selected LINC01011(4) for subsequent experiments (Figs. 8B-C). To confirm that LINC01011 is an ERS-related lncRNA, we induced ERS using tunicamycin (TM) and observed increased mRNA expression of binding immunoglobulin protein (BIP) and LINC01011 (Fig. 8D); this confirmed the association between LINC01011 and ERS. We conducted a correlation analysis between LINC01011 and ERS-related genes in the data from TCGA and visualized the results using a heatmap (Fig. 8E).To evaluate the effect of LINC01011 on the proliferative, migratory, and invasive capabilities of liver cancer cells, LINC01011 expression was knocked down, and the results were validated with clonogenic, CCK-8, Transwell, and scratch assays. As observed in clonogenic and CCK-8 assays, LINC01011 knockdown inhibited liver cancer cell proliferation (Figs. 8F-G), while the Transwell assay indicated that LINC01011 knockdown suppressed liver cancer cell invasion and migration (Figs. 8H-I). The scratch assay further demonstrated that reduced LINC01011 expression inhibited liver cancer cell migration (Figs. 9A-B).

Fig. 8
figure 8

Effects of LINC01011 silencing. (A) Relative LINC01011 expression in normal tissues and liver cancer tissues. (B) Relative LINC01011 expression in five liver cancer cell lines. (C) LINC01011 knockout efficiency in HepG2 and Hep3B cells. (D) Relative BIP/LINC01011 expression induced by TM. (E) The heatmap illustrating the correlation between LINC01011 and ERS-related genes in the TCGA-LIHC dataset. (F) Clone image after LINC01011 silencing in HepG2 and Hep3B cells. (G) CCK-8 assay for detecting Hep3B and HepG2 cell proliferation. (H-I) Imaging results for invasion and migration assays after LINC01011 silencing in Hep3B and HepG2 cells. Experiments were carried out at least thrice. Significance values: *p < 0.05, **p < 0.01, ***p < 0.001

Full size image
Fig. 9
figure 9

LINC01011 silencing suppressed migration, promoted apoptosis, and inhibited EMT. (A, B) Determination of liver cancer cells’ ability to migrate by wound healing assay. (C) WB assay for BCL2 and BAX protein expression after LINC01011 silencing. (D) E-cadherin level and N-cadherin level after LINC01011 silencing. Supplementary Figs. 1–4 illustrate full-length blots/gels. Experiments were carried out at least thrice. Significance values: *p < 0.05, **p < 0.01, ***p < 0.001

Full size image

Silencing LINC01011 induces liver cancer cell apoptosis and enhances endothelial-mesenchymal transition (EMT)

To examine how LINC01011 influences liver cancer cell apoptosis and EMT, we knocked down LINC01011 expression and conducted WB assay to quantify the levels of BAX, BCL2, E-cadherin, and N-cadherin protein expression. LINC01011 knockdown decreased BCL2 expression and increased Bax expression, suggesting that silencing LINC01011 promoted liver cancer cell apoptosis (Fig. 9C). Furthermore, E-cadherin and N-cadherin levels increased and decreased, respectively, following LINC01011 silencing, indicating suppression of EMT (Fig. 9D).

Discussion

Liver cancer is a prevalent lethal cancer and constitutes substantial global risk burden [12]. Despite advances and innovations in liver cancer treatment, the underlying mechanisms remain elusive, and patient outcomes are poor [13]. Therefore, it is essential to construct reliable predictive models for enhancing prognosis and recognizing key biological markers, which are practically relevant for treating liver cancer.

To construct our predictive model, we utilized LASSO regression and Cox regression analyses and detected 3 ERS-related lncRNAs with significant prognostic relevance. Patients were assigned to two risk categories according to median risk scores. ROC curve analysis, K-M survival analysis, and comparison of survival status demonstrated the robust performance of the prognostic model, which agreed with the findings of Shen et al. [14]. Additionally, TMB analysis highlighted that elevated TMB was associated with poor outcomes, indicating that a thorough investigation is required for examining the role of mutated tumor genes. Furthermore, according to TME analysis, higher immune cell infiltration and repeated immune checkpoint alterations were noted, offering critical insights for guiding therapeutic strategies.

The DEGs predominantly influence pathways such as the PI3K-Akt signaling cascade, extracellular matrix (ECM) receptor interaction, and cytokine-cytokine receptor interactions. Cytokine-cytokine receptor interactions involve the binding of small signaling proteins, released by immune and tissue cells, to their specific receptors. This process controls cellular growth, differentiation, and immune regulation, contributing to the development of multiple illnesses, including autoimmune disorders and malignancies [15].During tumor formation and progression, these interactions contribute to cancer development by facilitating immune evasion.

The ECM functions as a framework for preserving tissue stability. It functions as a growth factor and cytokine reservoir and supports cancer and stromal cell growth. A change in environmental conditions disrupts the ECM and releases the stored molecules, which attach to their corresponding receptors and trigger a signaling cascade that promotes disease progression. Previous studies have highlighted that alterations in the rigidity of the ECM influence tumor development, particularly in liver and breast cancers [16, 17]. ECM degradation and alteration in its rigidity are associated with the growth, diffusion, and angiogenesis capability of tumor cells, thus revealing their critical role in cancer progression.

The PI3K-Akt pathway is tightly regulated. Various external stimuli, including hormones and ECM components, can activate this pathway. By activating downstream molecules in a molecular cascade, this pathway initiates various biological processes that promote disease progression [18]. Because overactivation of this pathway enhances tumor growth and therapy resistance, it is a critical research subject.

High-risk group exhibited lower TME scores, indicating reduced tumor cell purity and higher complexity of cellular composition. This observation aligned with the findings of Hoogstrate et al. on glioma tumor purity [19]. This result may be attributed to differences in tumor heterogeneity, which frequently leads to poor treatment outcomes. Immune infiltration analysis showed that high-risk patients displayed greater levels of immune cells and checkpoint proteins, including aDCs, Tregs, CD8+ T cells, CD40LG, CD86, LGALS9, HHLA2, NRP1, CD200R1, VTCN1, CTLA4, PDCD1LG2, TNFRSF18, and LAG3. Dendritic cells are specialized antigen-presenting cells (APCs) that can effectively activate T cells [20]. CD8+ T cells effectively eliminate intracellular infections and suppress malignant tumor progression, providing long-term immune protection [21]. Tumor-associated macrophages promote tumor advancement by enhancing migration, facilitating vascular infiltration, and suppressing anti-tumor immunity [22]. Tregs have immunosuppressive functions [23], and CTLA-4 is crucial for Treg activity [24]. Tregs suppress CD80/CD86 expression on APCs through CTLA-4-mediated phagocytosis [25]. These findings suggest that immunotherapy may have potential value in high-risk populations.

Drug sensitivity analysis showed distinct responses to medications between the two risk groups, which can assist clinicians in formulating personalized treatment strategies for individual patients. However, liver cancer patients typically require combination therapies; therefore, examining liver cancer patients’ sensitivity to a single drug is inappropriate. In clinical practice, it is crucial to comprehensively consider drug sensitivity, tolerance, and compatibility [26]. Furthermore, immune cell infiltration as well as mutations in immune checkpoint proteins across the risk groups should also be considered. By integrating individual patient factors, this approach can more effectively guide the selection of therapeutic agents.

Based on the prognostic capabilities of the model, we consider that the three lncRNAs used in this model have prognostic value for liver cancer patients and deserve further investigations.

Regarding these three lncRNAs, Huamei Wei et al. demonstrated that m6A modification of AC026356.1 facilitated hepatocellular carcinoma progression [27]. Sun et al. identified AC026412.3 as a part of a senescence-associated model in liver cancer; the authors also showed liver cancer cells exhibited higher levels of AC026412.3 than normal liver cells [28]. The function of AC026412.3 in the prognosis of liver cancer has also been analyzed by other studies [29,30,31]. Based on these findings, we chose to further investigate LINC01011.

After inducing ERS with TM, we observed an increase in LINC01011 expression, confirming that LINC01011 is an ERS-associated lncRNA. Silencing LINC01011 decreased the proliferative, invasive, and migratory capabilities of liver cancer cells. Knockdown of LINC01011 promoted apoptosis and inhibited EMT. As suggested previously, lncRNAs may influence cancer cell behavior by inhibiting microRNAs [32]. lncRNAs may also regulate related gene expression by participating in chromatin remodeling and modifications or by directly influencing post-transcriptional regulation of mRNAs, thereby affecting cancer cell functions [33, 34].

This research relied on the TCGA database data, potentially limiting its broader applicability. In future studies, we will incorporate validation using clinical samples to demonstrate the prognostic capability of the prognostic model. The mechanisms underlying biological changes post-LINC01011 silencing remain unclear, thus necessitating further in vivo validation. Additionally, the absence of clinical validation restricts the scope of the conclusions.

In summary, we developed a risk model by incorporating three ERS-related lncRNAs and validated its prognostic capacity. Functional assays revealed LINC01011 as a valuable biological target for further investigations. Interfering with LINC01011 expression may inhibit the malignancy of liver cancer cells.

Data availability

No datasets were generated or analysed during the current study.

References

  1. Siegel RL, Kratzer TB, Giaquinto AN et al. Cancer statistics, 2025[J/OL]. CA: a cancer journal for clinicians, 2025, 75(1): 10–45. https://doiorg.publicaciones.saludcastillayleon.es/10.3322/caac.21871

  2. Bray F, Laversanne M, Sung H et al. Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries[J/OL]. CA: a cancer journal for clinicians, 2024, 74(3): 229–63. https://doiorg.publicaciones.saludcastillayleon.es/10.3322/caac.21834

  3. Huang DQ, Mathurin P, Cortez-Pinto H, et al. Global epidemiology of alcohol-associated cirrhosis and HCC: trends, projections and risk factors[J/OL]. Nat Rev Gastroenterol Hepatol. 2023;20(1):37–49. https://doiorg.publicaciones.saludcastillayleon.es/10.1038/s41575-022-00688-6.

    Article  PubMed  Google Scholar 

  4. Chen X, Shi C, He M, et al. Endoplasmic reticulum stress: molecular mechanism and therapeutic targets[J/OL]. Signal Transduct Target Therapy. 2023;8(1):352. https://doiorg.publicaciones.saludcastillayleon.es/10.1038/s41392-023-01570-w.

    Article  CAS  Google Scholar 

  5. Urra H, Dufey E, Avril T, et al. Endoplasmic reticulum stress and the hallmarks of Cancer[J/OL]. Trends Cancer. 2016;2(5):252–62. https://doiorg.publicaciones.saludcastillayleon.es/10.1016/j.trecan.2016.03.007.

    Article  PubMed  Google Scholar 

  6. Song Z, Su M, Li X, et al. A novel Endoplasmic reticulum stress-related LncRNA signature for prognosis prediction and immune response evaluation in stomach adenocarcinoma[J/OL]. BMC Gastroenterol. 2023;23(1):432. https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12876-023-03001-0.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  7. Bhan A, Soleimani M, Mandal SS. Long noncoding RNA and cancer: A new Paradigm[J/OL]. Cancer Res. 2017;77(15):3965–81. https://doiorg.publicaciones.saludcastillayleon.es/10.1158/0008-5472.CAN-16-2634.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  8. Wang X, Kang M, Liu C, et al. Current state and progress of research on the role of LncRNA in HBV-Related liver Cancer[J/OL]. Front Cell Infect Microbiol. 2021;11:714895. https://doiorg.publicaciones.saludcastillayleon.es/10.3389/fcimb.2021.714895.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  9. Tang GH, Chen X, Ding JC, et al. LncRNA LUCRC regulates colorectal cancer cell growth and tumorigenesis by targeting Endoplasmic reticulum stress Response[J/OL]. Front Genet. 2019;10:1409. https://doiorg.publicaciones.saludcastillayleon.es/10.3389/fgene.2019.01409.

    Article  CAS  PubMed  Google Scholar 

  10. Li T, Zhang X, Cheng L, et al. Modulation of LncRNA H19 enhances resveratrol-inhibited cancer cell proliferation and migration by regulating Endoplasmic reticulum stress[J/OL]. J Cell Mol Med. 2022;26(8):2205–17. https://doiorg.publicaciones.saludcastillayleon.es/10.1111/jcmm.17242.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  11. Liberzon A, Birger C, Thorvaldsdóttir H, et al. The molecular signatures database (MSigDB) hallmark gene set collection. Cell Syst. 2015;1(6):417–25. https://doiorg.publicaciones.saludcastillayleon.es/10.1016/j.cels.2015.12.004.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  12. Arnold M, Abnet CC, Neale RE, et al. Global burden of 5 major types of Gastrointestinal Cancer[J/OL]. Gastroenterology. 2020;159(1):335–e34915. https://doiorg.publicaciones.saludcastillayleon.es/10.1053/j.gastro.2020.02.068.

    Article  PubMed  Google Scholar 

  13. Anwanwan D, Singh SK, Singh S, et al. Challenges in liver cancer and possible treatment approaches[J/OL]. Biochim Et Biophys Acta Reviews Cancer. 2020;1873(1):188314. https://doiorg.publicaciones.saludcastillayleon.es/10.1016/j.bbcan.2019.188314.

    Article  CAS  Google Scholar 

  14. Shen X, Wu S, Yang Z, et al. Establishment of an Endoplasmic reticulum stress-associated LncRNAs model to predict prognosis and immunological characteristics in hepatocellular carcinoma[J/OL]. PLoS ONE. 2023;18(8):e0287724. https://doiorg.publicaciones.saludcastillayleon.es/10.1371/journal.pone.0287724.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  15. He K, Meng X, Su J, et al. Oleanolic acid inhibits the tumor progression by regulating Lactobacillus through the cytokine-cytokine receptor interaction pathway in 4T1-induced mice breast cancer model[J/OL]. Heliyon. 2024;10(5):e27028. https://doiorg.publicaciones.saludcastillayleon.es/10.1016/j.heliyon.2024.e27028.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  16. Liu Z, Mo H, Liu R, et al. Matrix stiffness modulates hepatic stellate cell activation into tumor-promoting myofibroblasts via E2F3-dependent signaling and regulates malignant progression. Cell Death Dis. 2021;12(12):1134. https://doiorg.publicaciones.saludcastillayleon.es/10.1038/s41419-021-04418-9.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  17. Patwardhan S, Mahadik P, Shetty O, Sen S. ECM stiffness-tuned exosomes drive breast cancer motility through thrombospondin-1. Biomaterials. 2021;279:121185. https://doiorg.publicaciones.saludcastillayleon.es/10.1016/j.biomaterials.2021.121185.

    Article  CAS  PubMed  Google Scholar 

  18. Glaviano A, Foo ASC, Lam HY, et al. PI3K/AKT/mTOR signaling transduction pathway and targeted therapies in cancer[J/OL]. Mol Cancer. 2023;22(1):138. https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12943-023-01827-6.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  19. Hoogstrate Y, Draaisma K, Ghisai SA, et al. Transcriptome analysis reveals tumor microenvironment changes in glioblastoma[J/OL]. Cancer Cell. 2023;41(4):678–e6927. https://doiorg.publicaciones.saludcastillayleon.es/10.1016/j.ccell.2023.02.019.

    Article  CAS  PubMed  Google Scholar 

  20. Del Prete A, Salvi V, Soriani A, et al. Dendritic cell subsets in cancer immunity and tumor antigen sensing. Cell Mol Immunol. 2023;20(5):432–47. https://doiorg.publicaciones.saludcastillayleon.es/10.1038/s41423-023-00990-6.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  21. Reina-Campos M, Scharping NE, Goldrath AW. CD8 + T cell metabolism in infection and cancer[J/OL]. Nat Rev Immunol. 2021;21(11):718–38. https://doiorg.publicaciones.saludcastillayleon.es/10.1038/s41577-021-00537-8.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  22. Anderson NR, Minutolo NG, Gill S, et al. Macrophage-Based approaches for cancer Immunotherapy[J/OL]. Cancer Res. 2021;81(5):1201–8. https://doiorg.publicaciones.saludcastillayleon.es/10.1158/0008-5472.CAN-20-2990.

    Article  CAS  PubMed  Google Scholar 

  23. Dikiy S, Rudensky AY. Principles of regulatory T cell function[J/OL]. Immunity. 2023;56(2):240–55. https://doiorg.publicaciones.saludcastillayleon.es/10.1016/j.immuni.2023.01.004.

    Article  CAS  PubMed  Google Scholar 

  24. Tekguc M, Wing JB, Osaki M, et al. Treg-expressed CTLA-4 depletes CD80/CD86 by trogocytosis, releasing free PD-L1 on antigen-presenting cells[J/OL]. Proc Natl Acad Sci USA. 2021;118(30):e2023739118. https://doiorg.publicaciones.saludcastillayleon.es/10.1073/pnas.2023739118.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  25. Tekguc M, Wing JB, Osaki M, Long J, Sakaguchi S. Treg-expressed CTLA-4 depletes CD80/CD86 by trogocytosis, releasing free PD-L1 on antigen-presenting cells. Proc Natl Acad Sci U S A. 2021;118(30):e2023739118. https://doiorg.publicaciones.saludcastillayleon.es/10.1073/pnas.2023739118.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  26. Kamen DE, Crotts G, Narasimhan C, et al. An intercompany perspective on compatibility and In-use stability studies to enable administration of biopharmaceutical drug Products[J/OL]. J Pharm Sci. 2022;111(4):1092–103. https://doiorg.publicaciones.saludcastillayleon.es/10.1016/j.xphs.2021.09.043.

    Article  CAS  PubMed  Google Scholar 

  27. Wei H, Yang J, Lu R, et al. m6A modification of AC026356.1 facilitates hepatocellular carcinoma progression by regulating the IGF2BP1-IL11 axis[J/OL]. Sci Rep. 2023;13(1):19124. https://doiorg.publicaciones.saludcastillayleon.es/10.1038/s41598-023-45449-w.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  28. Sun R, Wang X, Chen J, et al. Development and validation of a novel cellular senescence-related prognostic signature for predicting the survival and immune landscape in hepatocellular carcinoma[J/OL]. Front Genet. 2022;13:949110. https://doiorg.publicaciones.saludcastillayleon.es/10.3389/fgene.2022.949110.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  29. Ye W, Li H, Zhao J, et al. Graphene therapy-related LncRNAs as prognostic and immune microenvironmental biomarkers in hepatocellular carcinoma[J/OL]. Translational Oncol. 2024;43:101915. https://doiorg.publicaciones.saludcastillayleon.es/10.1016/j.tranon.2024.101915.

    Article  CAS  Google Scholar 

  30. Song D, Wang X, Wang Y, et al. Integrated analysis of N1-Methyladenosine methylation Regulators-Related LncRNAs in hepatocellular Carcinoma[J/OL]. Cancers. 2023;15(6):1800. https://doiorg.publicaciones.saludcastillayleon.es/10.3390/cancers15061800.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  31. Pan Q, Yi C, Zhang Y. Overall survival signature of 5-Methylcytosine regulators related long Non-Coding RNA in hepatocellular Carcinoma[J/OL]. Front Oncol. 2022;12:884377. https://doiorg.publicaciones.saludcastillayleon.es/10.3389/fonc.2022.884377.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  32. Jiang H, Pang H, Wu P, et al. LncRNA SNHG5 promotes chondrocyte proliferation and inhibits apoptosis in osteoarthritis by regulating miR-10a-5p/H3F3B axis[J/OL]. Connect Tissue Res. 2021;62(6):605–14. https://doiorg.publicaciones.saludcastillayleon.es/10.1080/03008207.2020.1825701.

    Article  CAS  PubMed  Google Scholar 

  33. Yang H, Hu Y, Weng M, et al. Hypoxia inducible lncRNA-CBSLR modulates ferroptosis through m6A-YTHDF2-dependent modulation of CBS in gastric cancer[J/OL]. J Adv Res. 2022;37:91–106. https://doiorg.publicaciones.saludcastillayleon.es/10.1016/j.jare.2021.10.001.

    Article  CAS  PubMed  Google Scholar 

  34. Shaath H, Vishnubalaji R, Elango R, et al. Long non-coding RNA and RNA-binding protein interactions in cancer: experimental and machine learning approaches[J/OL]. Sem Cancer Biol. 2022;86(Pt 3):325–45. https://doiorg.publicaciones.saludcastillayleon.es/10.1016/j.semcancer.2022.05.013.

    Article  CAS  Google Scholar 

Download references

Funding

Financial support was provided by the National Natural Science Foundation of China (82072751).

Author information

Authors and Affiliations

  1. Department of Oncology, the First Affiliated Hospital of Anhui Medical University, 81 Meishan Road, Shushan District, Hefei, Anhui, 230000, China

    Xiao Du, Anqi Wang & Guoping Sun

  2. Cheeloo College of Medicine, Shandong University, Jinan, Shandong, 250000, China

    Ning Wei

  3. Department of Radiology, Division of Life Sciences and Medicine, the First Affiliated Hospital of USTC, University of Science and Technology of China, Hefei, Anhui, 230000, China

    Ning Wei

Authors
  1. Xiao Du
    View author publications

    You can also search for this author inPubMed Google Scholar

  2. Ning Wei
    View author publications

    You can also search for this author inPubMed Google Scholar

  3. Anqi Wang
    View author publications

    You can also search for this author inPubMed Google Scholar

  4. Guoping Sun
    View author publications

    You can also search for this author inPubMed Google Scholar

Contributions

All authors: study inception and design. Ning Wei: material preparation, data collection, and performance of analyses. Anqi Wang: data screening and verification. Xiao Du: writing of the initial draft of the manuscript. Guoping Sun: manuscript editing and supervising the investigation. All authors have made major contributions to the present research and have read and agreed with submission of the manuscript.

Corresponding author

Correspondence to Guoping Sun.

Ethics declarations

Ethical approval

Not applicable.

Consent to participate

Not applicable.

Consent to publish

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary Material 1

Supplementary Material 2

Supplementary Material 3

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Du, X., Wei, N., Wang, A. et al. Liver cancer-specific prognostic model developed using endoplasmic reticulum stress-related LncRNAs and LINC01011 as a potential therapeutic target. BMC Med Genomics 18, 71 (2025). https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12920-025-02142-3

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12920-025-02142-3

Keywords

BMC Medical Genomics

ISSN: 1755-8794

Contact us