Subtype cluster analysis unveiled the correlation between m6A- and cuproptosis-related lncRNAs and the prognosis, immune microenvironment, and treatment sensitivity of esophageal cancer

Esophageal cancer (EC) is a highly aggressive malignancy with a poor prognosis, representing a significant global health challenge (1). Despite advances in treatment modalities—including surgery, chemotherapy, radiotherapy, molecular targeted therapy, and various combination approaches—both the morbidity and mortality rates associated with EC remain alarmingly high (2, 3). These therapeutic advancements have yet to achieve substantial improvements in long-term outcomes, as the overall prognosis for EC patients remains grim (4). Consequently, EC continues to pose a serious threat to public health worldwide, underscoring the urgency of developing more effective therapeutic strategies (5, 6). Addressing these challenges requires an integrated effort to better understand the disease’s molecular mechanisms and to identify innovative treatments capable of improving patient survival and quality of life.

Copper is an essential cofactor required for the survival and function of all organisms, playing critical roles in various biochemical processes. However, when copper concentrations exceed the levels maintained by evolved homeostatic systems, it becomes toxic (7). This toxicity is primarily mediated through the disruption of the tricarboxylic acid (TCA) cycle, where copper directly binds to the lipid components of TCA cycle enzymes, inducing copper-dependent cell death (cuproptosis). In cells with active TCA cycles, elevated levels of lipid-acylated TCA enzymes and direct copper binding to their lipid acyl moieties result in the accumulation of lipid-acylated proteins, depletion of Fe-S cluster-containing proteins, and activation of HSP70. These events collectively lead to protein stress and ultimately cellular demise (7). Recent studies have identified cuproptosis-related genes (CRGs) as key determinants in predicting prognosis for certain cancers, including colorectal cancer and osteosarcoma, using risk prognostic models (8–12). These findings underscore the potential of CRGs as novel biomarkers and therapeutic targets. However, the role of CRGs in EC has not been further explored, representing a critical gap in the understanding of copper-induced cellular toxicity and its implications for EC prognosis and treatment. Further investigation is warranted to elucidate the function of CRGs in EC and their potential utility in clinical practice.

N6-methyladenosine (m6A) represents the most prevalent internal modification of mRNA in mammalian cells, playing a pivotal role in post-transcriptional regulation (13). First identified in the 1970s (14), m6A has since been recognized as a crucial regulator of diverse cellular and biological processes (14). This modification influences the structural conformation, stability, degradation, and cellular interactions of mRNA, thereby modulating key processes such as splicing, translation, nuclear export, and RNA decay (15). The implications of m6A methylation extend far beyond fundamental biology, with growing evidence highlighting its critical role in human diseases. Aberrant m6A modification has emerged as a hallmark of cancer, with methylation patterns of m6A-related genes offering potential as distinctive diagnostic biomarkers and therapeutic targets (16). Moreover, m6A plays a vital role in regulating tumor metabolism, further underscoring its significance in cancer progression and treatment strategies (17). As research continues to unravel the complexities of m6A modifications, their potential to transform cancer diagnostics and therapeutics becomes increasingly apparent. Exploring the intricate interplay between m6A methylation and disease processes will likely yield novel insights and open new avenues for precision medicine.

The regulatory mechanisms involving m6A and cuproptosis have garnered increasing attention in cancer biology due to their multifaceted roles. m6A modifications have demonstrated significant potential not only as prognostic biomarkers but also as key regulators of tumor cell proliferation and immune modulation within the tumor microenvironment (18–20). Similarly, cuproptosis-associated characteristics have been identified as predictive indicators for the prognosis and immune response across various cancer types (21, 22). Importantly, both m6A and cuproptosis are intricately linked to cancer prognosis and are actively involved in shaping the tumor immune microenvironment. Their contributions to immune regulation underscore their potential as therapeutic targets, opening avenues for developing novel immunotherapeutic strategies in oncology. Previous finding underscore the potential of RiskScore system comprising ten m6A/m5C-related lncRNAs as effective biomarkers for predicting survival outcomes, characterizing the immune landscape, and assessing response to immunotherapy in esophageal squamous cell carcinoma (ESCC) (23). Study identified that m6A-mediated modification of the autophagy-related gene ATG10 plays a critical role in inducing cuproptosis in kidney chromophobe, revealing a novel intersection between autophagy, m6A methylation, and cuproptosis in cancer biology (24). Previous investigations have extensively examined the prognostic implications of m6A- and cuproptosis-related long non-coding RNAs (m6aCRLncs) across a broad spectrum of malignancies, including hepatocellular carcinoma, head and neck squamous cell carcinoma, gastrointestinal cancers, and clear cell renal cell carcinoma (25–28). These studies collectively underscore the significant value of m6aCRLncs as robust predictors of cancer prognosis, further highlighting their potential to inform clinical decision-making and therapeutic strategies by integrating molecular markers with patient outcomes.

Despite these advancements, the development of a risk prognostic model specifically focused on m6aCRLncs in EC remains unexplored. Given the significant prognostic and therapeutic implications of such models in other malignancies, investigating the role of m6aCRLncs in EC could provide valuable insights. This gap in research underscores the need for future studies to assess the interplay between m6A methylation, cuproptosis, and lncRNAs in EC, potentially unveiling novel biomarkers and therapeutic targets for this aggressive disease. This study successfully established a novel risk prognostic model centered on m6aCRLncs, providing a comprehensive framework for evaluating their critical role in EC. The model highlights the prognostic significance of m6aCRLncs and their potential involvement in modulating the immune microenvironment. By elucidating the interplay between m6aCRLncs and key immunological processes, this research offers valuable insights into their utility not only as predictive biomarkers but also as potential therapeutic targets in EC. Such findings underscore the importance of m6aCRLncs in advancing precision oncology and optimizing treatment strategies for patients with EC.

To enhance the clarity and comprehensibility of our study, a flowchart was constructed, as illustrated in Figure 1.

This study developed a novel prognostic risk model for EC, leveraging five m6aCRLncs (ELF3-AS1, HNF1A-AS1, LINC00942, LINC01389, and MIR181A2HG) identified as high-risk factors for EC patients. Comprehensive validation demonstrated that the model effectively predicts patient survival outcomes and provides insights into the immune microenvironment of EC. Additionally, subtype clustering and correlation analyses with clinical features revealed significant differences in cluster composition, clinical stage, and N stage between high- and low-risk groups. Immune profiling further highlighted disparities between these groups, with naive B cells, resting CD4 T cells, and plasma cells positively correlating with risk scores, whereas macrophages M0 and M1 displayed negative correlations. Differential expression of immune checkpoint-related genes, including TNFRSF14, TNFSF15, TNFRSF18, LGALS9, CD44, HHLA2, and CD40, provided additional mechanistic insights. Lastly, drug sensitivity analysis identified nine therapeutic agents with potential efficacy for EC patients, including Bleomycin, Cisplatin, Cyclopamine, PLX4720, Erlotinib, Gefitinib, RO.3306, XMD8.85, and WH.4.023, offering promising avenues for personalized treatment strategies.

In this study, we identified seven m6aCRLncs significantly associated with the prognosis of EC through univariate Cox regression analysis. These m6aCRLncs included ELF3-AS1, HNF1A-AS1, JPX, LINC00942, LINC01389, MIR181A2HG, and PPP1R26-AS1. As illustrated in Figure 3A, all seven m6aCRLncs exhibited hazard ratio (HR) greater than 1, indicating their classification as high-risk m6aCRLncs implicated in the progression and pathogenesis of EC. Differential expression analysis was subsequently conducted to compare these m6aCRLncs between the control group (comprising 11 normal samples) and the tumor group (comprising 159 EC patients), as shown in Figures 3B, C. The results demonstrated that ELF3-AS1, JPX, LINC00942, LINC01389, MIR181A2HG, and PPP1R26-AS1 were significantly upregulated in EC patients, whereas HNF1A-AS1 showed reduced expression in the EC group. To further refine prognostic relevance, a risk-prognosis model was developed, which identified five key m6aCRLncs (ELF3-AS1, HNF1A-AS1, LINC00942, LINC01389, and MIR181A2HG) as significant contributors to patient stratification. As depicted in Figures 6E, 7E, and 8E, these five m6aCRLncs exhibited higher expression levels in the high-risk EC patient group compared to the low-risk group. Notably, four of these m6aCRLncs (ELF3-AS1, LINC00942, LINC01389, and MIR181A2HG) were consistently overexpressed in both the overall EC cohort and the high-risk subgroup, suggesting their critical roles in EC pathobiology. The findings were further validated through PCR analysis, which confirmed the elevated expression of ELF3-AS1, LINC00942, LINC01389, and MIR181A2HG in EC patients compared to controls, as shown in Figure 12. These results collectively underscore the potential of these m6aCRLncs as biomarkers for prognosis and as targets for therapeutic intervention in EC.

In contrast to the four m6aCRLncs (ELF3-AS1, LINC00942, LINC01389, and MIR181A2HG), HNF1A-AS1 exhibits a unique expression pattern in EC patients. Although HNF1A-AS1 is more highly expressed in the high-risk group compared to the low-risk group of EC patients, it shows decreased expression in the tumor group (comprising 159 EC samples) compared to the control group (11 normal samples). It is important to note that the high- and low-risk groups within the risk prognosis model consist solely of EC patients, and the comparisons do not involve normal controls. Consequently, the elevated expression of HNF1A-AS1 in high-risk EC patients relative to the low-risk group does not imply that its expression is generally higher in EC patients when compared to healthy individuals. Differential expression analysis revealed lower expression levels of HNF1A-AS1 in EC patients relative to normal controls, although its expression remained significantly different between the high-risk and low-risk subgroups, with higher levels observed in high-risk patients. However, the PCR validation results did not show statistically significant differences in HNF1A-AS1 expression between EC patients and controls. While HNF1A-AS1 appeared to be moderately overexpressed in EC patients, as depicted in Figure 12, this observation lacked statistical significance. Such discrepancies between bioinformatics predictions and PCR validation are not uncommon and reflect the inherent limitations of computational analysis, particularly when combined with experimental methods. Importantly, four of the five m6aCRLncs included in the risk-prognosis model (ELF3-AS1, LINC00942, LINC01389, and MIR181A2HG) were successfully validated by PCR, which sufficiently supports the overall findings and biological significance of the model. The inconsistent results for HNF1A-AS1 may be influenced by the disparity in sample sizes between the two groups analyzed. Specifically, the differential expression analysis compared 11 normal samples with 159 EC samples, and the imbalanced sample sizes may have contributed to variations in the statistical outcomes for HNF1A-AS1. Unfortunately, the current dataset did not allow for analyses using a more balanced or larger sample size, representing a limitation of this study. Addressing this limitation in future research through studies with larger and more evenly distributed sample sizes will be crucial to better understand the expression patterns and biological roles of HNF1A-AS1 in EC prognosis and pathogenesis.

This study identified ELF3-AS1, LINC00942, LINC01389 and MIR181A2HG as being significantly associated with tumor prognosis of EC patients, thereby reinforcing the validity of the findings. The prognostic value of HNF1A-AS1 in EC patients needs further study. ELF3-AS1 has been strongly linked to the prognosis of glioma and hepatocellular carcinoma (37–39) and has been shown to accelerate gastric cancer progression through binding to hnRNPK (40). LINC00942 has been linked to prognosis and immune responses in hepatocellular and bladder cancers (41–45) and promotes METTL14-mediated m6A methylation in breast cancer (46). Although the role of LINC01389 in tumor prognosis remains uncertain, it has been shown to participate in the epithelial-mesenchymal transition in stomach cancer (47). MIR181A2HG is associated with prognostic prediction and immunotherapy response in bladder cancer (48) and serves as a diagnostic marker for thyroid cancer (49). HNF1A-AS1 is implicated in osteosarcoma prognosis and tumorigenesis (50) and plays roles in the progression of gastric cancer and glioblastoma (50–52). Specifically, it acts as a competitive endogenous RNA in gastric cancer by sponging miR-30b-3p (34) and, when regulated by HNF1α, mitigates hepatocellular carcinoma malignancy by enhancing SHP-1 phosphatase activity (53). Importantly, HNF1A-AS1 is also involved in the development of EC. Studies reveal its role in regulating proliferation and migration in esophageal adenocarcinoma (EAC) cells (54) and promoting growth and metastasis in esophageal squamous cell carcinoma (ESCC) by sponging miR-214 to upregulate SOX-4 expression (55). These findings highlight the multifaceted roles of these m6aCRLncs in cancer progression and their potential as biomarkers or therapeutic targets.

Tumor immunotherapy represents a promising and innovative therapeutic strategy for EC (6). EC cells are characterized by a rich repertoire of tumor antigens, yet they have developed sophisticated mechanisms to evade anti-tumor immune responses. These mechanisms include activation of immune checkpoints, secretion of immunosuppressive factors, and negative regulation of immune cell activity. The immune landscape within the tumor microenvironment significantly influences cancer progression, patient survival, and treatment efficacy (6). Immune checkpoints play a pivotal role in maintaining self-tolerance and preventing the onset of inflammatory disorders. However, cancer cells can exploit these pathways to induce T-cell exhaustion and impair immune function. In recent years, immunotherapy targeting immune checkpoints has seen remarkable advancements, providing new hope for improving clinical outcomes in EC (56).

The genetic variant rs2234167 within the TNFRSF14 locus has not been associated with the risk of ESCC (57). However, this study revealed that TNFRSF14 expression was upregulated in high-risk EC patients, suggesting its potential involvement in the risk of EC. Additionally, CD44 was found to be downregulated in high-risk EC patients, indicating its possible role in the immunotherapy of EC. Previous studies have identified CD44 as a novel biomarker for EC patients undergoing neoadjuvant chemoradiotherapy and highlighted its utility, in combination with HER2, in enhancing the predictive accuracy of 18F-FDG PET-based clinico-radiomic models for treatment response (58, 59). The immune checkpoint molecule HHLA2 has been shown to predict survival and immune characteristics in patients with ESCC (60). This study reinforces the prognostic and immunotherapeutic potential of HHLA2 in EC. Furthermore, the expression of CD40 in human ESCC has been linked to tumor progression and lymph node metastasis (61). Currently, limited data exists regarding the roles of TNFSF15, TNFRSF18, and LGALS9 in EC. This study identifies these molecules as being associated with the prognosis and immune microenvironment of EC, providing new directions for future research on EC prognosis and immunotherapeutic strategies.

Bleomycin and Cisplatin are established chemotherapeutic agents for EC (62, 63). TNFAIP8 has been implicated in promoting Cisplatin resistance via interaction with TAF-Iα, thereby contributing to the malignant progression of EC (64). Cyclopamine exerts its anti-tumor effects by inhibiting glioma-associated oncogene protein-1, a key marker of EC progression, effectively suppressing the growth of EC cells (65). Furthermore, treatment with KAAD-Cyclopamine or neutralizing antibodies targeting Shh has been shown to reduce EC cell proliferation and induce apoptosis (66). For patients with ESCC who are intolerant to chemoradiotherapy, Erlotinib in combination with radiotherapy has demonstrated therapeutic efficacy (67). Similarly, Gefitinib has been shown to enhance survival and improve quality of life in advanced-stage EC patients who have failed first-line chemotherapy (68). Among the drugs identified as sensitive in this study, several are known to exhibit therapeutic potential in EC, underscoring the robustness of the study’s findings. The roles of PLX4720, RO.3306, XMD8.85, and WH.4.023 in EC have not been well-documented. These agents may represent promising candidates for further exploration as potential therapeutic options for EC.

This study, while offering meaningful insights, is not without its limitations. First, the relatively small sample size of tumor specimens, ​which may constrain the statistical power and generalizability of the findings. A larger, more diverse cohort is essential in future research to ensure the robustness and reproducibility of the conclusions. Expanding the sample size would not only strengthen the statistical validity of the prognostic model but also enhance its applicability across varied patient populations. Furthermore, a larger cohort would provide an opportunity to validate the identified biomarkers in independent datasets, thereby increasing the reliability of the proposed risk stratification framework for clinical implementation. Second, although this research identifies key m6aCRLncs with prognostic and immune-regulatory roles in EC, their underlying biological mechanisms remain largely unexplored. Comprehensive functional analyses are required to elucidate how these m6aCRLncs contribute to tumor progression, immune microenvironment modulation, and therapy resistance. This would provide deeper insights into their potential as biomarkers or therapeutic targets. Addressing these limitations in subsequent studies will be crucial to confirming the translational potential of the findings and further advancing the field of EC research.

In conclusion, this study successfully established a novel prognostic model for EC based on five m6aCRLncs, offering a comprehensive approach to risk stratification. These five m6aCRLncs demonstrated significant potential in predicting immune efficacy and drug sensitivity, as evidenced by analyses of the tumor microenvironment, immune correlations, and drug response patterns. The integration of these biomarkers into a prognostic framework not only provides valuable insights into the immune landscape of EC but also highlights their utility in identifying potential therapeutic options. The findings of this research hold promising implications for predicting patient survival and optimizing clinical management strategies, and lay a theoretical foundation for more personalized and effective treatment protocols for EC patients.

The author(s) declare financial support was received for the research, authorship, and/or publication of this article. This work was financially supported by the Cultivation Project of Xi’an Health and Construction Commission (No. 2024ms14) and Innovation Capacity Support Plan Project of Shaanxi Province (No. 2024RS-CXTD-86).

Publicly available datasets were analyzed in this study. This data can be found here: https://portal.gdc.cancer.gov/analysis_page?app=Downloads↗.

The data used in this article comes from TCGA databases, which are publicly available and can be reused indefinitely through open access. No ethical review was required for this study.

MZ: Formal analysis, Methodology, Resources, Validation, Funding acquisition, Writing – original draft, Writing – review & editing. YS: Formal analysis, Methodology, Resources, Validation, Writing – original draft. PW: Formal analysis, Methodology, Resources, Writing – original draft. XS: Data curation, Validation, Writing – original draft. PY: Data curation, Validation, Writing – original draft. PA: Software, Visualization, Writing – original draft. WJ: Software, Writing – original draft. LL: Software, Visualization, Writing – original draft. ZY: Conceptualization, Funding acquisition, Investigation, Project administration, Supervision, Writing – review & editing. MY: Conceptualization, Investigation, Project administration, Resources, Supervision, Validation, Writing – review & editing.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

The author(s) declare that no Generative AI was used in the creation of this manuscript.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fimmu.2025.1539630/full#supplementary-material↗