Identification and analysis of endoplasmic reticulum stress-related genes in neuroblastoma and construction of a prognostic gene signature

Introduction

Neuroblastoma (NB), a neuroendocrine tumor that occurs in the developing sympathetic nervous system, is the most common extracranial solid tumor in children, accounting for 7–8% of childhood malignancies and for approximately 15% of all childhood cancer deaths (1,2). Mounting evidence indicates that the prognosis of patients with NB ranges from spontaneous regression to fatal outcomes depending on risk factors, including age, disease stage, and genetic aspects such as amplification of the MYCN gene (3). Patients with low-risk NB, who have undergone only surgery or demonstrate spontaneous regression or maturation without treatment, have long-term survival rates of up to 90% (4,5). At the same time, for patients with high-risk NB, the long-term survival rate is only 30–40% despite intensive multimodal treatments with high-dose chemotherapy, radiotherapy, and stem cell transplantation (3,6). Early diagnosis and timely treatment of patients with high-risk NB can increase their long-term survival and improve the overall prognosis. Therefore, it is necessary to have reliable diagnostic and prognostic biomarkers for early identification of patients with high-risk NB.

NB originates from neural crest-derived progenitor cells and is marked by profound transcriptional dysregulation, metabolic reprogramming, and high proliferative activity, particularly in tumors with MYCN amplification. These features impose an excessive burden on protein synthesis and folding machinery, rendering NB cells especially vulnerable to disturbances in endoplasmic reticulum (ER) homeostasis. ER stress occurs when the protein-folding capacity of the ER is overwhelmed, leading to activation of adaptive or apoptotic signaling pathways (7). Increasing evidence suggests that ER stress plays a critical role in cancer progression by influencing tumor cell growth, proliferation, metastasis, invasion, angiogenesis, and resistance to radiotherapy (8-10). Cancer cells characterized by abnormal transcription and metabolism produce a hostile microenvironment, which triggers ER stress via oxygen and nutrient deprivation, thus disrupting correct protein folding not only in tumor cells but also in stromal cells and infiltrating leukocytes (11). Although prolonged or unresolved ER stress can lead to apoptosis of cancer cells (7), adaptive ER stress responses may paradoxically promote tumor growth, angiogenesis, immune escape, and resistance to chemo- and radiotherapy (11). Notably, NB cells exhibit features such as developmental arrest and metabolic dependency that may favor reliance on ER stress-associated signaling pathways.

It has been shown that in NB, some drugs such as S(+)-ibuprofen and ABTL0812 affect the expression of ER stress-related genes and regulate tumor cell apoptosis (12,13). The global landscape of ER stress-related genes in NB and their clinical relevance remain largely unexplored. In particular, it is unclear whether aberrant expression patterns of ER stress-associated genes define distinct biological subtypes of NB or can be leveraged to predict patient outcomes. Therefore, we hypothesized that aberrant expression of ER stress-related genes might have a prognostic value and present a potential therapeutic target in NB.

In this study, we identified ER stress-related genes differentially expressed in NB cells and performed their functional annotation, followed by protein-protein interaction (PPI) network construction. ER stress-related genes associated with the survival of patients with NB were identified. In addition, the developed NB prognostic model based on the ER stress-related gene signature presents new biomarkers and potential therapeutic targets in NB and should be verified in future studies. We present this article in accordance with the TRIPOD reporting checklist (available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-1-2623/rc).

Methods

Datasets and data collection

In this study, we analyzed two public RNA-microarray datasets (GSE66586 and GSE78061) from Gene Expression Omnibus (GEO; http://www.ncbi.nlm.nih.gov/geo/). The GSE78061 dataset includes the expression data for 25 NB and 4 control cell lines (14), and the GSE66586 dataset contains expression profiles of 8 NB and 2 control cell lines (15). In both datasets, the gene expression data were generated using the GPL6244 platform (Affymetrix Human Gene 1.0 ST Array; Agilent Technologies, Palo Alto, CA, USA).

ER stress-related genes were extracted from the GeneCards database (https://www.genecards.org/), and those with a relevance score of ≥7 were selected.

Therapeutically Applicable Research to Generate Effective Treatments (TARGET) is a database containing high-resolution genomic data and clinical information related to childhood cancers. We downloaded the data on 249 patients with NB from TARGET, including microarray-seq messenger RNA (mRNA) expression and clinical information; for four patients, clinical information was missing, and they were excluded from analysis. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.

Analysis and annotation of differentially expressed genes (DEGs)

The GEOquery R package was used to download mRNA expression profiles from GEO. The t-test from R package “limma” was used to identify DEGs in NB and control cells (16,17), which were further screened based on fold change (FC) and adjusted P values: |log₂FC| ≥1 and P<0.05. Heatmaps and volcano plots were produced using R package “ggplot2”. The ER stress‐related genes overlapping with DEGs were identified using R package “Venn” and considered as ER stress‐related DEGs.

R package “clusterProfiler” was used to process the data and visualize the results of Gene Ontology (GO; http://www.geneontology.org) and Kyoto Encyclopedia of Genes and Genomes (KEGG; http://www.genome.ad.jp/kegg/) enrichment analyses.

Construction of the PPI network

The PPI network for ER stress-related genes was generated using STRING (version 11.5; https://string-db.org), which detects physical and functional interactions between proteins. The results were uploaded into Cytoscape (version 3.9.1) to establish a network model, and hub genes were selected using the CytoHubba plugin in Cytoscape. The top 10 ER stress-related genes with the highest maximal clique centrality (MCC) scores were defined as hub genes.

Association between ER stress-related genes and survival

To further clarify the relationship between ER stress‐related genes and NB prognosis, we conducted Kaplan-Meier survival analysis of the TARGET database used R packages “survival” and “survminer”. The results were statistically analyzed using log-rank test and considered significant at P<0.05.

Construction and validation of the prognostic risk model

The ER stress‐related genes were then integrated into an overall survival (OS) based least absolute shrinkage and selection operator (LASSO) Cox regression model implemented with R package “glmnet” (18). The related survival and genomic information on patients with NB was extracted from the TARGET database. To obtain the best penalty parameter lambda, 10‐fold cross‐validation (10 FCV) was performed in LASSO analysis (19). The risk score of the prognostic model was calculated according to the expression of ER stress-related genes and the corresponding coefficients using the following formula:

Riskscore=∑in(coefi×Expi)[1]

Patients were clustered into low- and high-risk groups based on the best cutoff value of the risk score in the TARGET database, which was calculated using R package “survminer”. The TARGET clinical cohort was used to evaluate the predictive power of the risk model. Kaplan-Meier survival curves and log-rank tests were used to compare the differences in OS between the two risk groups. The sensitivity and specificity of the prognostic risk model were evaluated based on receiver operating characteristic (ROC) curves generated using R package “pROC”.

RNA isolation and quantitative real-time polymerase chain reaction (qRT-PCR)

Total RNA of cells was extracted using the TRIzol reagent (Invitrogen, Carlsbad, CA, USA) and was reverse transcribed into cDNA using PrimeScript RT reagent Kit (Takara, Shiga, Japan). qRT-PCR was conducted using iTaq Universal SYBR Green Supermix (Bio-Rad Laboratories, Hercules, CA, USA) on an iCycler RT-PCR Detection System (Bio-Rad Laboratories). Each test for each sample is performed in triplicate. Besides, the actin was used as an internal reference, and the qRT-PCR data were normalized using the 2^−ΔΔCt method. The primer sequences for RT-PCR were as follows: GAPDH-forward (F) 5'-ATGTTCGTCATGGGTGTGAA-3', GAPDH-reverse (R) 5'-GTCTTCTGGGTGGCAGTGAT-3'; MAPK8-F 5'-TCGCTACTACAGAGCACCCG-3', MAPK8-R 5'-TGGGAACAAAACACCACCTTTG-3'.

Statistical analysis

All statistical analyses were performed using R software. In the bioinformatics analysis section, the t-test was used to identify DEGs between groups (|log₂FC| ≥1, adjusted P<0.05), and the hypergeometric test was applied for GO and KEGG enrichment analyses. Hub genes were selected based on MCC scores. Survival differences between risk groups were examined via Kaplan-Meier analysis with log-rank tests. A LASSO Cox regression model with 10‑fold cross‑validation was constructed to develop a prognostic risk score for overall survival, and model performance was assessed using ROC curves. A P value <0.05 was considered statistically significant.

Results

DEGs in NB cells compared to control cells

To identify DEGs in NB and control cells, we analyzed two microarray mRNA expression profiles extracted from the GEO database (GSE66586 and GSE78061). In the GSE66586 dataset, a total of 3,365 genes were significantly differentially expressed between NB and control cells (P<0.05, |log₂FC| ≥1), including 1,567 downregulated and 1,798 upregulated genes (Figure 1A). In the GSE78061 dataset, 2,351 genes were significantly differentially expressed between NB and control cells (P<0.05, |log₂FC| ≥1), including 1,242 downregulated and 1,109 upregulated genes (Figure 1B). Notably, NB and control cells were clustered separately in both datasets (Figure 1C,1D). Overall, 727 and 576 genes were downregulated and upregulated, respectively, in NB compared to control cells, which coincided in the two datasets (Figure 2A,2B).

Figure 1 Volcano plots of DEGs and PCA of two GEO datasets. (A,B) Volcano plots of GSE66586 (A) and GSE78061 (B). Blue, pink, and gray points indicate downregulated and upregulated DEGs and stably expressed genes, respectively. (C,D) PCA of GSE66586 (C) and GSE78061 (D). Blue and pink points indicate normal and tumor cells, respectively. DEGs, differentially expressed genes; GEO, Gene Expression Omnibus; PCA, principal component analysis.

Figure 2 Identification of ER stress-related genes in NB. Venn diagrams of 58 downregulated (A) and 15 upregulated (B) DEGs in NB cells from GSE66586 and GSE78061 datasets that overlapped with ER stress‐related genes from GeneCards. DEGs, differentially expressed genes; ER, endoplasmic reticulum; NB, neuroblastoma.

ER stress-related DEGs in NB cells

From the GeneCards database, we extracted 785 ER stress-related genes with a relevance score ≥7 (table available at https://cdn.amegroups.cn/static/public/tcr-2025-1-2623-1.pdf). Overall, 73 DEGs overlapped with ER stress-related genes (Table S1), including 58 downregulated (Figure 2A) and 15 upregulated (Figure 2B) genes.

To understand the molecular functions and pathways of ER stress-related genes in NB, we performed GO enrichment analysis. Among GO biological processes, ER stress-related DEGs were mainly involved in wound healing, intrinsic apoptotic signaling pathways, and cellular responses to abiotic and environmental stimuli, and chemical stress (Figure 3). Notably, most of the ER stress-related DEGs were downregulated in NB. Thus, all 18 genes enriched in wound healing, including CAV1, SERPINE1, and GJA1, were downregulated, whereas among 15 genes enriched in the intrinsic apoptotic signaling pathway, 11 were downregulated and 4 upregulated, and among 15 genes involved in cellular response to abiotic stimuli, 13 were downregulated and 2 upregulated (Figure 4A, Table S2). GO-based cellular component analysis revealed that ER stress-related DEGs in NB cells were associated with membrane rafts and microdomains (Figure 3); thus, 14 genes (13 downregulated and 1 upregulated) enriched in the membrane raft category were also related to the membrane microdomain category (Figure 4B, Table S2). GO analysis of molecular functions showed that ER stress-related DEGs in NB were significantly enriched in peptidase regulator activity and protease binding (Figure 3) and that most of them were downregulated: all 7 genes associated with the former category and 5 out of 6 genes associated with the latter category (Figure 4C, Table S2).

Figure 3 Functional enrichment analysis of ER stress-related genes in NB. Analysis of GO categories of BP, CC, and MF. BP, biological process; CC, cellular component; ER, endoplasmic reticulum; GO, Gene Ontology; MF, molecular function; NB, neuroblastoma.

Figure 4 ER stress-related genes linked to biological processes, cellular components, and molecular functions. ER stress-related genes belonging to biological process (A), cellular component (B), and molecular function (C) categories are shown. ER, endoplasmic reticulum.

We further conducted KEGG enrichment analysis and found that the ER stress-related genes upregulated in NB cells were significantly enriched in apoptosis and pathways of neurodegeneration (Figure 5A), and those downregulated were significantly enriched in the pathways of proteoglycans in cancer and PI3K-Akt signaling (Figure 5B).

Figure 5 KEGG pathway enrichment analysis of ER stress-related genes in NB. The top 10 KEGG pathways of upregulated (A) and downregulated (B) ER stress-related genes in NB. ER, endoplasmic reticulum; KEGG, Kyoto Encyclopedia of Genes and Genomes; NB, neuroblastoma.

PPI network construction and determination of hub genes

To examine the correlation among the 73 ER stress-related genes differentially expressed in NB cells, we performed PPI analysis using STRING. A total of 44 overlapping ER stress-related DEGs were identified, visualized, and sorted according to their connection degree scores (Figure 6A,6B). The results indicated that CAV1, FN1, ITGB1, MAPK8, TGFB1, COL1A1, VCL, BCL2L11, PXN, and THBS1 had a high degree of connectivity and, therefore, were selected as hub genes (Figure 6B).

Figure 6 PPI network and identification of ER stress-related hub genes in NB. (A) PPI network constructed for the 73 ER stress-related genes using the STRING database and Cytoscape. Pink and blue nodes represent upregulated and downregulated genes, respectively, and edges represent the interactions between the nodes. (B) The 10 hub genes identified using CytoHubba. ER, endoplasmic reticulum; NB, neuroblastoma; PPI, protein-protein interaction.

Association of ER stress-related DEGs with patient survival

Next, we explored the prognostic value of each ER stress-related DEG in NB based on Kaplan-Meier survival analysis with the log-rank test. In total, the data on 245 patients with complete clinical information were evaluated. Among the 73 ER stress-related DEGs, 10 showed association with patient outcomes. Thus, patients with the high expression of MAPT, MAP2, PRKCD, C9orf72, and MAPK8, and low expression of PIEZO1, RRBP1, JPH1, FLNB, and FN1 had significantly prolonged survival (Figure 7). Notably, FN1 and MAPK8, identified as hub ER stress-related genes, were associated with NB prognosis.

Figure 7 Kaplan-Meier survival curves depending on the expression of ER stress-related genes. ER, endoplasmic reticulum.

Construction of a prognostic gene signature based on ER stress-related genes

To construct an OS model, we subjected all 71 ER stress-related DEGs (2 ER stress-related DEGs were not provided in the TARGET database) to LASSO Cox regression analysis (Figure 8A). The LASSO Cox regression model showed the optimal performance with 10 genes (Figure 8B, Table 1). The risk score was computed as: (−0.1160 × MAP2) + (−0.0994 × PRKCD) + (0.0721 × MAPK8IP1) + (0.0617 × JPH1) + (0.0497 × TPP1) + (−0.0330 × BCL2) + (−0.0223 × EDN1) + (0.0211 × THBS1) + (−0.0206 × MAPK8) + (0.0089 × SERPINA3), where gene symbols indicate gene expression. MAPK8 was identified as a potential hub gene associated with NB risk. To preliminarily validate its expression pattern, we compared MAPK8 mRNA levels between a highly aggressive NB cell line, SK-N-BE(2), and a less aggressive cell line, SK-N-AS (20-22). Notably, MAPK8 expression was markedly decreased in SK-N-BE(2) cells relative to SK-N-AS cells, which is consistent with the trend observed in our risk model (Figure S1). According to the R package “survminer”, the best cutoff value for the risk score was 0.53 (Figure 8C); accordingly, it was used to assign patients into high- and low-risk groups. The results indicated that patients in the high-risk group had significantly shorter survival time [hazard ratio (HR) =4.57; 95% confidence interval (CI): 2.86–7.28; Figure 8D] and that patient mortality was significantly increased with the increase in the risk score (Figure 8E). ROC curve analysis of the sensitivity and specificity of the prognostic model showed that the area under the curve of OS was 0.75, whereas batch survival analysis revealed that the 10 ER stress-related signature genes were closely associated with the outcome (Figure 8F).

Figure 8 ER stress-related genes included in a risk model for NB. (A) LASSO coefficients of the 71 ER stress-related genes in NB. (B) Selection of the optimal lambda penalization coefficient in the LASSO model. (C) Selection of the optimal cutoff value for high- and low-risk groups. (D) Survival of patients from the TARGET database classified into high- and low-risk groups according to the 10-gene prognostic signature. (E) Risk score distribution and survival time (months) in the TARGET database. (F) ROC curves of the sensitivity and specificity of the prognostic signature in NB according to the TARGET database. AUC, area under the curve; ER, endoplasmic reticulum; LASSO, least absolute shrinkage and selection operator; NB, neuroblastoma; ROC, receiver operating characteristic; TARGET, Therapeutically Applicable Research to Generate Effective Treatments.

We further explored the immune infiltration status between the high- and low-risk groups using the ESTIMATE algorithm. No significant differences were observed in ImmuneScore, StromalScore, or ESTIMATEScore between the two groups, suggesting that the prognostic model may not be primarily associated with overall immune or stromal content (Figure S2A). To further characterize the tumor immune microenvironment at a finer resolution, we subsequently quantified the relative abundance of individual immune cell populations. Notably, differential infiltration patterns of specific immune cell subsets were identified between the high- and low-risk groups (Figure S2B).

Discussion

NB is a malignant embryonal neuroendocrine tumor occurring in the sympathetic nervous system, which usually has signs and symptoms of a metastatic tumor (23). Several studies have suggested that ER stress could potentially promote cancer progression, including cancer cell survival and metastasis, treatment resistance, and angiogenesis, ultimately affecting patient prognosis (8,10). However, the underlying mechanisms, such as the functional activity of ER stress-related genes in NB, remain unexplored, and it is unclear whether these genes could be used as predictive biomarkers and therapeutic targets in NB. Therefore, the aim of the current study was to identify ER stress-related genes differentially expressed in NB cells compared to control cells, analyze how their functions and the associated molecular pathways could be involved in NB pathogenesis, and establish a prognostic gene signature.

Using GEO databases, we identified 73 ER stress-related DEGs in NB cells, which were mainly involved in wound healing, intrinsic apoptotic signaling pathways, and cellular responses to abiotic stimuli. These results are consistent with previous studies on cancer mechanisms. Thus, it has been shown that tumor progression and wound healing processes share many common molecular pathways (24); therefore, tumors could be considered as “wounds that do not heal” (25). The biological processes underlying wound healing are normally associated with epithelial-to-mesenchymal transition (EMT) (25), which is a key phenomenon in cancer development, and it has been suggested that ER stress is associated with EMT, cancer cell motility, metastasis, and tumor invasiveness (26). In addition, apoptotic pathways are often suppressed in cancer through upregulation of anti-apoptotic (e.g., BCL2) and downregulation of pro-apoptotic (e.g., CAV1) factors (27-29), which is consistent with our findings that the expression of BCL2 and CAV1 genes was significantly increased and decreased, respectively, in NB cells. Cancer cell apoptosis has been established as a promising target for anticancer therapy; thus, in breast cancer, the knockdown of BCL2-associated factors has been shown to reduce the resistance to endocrine therapy and promote tumor cell death (29). It has also been reported that responses to abiotic stimuli and stress are deregulated in NB, increasing its resistance to cisplatin (30).

Our analysis of GO cellular components indicated that the ER stress-related genes differentially expressed in NB, including FAS and CAV1, were mainly associated with membrane rafts and microdomains. Previous studies indicate that the formation of membrane rafts and microdomains is linked to cancer metastasis and progression (31,32) and that FAS is a key factor in this process, because its downregulation or mutation is correlated with the development of hematologic malignancies and their resistance to chemotherapy (33). At the molecular level, the identified ER stress-related DEGs were mainly involved in the regulation of peptidase activity and protease binding, which play important roles in tumor growth and invasion of surrounding tissues through degradation of pericellular proteins (34). Furthermore, KEGG enrichment analysis showed that the ER stress-related DEGs were related to the PI3K-Akt signaling pathway, a key molecular mechanism involved in apoptosis and multidrug resistance of cancers (35,36). Cumulatively, our functional analysis of the identified ER stress-related genes indicates that they may serve as new candidate targets in NB, which may aid in reducing tumor metastasis and increasing therapeutic efficacy.

The PPI network of ER stress-related DEGs revealed 10 hub genes with a high degree of interaction, including CAV1, FN1, ITGB1, MAPK8, TGFB1, COL1A1, VCL, BCL2L11, PXN, and THBS1. Among them, FN1 and MAPK8 were associated with the prognosis of patients with NB according to Kaplan-Meier survival analysis, which is consistent with previous studies implicating these genes in tumor progression (37-39). FN1, a key extracellular matrix component and an EMT-associated gene, has been associated with a poor prognosis in several malignancies, including gastric adenocarcinoma and non-small cell lung cancer (37,40). MAPK8, a stress-activated kinase within the MAPK signaling cascade, plays a critical role in regulating apoptosis and autophagy under stress conditions and has been implicated in stress-responsive signaling in NB cells (39,41). Previous studies have reported that increased MAPK8 activity may promote autophagy-mediated degradation of FN1, thereby suppressing EMT and improving prognosis in certain cancer types (38,42), whereas MAPK8 downregulation is associated with the inhibition of autophagy and poor prognosis for patients with non-small cell lung cancer (43). Consistent with previous reports on MAPK8 and FN1 expression in cancer, we observed the opposite changes in the transcription of MAPK8 and FN1 genes in NB cells, when the former was upregulated and the latter downregulated compared to normal cells. Importantly, both alterations were significantly associated with patient prognosis, suggesting that MAPK8 and FN1 may play distinct yet complementary roles in NB progression. Given the developmental origin and stress-adaptive nature of NB, these findings raise the possibility that MAPK8-mediated stress and autophagy signaling, together with FN1-associated interactions, may differentially influence disease behavior in an ER stress-dependent manner. However, the precise mechanisms linking MAPK8 and FN1 to ER stress regulation and NB progression remain to be elucidated and warrant further functional and mechanistic investigation.

An important result of our study is the construction of an NB prognostic model based on an ER stress-related gene signature through comprehensive search of the TARGET and GEO databases. After the identification of ER stress-related DEGs in NB, we applied the LASSO Cox regression model to generate a 10-gene signature and an NB prognostic model, which was validated by analyzing clinical information of patients with NB extracted from the TARGET database. ROC curve and Kaplan-Meier survival analyses showed that there were statistically significant differences between patients with high and low NB risk scores, confirming the prognostic value of the ER stress-related 10-gene signature in NB. The 10 genes included in the model, MAP2, PRKCD, MAPK8IP1, JPH1, TPP1, BCL2, EDN1, THBS1, MAPK8, and SERPINA3, have been previously shown to be functionally linked to cancer progression. Thus, high expression of SERPINA3 and low expression of BCL2 and MAP2 are associated with poor prognosis and recurrence in patients with central nervous system tumors (44-46); furthermore, SERPINA3 knockdown results in the inhibition of cancer cell proliferation, invasion, migration, and transformation to a mesenchymal phenotype (47). TPP1 is a telomere-binding protein responsible for the recruitment of telomerase and telomere elongation, which is essential for cancer cell survival (48), and the inhibition of TPP1-telomerase interaction could decrease proliferation and increase apoptosis of lung cancer cells, suggesting a potential therapeutic strategy (49). THBS1, an extracellular matrix protein, promotes cell migration and invasion of oral squamous cell carcinoma (50), and its high expression in esophageal cancer indicates poor prognosis (51). These observations strongly suggest a prognostic potential of these genes in cancer, providing further rationale for including the 10 ER stress-related genes into our NB predictive signature.

Recent studies have proposed NB prognostic models based on distinct biological programs, including pyroptosis-related gene sets, nuclear export-associated markers, and epitranscriptomic regulators, highlighting the relevance of cell fate regulation, stress response, and transcriptional control in NB prognosis (52-54). Compared with these existing approaches, which focus on specific death pathways or individual regulatory processes, our model is based on the coordinated expression patterns of multiple ER stress-related genes. Rather than relying on a single prognostic marker or a specific cell death program, our approach highlights a coherent stress-adaptive biological axis that may be relevant to NB pathophysiology, reflecting the combined effects of ER proteostasis imbalance, autophagy regulation, and tumor-microenvironment interactions. This biologically focused framework may help improve the interpretability of our findings and suggests that ER stress-related transcriptional programs could represent one dimension of prognostic heterogeneity in NB.

However, there are still some limitations in this study. The data in this study were obtained from public databases, and no clinical specimens were collected for experimental verification. Furthermore, in our study, due to the inclusion of multiple data sets, and fail to fully remove the batch effect in such cases. Despite these shortcomings, our results also provided valuable information about prognostic predictions for patients with NB.

Conclusions

In conclusion, the results of this study suggest that in NB, ER stress-related genes are mainly involved in the regulation of cancer cell autophagy, metastasis, and drug resistance. Among the ER stress-related genes, FN1 and MAPK8 are the hub genes with independent contribution to NB prognosis. We developed a reliable prognostic model based on 10 ER stress-related DEGs, MAP2, PRKCD, MAPK8IP1, JPH1, TPP1, BCL2, EDN1, THBS1, MAPK8, and SERPINA3, among which MAPK8 and THBS1 were the hub genes. The findings suggest that ER stress-related gene expression patterns are closely linked to prognostic heterogeneity in NB. While further experimental validation and clinical studies are required, MAPK8, FN1, and THBS1 may represent candidate prognostic biomarkers and provide a rationale for future investigations into ER stress-related therapeutic strategies in NB.

Acknowledgments

We would like to thank the TARGET and GEO databases for their contributions.

Footnote

Reporting Checklist: The authors have completed the TRIPOD reporting checklist. Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-1-2623/rc

Peer Review File: Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-1-2623/prf

Funding: This study was supported by the National Key R&D Program of China (Nos. 2021YFC2701704 and 2021YFC2701700), the National Natural Science Foundation of China (No. 82271749), and the Wujieping Medical Foundation (No. 320.6750.2025-23-10).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-1-2623/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.

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. Matthay KK, Maris JM, Schleiermacher G, et al. Neuroblastoma. Nat Rev Dis Primers 2016;2:16078. [Crossref] [PubMed]
2. Zafar A, Wang W, Liu G, et al. Molecular targeting therapies for neuroblastoma: Progress and challenges. Med Res Rev 2021;41:961-1021. [Crossref] [PubMed]
3. Sharp SE, Gelfand MJ, Shulkin BL. Pediatrics: diagnosis of neuroblastoma. Semin Nucl Med 2011;41:345-53. [Crossref] [PubMed]
4. Perez CA, Matthay KK, Atkinson JB, et al. Biologic variables in the outcome of stages I and II neuroblastoma treated with surgery as primary therapy: a children's cancer group study. J Clin Oncol 2000;18:18-26. [Crossref] [PubMed]
5. Hero B, Simon T, Spitz R, et al. Localized infant neuroblastomas often show spontaneous regression: results of the prospective trials NB95-S and NB97. J Clin Oncol 2008;26:1504-10. [Crossref] [PubMed]
6. Matthay KK, Villablanca JG, Seeger RC, et al. Treatment of high-risk neuroblastoma with intensive chemotherapy, radiotherapy, autologous bone marrow transplantation, and 13-cis-retinoic acid. Children's Cancer Group. N Engl J Med 1999;341:1165-73.
7. Tabas I, Ron D. Integrating the mechanisms of apoptosis induced by endoplasmic reticulum stress. Nat Cell Biol 2011;13:184-90. [Crossref] [PubMed]
8. Lee AS. GRP78 induction in cancer: therapeutic and prognostic implications. Cancer Res 2007;67:3496-9. [Crossref] [PubMed]
9. Zhang Q, Guan G, Cheng P, et al. Characterization of an endoplasmic reticulum stress-related signature to evaluate immune features and predict prognosis in glioma. J Cell Mol Med 2021;25:3870-84. [Crossref] [PubMed]
10. Urra H, Dufey E, Avril T, et al. Endoplasmic Reticulum Stress and the Hallmarks of Cancer. Trends Cancer 2016;2:252-62. [Crossref] [PubMed]
11. Chen X, Cubillos-Ruiz JR. Endoplasmic reticulum stress signals in the tumour and its microenvironment. Nat Rev Cancer 2021;21:71-88. [Crossref] [PubMed]
12. París-Coderch L, Soriano A, Jiménez C, et al. The antitumour drug ABTL0812 impairs neuroblastoma growth through endoplasmic reticulum stress-mediated autophagy and apoptosis. Cell Death Dis 2020;11:773. [Crossref] [PubMed]
13. Ikegaki N, Hicks SL, Regan PL, et al. S(+)-ibuprofen destabilizes MYC/MYCN and AKT, increases p53 expression, and induces unfolded protein response and favorable phenotype in neuroblastoma cell lines. Int J Oncol 2014;44:35-43. [Crossref] [PubMed]
14. Hart LS, Rader J, Raman P, et al. Preclinical Therapeutic Synergy of MEK1/2 and CDK4/6 Inhibition in Neuroblastoma. Clin Cancer Res 2017;23:1785-96. [Crossref] [PubMed]
15. Gu L, Chu P, Lingeman R, et al. The Mechanism by Which MYCN Amplification Confers an Enhanced Sensitivity to a PCNA-Derived Cell Permeable Peptide in Neuroblastoma Cells. EBioMedicine 2015;2:1923-31. [Crossref] [PubMed]
16. Ritchie ME, Phipson B, Wu D, et al. limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res 2015;43:e47. [Crossref] [PubMed]
17. Gautier L, Cope L, Bolstad BM, et al. affy--analysis of Affymetrix GeneChip data at the probe level. Bioinformatics 2004;20:307-15. [Crossref] [PubMed]
18. Goeman JJ. L1 penalized estimation in the Cox proportional hazards model. Biom J 2010;52:70-84. [Crossref] [PubMed]
19. Friedman J, Hastie T, Tibshirani R. Regularization Paths for Generalized Linear Models via Coordinate Descent. J Stat Softw 2010;33:1-22.
20. Ackermann S, Goeser F, Schulte JH, et al. Polo-like kinase 1 is a therapeutic target in high-risk neuroblastoma. Clin Cancer Res 2011;17:731-41. [Crossref] [PubMed]
21. Lin YH, Jiang JH, Chuang HC, et al. The Involvement of CSRP1 in Neuroblastoma Differentiation and Apoptosis Impacting Tumor-Suppressive Therapeutic Responses. FASEB J 2025;39:e70521. [Crossref] [PubMed]
22. Su M, Ma Y, Liu X, et al. Valproate Releases the Latent Differentiation Capacity of Neuroblastoma by Targeting SP1/CDK4 Signaling. Neurochem Res 2026;51:28. [Crossref] [PubMed]
23. Park JA, Cheung NV. Targets and Antibody Formats for Immunotherapy of Neuroblastoma. J Clin Oncol 2020;38:1836-48. [Crossref] [PubMed]
24. Cirri P, Chiarugi P. Cancer associated fibroblasts: the dark side of the coin. Am J Cancer Res 2011;1:482-97.
25. Rybinski B, Franco-Barraza J, Cukierman E. The wound healing, chronic fibrosis, and cancer progression triad. Physiol Genomics 2014;46:223-44. [Crossref] [PubMed]
26. Tiwari N, Gheldof A, Tatari M, et al. EMT as the ultimate survival mechanism of cancer cells. Semin Cancer Biol 2012;22:194-207. [Crossref] [PubMed]
27. Li S, Chen Y, Zhang Y, et al. Shear stress promotes anoikis resistance of cancer cells via caveolin-1-dependent extrinsic and intrinsic apoptotic pathways. J Cell Physiol 2019;234:3730-43. [Crossref] [PubMed]
28. Pfeffer CM, Singh ATK. Apoptosis: A Target for Anticancer Therapy. Int J Mol Sci 2018;19:448. [Crossref] [PubMed]
29. Williams MM, Lee L, Werfel T, et al. Intrinsic apoptotic pathway activation increases response to anti-estrogens in luminal breast cancers. Cell Death Dis 2018;9:21. [Crossref] [PubMed]
30. Rodrigo MAM, Buchtelova H, Jimenez AMJ, et al. Transcriptomic Landscape of Cisplatin-Resistant Neuroblastoma Cells. Cells 2019;8:235. [Crossref] [PubMed]
31. Murai T. The role of lipid rafts in cancer cell adhesion and migration. Int J Cell Biol 2012;2012:763283. [Crossref] [PubMed]
32. Luo X, Cheng C, Tan Z, et al. Emerging roles of lipid metabolism in cancer metastasis. Mol Cancer 2017;16:76. [Crossref] [PubMed]
33. Villa-Morales M, Cobos MA, González-Gugel E, et al. FAS system deregulation in T-cell lymphoblastic lymphoma. Cell Death Dis 2014;5:e1110. [Crossref] [PubMed]
34. Scornik OA, Botbol V. Effects on Mammals of the Aminopeptidase Inhibitor Bestatin. In: Hooper NM, Lendeckel U, editors. Aminopeptidases in Biology and Disease. Boston: Springer US; 2004:271-307.
35. Liu R, Chen Y, Liu G, et al. PI3K/AKT pathway as a key link modulates the multidrug resistance of cancers. Cell Death Dis 2020;11:797. [Crossref] [PubMed]
36. Giussani P, Tringali C, Riboni L, et al. Sphingolipids: key regulators of apoptosis and pivotal players in cancer drug resistance. Int J Mol Sci 2014;15:4356-92. [Crossref] [PubMed]
37. Cai X, Liu C, Zhang TN, et al. Down-regulation of FN1 inhibits colorectal carcinogenesis by suppressing proliferation, migration, and invasion. J Cell Biochem 2018;119:4717-28. [Crossref] [PubMed]
38. Liu X, Meng L, Li X, et al. Regulation of FN1 degradation by the p62/SQSTM1-dependent autophagy-lysosome pathway in HNSCC. Int J Oral Sci 2020;12:34. [Crossref] [PubMed]
39. Öztaş E, Kara M, Boran T, et al. Cellular Stress Pathways Are Linked to Acetamiprid-Induced Apoptosis in SH-SY5Y Neural Cells. Biology (Basel) 2021;10:820. [Crossref] [PubMed]
40. Chen J, Rao B, Huang Z, et al. Circular RNA hsa_circ_0050386 suppresses non-small cell lung cancer progression via regulating the SRSF3/FN1 axis. J Transl Med 2024;22:47. [Crossref] [PubMed]
41. Eleveld TF, Schild L, Koster J, et al. RAS-MAPK Pathway-Driven Tumor Progression Is Associated with Loss of CIC and Other Genomic Aberrations in Neuroblastoma. Cancer Res 2018;78:6297-307. [Crossref] [PubMed]
42. He W, Wang Q, Xu J, et al. Attenuation of TNFSF10/TRAIL-induced apoptosis by an autophagic survival pathway involving TRAF2- and RIPK1/RIP1-mediated MAPK8/JNK activation. Autophagy 2012;8:1811-21. [Crossref] [PubMed]
43. Zhang W, Sun J, Luo J. High Expression of Rab-like 3 (Rabl3) is Associated with Poor Survival of Patients with Non-Small Cell Lung Cancer via Repression of MAPK8/9/10-Mediated Autophagy. Med Sci Monit 2016;22:1582-8. [Crossref] [PubMed]
44. Nimbalkar VP, Kruthika BS, Sravya P, et al. Differential gene expression in peritumoral brain zone of glioblastoma: role of SERPINA3 in promoting invasion, stemness and radioresistance of glioma cells and association with poor patient prognosis and recurrence. J Neurooncol 2021;152:55-65. [Crossref] [PubMed]
45. Bisceglia M, Galliani C, Giannatempo G, et al. Solitary fibrous tumor of the central nervous system: a 15-year literature survey of 220 cases (August 1996-July 2011). Adv Anat Pathol 2011;18:356-92. [Crossref] [PubMed]
46. Yi R, Feng J, Yang S, et al. miR-484/MAP2/c-Myc-positive regulatory loop in glioma promotes tumor-initiating properties through ERK1/2 signaling. J Mol Histol 2018;49:209-18. [Crossref] [PubMed]
47. Xu P, Zhang G, Hou S, et al. MAPK8 mediates resistance to temozolomide and apoptosis of glioblastoma cells through MAPK signaling pathway. Biomed Pharmacother 2018;106:1419-27. [Crossref] [PubMed]
48. Abreu E, Aritonovska E, Reichenbach P, et al. TIN2-tethered TPP1 recruits human telomerase to telomeres in vivo. Mol Cell Biol 2010;30:2971-82. [Crossref] [PubMed]
49. Zhu J, Liu W, Chen C, et al. TPP1 OB-fold domain protein suppresses cell proliferation and induces cell apoptosis by inhibiting telomerase recruitment to telomeres in human lung cancer cells. J Cancer Res Clin Oncol 2019;145:1509-19. [Crossref] [PubMed]
50. Pal SK, Nguyen CT, Morita KI, et al. THBS1 is induced by TGFB1 in the cancer stroma and promotes invasion of oral squamous cell carcinoma. J Oral Pathol Med 2016;45:730-9. [Crossref] [PubMed]
51. Zhou ZQ, Cao WH, Xie JJ, et al. Expression and prognostic significance of THBS1, Cyr61 and CTGF in esophageal squamous cell carcinoma. BMC Cancer 2009;9:291. [Crossref] [PubMed]
52. Liu X, Xu Z, Yin H, et al. Immune-oncology targets and therapeutic response of cell pyroptosis-related genes with prognostic implications in neuroblastoma. Discov Oncol 2024;15:661. [Crossref] [PubMed]
53. Pan LJ, Chen JL, Wu ZX, et al. Exportin-T: A Novel Prognostic Predictor and Potential Therapeutic Target for Neuroblastoma. Technol Cancer Res Treat 2021;20:15330338211039132. [Crossref] [PubMed]
54. Wang Z, Cheng H, Xu H, et al. A five-gene signature derived from m6A regulators to improve prognosis prediction of neuroblastoma. Cancer Biomark 2020;28:275-84. [Crossref] [PubMed]