Migrasome-related long non-coding RNAs orchestrate immune microenvironment and serve as a novel prognostic model in clear cell renal cell carcinoma

Introduction

Renal cell carcinoma (RCC) ranks as the 14th most common malignancy globally, with over 400,000 new diagnoses reported in 2020 (1). In 2022, approximately 434,840 new cases of RCC were diagnosed worldwide, resulting in an estimated 155,953 deaths globally (2). Despite advancements in surgical and targeted therapies, the 5-year survival rate for advanced clear cell RCC (ccRCC) remains below 20%, emphasizing the urgent need for novel diagnostic biomarkers and therapeutic strategies. Migrasomes are recently identified extracellular vesicles that form during cell migration and carry bioactive contents such as proteins and RNAs (3). They are generated in highly migratory cancer cells, including cervical cancer (4), hepatocellular carcinoma (5), and pancreatic cancer (6), with their formation strongly dependent on cellular motility (7). Emerging evidence suggests that cytokines and chemokines within migrasomes play pivotal roles in tumor development. For instance, migrasomes derived from pancreatic cancer cells can foster an immunosuppressive microenvironment that facilitates tumor progression (8). Based on these findings, we hypothesize that migrasome activity may influence cancer cell invasion, migration, and tumorigenesis. During metastasis, circulating tumor cells may release migrasomes into the bloodstream, potentially altering the immune microenvironment (7). In pancreatic cancer, migrasomes have been shown to promote immune cell recruitment via chemokines such as C-X-C motif chemokine ligand 5 (CXCL5), thereby contributing to tumor progression (5). Given that CXCL5, a key regulator of cancer immunity, is highly expressed in multiple malignancies and is frequently associated with migrasome content (9), it is plausible that migrasomes may serve as predictive biomarkers for immunotherapy response (10). However, their role in ccRCC remains elusive. Given their potential involvement in shaping the tumor microenvironment (TME), investigating migrasome-related mechanisms in ccRCC may uncover novel diagnostic and therapeutic opportunities.

In this study, we focused on migrasome-related long non-coding RNAs (MRLs) in ccRCC and explored their associations with clinical prognosis, tumor immune status, and tumor mutation burden (TMB). We constructed a co-expression network to identify MRLs correlated with migrasome-related genes (MRGs), and developed a prognostic model based on selected long non-coding RNA (lncRNA) expression profiles. Additionally, comprehensive analyses of TME and TMB were performed to assess the broader implications of migrasomes in ccRCC. We further delineated cell-type-specific signaling landscapes using single-cell RNA sequencing (scRNA-seq) datasets. Our findings offer new insights into the signaling roles of migrasomes and their potential utility in improving ccRCC prognostication and treatment. We present this article in accordance with the TRIPOD reporting checklist (available at https://tau.amegroups.com/article/view/10.21037/tau-2025-541/rc).

Methods

Data acquisition and sample information

Patient data, including gene expression profiles, clinical information, and somatic mutation data, were derived from The Cancer Genome Atlas (TCGA) database (https://portal.gdc.cancer.gov/). A total of 355 samples, comprising tumor and adjacent normal tissues, were included (table available at https://cdn.amegroups.cn/static/public/10.21037tau-2025-541-1.xls). RNA sequencing (RNA-seq) data for messenger RNA (mRNA) and lncRNAs were retrieved and processed. Gene classification into mRNAs and lncRNAs was performed based on Ensembl gene annotation, and all expression values were normalized accordingly. Expression levels of MRGs were extracted from all samples for subsequent analyses. Somatic mutation data were used to calculate TMB. All datasets were standardized and preprocessed using the “TCGAbiolinks” R package, with the following inclusion criteria: (I) retention of only primary solid tumor and solid normal tissue samples; (II) exclusion of samples with incomplete clinical information [e.g., missing overall survival (OS) time, tumor stage, or tumor grade]; (III) removal of low-expression genes; and (IV) only the intersecting samples with both gene expression data and clinical information available were retained.

Development and validation of the MRLs prognostic model

The co-expression patterns between MRLs and MRGs were investigated and visualized using a Sankey diagram. The “ggalluvial” R package was employed to generate visual representations of the co-expression network and the Sankey diagram (11). Univariate Cox proportional hazards regression was employed to identify MRLs that were associated with patient prognosis. Consequently, a least absolute shrinkage and selection operator (LASSO) regression model was developed. The model’s robustness was assessed through cross-validation. A risk score for each patient was calculated based on the expression levels of five selected MRLs. Patients were then stratified into high-risk and low-risk groups according to the median risk score. To further evaluate the prognostic value of the prognostic model across clinical subgroups, ccRCC patients were stratified by age (<65 and ≥65 years), sex (female and male), tumor stage (stage I–II and stage III–IV), and tumor grade (G1–2 and G3–4). Kaplan-Meier (K-M) survival analysis was performed to compare OS within each subgroup. OS was defined as the time from diagnosis to death from any cause or last follow-up, and was measured in months.

Comparison of mutational profiles between high- and low-risk groups

The somatic mutation profiles of the 18 most frequently mutated genes in ccRCC samples were investigated. Differences in mutation frequency between the high- and low-risk groups were compared to explore potential associations between genetic alterations and patient prognosis.

Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses

ccRCC patients were divided into high- and low-risk groups based on their calculated risk scores. Differentially expressed genes (DEGs) between the two groups were identified with thresholds of |log₂fold change (FC)| ≥1. GO and KEGG pathway enrichment analyses were conducted using the “clusterProfiler” and “org.Hs.eg.db” R packages to further investigate the potential biological functions and signaling pathways associated with these DEGs.

Gene set enrichment analysis (GSEA)

To assess differences in pathway activity between risk groups, GSEA was performed. The “GSEABase” R package was used to calculate enrichment scores for immune-related signaling pathways and estimate immune cell infiltration. Data visualization was performed using the “ggpubr” R package to generate box plots comparing pathway activity levels.

Analysis of the potential clinical significance of the prognostic model

The clinical relevance of the prognostic model was assessed by calculating the TMB based on somatic mutation data (12). The optimal cutoff value for TMB was determined using the “survminer” R package, and ccRCC patients were subsequently stratified into low- and high-TMB groups. Similarly, to assess the combined impact of risk score and TMB, patients were further categorized into four subgroups based on both metrics, and OS was compared across these groups.

TME and immune infiltration analysis

To investigate the characteristics of TME, stromal score, immune score, and ESTIMATE score were calculated for each sample using the ESTIMATE R package (13). Furthermore, the “CIBERSORT” R package was employed to quantify the relative proportions of 22 immune cell types within each tumor sample.

Drug sensitivity analysis

Drug sensitivity was evaluated using the oncoPredict algorithm, which estimates individual patient responses to a broad spectrum of anti-cancer agents by predicting half-maximal inhibitory concentration (IC₅₀) values based on gene expression data. The reference drug sensitivity profiles were derived from the Cancer Cell Line Encyclopedia (CCLE), which includes a wide range of anti-cancer compounds, including conventional chemotherapeutic agents and molecularly targeted drugs.

scRNA‑seq analysis

The scRNA-seq datasets GSE131685, GSE139555, GSE140989, GSE152938, and GSE156632, containing 104,518 cells from tumor samples and 111,637 cells from healthy kidney tissues, were processed utilizing the standard Seurat workflow to evaluate cellular characteristics at the single-cell level. Stringent quality thresholds were applied to filter low-quality cells: cells with fewer than 400 or more than 5,000 detected genes, or over 30% of transcripts mapped to mitochondrial genes were excluded. Following quality filtering, the SCTransform function was employed for normalization and batch effect correction across samples. Cell type annotation was performed based on canonical markers and prior published studies (14,15). Data visualization was carried out using Uniform Manifold Approximation and Projection (UMAP).

Experimental validation via reverse transcription-quantitative polymerase chain reaction (RT-qPCR)

Tissue samples from five ccRCC patients were collected, and total RNA was extracted using the HiPureUniversal RNA Kit (Magen, Shanghai, China). RNA purity and concentration were assessed before being reverse-transcribed into complementary DNA (cDNA) using the PrimeScript™ RT kit (Takara, Dalian, China). RT-qPCR was performed using the PowerUpTMSYBRTMGreen Master MIX (Thermo Fisher, Waltham, Massachusetts, USA) on a LightCycler^® 96 Instrument (Roche Diagnostics GmbH, Basel, Switzerland). β-actin primer served as the internal control. Clinical characteristics of the included patients are summarized in Table S1, and primer sequences used are provided in Table S2. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by The First Affiliated Hospital of Guangxi Medical University (No. 2024-E805-01). The participants provided their written informed consent to participate in this study.

Statistical analysis

Gene expression levels between normal and tumor samples were compared using one-way analysis of variance. Immune cell infiltration levels and associated functional activity scores were compared between the high- and low-risk groups using the Wilcoxon rank-sum test. Statistical significance was defined as P<0.05.

Results

Co-expression analysis of MRGs and MRLs and prognostic model construction

The workflow of this study is illustrated in Figure 1. A co-expression analysis of MRGs and MRLs was conducted to explore their regulatory relationships, which later contributed to the development of a prognostic model for ccRCC patients. Figure 2A illustrates the co-expression network, highlighting significant associations between MRGs, including CPQ, EOGT, EPCIP, ITGA5, ITGB1, NDST1, PIGK, PKD1, PKD2, and TSPAN4 and various MRLs (P<0.001), suggesting that these MRLs may play a regulatory role in migrasomes function.

Figure 1 Flow chart of the study. GEO, Gene Expression Omnibus; KIRC, kidney renal clear cell carcinoma; LASSO, least absolute shrinkage and selection operator; lncRNA, long non-coding RNA; MRG, migrasome-related gene; RNA-seq, RNA sequencing; scRNA-seq, single-cell RNA sequencing; TCGA, The Cancer Genome Atlas.

Figure 2 Construction of the MRLs-MRGs co-expression network. (A) Sankey diagram illustrating the co-expression relationships between the MRGs and MRLs. (B) Univariate Cox regression identifying 22 MRLs significantly associated with OS. (C) Cross-validation was used to identify the optimal λ value in LASSO regression. (D) LASSO Cox regression used to construct the prognostic model based on OS-related lncRNAs. (E) Heatmap showing the correlation between prognostic MRLs and MRGs. CI, confidence interval; LASSO, least absolute shrinkage and selection operator; lncRNA, long non-coding RNA; MRG, migrasome-related gene; MRL, migrasome-related long non-coding RNA; OS, overall survival.

Subsequently, univariate Cox regression analysis was performed to identify MRLs associated with patient prognosis (Figure 2B). The resulting forest plot categorized MRLs as risk factors [hazard ratio (HR) >1] or protective factors (HR <1). To quantify prognostic risk, a model was constructed using the LASSO Cox regression method based on the minimum λ criterion (Figure 2C,2D). Five prognostic MRLs—EMX2OS, AC106897.1, AC087645.2, AC121338.2, and C5orf66—were ultimately selected to develop the prognostic model. The risk score was calculated as follows: risk score = (−0.39170 × EMX2OS expression score) + (1.28205 × AC106897.1 expression score) + (0.64942 × AC087645.2 expression score) + (−0.54445 × AC121338.2 expression score) + (2.50078 × C5orf66 expression score). Figure 2E displays a heatmap of the correlation strengths between the selected MRLs in the prognostic model and MRGs, highlighting strong co-expression relationships that may reflect underlying regulatory interactions in ccRCC. Collectively, these findings revealed a complex MRL-MRG interaction network and underscored the prognostic potential of specific MRLs, providing insights that may inform future therapeutic strategies and personalized treatment approaches.

Risk stratified survival analysis

To assess the effectiveness of the prognostic model, K-M survival analysis was conducted to compare the survival outcomes between high- and low-risk ccRCC patients. The overall cohort was randomly divided into training and validation sets at a 1:1 ratio. For each patient, a risk score was calculated based on the established model, and samples in both sets were stratified into high- and low-risk groups using the median risk score as the cutoff. Survival analysis of the overall cohort (Figure 3A), training set (Figure 3B), and validation set (Figure 3C) revealed significant differences in OS between the two groups. Progression-free survival (PFS) of overall cohort was further evaluated and showed similarly worse outcomes in the high-risk group (Figure 3D). Taken together, the prognostic model can reliably distinguish high- and low-risk ccRCC patients, providing a robust tool for risk stratification.

Figure 3 Risk stratification based on the five-lncRNA prognostic model. (A) K-M curve showing OS differences in the overall TCGA-ccRCC cohort. (B) OS comparison between high- and low-risk groups in the training set. (C) OS comparison between risk groups in the validation set. (D) PFS comparison between high- and low-risk patients in the overall cohort. ccRCC, clear cell renal cell carcinoma; lncRNA, long non-coding RNA; K-M, Kaplan-Meier; OS, overall survival; PFS, progression-free survival; TCGA, The Cancer Genome Atlas.

To further evaluate the performance of the prognostic model, risk curve analyses were performed on the training, validation, and overall cohorts. Figure S1A displays the distribution of risk scores, where patients were ranked in ascending order and clearly stratified into high- and low-risk groups based on the median cutoff. Figure S1B presents the survival status and survival time for each patient, showing that higher risk scores are associated with increased mortality and shorter survival durations. Figure S1C depicts the expression patterns of the five MRLs in the prognostic model, revealing distinct expression profiles between high- and low-risk groups. These consistent trends across all cohorts further confirmed the robustness of the model in distinguishing between patient subgroups. Collectively, the results demonstrated that the prognostic model provided a reliable and reproducible method for risk stratification in ccRCC.

Independent prognostic value and predictive accuracy of the constructed model

To verify whether the constructed model can independently predict prognosis, both univariate and multivariate prognostic analyses were carried out. In the univariate analysis, each clinical variable was analyzed for its association with OS (Figure 4A). In the multivariate analysis, all clinical variables were included to assess whether the risk score remained a significant prognostic factor when adjusting for other clinical characteristics. Notably, the risk score showed a strong independent association with OS (P<0.001), indicating that the prognostic model was not confounded by other clinical variables (Figure 4B).

Figure 4 Evaluation of the independent prognostic value and predictive performance of the five-lncRNA prognostic model. (A) Univariate Cox regression analysis showing the association between survival outcomes and clinical variables. (B) Multivariate Cox regression analysis confirming the risk score as an independent prognostic factor for OS. (C) Time-dependent ROC curves of the risk score at 1, 3, and 5 years. (D-F) ROC curve comparisons of the risk score and clinical variables at 1 year (D), 3 years (E), and 5 years (F). AUC, area under the curve; CI, confidence interval; lncRNA, long non-coding RNA; OS, overall survival; ROC, receiver operating characteristic.

Next, receiver operating characteristic (ROC) curves for survival predictions are presented in Figure 4C, with each time point achieving an area under the curve (AUC) greater than 0.7, demonstrating the model’s high predictive accuracy at these intervals based on risk scores. The higher value of concordance index (C-index) also demonstrated the model’s strong discrimination in survival prediction (Figure S2A). The calibration curve shows that the predicted survival probabilities were highly consistent with the actual results (Figure S2B). To assess the comparative predictive value of the model against conventional clinical parameters, combined ROC analyses were performed (Figure 4D-4F). Among all factors, the 5-year AUC of the risk score model was the highest, suggesting that it outperformed other clinical features in predicting long-term survival. The established nomogram, which integrates clinical characteristics and risk scores to provide individualized predictions for survival (Figure S2C), offers a practical tool for personalized prognosis. Together, these results demonstrated that the constructed model served as an independent prognostic factor and offered superior predictive accuracy, especially in 5-year survival predictions, compared with other clinical characteristics. Furthermore, to validate the generalizability of the prognostic model across diverse clinical contexts, ccRCC patients were stratified by age, sex, tumor stage, and tumor grade. K-M survival analyses were conducted within each subgroup, and the model consistently demonstrated significant survival differences (P<0.05) between high- and low-risk groups across all strata (Figure S3A-S3D), confirming its broad prognostic utility. Next, principal component analysis (PCA) was performed to evaluate the discriminative ability of the lncRNAs used in model construction. Four PCA plots were generated based on the expression profiles of (I) all genes, (II) MRGs, (III) MRLs, and (IV) the five MRLs included in the final model (Figure S4A-S4D). Among these models, the PCA plot based on the five model-derived MRLs showed the most distinct separation, highlighting their strong capacity to distinguish risk status and underscoring their importance in the prognostic model.

Functional enrichment analysis results

To explore the underlying biological mechanisms associated with the identified risk groups, functional enrichment analyses on the DEGs were conducted. DEGs were shown in table available at https://cdn.amegroups.cn/static/public/10.21037tau-2025-541-2.xls. GO functional enrichment analysis revealed that DEGs were enriched in immune-related biological processes, including T cell activation, immune receptor activity, and cytokine receptor binding (Figure 5A). Furthermore, the KEGG pathway enrichment analysis revealed that DEGs were associated with pathways related to B cell activation, immunoglobulin production, and other immune-related signaling pathways (Figure 5B). GSEA was performed to assess global pathway-level differences. It was found that immunological pathways were enriched in the high-risk group (Figure 5C). In contrast, the low-risk group exhibited enrichment in pathways related to calcium signaling, epithelial cell signaling in Helicobacter pylori infection, actin cytoskeleton regulation, and Vibrio cholerae infection (Figure 5D), suggesting that cytoskeletal and microbial response signaling may be more prominent in the low-risk group. Collectively, these results suggest that the immune microenvironment plays a pivotal role in differentiating risk groups in ccRCC and may serve as a potential therapeutic target for future research (16).

Figure 5 Functional enrichment analysis of the high- and low-risk groups. (A) GO enrichment analysis of DEGs. (B) KEGG pathway enrichment analysis of DEGs. (C) GSEA results for the high-risk group. (D) GSEA results for the low-risk group. DEG, differentially expressed gene; GO, Gene Ontology; GSEA, gene set enrichment analysis; KEGG, Kyoto Encyclopedia of Genes and Genomes.

Immune cell analysis

To further understand the immune landscape in ccRCC and its potential impact on patient prognosis, we systematically analyzed immune cell infiltration and immune-related functions across risk groups. Analysis based on the ESTIMATE algorithm revealed that high-risk patients had significantly higher stromal scores, immune scores, and ESTIMATE scores, suggesting greater immune and stromal involvement within the TME (Figure 6A).

Figure 6 Characterization of the immune infiltration. (A) Comparison of stromal score, immune score, and ESTIMATE score between groups. (B) Stacked bar plot showing the relative proportions of 22 infiltrating immune cell types. (C) Boxplots comparing the abundance of individual immune cell populations. (D) Comparison of immune-related functional scores between groups. *, P<0.05; **, P<0.01; ***, P<0.001. aDC, activated dendritic cell; APC, antigen-presenting cell; CCR, C-C chemokine receptor; DC, dendritic cell; HLA, human leukocyte antigen; iDC, immature dendritic cell; IFN, interferon; MHC, major histocompatibility complex; NK, natural killer; pDC, plasmacytoid dendritic cell; Tfh, T follicular helper; Th, T helper; TIL, tumor-infiltrating lymphocyte; TME, tumor microenvironment; Treg, T regulatory cell.

Tumor progression in ccRCC patients may be driven or exacerbated by alterations in immune cell markers (17). Thus, the immune status was assessed to determine the immune cell composition in ccRCC samples, and the relative abundance of immune cells was compared between low- and high-risk patients. The proportions of 22 infiltrating immune cell types are presented in Figure 6B. High-risk patients exhibited higher infiltration of CD4⁺ memory activated and T follicular helper cells compared with low-risk patients (Figure 6C). Besides, the high-risk group demonstrated globally elevated immune activity, with higher scores across multiple immune functions, including antigen-presenting cell co-stimulation, cytolytic activity, macrophage activation, and type I/II interferon responses (Figure 6D). These findings are consistent with previous enrichment analyses, suggesting that enhanced but potentially dysregulated immune activity in high-risk patients may contribute to adverse outcomes.

Mutational profiles of the different risk groups exhibit distinct variations

To investigate the genetic alterations that may contribute to the risk stratification of ccRCC patients, we analyzed and highlighted the 18 most frequently mutated genes. The mutational landscape differed (Figure 7A,7B). For example, polybromo 1 (PBRM1) mutations occurred more frequently in the high-risk group, whereas von Hippel-Lindau tumor suppressor (VHL) mutations were more commonly found in the low-risk group. Previous studies have demonstrated that patients with high TMB are more likely to respond favorably to immune checkpoint inhibitor (ICI) therapy. Herein, both TMB scores (Figure 7C) and Tumor Immune Dysfunction and Exclusion (TIDE) scores (Figure 7D) were significantly higher in the high-risk group. Survival analysis further revealed that patients with higher TMB scores exhibited poorer survival outcomes (Figure 7E,7F).

Figure 7 Mutation and immunotherapy-related characteristics of high- and low-risk groups. (A,B) Somatic mutation landscape of the high- and low-risk groups, respectively. (C) Comparison of TMB levels between the two risk groups. (D) Comparison of TIDE scores between the two risk groups. (E) K-M survival analysis stratified by TMB level. (F) Survival analysis combining TMB level and risk group. ***, P<0.001. H, high; L, low; K-M, Kaplan-Meier; TIDE, Tumor Immune Dysfunction and Exclusion; TMB, tumor mutation burden.

Drug sensitivity

To evaluate potential therapeutic options for ccRCC patients, the IC₅₀ values of 198 anticancer compounds were estimated. Differential analysis identified seven drugs with significantly different sensitivities between high- and low-risk groups (Wilcoxon test, P<10⁻⁹), including AZ960, AZD7762, gemcitabine, MK-8776, OSI-027, vinblastine, and XAV939. Most of these drugs have also been reported to be associated with ICI-based therapies in previous studies (Figure S5). While gemcitabine and vinblastine are classical chemotherapeutic agents, the others are molecularly targeted drugs with reported relevance to ICI-based therapies. These findings suggest that the identified compounds may serve as predictive markers for treatment response and guide individualized therapy selection in ccRCC.

Validation of model-derived MRL expression

To further validate the expression pattern of the five MRLs in the prognostic model, we examined their expression levels in paired tumor and adjacent non-tumor tissues from ccRCC patients. C5orf66, AC106897.1, and AC087645.2, which were included in the model, were significantly upregulated in tumor tissues, whereas EMX2OS showed significantly lower expression in tumor samples. However, AC121338.2 did not exhibit a statistically significant expression difference between tumor and pericarcinous tissues (Figure 8). These findings support the potential relevance of the model-derived MRLs in ccRCC tumor biology.

Figure 8 Validation of the expression levels of model-derived MRLs in ccRCC tissues by RT-qPCR. **, P<0.01; ns, not significant (P>0.05). ccRCC, clear cell renal cell carcinoma; MRL, migrasome-related long non-coding RNA; mRNA, messenger RNA; RT-qPCR, reverse transcription-quantitative polymerase chain reaction.

Single-cell profiling of ccRCC

To elucidate the cellular composition and transcriptomic heterogeneity of ccRCC, we conducted scRNA analysis. Following stringent quality control (Figure S6A) and batch effect mitigation, 216,155 cells were retained for subsequent analysis. The characteristics of each sample are shown in Figure S6B,S6C. Unsupervised clustering delineated 59 clusters (Figure S6D), which were subsequently annotated into 17 major cell types (Figure 9A). These included endothelial cells, fibroblasts (Fib), distal tubule (DT), loop of Henle cells (LOH), principal cells (PC), podocytes (Podo), intercalated cells (IC), tumor cell, proximal tubule cells (PT), basophils (Baso), macrophages, monocytes, B cell, CD4⁺ T cells, CD8⁺ T cells, CD8⁺ natural killer T cells (CD8⁺ NKT), and natural killer T cells (NK). Characteristic biomarkers defined each cellular population, with their proportional distributions illustrated in Figure 9B. Notably, in ccRCC patients, the abundance of most cell types differed markedly between tumor tissues and adjacent normal tissues. Afterwards, Figure 9C visualizes the spatial distribution and tissue origins of these populations via UMAP. To further investigate tumor-specific transcriptional alterations, DEGs between tumor and adjacent tissues were analyzed across all 17 cell types. Functional enrichment revealed that most DEGs were enriched in pathways related to “human diseases” (Figure 9D). We next assessed the expression of MRGs across different cell types using AUCell enrichment scoring. As shown in Figure 9E, tumor cells exhibited notably higher MRG-associated AUC scores, suggesting a pivotal role for migrasome-related pathways in ccRCC progression. Furthermore, CellChat analysis revealed an intercellular communication pattern (pattern 3) prominently involving Fib and tumor cells (Figure 9F). VEGF signaling was identified as a key enriched pathway within this communication pattern. Given that vascular endothelial growth factor (VEGF) pathway activation is a hallmark of ccRCC, frequently resulting from VHL inactivation and subsequent hypoxia-inducible factor-induced upregulation of VEGF, this finding underscores its pivotal role in angiogenesis, tumor progression, and immune evasion, thereby reinforcing its value as a therapeutic target in ccRCC.

Figure 9 Single-cell transcriptomic characterization of ccRCC and migrasome-related signaling patterns. (A) UMAP visualization of 17 principal cellular types identified within public ccRCC scRNA-seq dataset. (B) Cluster-defining marker genes alongside proportional sample origins (%) across 17 cell types. (C) Compositional distribution of predominant cell types. (D) Bar plots displaying numbers of DEGs between adjacent normal tissues and ccRCC specimens within individual cell types. (E) Violin plots depicting density distributions of AUC values for MRG sets, stratified by distinct cell types. (F) Alluvial diagram depicting outgoing signaling patterns of target cells based on CellChat analysis. AUC, area under the curve; Baso, basophils; ccRCC, clear cell renal cell carcinoma; DEG, differentially expressed gene; DT, distal tubule; Endo, endothelial cells; Fib fibroblasts; IC, intercalated cells; KEGG, Kyoto Encyclopedia of Genes and Genomes; LOH, loop of Henle cells; Mac, macrophages; Mono, monocytes; MRG, migrasome-related gene; NK, natural killer T cells; NKT, natural killer T cells; PC, principal cells; Podo, podocytes; PT, proximal tubule cells; scRNA-seq, single-cell RNA sequencing; UMAP, Uniform Manifold Approximation and Projection.

Discussion

RCC is a highly heterogeneous cancer with several distinct subtypes (18,19). It progresses through a multistep process characterized by numerous genetic mutations and epigenetic changes. The most defining biological features of the disease include unregulated defective apoptotic mechanisms and the capacity for metastasis, which significantly contribute to its high mortality rate (20). At present, diagnosis and prognosis of ccRCC mainly depend on pathological evaluation of surgically removed tumor tissue and imaging studies. However, due to the heterogeneity of ccRCC and subjective variations in pathological interpretation, these assessments can yield inconsistent results. Furthermore, the lack of reliable biomarkers and limited understanding of the underlying mechanisms of ccRCC contribute to late-stage diagnoses in most patients.

LncRNAs have emerged as significant biomarkers in cancer diagnostics, no longer dismissed as mere transcriptional noise (20). Research has demonstrated that lncRNAs play crucial roles in the development of cancer by influencing multiple cellular functions, including cell proliferation, as well as the processes of invasion and metastasis (21). Several studies have identified lncRNAs as key contributors to various malignancies, actively promoting cancer progression (22,23). lncRNA carbonic anhydrase 3-AS1 has been identified as a tumor suppressor in colorectal cancer, reducing cancer cell viability and invasiveness via modulation of the miR-93/phosphatase and tensin homolog axis (24). Furthermore, in gastric cancer, the downregulation of long intergenic non-coding RNA 00152 has been shown to impede tumor development by regulating miR-193b-3p, thereby hindering cancer progression. These examples underscore the critical and diverse roles of lncRNAs in various cancer types (25). In addition to lncRNAs, protein-coding genes such as the enhancer of rudimentary homolog (ERH) have also been shown to play essential roles in tumor development by participating in processes like DNA damage repair, cell cycle regulation, and mRNA splicing, further broadening our understanding of cancer-related regulatory mechanisms (26).

Immunotherapy has emerged as a transformative approach in cancer treatment (27). According to previous studies, migrasomes have been implicated in metastasis, where cancer cells produce these vesicles during their interaction with endothelial cells while circulating in the bloodstream (28). They are present in the TME and are involved in key signaling pathways in cancer biology. For example, Zhang et al. show that migrasomes are enriched in VEGF and that these VEGF-rich migrasomes correlate with angiogenic activity and invasiveness in hepatocellular carcinoma (29). Also, previous review highlights data from chorioallantoic membrane models where monocyte-derived migrasomes carrying VEGF and CXCL12 are transferred to endothelial/angiogenesis regions, contributing to vessel formation (30). Regarding the mechanistic link between VEGF and migrasomes, VEGF signaling plays a key role in tumor progression by promoting angiogenesis, enhancing tumor cell survival and migration, and modulating the tumor immune microenvironment. Tumor cells secrete VEGF, which activates VEGF receptor 2 (VEGFR-2) on endothelial cells to stimulate neovascularization, meeting the tumor’s oxygen and nutrient demands—a process often referred to as the “angiogenic switch”, essential for the transition from microscopic lesions to larger tumors (31). In addition, VEGF binding to VEGFR-2 activates downstream pathways, such as PI3K/Akt and MAPK, promoting tumor cell proliferation, survival, and migration, facilitating apoptosis evasion and distant metastasis (32). Moreover, VEGF binding to VEGFR-1 can also suppress anti-tumor immune responses by reducing effector T cell infiltration and increasing the proportion of immunosuppressive cells, thereby promoting immune escape (33). Consequently, VEGF signaling has multifaceted roles in tumor development and metastasis and is an established therapeutic target, with antibodies against VEGF or VEGFR (e.g., bevacizumab) approved for multiple cancer types. Based on our findings, we speculate that migrasomes may transport VEGF or VEGF-regulating lncRNAs produced by tumor cells to neighboring cells (such as Fib), thereby promoting local angiogenesis and tumor-stroma interactions, which could contribute to tumor progression. Nevertheless, the specific relationship between migrasomes and VEGF signaling pathway in ccRCC remains understudied, and further exploration is needed.

The current study has demonstrated that MRLs may play a pivotal role in ccRCC pathology, particularly by influencing immune-related pathways. Functional enrichment analyses showed that immune functions were significantly enriched. These findings suggest that MRLs may regulate immune cell infiltration and activity within the TME, influencing ccRCC progression and patient outcomes. Additionally, KEGG pathway analysis identified key immunological pathways, including leukocyte-mediated immunity and immunoglobulin production, further supporting the idea that MRLs may modulate the immune landscape in ccRCC. Interestingly, beyond immune-related pathways, recent studies in bladder cancer have identified ERH as a key protein interacting with translation-related regulators such as EIF2a, influencing the EIF2a-ATF4/CHOP axis and thus modulating cellular stress responses and protein synthesis control (34). While this mechanism has not yet been explored in ccRCC, it raises the possibility of a broader role for ERH across different tumor contexts.

The five risk lncRNAs (EMX2OS, AC106897.1, AC087645.2, AC121338.2, and C5orf66) exhibited differential expression between high- and low-risk groups and involvement in ccRCC progression through immune microenvironment regulation. According to previous studies, the lowly expressed EMX2 opposite strand/antisense RNA (EMX2OS) is involved in various cancers by inhibiting T-cell proliferation and promoting immune evasion (35). AC106897.1, highly expressed in the high-risk group, promotes regulatory T cell infiltration and enhances immune suppression (36). AC087645.2 was also upregulated in the high-risk group; however, no prior studies have specifically reported on these lncRNAs. Based on previous studies, AC121338.2 is highly expressed in tumor cells (37), and this expression pattern is associated with a longer survival time in ccRCC (38). However, its biological function remains unclear, warranting further investigation. C5orf66, which is highly expressed in the high-risk group, has been linked to increased TMB, potentially impairing immune recognition and tumor cell elimination (39). Moreover, the oncogenic role of C5orf66 has been previously reported in other malignancies, such as triple-negative breast cancer (40), osteosarcoma cell proliferation (41), where it was shown to promote tumor progression. Although some of these lncRNAs have been preliminarily studied in other tumor types, their specific functional roles in ccRCC remain largely unexplored and warrant further investigation.

Moreover, our data showed that immune pathways were predominantly enriched in high-risk ccRCC patients, which may contribute to the aggressive nature of the disease in these individuals. The potential mechanisms underlying the observation that enhanced immune pathways in high-risk ccRCC patients may be accompanied by immune dysfunction, particularly T cell exhaustion. This dysfunctional state may impair the immune system’s ability to eliminate tumor cells, thereby contributing to poor prognosis. Recent studies have supported this notion. For example, Zhang et al. demonstrated that the lncRNA MIAT promotes exhaustion of tumor-infiltrating CD8⁺ T cells by modulating immune checkpoint pathways, highlighting its role in immune evasion (42). In addition, Xie et al. reported that immune escape in ccRCC is largely driven by T cell exhaustion, with immunosuppressive cells in the TME, such as regulatory T cells and myeloid-derived suppressor cells, further exacerbating this process (43). The immune-related functions, including antigen presentation and phagocytic activity, are crucial for recognizing and eliminating tumor cells. Supporting this, functional analyses in bladder cancer have shown that ERH regulates apoptotic and inflammatory signaling pathways, including NF-κB, TLR, TNF, and TGF-β, suggesting a role in modulating both tumor survival and immune response networks (44). Although these findings are based on bladder cancer models, the conserved nature of these pathways suggests potential relevance to ccRCC. These findings collectively underscore the potential of immune-associated regulatory molecules, including MRLs, as promising targets for immunotherapy, offering a new avenue for therapeutic intervention in ccRCC by restoring immune function and improving patient prognosis.

Importantly, these findings also carry notable clinical implications. At present, lncRNA expression detection has not yet been incorporated into routine hematological assays for tumor diagnosis. However, in research and exploratory clinical applications, circulating lncRNAs in peripheral blood (including those in plasma, serum, or exosomes) have been widely investigated. Numerous studies have reported clinically relevant circulating lncRNAs with diagnostic and prognostic potential. For example, HOTAIR, MALAT1, and H19 have shown diagnostic and prognostic value in breast (45-47), liver (48,49), and various diseases (50). Moreover, the stability of lncRNAs in circulation and exosomes highlights their promise as liquid biopsy biomarkers (51). Hence, the aim of our study is to provide a theoretical basis for the potential future application of five lncRNAs in the clinical application of ccRCC.

Although our study obtained valuable insights, there are some limitations that need to be acknowledged. First, our study relied on publicly available datasets for disease cohorts, highlighting the need for additional, independent external datasets to further validate the accuracy and reliability of the prognostic models. Second, the expression of model-derived MRLs was validated in five patients, serving as preliminary proof-of-concept data, and further validation in larger, independent cohorts, as well as at the protein level using immunohistochemistry and Western blotting, is required to strengthen the findings. Although our findings implicate VEGF in regulating migrasome-associated patterns, functional validation will be required in future studies to delineate the causal role of VEGF signaling pathway. Moreover, integrating in vivo data is crucial to gain a deeper understanding of the mechanisms by which MRLs contribute to the pathophysiology of ccRCC.

Collectively, by elucidating their association with immune infiltration, tumor progression, and therapeutic sensitivity, these MRLs may serve as a better therapeutic strategy for ccRCC patients.

Conclusions

In conclusion, this study offers novel insights into the prognostic value of MRLs in ccRCC. Our findings highlight the critical role of these lncRNAs, especially those associated with immune-related pathways, in tumor progression and patient prognosis. By constructing a robust prognostic model, we successfully classified ccRCC patients into high- and low-risk groups, with significant survival differences, highlighting the clinical relevance of MRLs. This study proposes a novel framework that not only sheds light on the complexities of ccRCC heterogeneity but also provides a valuable understanding of the TME in ccRCC.

Acknowledgments

We thank Home for Researchers editorial team (https://www.home-for-researchers.com/) for language editing service.

Footnote

Reporting Checklist: The authors have completed the TRIPOD reporting checklist. Available at https://tau.amegroups.com/article/view/10.21037/tau-2025-541/rc

Data Sharing Statement: Available at https://tau.amegroups.com/article/view/10.21037/tau-2025-541/dss

Peer Review File: Available at https://tau.amegroups.com/article/view/10.21037/tau-2025-541/prf

Funding: This research was supported by the Youth Program of Scientific Research Foundation of Guangxi Medical University Cancer Hospital (No. 2024-01), the Youth Science Foundation of Guangxi Medical University (No. GXMUYSF202546), the Youth Program of Scientific Research Foundation of Guangxi Medical University Cancer Hospital (No. 2023-02), the Guangxi Research Foundation for Basic Capability Improvement of Young and Middle-aged Teachers in Universities (No. 2024KY0124), the Innovation Project of Guangxi Graduate Education (No. YCSW2025276), and the Health Commission of the Guangxi Zhuang Autonomous Region (No. Z-A20230721).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tau.amegroups.com/article/view/10.21037/tau-2025-541/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by The First Affiliated Hospital of Guangxi Medical University (No. 2024-E805-01). The participants provided their written informed consent to participate in this study.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Bukavina L, Bensalah K, Bray F, et al. Epidemiology of Renal Cell Carcinoma: 2022 Update. Eur Urol 2022;82:529-42. [Crossref] [PubMed]
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. CA Cancer J Clin 2024;74:229-63. [Crossref] [PubMed]
3. Yu S, Yu L. Migrasome biogenesis and functions. FEBS J 2022;289:7246-54. [Crossref] [PubMed]
4. Grau JF, Farinas-Madrid L, Garcia-Duran C, et al. Advances in immunotherapy in cervical cancer. Int J Gynecol Cancer 2023;33:403-13. [Crossref] [PubMed]
5. Yu J, Yu L. A decade of migrasome research: biogenesis, physiological functions, and disease implications. Cell Res 2025;35:629-41. [Crossref] [PubMed]
6. Zhang R, Peng J, Zhang Y, et al. Pancreatic cancer cell-derived migrasomes promote cancer progression by fostering an immunosuppressive tumor microenvironment. Cancer Lett 2024;605:217289. [Crossref] [PubMed]
7. Zhang Y, Chen W, Zhu J, et al. The biogenesis and biological roles of migrasomes in human diseases. Cell Death Discov 2025;11:296. [Crossref] [PubMed]
8. Zhang R, Liu Q, Peng J, et al. CXCL5 overexpression predicts a poor prognosis in pancreatic ductal adenocarcinoma and is correlated with immune cell infiltration. J Cancer 2020;11:2371-81. [Crossref] [PubMed]
9. Liu X, Liu D, Zhao H, et al. Biogenesis, functional roles, and pathological implications of migrasomes. Cell Death Dis 2025;16:629. [Crossref] [PubMed]
10. Chen R, Dong L, Wei T, et al. Migrasome-related lncRNAs predict prognosis and immune response of clear cell renal cell carcinoma. Transl Androl Urol 2025;14:1214-29. [Crossref] [PubMed]
11. Sun Z, Jing C, Xiao C, et al. An autophagy-related long non-coding RNA prognostic signature accurately predicts survival outcomes in bladder urothelial carcinoma patients. Aging (Albany NY) 2020;12:15624-37. [Crossref] [PubMed]
12. Wang C, Chen S, Li S, et al. A Prognostic Model for Predicting Tumor Mutation Burden and Tumor-Infiltrating Immune Cells in Bladder Urothelial Carcinoma. Front Genet 2022;13:708003. [Crossref] [PubMed]
13. Chen S, Sun Y, Zhu X, et al. Prediction of Survival Outcome in Lower-Grade Glioma Using a Prognostic Signature with 33 Immune-Related Gene Pairs. Int J Gen Med 2021;14:8149-60. [Crossref] [PubMed]
14. Song X, Zhu Y, Geng W, et al. Spatial and single-cell transcriptomics reveal cellular heterogeneity and a novel cancer-promoting Treg cell subset in human clear-cell renal cell carcinoma. J Immunother Cancer 2025;13:e010183. [Crossref] [PubMed]
15. Huang J, Pang D, Fan C, et al. Integrative transcriptomic and machine learning analyses identify HDAC9 as a key regulator of mitochondrial dysfunction and senescence-associated inflammation in diabetic nephropathy. Front Immunol 2025;16:1627173. [Crossref] [PubMed]
16. Song Q, Zheng Y, Wu J, et al. PTP4A3 Is a Prognostic Biomarker Correlated With Immune Infiltrates in Papillary Renal Cell Carcinoma. Front Immunol 2021;12:717688. [Crossref] [PubMed]
17. Raghubar AM, Roberts MJ, Wood S, et al. Cellular milieu in clear cell renal cell carcinoma. Front Oncol 2022;12:943583. [Crossref] [PubMed]
18. Akhtar M, Al-Bozom IA, Al Hussain T. Papillary Renal Cell Carcinoma (PRCC): An Update. Adv Anat Pathol 2019;26:124-32. [Crossref] [PubMed]
19. Ni J, Cui N, Wang Y, et al. Case Report: Bilateral Renal Cell Carcinoma With Different Histological and Morphological Features, Clear Cell and Cystic Thyroid-Like Follicular Subtype. Front Oncol 2021;11:659706. [Crossref] [PubMed]
20. Qi M, Tu J, He R, et al. NEDD4L Suppresses Proliferation and Promotes Apoptosis by Ubiquitinating RAC2 Expression and Acts as a Prognostic Biomarker in Clear Cell Renal Cell Carcinoma. Int J Mol Sci 2024;25:11933. [Crossref] [PubMed]
21. Montero JJ, Trozzo R, Sugden M, et al. Genome-scale pan-cancer interrogation of lncRNA dependencies using CasRx. Nat Methods 2024;21:584-96. [Crossref] [PubMed]
22. Ma L, Liu X, Roopashree R, et al. Long non-coding RNAs (lncRNAs) in cancer development: new insight from STAT3 signaling pathway to immune evasion. Clin Exp Med 2025;25:53. [Crossref] [PubMed]
23. Ge X, Shen Z, Yin Y. Comprehensive review of LncRNA-mediated therapeutic resistance in non-small cell lung cancer. Cancer Cell Int 2024;24:369. [Crossref] [PubMed]
24. Wei H, Yang Z, Lin B. Overexpression of long non coding RNA CA3-AS1 suppresses proliferation, invasion and promotes apoptosis via miRNA-93/PTEN axis in colorectal cancer. Gene 2019;687:9-15. [Crossref] [PubMed]
25. Wang H, Chen W, Yang P, et al. Knockdown of linc00152 inhibits the progression of gastric cancer by regulating microRNA-193b-3p/ETS1 axis. Cancer Biol Ther 2019;20:461-73. [Crossref] [PubMed]
26. Pang K, Li ML, Hao L, et al. ERH Gene and Its Role in Cancer Cells. Front Oncol 2022;12:900496. [Crossref] [PubMed]
27. Baraibar I, García Durán Sr C, Nieto P, et al. Single cell characterization of persistent cells upon immunotherapy treatment in colorectal and endometrial tumors: The SERPENTINE trial. J Clin Oncol 2024;42:3139.
28. Zhang C, Li T, Yin S, et al. Monocytes deposit migrasomes to promote embryonic angiogenesis. Nat Cell Biol 2022;24:1726-38. [Crossref] [PubMed]
29. Zhang K, Zhu Z, Jia R, et al. CD151-enriched migrasomes mediate hepatocellular carcinoma invasion by conditioning cancer cells and promoting angiogenesis. J Exp Clin Cancer Res 2024;43:160. [Crossref] [PubMed]
30. Zhang Q, Su J, Li Z, et al. Migrasomes as intercellular messengers: potential in the pathological mechanism, diagnosis and treatment of clinical diseases. J Nanobiotechnology 2025;23:302. [Crossref] [PubMed]
31. Liu ZL, Chen HH, Zheng LL, et al. Angiogenic signaling pathways and anti-angiogenic therapy for cancer. Signal Transduct Target Ther 2023;8:198. [Crossref] [PubMed]
32. Shaw P, Dwivedi SKD, Bhattacharya R, et al. VEGF signaling: Role in angiogenesis and beyond. Biochim Biophys Acta Rev Cancer 2024;1879:189079. [Crossref] [PubMed]
33. Kang Y, Li H, Liu Y, et al. Regulation of VEGF-A expression and VEGF-A-targeted therapy in malignant tumors. J Cancer Res Clin Oncol 2024;150:221. [Crossref] [PubMed]
34. Pang K, Hao L, Shi Z, et al. Comprehensive gene expression analysis after ERH gene knockdown in human bladder cancer T24 cell lines. Gene 2020;738:144475. [Crossref] [PubMed]
35. Wang Z, Zhang C, Chang J, et al. LncRNA EMX2OS, Regulated by TCF12, Interacts with FUS to Regulate the Proliferation, Migration and Invasion of Prostate Cancer Cells Through the cGMP-PKG Signaling Pathway. Onco Targets Ther 2020;13:7045-56. [Crossref] [PubMed]
36. Xie Q, Yu H, Liu Z, et al. Identification and characterization of the ferroptosis-related ceRNA network in irreversible pulpitis. Front Immunol 2023;14:1198053. [Crossref] [PubMed]
37. Zhou X, Yao L, Zhou X, et al. Pyroptosis-Related lncRNA Prognostic Model for Renal Cancer Contributes to Immunodiagnosis and Immunotherapy. Front Oncol 2022;12:837155. [Crossref] [PubMed]
38. Feng K, Zhou S, Sheng Y, et al. Disulfidptosis-Related LncRNA Signatures for Prognostic Prediction in Kidney Renal Clear Cell Carcinoma. Clin Genitourin Cancer 2024;22:102095. [Crossref] [PubMed]
39. Zhou Q, Li H, Jing J, et al. Evaluation of C5orf66-AS1 as a Potential Biomarker for Predicting Early Gastric Cancer and Its Role in Gastric Carcinogenesis. Onco Targets Ther 2020;13:2795-805. [Crossref] [PubMed]
40. Zhu S, Sun J, Liu X, et al. CTCF-Induced lncRNA C5orf66-AS1 Facilitates the Progression of Triple-Negative Breast Cancer via Sponging miR-149-5p to Up-Regulate CTCF and CTNNB1 to Activate Wnt/β-Catenin Pathway. Mol Cell Biol 2022;42:e0018821. [Crossref] [PubMed]
41. Zhang H, Song J. Knockdown of lncRNA C5orf66-AS1 inhibits osteosarcoma cell proliferation and invasion via miR-149-5p upregulation. Oncol Lett 2021;22:757. Erratum in: Oncol Lett 2025;29:254. [Crossref] [PubMed]
42. Zhang MX, Jing LY, Tan HT, et al. MIAT promotes tumor-infiltrating CD8(+) T-cell exhaustion and malignant progression of renal cell carcinoma via activating JAK3/STAT3 pathway. J Immunother Cancer 2025;13:e011162. [Crossref] [PubMed]
43. Xie Z, Peng S, Wang J, et al. Multi-omics analysis reveals the role of ribosome biogenesis in malignant clear cell renal cell carcinoma and the development of a machine learning-based prognostic model. Front Immunol 2025;16:1602898. [Crossref] [PubMed]
44. Pang K, Dong Y, Hao L, et al. ERH Interacts With EIF2α and Regulates the EIF2α/ATF4/CHOP Pathway in Bladder Cancer Cells. Front Oncol 2022;12:871687. [Crossref] [PubMed]
45. Zhang C, Lu YJ, Wang M, et al. Characterisation of APOBEC3B-Mediated RNA editing in breast cancer cells reveals regulatory roles of NEAT1 and MALAT1 lncRNAs. Oncogene 2024;43:3366-77. [Crossref] [PubMed]
46. Wu W, Yao Z, Chen Y, et al. HOTAIR, a ferroptosis-related gene, promotes malignant behavior of breast cancer via sponging miR-206. Discov Oncol 2025;16:948. [Crossref] [PubMed]
47. Yang QH, Fu YQ, Feng WL, et al. LncRNA-MALAT1 promotes triple-negative breast cancer progression and function as ceRNA to target REEP5 by sponging miR-106a-5p. Eur J Med Res 2025;30:159. [Crossref] [PubMed]
48. Cheng M, Wang B, Duan L, et al. HOTAIR Knockdown Increases the Sensitivity of Hepatocellular Carcinoma Cells to Sorafenib by Disrupting miR-145-5p/HK2 Axis-Mediated Mitochondrial Function and Glycolysis. Front Biosci (Landmark Ed) 2025;30:37368. [Crossref] [PubMed]
49. Shi CJ, Pang FX, Lei YH, et al. 5-methylcytosine methylation of MALAT1 promotes resistance to sorafenib in hepatocellular carcinoma through ELAVL1/SLC7A11-mediated ferroptosis. Drug Resist Updat 2025;78:101181. [Crossref] [PubMed]
50. Gupta SC, Awasthee N, Rai V, et al. Long non-coding RNAs and nuclear factor-κB crosstalk in cancer and other human diseases. Biochim Biophys Acta Rev Cancer 2020;1873:188316. [Crossref] [PubMed]
51. Wei J, Wang X, Dong D, et al. Extracellular vesicle-derived lncRNA-GC1 serves as a novel biomarker for predicting and monitoring the immunotherapeutic outcomes of patients with gastric cancer. BMC Med 2025;23:499. [Crossref] [PubMed]