Bulk and single-cell transcriptomics of stomach adenocarcinoma provide prognostic insights via the application of a novel gene signature associated with manganese metabolism

Introduction

Gastric cancer (GC) is a malignancy originating from the epithelial cells of the stomach mucosa, and represents a substantial health challenge worldwide (1,2). It is the fifth most prevalent cancer and the fourth leading cause of cancer-related deaths worldwide (3). The etiology of GC is complex, and it demonstrates significant heterogeneity (4). Among the various subtypes of GC, stomach adenocarcinoma (STAD) is the most frequently encountered subtype in clinical settings, constituting approximately 95% of all GC cases (1-4). Despite the adoption of aggressive multimodal therapeutic strategies, patients with advanced GC often face a poor prognosis, with more than half experiencing local recurrence and distant metastases (5). This challenging scenario underscores the limitations of current clinical strategies in managing complex diseases with diverse molecular foundations, such as GC. Thus, the precise subtyping and comprehensive molecular characterization of GC heterogeneity are essential for achieving targeted therapeutic outcomes.

Recent advancements in tumor molecular biology have enabled the development of several molecular classifications for GC. The Cancer Genome Atlas (TCGA) Research Network has identified four distinct molecular subtypes of GC: Epstein-Barr virus positive tumors, microsatellite instability (MSI) tumors, genomically stable tumors, and chromosomal instability tumors (6). Cristescu et al. further refined these classifications into the following four subtypes based on microsatellite stability, epithelial-mesenchymal transition (EMT), and TP53 mutations: MSI, MSS/EMT, MSS/TP53⁺, and MSS/TP53^– (7). Additionally, Wu et al. categorized GC into two subtypes of Cluster 1 (C1) and Cluster 2 (C2) based on immune infiltration profiles (8). These classifications are pivotal for patient stratification, and for advancing both clinical and preclinical research.

Manganese (Mn) is an essential micronutrient that significantly contributes to human physiological functions by participating in various biological processes (9). Recent study has shown that Mn is critical for both tumor development and the modulation of the immune response (10). The liver predominantly filters Mn ions, with minimal excretion, through a metabolic pathway involving multiple genes that are closely associated with oncogenesis and tumor progression (11). Research has identified manganese metabolism-related genes (MMRGs) as potential predictive markers of the survival and treatment responses of patients with GC (12). Among these, Mn has garnered attention for its crucial role as a co-factor for multiple enzymes, including manganese superoxide dismutase (MnSOD), which is pivotal in managing oxidative stress and supporting tumor cell survival. Mn also influences immune responses, angiogenesis, and cellular signaling pathways. Dysregulated Mn metabolism has been found to be associated with neurotoxicity and certain cancers, yet its role in stomach adenocarcinoma remains poorly explored.

As a vital trace element, Mn is integral to numerous metabolic pathways and the structural configuration of various enzymes, and thus plays a crucial role in renal function. Mn-SOD is responsible for neutralizing reactive oxygen species (ROS), thereby protecting mitochondria from oxidative stress and maintaining renal health (13). Additionally, Mn can induce oxidative stress and increase ROS levels, leading to cellular damage and promoting apoptosis in cancer cells (14). Research has shown that Mn may adversely affect mitochondrial function, thereby affecting energy metabolism and potentially contributing to the progression of renal cancer (15). Nonetheless, the association between Mn metabolism and GC, specifically STAD, is not yet well understood. Consequently, it is both imperative and timely to explore the role of Mn metabolism in the context of STAD.

In this study, we analyzed the MMRG data and clinical information of GC patients from TCGA database. We performed a clustering analysis using the MMRGs, and identified two subtypes, C1 and C2. We then compared their clinical characteristics, immunotherapy responses, and molecular features. STEP regression analyses revealed the key genes linked to nicotine metabolism and GC prognosis, providing new insights into the molecular classification of GC. We present this article in accordance with the TRIPOD reporting checklist (available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-1826/rc).

Methods

Data collection and preprocessing

STAD is the most prevalent pathological variant of GC, accounting for nearly 95% of all clinically documented GC cases. Thus, this study focused exclusively on STAD as its main topic of investigation. We obtained the data from the TCGA database (https://portal.gdc.cancer.gov). After downloading the STAR counts data and the corresponding clinical information of STAD tumor, transcripts per million (TPM) format data were extracted and normalized using log2 (TPM+1). Finally, 370 samples with both RNA-sequencing (RNAseq) data and clinical information were retained for subsequent analysis. Moreover, a total of 1,917 MMRGs were downloaded from the GeneCards database (https://www.genecards.org/) using the key word “manganese metabolism”. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.

Consistent clustering analysis based on MMRGs in STAD

Using the expression profiles of the MMRGs, the STAD samples were systematically clustered using a robust clustering algorithm. We performed consensus analysis using the R package ConsensusClusterPlus (v1.54.0), with the maximum number of clusters set to 6, repeating 100 times by extracting 80% of the total samples, and setting clusterAlg = “hc” and innerLinkage = “ward.D2”. The optimal number of clusters was identified by analyzing the cumulative distribution function curve of the consistency score, alongside the features depicted in the heat map of the consistency matrix. The ideal number of clusters, denoted as “k”, was established at the point where the consistency score reached its peak, corresponding to a significant difference in overall survival (OS) among the delineated subtypes.

Identifying the DEGs of the two STAD subtypes

The R package “DESeq2” was used to identify the differentially expressed genes (DEGs) across the various clusters of STAD. Using the “DESeq2” package in the R environment, we successfully identified the DEGs that differentiated the STAD clusters. A statistically significant distinction was established using a threshold of P<0.05 and |log2 fold change| >1. To illustrate the DEGs in STAD, a volcano plot was generated using the “ggplot2” package in R. Additionally, a heatmap representing the DEGs was created using the “pheatmap” package (version 1.0.12) in R. Furthermore, Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses were conducted using the “clusterProfiler” package in R.

Detection of immune activity in the two MMRG-related clusters of STAD

A comparative investigation was conducted to analyze the immune activity between the two clusters associated with the MMRGs by measuring the expression levels of eight immune checkpoint genes (i.e., CD274, PDCD1, PDCD1LG2, CTLA4, LAG3, HAVCR2, TIGIT, and SIGLEC15). To illustrate the findings, heatmaps and box plots were created independently using the R packages “pheatmap” and “ggplot2”. A P value lower than 0.05 was considered statistically significant.

Development of a prognostic model based on MMRGs

The “glmnet” package in R was employed to perform the Cox regression analysis to evaluate the prognostic significance of the MMRGs in TCGA-STAD cohort. Initially, a multifactorial Cox regression model was implemented for the data analysis, which was followed by an iterative refinement process using the “step” function. Ultimately, the most appropriate model was identified as the final representation. The minimum criteria were applied to determine the conditions for variables with non-zero coefficients. A risk score was calculated by summing the expression levels of each gene, weighted according to their corresponding coefficients. Using the median risk scores, the patients in TCGA-STAD cohort were categorized into low- and high-risk groups. In addition, the “GGalluvial” package in R was used to investigate the associations among different types, OS status, and risk scores.

Single-cell sequencing analysis of STAD

We used the Tumor Immune Single-cell Hub 2 (TISCH2) platform (http://tisch.comp-genomics.org/) to perform an analysis at the single-cell level of the genes linked to the MMRGs. The data underwent batch effect and standardization processing. The cell types were classified using the GSE134520 dataset from the Gene Expression Omnibus (GEO) database. The differences in gene expression among the various clusters were depicted using a heatmap.

Statistical analysis

The statistical analyses were performed using R software (version 4.4.0) (https://www.r-project.org/) and Statistical Product and Service Solutions (SPSS; version 19.0). Statistical significance was set at a threshold of adjusted P value for all the tests.

Results

Consensus clustering analysis of the MMRGs in STAD

To identify the prognostic biomarkers linked to STAD, a univariate Cox regression analysis was initially conducted. This analysis successfully identified 829 genes that exhibited significant associations with patient survival in STAD. The 20 most prominent genes are illustrated in Figure 1A. Using a Venn diagram, 115 prognostic MMRGs were ultimately identified (Figure 1B).

Figure 1 Consensus clustering analysis of MMRGs in STAD. (A) The prognostic genes in STAD were identified by univariate Cox regression analysis. The numerical value outside the parentheses indicates the point estimate of the HR, while the value within the parentheses denotes the 95% CI. (B) The identification of the prognostic MMRGs was achieved via the application of a Venn diagram. In total, 829 prognostic genes were identified by the univariate Cox analysis, and 1,917 were documented in this investigation. Further, 115 pivotal MMRGs were identified. (C) A principal component analysis was conducted to distinguish between the two subtypes of STAD. (D) The consensus clustering analysis uncovered two discrete gene subtypes (k=2), along with their corresponding correlation regions. (E) A heatmap was generated to depict the distribution of the two gene subtypes in STAD. (F) A survival analysis was conducted to evaluate the prognostic significance of the two identified subtypes in STAD. C1, Cluster 1; C2, Cluster 2; CI, confidence interval; HR, hazard ratio; MMRGs, manganese metabolism-related genes; PC, principal component; STAD, stomach adenocarcinoma.

The expression of the 115 MMRGs in the STAD cohort was evaluated using the R package ConsensusClusterPlus. The results indicated that the highest consistency was achieved when the clustering parameter “k” was set to 2 (Figure 1C-1E). Following this, survival differences among the identified subtypes were assessed by Kaplan-Meier analysis, which explored the prognostic variations between the molecular subtypes. The results revealed that the C1 subtype had a significantly better survival probability than the C2 subtype (Figure 1F).

Variations in clinical characteristics between the two STAD subtypes

We examined the expression patterns of the key MMRGs in relation to various clinical characteristics, including age (Figure 2A), gender (Figure 2B), Post-Treatment Nodal Margin (pTNM)_stage (Figure 2C), tumor grade (Figure 2D), and radiation therapy (Figure 2E). A significant difference was found between the two subtypes in terms of age (Figure 2A) and tumor grade (Figure 2D). Conversely, no significant difference was found between the two subtypes in terms of gender (Figure 2B), pTNM_stage (Figure 2C), and radiation therapy (Figure 2E).

Figure 2 Immunotherapy response between the two STAD subtypes. (A) A significant difference was found between the C1 and C2 groups in terms of age. (B) A significant difference was found between the C1 and C2 groups in terms of gender. (C) A significant difference was found between the C1 and C2 groups in terms of pTNM_stage classification. (D) A significant difference was found between the C1 and C2 groups in terms of grade classification. (E) A significant difference was found between the C1 and C2 groups in terms of radiation therapy. C1, Cluster 1; C2, Cluster 2; CI, confidence interval; pTNM, pathological tumor-node-metastasis; STAD, stomach adenocarcinoma.

Immunotherapy response of the two STAD subtypes

We evaluated the immune response between the two distinct patient clusters associated with Mn metabolism in STAD. The box plot analysis revealed a significant difference in the populations of immune cells, including a reduced presence of B cells, CD4⁺ T cells, endothelial cells, and macrophages in the C1 samples compared with the C2 samples (Figure 3A). Recently, the efficacy of immunotherapy in combating various solid tumors has been established (16). Given the significance of immune checkpoints in the context of immunotherapy, we examined the relationship between the MMRGs and immune checkpoint genes. A majority of immune checkpoints, particularly CTLA4, HAVCR2, IGSF8, ITPRIPL1, LAG3, PDCD1, PDCD1LG2, and TIGIT, displayed elevated expression levels in the C2 group relative to the C1 group (Figure 3B).

Figure 3 Evaluation of the molecular characteristics of the STAD subtypes based on the MMRGs. (A) A comparative analysis of six distinct immune cell types was conducted between the two groups associated with the MMRGs in STAD. (B) The expression profiles of the genes that encode the immune checkpoint inhibitors in STAD were systematically evaluated and compared between the two identified clusters: C1 and C2. *, P<0.05; **, P<0.01; ****, P<0.0001; ns, non-significant. C1, Cluster 1; C2, Cluster 2; EPIC, estimating the proportions of immune and cancer cells; MMRGs, manganese metabolism-related genes; NK, natural killer; STAD, stomach adenocarcinoma.

Evaluation of the molecular characteristics of STAD subtypes based on MMRGs

To extend our understanding of the distinct molecular regulatory mechanisms that characterize the various subtypes, we conducted a comparative analysis of the gene expression levels, focusing on C1 and C2. The distribution of the DEGs in these two categories is shown in Figure 3A. Our analysis revealed that 13 genes were upregulated, and 1,146 genes were downregulated (Figure 4A). The heatmap in Figure 4B (in which red signifies heightened expression, and blue indicates diminished expression) further elucidated the expression patterns of the DEGs across different subgroups.

Figure 4 Variations in the clinical characteristics between the two STAD subtypes. (A) Volcano plot showing that 13 DEGs exhibited an increase in expression, while 1,146 DEGs exhibited a decrease (P value <0.05, and a |log2 fold change| >1). (B) Heatmap showing the expression profiles of the 1,159 DEGs between C1 and C2. (C) KEGG pathways significantly enriched for the DEGs. (D) Biological processes enriched in relation to the DEGs. Akt, serine/threonine kinase B; C1, Cluster 1; C2, Cluster 2; DEGs, differentially expressed genes; ECM, extracellular matrix; GO, Gene Ontology; PI3K, phosphatidylinositol 3-kinase; KEGG, Kyoto Encyclopedia of Genes and Genomes; MMRGs, manganese metabolism-related genes; STAD, stomach adenocarcinoma.

After identifying 1,159 DEGs, KEGG and GO enrichment analyses were performed to examine the biological functions associated with these genes. The KEGG analysis revealed that the following functions were associated with these genes: ECM-receptor interaction, focal adhesion, hypertrophic cardiomyopathy, PI3K-Akt signaling pathway, dilated cardiomyopathy, protein digestion and absorption, vascular smooth muscle contraction, malaria, and proteoglycans in cancer (Figure 4C). The results of the GO analysis showed that the following biological processes were associated with the genes: extracellular matrix (ECM) organization, extracellular structure organization, cell-substrate adhesion, muscle system process, muscle contraction, regulation of cellular response to growth factor stimulus, and cell-matrix adhesion (Figure 4D).

Prognostic signature development of two MMRG-related groups in STAD

We then conducted an extensive evaluation of the prognostic relevance of the MMRGs. Each individual diagnosed with STAD was allocated a MMRG score, and subsequently categorized into high- and low-MMRG groups based on the determined optimal cut-off value (Figure 5A). This thorough analysis identified seven prognostic markers: APOD, AKR1B1, CGB5, GAMT, VTN, SERPINE1, and GPX3 (Figures 5A). Patients with a low MMRG score exhibited significantly improved OS compared to those with a high MMRG score in TCGA-STAD dataset (Figure 5B). Then, the effectiveness of the prognostic model was evaluated by a receiver operating characteristic (ROC) curve analysis, which revealed area under the curve (AUC) values of 0.633 for 1 year, 0.716 for 3 years, and 0.724 for 5 years (Figure 5C).

Figure 5 Prognostic signature development of the two MMRG-related groups in STAD. (A) An assessment of the prognostic gene signature was carried out using The Cancer Genome Atlas cohort. Corresponding heatmap showing the expression profiles of the prognostic genes, which were classified into low- and high-risk categories. (B) A KM survival analysis was conducted to evaluate the prognostic signature. (C) A time-dependent receiver operating characteristic analysis of the gene signature was performed. AUC, area under the curve; CI, confidence interval; HR, hazard ratio; KM, Kaplan-Meier; MMRGs, manganese metabolism-related genes; STAD, stomach adenocarcinoma.

Cellular classification and expression patterns of MMRGs in GC

An unsupervised hierarchical clustering analysis was conducted using the single-cell RNA-sequencing GSE134520 dataset. Nine distinct cell subtypes were identified in STAD (Figure 6A). The expression levels of two specific genes, APOD and GPX3, were found to be significantly elevated in both the fibroblast and myofibroblast subtypes of STAD (Figure 6B,6C). The box plots showed that AKR1B1 and GAMT were expressed across all the nine identified cell subtypes (Figure 6D,6E), while VTN and SERPINE1 were lowly expressed across the same nine subtypes (Figure 6F,6G).

Figure 6 Cellular classification and expression patterns of the MMRGs in STAD. (A) Cell types and distributions in STAD. (B) The expression levels of APOD were analyzed in nine different cell subtypes in STAD. (C) The expression levels of GPX3 were analyzed across the various cell subtypes in STAD. (D) The expression levels of AKR1B1 were analyzed across the different cell subtypes in STAD. (E) The expression levels of GAMT were analyzed across the different cell subtypes in STAD. (F) The expression levels of VTN were analyzed across the different cell subtypes in STAD. (G) The expression levels of SERPINE1 were analyzed across the different cell subtypes in STAD. DC, dendritic cell; MMRGs, manganese metabolism-related genes; STAD, stomach adenocarcinoma.

Discussion

Mn is a vital micronutrient involved in key human physiological processes and immune response regulation. It is primarily filtered by the liver, with limited excretion through a pathway linked to cancer-related genes. MMRGs could serve as potential prognostic markers for GC outcomes. As an essential trace element, Mn supports metabolic pathways and enzyme structure, crucial for kidney function. Mn-SOD protects mitochondria by neutralizing ROS, but Mn can also increase ROS, potentially causing cellular damage. Recent research has highlighted the potential role of Mn metabolism in modulating the immune microenvironment of GC, offering new avenues for therapeutic intervention (17). The study provides a comprehensive analysis of Mn metabolism- and immune-related genes (MIRGs) in GC, identifying a set of genes that can predict patient prognosis and inform treatment strategies (17). This research underscores the significance of Mn in enhancing the anti-tumor immune response, particularly when combined with immune checkpoint inhibitors, suggesting a synergistic potential that could improve treatment outcomes for GC patients.

In this study, a comprehensive clustering analysis of GC was conducted using 115 prognostic MMRGs, which enabled the identification of two distinct subtypes, labeled C1 and C2. A Kaplan-Meier analysis subsequently revealed that the C1 subtype had a higher survival probability than the C2 subtype. Furthermore, a significant difference was found between the two subtypes in terms of age and tumor grade. The box plot analysis revealed a significant difference in the populations of immune cells, indicating a reduced presence of B cells, CD4⁺ T cells, endothelial cells, and macrophages in the C1 samples compared with the C2 samples. A majority of immune checkpoints, particularly CTLA4, HAVCR2, IGSF8, ITPRIPL1, LAG3, PDCD1, PDCD1LG2, and TIGIT, displayed elevated expression levels in the C2 group relative to the C1 group. Our analysis revealed that 13 genes were upregulated, and 1,146 genes were downregulated in the C1 group. The KEGG analysis revealed that the relevant functions included the ECM-receptor interaction, focal adhesion, PI3K-Akt signaling pathway, and proteoglycans in cancer. The GO analysis further revealed that the associated biological processes included ECM organization, cell-substrate adhesion, regulation of cellular response to growth factor stimulus, and cell-matrix adhesion. We then conducted an assessment of the prognostic relevance of the MMRGs in patients diagnosed with STAD. Each patient was assigned a score based on their MMRGs, allowing for their stratification into high- and low-MMRG groups. This analysis identified eight prognostic markers: APOD, AKR1B1, CGB5, GAMT, VTN, SERPINE1, and GPX3. The expression levels of two specific genes, APOD and GPX3, were found to be significantly elevated in both the fibroblast and myofibroblast subtypes of STAD. The box plots showed that AKR1B1 and GAMT were expressed across all the nine identified cell subtypes, while VTN and SERPINE1 were lowly expressed across the same nine subtypes.

AKR1B1 plays a crucial role in GC by promoting cancer progression via various mechanisms. It inhibits ferroptosis by interacting with STAT3 to activate SLC7A11, aiding in GC cell proliferation and survival (18). AKR1B1 also affects immune regulation in the tumor microenvironment by influencing macrophage polarization and immune infiltration, and has been linked to poor outcomes (19). The ZNF521/EBF1 axis regulates AKR1B1, highlighting its role in GC cell proliferation and migration (20). These insights suggest that targeting AKR1B1 and its pathways could serve as a potential approach to GC therapy. SERPINE1 is significant in GC research due to its role in enhancing tumor growth, spread, and EMT, which are crucial for metastasis (21). It also influences immune modulation by promoting M2 macrophage polarization, leading to an immunosuppressive environment (22). High SERPINE1 expression has been shown to be correlated with a poor prognosis; thus, it could serve as a potential prognostic biomarker and therapeutic target in GC (23). GAMT is a promising early-stage GC diagnostic marker due to its altered expression in GC tissues and its role in cellular processes affecting cancer cell growth (24). Its involvement in disulfidptosis, a type of programmed cell death, offers new molecular insights and potential therapeutic targets for GC (25).

There are several limitations in this study, primarily due to its retrospective nature, necessitating validation through prospective studies. The use of historical data may have introduced biases affecting reproducibility. Despite challenges of relying solely on TCGA data, its comprehensive and high-quality nature offers a solid foundation for our analysis. Future research should include functional experiments with these genes to further understand their role in GC.

Conclusions

The roles of MMRGs as biomarkers and therapeutic targets offer opportunities for improving GC diagnosis and treatment. Future research should seek to further elucidate the molecular mechanisms involving these molecules and develop targeted therapies that can effectively modulate their activity in GC.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the TRIPOD reporting checklist. Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-1826/rc

Peer Review File: Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-1826/prf

Funding: None.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-1826/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. Ajani JA, D'Amico TA, Bentrem DJ, et al. Gastric Cancer, Version 2.2022, NCCN Clinical Practice Guidelines in Oncology. J Natl Compr Canc Netw 2022;20:167-92. [Crossref] [PubMed]
2. Wang FH, Zhang XT, Li YF, et al. The Chinese Society of Clinical Oncology (CSCO): Clinical guidelines for the diagnosis and treatment of gastric cancer, 2021. Cancer Commun (Lond) 2021;41:747-95. [Crossref] [PubMed]
3. Sung H, Ferlay J, Siegel RL, et al. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J Clin 2021;71:209-49. [Crossref] [PubMed]
4. Li L, Wang X. Identification of gastric cancer subtypes based on pathway clustering. NPJ Precis Oncol 2021;5:46. [Crossref] [PubMed]
5. Chen T, Xu XY, Zhou PH. Emerging molecular classifications and therapeutic implications for gastric cancer. Chin J Cancer 2016;35:49. [Crossref] [PubMed]
6. Comprehensive molecular characterization of gastric adenocarcinoma. Nature 2014;513:202-9. [Crossref] [PubMed]
7. Cristescu R, Lee J, Nebozhyn M, et al. Molecular analysis of gastric cancer identifies subtypes associated with distinct clinical outcomes. Nat Med 2015;21:449-56. [Crossref] [PubMed]
8. Wu D, Feng M, Shen H, et al. Prediction of Two Molecular Subtypes of Gastric Cancer Based on Immune Signature. Front Genet 2021;12:793494. [Crossref] [PubMed]
9. Aschner M, Erikson K. Manganese. Adv Nutr 2017;8:520-1. [Crossref] [PubMed]
10. Zhang K, Qi C, Cai K. Manganese-Based Tumor Immunotherapy. Adv Mater 2023;35:e2205409. [Crossref] [PubMed]
11. Rozenberg JM, Kamynina M, Sorokin M, et al. The Role of the Metabolism of Zinc and Manganese Ions in Human Cancerogenesis. Biomedicines 2022;10:1072. [Crossref] [PubMed]
12. Han X, Leng C, Zhao S, et al. Development and verification of a manganese metabolism- and immune-related genes signature for prediction of prognosis and immune landscape in gastric cancer. Front Immunol 2024;15:1377472. [Crossref] [PubMed]
13. Padoan F, Guarnaroli M, Brugnara M, et al. Role of Nutrients in Pediatric Non-Dialysis Chronic Kidney Disease: From Pathogenesis to Correct Supplementation. Biomedicines 2024;12:911. [Crossref] [PubMed]
14. Xie Y, Liu F, Zhang X, et al. Benefits and risks of essential trace elements in chronic kidney disease: a narrative review. Ann Transl Med 2022;10:1400. [Crossref] [PubMed]
15. Smith MR, Fernandes J, Go YM, et al. Redox dynamics of manganese as a mitochondrial life-death switch. Biochem Biophys Res Commun 2017;482:388-98. [Crossref] [PubMed]
16. Liu S, Cui M, Zang J, et al. SLC6A3 as a potential circulating biomarker for gastric cancer detection and progression monitoring. Pathol Res Pract 2021;221:153446. [Crossref] [PubMed]
17. Liu Y, Ye H, Zhang R, et al. A manganese metabolism-related gene signature stratifies prognosis and immunotherapy efficacy in kidney cancer. Discov Oncol 2025;16:1242. [Crossref] [PubMed]
18. Yang K, Zhang X, Long F, et al. AKR1B1 Inhibits Ferroptosis and Promotes Gastric Cancer Progression via Interacting With STAT3 to Activate SLC7A11. Cell Biol Int 2025;49:374-83. [Crossref] [PubMed]
19. Zhao Z, Hao Z, Zhang Z, et al. Bioinformatics Analysis Reveals the Vital Role of AKR1B1 in Immune Infiltration and Clinical Outcomes of Gastric Cancer. DNA Cell Biol 2023;42:372-89. [Crossref] [PubMed]
20. Cheng Y, Ni YJ, Tang LM. ZNF521/EBF1 axis regulates AKR1B1 to promote the proliferation, migration, and invasion of gastric cancer cells. Kaohsiung J Med Sci 2023;39:244-53. [Crossref] [PubMed]
21. Yang JD, Ma L, Zhu Z. SERPINE1 as a cancer-promoting gene in gastric adenocarcinoma: facilitates tumour cell proliferation, migration, and invasion by regulating EMT. J Chemother 2019;31:408-18. [Crossref] [PubMed]
22. Ye Z, Yi J, Jiang X, et al. Gastric cancer-derived exosomal let-7 g-5p mediated by SERPINE1 promotes macrophage M2 polarization and gastric cancer progression. J Exp Clin Cancer Res 2025;44:2. [Crossref] [PubMed]
23. Feng L, Li G, Li D, et al. Cuproptosis-related gene SERPINE1 is a prognostic biomarker and correlated with immune infiltrates in gastric cancer. J Cancer Res Clin Oncol 2023;149:10851-65. [Crossref] [PubMed]
24. Gong C, Zhou M, Hu Y, et al. Elastic net-based identification of GAMT as potential diagnostic marker for early-stage gastric cancer. Biochem Biophys Res Commun 2022;591:7-12. [Crossref] [PubMed]
25. Wu W, Tao G, Chen Z, et al. Identification of disulfidptosis-related subtypes in gastric cancer and GAMT is a key gene during disulfidptosis. Sci Rep 2025;15:111. [Crossref] [PubMed]