The mRNA MOXD1: Link to oxidative stress and prognostic significance in gastric cancer

Youming Xiao; Xiqing Zhu; Cong Wang; Hongyu Gao; Zenghui Hao; Haibin Song; Zhaozhu Li

doi:10.1515/med-2025-1271

Artikel Open Access

The mRNA MOXD1: Link to oxidative stress and prognostic significance in gastric cancer

Youming Xiao , Xiqing Zhu , Cong Wang , Hongyu Gao , Zenghui Hao , Haibin Song und Zhaozhu Li

Veröffentlicht/Copyright: 19. September 2025

Veröffentlicht von

Veröffentlichen auch Sie bei De Gruyter Brill

Manuskript einreichen Informationen für Autor*innen Erkunden Sie dieses Fachgebiet

Aus der Zeitschrift Open Medicine Band 20 Heft 1

Abstract

Purpose

Oxidative stress (OS) plays a key role in gastric cancer (GC). The purpose of this study was to investigate the role of the mRNA monooxygenase DBH-like 1 (MOXD1) in OS and evaluate its prognostic significance in GC.

Methods

An OS risk score was constructed by unsupervised clustering analysis, the log-rank test, and least absolute shrinkage and selection operator–Cox analysis of OS-related genes. The Pearson correlation between MOXD1 expression and the OS risk score was evaluated. Correlations between MOXD1 expression and clinicopathological features in the training cohort were compared. CIBERSORT, ssGSEA, and ESTIMATE were used to analyze the effects of MOXD1 on the immune microenvironment. Gene Ontology, Kyoto Encyclopedia of Genes and Genomes, and gene set enrichment analysis were used to elucidate the biological functions of the mRNAs. Immunohistochemistry for MOXD1 was performed on patient tissue microarray (TMA) samples. Cox regression, log-rank tests, and chi-square analyses were used to investigate the clinicopathological features of the TMAs and associated MOXD1 expression levels. A stable knockdown cell line was constructed in HGC-27 GC cells and investigated using cell counting kit-8 and Transwell assays.

Results

The OS risk score was an independent prognostic factor for GC in the training cohort and was successfully combined with age and pTNM stage to construct a nomogram. MOXD1 expression was positively correlated with the OS risk score and was highly expressed in patients with GC. MOXD1 expression and the metastatic lymph node ratio in TMAs were found to be independent prognostic risk factors for GC. MOXD1 knockdown inhibited the proliferation and invasion of HGC-27 cells.

Conclusion

The mRNA MOXD1 is a biomarker for both OS and GC. MOXD1 expression can be used to evaluate GC prognosis and guide treatment.

Keywords: gastric cancer; MOXD1; mRNA; oxidative stress; prognosis

1 Introduction

Gastric cancer (GC) is the fifth most common malignant tumor in the world and the fifth most common cause of cancer death [1]. Owing to the highly heterogeneous character of this malignant tumor, GC patients encounter great difficulties in obtaining a timely prognosis and receiving comprehensive treatment. With developments in sequencing technology, GC has been characterized in detail at the transcriptomic level, and the involvement of multiple biological pathways has been verified. In particular, oxidative stress (OS), lipid metabolism, and ferroptosis have been demonstrated to play important roles in GC pathobiology [2,3,4]. Moreover, new molecular targets for GC are constantly being identified, improving our understanding of GC etiology and pathogenesis and ultimately providing new guidance for treatment [5,6,7].

OS occurs when an imbalance between oxidants and antioxidants causes an increase in the level of reactive oxygen species (ROS). ROS are produced by mitochondria during high oxygen consumption, and they can destroy the structure of intracellular proteins and nucleic acids, thus disrupting cell homeostasis and potentially inducing tumorigenesis [8,9]. In thyroid tumors, Dogan et al. reported that OS was significantly greater at the tumor edge than at the tumor center (and also significantly greater than that in healthy thyroid tissue) [10], suggesting that tumorigenesis in thyroid cancer patients is related to an increase in OS. Moreover, OS is known to affect the proliferation, migration, and invasion of tumor cells, processes that are closely related to the occurrence and development of many malignant tumors, including GC, breast cancer, and bladder cancer [11,12,13]. On the basis of these considerations, Wang et al. developed a prognostic scoring system based on OS and ferroptosis that could better stratify patients with colorectal cancer [14]. OS is also known to influence the immune microenvironment of tumors. For example, high ROS levels are frequently accompanied by an increase in M2 macrophages and the inhibition of dendritic cells, thus promoting tumor progression [15,16]. Additionally, OS is associated with the proliferation of tumor-associated fibroblasts [17]. OS has also been shown to perturb numerous biological pathways, including the MAPK/JNK/ERK pathway and the transforming growth factor-β signaling pathway [17,18]. In conclusion, the role of OS in the occurrence and development of tumors is complex, and this role is worth exploring further and is likely to be richly targeted.

In Jee et al.’s animal experiments, MOXD1 was found to be highly correlated with monooxygenase and oxidoreductase activity. Further research is needed to determine whether MOXD1 regulates OS levels in the tumor microenvironment [19]. In the present study, the role of mRNA and protein levels of monooxygenase DBH-like 1 (MOXD1) in OS and GC was investigated. The gene for MOXD1 is located on chromosome 6. MOXD1 is a member of the copper monooxygenase family and is known to modify proteins with copper ion binding and oxidoreductase activities [20]. In chronic obstructive pulmonary disease and diabetic kidney disease, MOXD1 expression was found to be associated with disease occurrence, providing indirect evidence that MOXD1 plays a role in chronic inflammation [21,22]. MOXD1 is also a known senescence-related molecule, further confirming its relationship with immunity [23]. Additionally, MOXD1 is a potentially important prognostic biomarker in bladder cancer, high-grade serous ovarian cancer (HGSOC), and hepatocellular carcinoma [24,25,26]. Shi et al. reported that MOXD1 can bind to β3GnT2 and affect the protein glycosylation process. MOXD1 knockdown induced endoplasmic reticulum (ER) stress, triggering the ER‒mitochondrial apoptosis pathway and modulating the progression of glioblastoma (GBM) [27]. Although MOXD1 has been demonstrated to be a biomarker for early GC [28], its biological function and prognostic significance require further study.

2 Materials and methods

2.1 GC data

The GC data were obtained from 269 tumor patients who underwent radical gastrectomy at Harbin Medical University (HMU) Cancer Hospital. From this cohort (HMU-GC), tumor tissue samples, paratumorous normal tissue samples, and clinical data were collected and collated. The mRNA data from GSE15459 [29] supplementary (Table S2-4), GSE62254 [30] supplementary (Table S5), and The Cancer Genome Atlas (TCGA)-stomach adenocarcinoma (STAD) in Gene Expression Omnibus (GEO) and TCGA were additionally included in our analyses.

2.2 Data processing

First, the HMU-GC and TCGA-STAD data were converted into transcripts per kilobase million values. To increase the sample size, improve the power of statistical analysis, and improve the universality and reliability of research results, the ComBat algorithm [31] in the sva package was used to combine these two independent mRNA data into one training cohort (the HMU-TCGA cohort). This can eliminate systematic bias caused by different experimental conditions, sample processing or measuring equipment. Compared with other methods of integrating gene expression data, combat can provide robust batch effect correction in the case of small samples. The original CEL files for the GSE15459 and GSE62254 datasets were downloaded from GEO and independently merged into a validation cohort using the ComBat algorithm.

2.3 Structure of the OS risk score

Following the methods of Liu et al. [32], we first downloaded the “GOBP_RESPONSE_TO_OXIDATIVE_STRESS” gene set in MSigDB and intersected it with the training cohort to obtain 392 mRNAs (Additional file 1). Unsupervised clustering analysis using the ConsensusClusterPlus package in R programming language [33] showed that dividing the training queue into three groups has prognostic significance. OS-related genes (OSRGs) with prognostic significance were then analyzed in the C1 and C3 groups, and the tumor and paratumorous normal tissues were compared via Limma. Genes with |fold change| > 1.5 and a false discovery rate (FDR) < 0.05 were subsequently selected. Finally, a least absolute shrinkage and selection operator–Cox analysis based on the glmnet package was applied. The OS risk score was determined as follows:

risk score = ∑ i = 1 n ( β i × Exp i ) .

A heat-map was used to provide a graphic illustration of gene expression and prognosis in the OS risk score. A log-rank test was used for the Kaplan‒Meier (K‒M) survival curve. Univariate and multivariate analyses of OS risk score, age, sex, and pTNM stage were performed (based on Cox regression). Independent prognostic factors in the training cohort were combined to construct a nomogram using the rms and survival packages. The performance of the nomogram was evaluated with a time-dependent receiver operating characteristic (timeROC) curve generated using the timeROC package, a calibration curve generated using the rms and survival packages, and a decision curve analysis (DCA) curve generated using the rmda package. The ggplot2 package was used for visualization.

2.4 Immune microenvironment analysis

The numbers of infiltrating immune cells in each tumor sample were estimated using ssGSEA and CIBERSORT. The immune score, stromal score, ESTIMATE score, and tumor purity were calculated using ESTIMATE. Tumor immune dysfunction and exclusion (TIDE) scores were calculated for patients using the TIDE database (http://tide.dfci.harvard.edu/). This surrogate biomarker was applied to evaluate patient suitability for immune checkpoint therapy (patients with a high TIDE score are not suitable for immunotherapy) [34]. The expression of MOXD1 in different cell subsets of the immune microenvironment was explored using the GC public single-cell database (STAD-GSE134520 and STAD-GSE167297) on the website of the Tumor Immune Single-cell Hub (http://tisch.comp-genomics.org/).

2.5 Bioinformatics analyses

For the Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis, the clusterProfiler package was used. ID conversion was performed using the org.Hs.eg.db package. The clusterProfiler package was also used in the gene set enrichment analysis (GSEA) to explore potential biological pathways. The reference genome was the hallmark gene set, and the selection conditions were as follows: |normalized enrichment score| > 1; nominal P value < 0.05; and FDR Q value < 0.25. The protein‒protein interaction (PPI) network was generated using STRING (version 12.5) (https://cn.string-db.org/). Somatic mutation data from the TCGA-STAD dataset were analyzed by the GDCquery_Maf() function (pipelines = “mutect2”) in the biolinks package, and the maftools package was used for identification and visualization [35,36].

2.6 Drug analysis

On the basis of the Profiling Relative Inhibition Simultaneously in Mixtures public drug susceptibility database [37], the area under the receiver operating characteristic curve (AUC) of each sample was estimated by ridge regression, and the chemotherapy response was predicted using the pRRophetic package. The prediction accuracy was evaluated by 10× cross-validation using the training cohort. Patients with a low AUC value show greater sensitivity to treatment.

2.7 Immunohistochemistry

The tissue microarrays (TMAs) were dewaxed, dehydrated using a gradient series, and rinsed with 3× concentrated phosphate-buffered solution (PBS) (5 min each) (DW0300, Dowobio, Shanghai, China), after which the antigens were retrieved in sodium citrate buffer (pH = 6) for 3 min at 120°C (DW2215, Dowobio, Shanghai, China). Finally, the TMAs were rinsed with 3× concentrated PBS (5 min each time), incubated with 3% H₂O₂ for 30 min (MM0750-500ML, MKbio, Shanghai, China), and then processed by immunostaining. First, the sections were blocked in goat serum for 1 h (Boster, USA). Next, a diluted anti-MOXD1 antibody (Bioss, bs-17733R; 1:150) was added, and the TMAs were then incubated overnight at 4°C. After the sections were rinsed with 3× concentrated PBS (5 min each time), goat anti-rabbit IgG (BA1039, Boster, USA) was added dropwise, and the TMAs were incubated at 37°C for 40 min. The color reaction was then developed using diaminobenzidine (AR1000, Boster, USA) staining. Finally, the sections were counterstained with hematoxylin (for nuclear staining) (MM1010-500ML, MKbio, Shanghai, China) and then viewed under a microscope. All the samples were examined by two pathologists, who evaluated the degree of positive cell staining (using a dividing line of 50%).

Immunohistochemical staining (IHC) is a commonly used method to evaluate the expression level of human epidermal growth factor receptor 2 (HER2): IHC 0: no staining or ≤10% of invasive cancer cells showed incomplete and weak cell membrane staining; IHC 1+: >10% of invasive cancer cells showed incomplete and weak cell membrane staining; IHC 2+: >10% of invasive cancer cells showed weak to moderate intensity of complete cell membrane staining or <10% of invasive cancer cells showed strong and complete cell membrane staining; IHC 3+: >10% of invasive cancer cells showed strong, complete, and uniform cell membrane staining. HER2 threshold is defined as: IHC score of 0 or 1+ is negative, IHC score of 2+, and further fish test is usually required to evaluate HER2 status by detecting the amplification of the HER2 gene. her2/cep17 ratio less than 2.0 or HER2 copy number less than 4 is negative, and her2/cep17 ratio greater than 2.0 or HER2 copy number greater than 6 is positive.

2.8 Cell line and transfection

GC HGC-27 cells were obtained from the Procell Life Science & Technology Co., Ltd. (Wuhan, China). The cells were cultured in a humidified incubator at 37°C (with 5% CO₂) with RPMI-1640 supplemented with 20% fetal bovine serum and 1% penicillin/streptomycin solution (Procell Life Science & Technology Co., Ltd., Wuhan, China). To obtain a stable MOXD1-knockdown cell line, HGC-27 cells were infected with MOXD1-interfering and control viruses (OBiO Technology Corp., Ltd., Shanghai, China; https://www.obiosh.com/) at a concentration of 15 µg/mL supplementary (Table S1). The cells were incubated with the infection mixture for 24 h, after which the infection mixture was exchanged for fresh medium. To isolate stably transfected cells, the infected cells were screened on a gradient of puromycin (Dalian Bergolin Biotechnology Co., Ltd., Dalian, China), and this procedure was continued until no cell death occurred. The resulting stable cell lines were designated HGC-NC, HGC-MOXD1sh1, and HGC-MOXD1sh2.

2.9 RNA extraction and quantitative real-time PCR (qPCR)

TRIzol (Invitrogen, USA) was used to extract total RNA from the HGC-NC, HGC-MOXD1sh1, and HGC-MOXD1sh2 cell lines. The PrimeScript RT Reagent Kit (TaKaRa, China) was used for reverse transcription of each total RNA sample to obtain cDNA. qPCR was performed on a LightCyler 96 Roche system using SYBR PreMix Ex Tap II (TaKaRa, China) according to the manufacturer’s instructions. The results were analyzed using the 2^−ΔΔCt method, with GAPDH used as an internal reference. The primer sequences for both MOXD1 and GAPDH are included in Additional file 1.

2.10 Western blot analysis

HGC-NC, HGC-MOXD1sh1, and HGC-MOXD1sh2 cells were lysed on ice for 30 min in RIPA buffer (Beyotime Biotechnology, Shanghai, PR China) containing phosphatase and protein inhibitors. The total protein concentrations of the resulting samples were quantified via a biochemical acid protein detection kit (Beyotime Biotechnology, Shanghai, China). After the protein samples were boiled for 10 min, 20 µg of each protein sample was loaded onto a 12% gel for separation by sodium dodecyl sulfate‒polyacrylamide gel electrophoresis. Next, the resolved proteins were transferred to a 0.22 µm polyvinylidene fluoride (PVDF) membrane (Millipore, USA). After transfer, the PVDF membrane was blocked with 5% skim milk powder for 2 h and then incubated overnight with primary antibody (at 4°C). The primary antibodies used were anti-MOXD1 (Bioss, bs-17733R; 1:1,000) and anti-GAPDH (Abcam, ab8245; 1:5,000). The membrane was then washed and incubated with a diluted anti-rabbit IgG (H + L) secondary antibody (ProteinTech, SA00001-2; 1:5,000) for 1 h. After further washing, the PVDF membrane was developed with enhanced chemiluminescence (ECL) reagent (Meilunbio, Dalian, China).

2.11 Cell proliferation measurement

A cell counting kit-8 (CCK-8) (Dalian Bergolin Biotechnology Co., Ltd., Dalian, China) was used to determine cell proliferation. HGC-NC, HGC-MOXD1sh1, and HGC-MOXD1sh2 cell lines were inoculated into 96-well plates (3 × 10³ cells per well). After adherence, a 10% CCK-8 mixture was added to the cells, and the plates were incubated for 45 min (away from direct light). Finally, fluorescence detection was performed at 450 nm on a miniature flat plate reader (at low speed). This experiment was repeated at 24, 48, and 72 h.

2.11.1 Cell migration assay

HGC-NC, HGC-MOXD1sh1, and HGC-MOXD1sh2 cell layers were trypsinized, washed with 3× concentrated PBS, and then mixed with serum-free medium. Next, 200 µL of each cell suspension (8 × 10⁴ cells) was transferred to the upper chamber of the Transwell inserts, and 800 µL of complete culture medium was added to the lower chamber. After incubation for 24 h, the migrated cells were stained with 0.5% crystal violet for 30 min. The dishes were then wiped clean and sealed for subsequent observation.

2.12 Statistical analyses and bioinformatics analyses

Continuous variable data were analyzed via the Mann‒Whitney U test or Wilcoxon signed rank sum test. Chi-square analysis was used to analyze the correlations between clinicopathological features. Correlations between two continuous variables were measured via the Pearson test. The hazard ratio (HR) and 95% confidence interval (CI) were calculated using a Cox regression model. The K‒M survival curve was drawn by the log-rank test. For all the statistical analyses, a two-tailed P value of <0.05 was used to assess statistical significance. R statistical software (v4.0.2) was used. For GSEA, visualize the enrichment analysis results using the ggplot2 package. Immune infiltration algorithm: based on the CIBERSORT (CIBERSORT; R script analysis) core algorithm, utilizing the CIBERSORT Tx website (https://cibersortx.stanford.edu/), calculate the markers of 22 immune cells provided. Heatmap: Visualize heatmaps using the ComplexHeatmap package. The Wilcoxon rank sum test was used for differential analysis.

Informed consent: All study participants or their legal guardians provided signed informed consent forms.
Ethical approval: All procedures followed were performed according to the ethical standards of the Human Subjects Responsibility Committee (institutions and countries), as well as the 1964 Helsinki Declaration and subsequent editions (ethics approval number: KY2021-09).

3 Results

3.1 The prognostic significance of OS in patients with GC

The course of this study is illustrated in the flowchart shown in Figure 1. The training cohort was split into three groups (C1, C2, and C3) on the basis of the results obtained from unsupervised clustering analysis and survival analysis (Figure 2a and b; Figure S1a–c). As demonstrated by the survival curve (Figure 2b), the median survival times of the C1 and C3 groups were significantly different (51.1 months vs. 29.7 months). The differentially expressed genes (DEGs) obtained from comparisons of tumor and paratumorous normal tissue samples from the C1 and C3 groups were then identified (Figure 2c). In total, 12 OSRGs were identified using this method. After LASSO Cox regression analysis, two key genes, APOD and CYP1B1, were identified (Figure 2d; Figure S1d and e). The OS risk score was calculated to be 2.64243451531501 e − 05 *APOD + 0.00237358944377552*CYP1B1. All patients with high expression of APOD and CYP1B1 (in both the training and validation cohorts) had a poor prognosis (all P < 0.05) (Figure S2a–d).

Figure 1

Flow chart of the research.

Figure 2

(a) Consensus heatmap of the three groups in the training cohort. (b) K‒M survival curves of the three groups in the training cohort. (c) Venn diagram showing the intersection of the oxidative stress (OS) gene set. (d) Heatmap of APOD and CYP1B1 mRNA expression and patient prognosis. (e) K‒M survival curve of the OS risk score in the training cohort. (f) K‒M survival curve of the OS risk score in the validation cohort.

On the basis of their OS risk score, patients in the training cohort were organized into either a high OS risk score group or a low OS risk score group. As shown by the survival curve, the survival times of patients in the low-OS risk score group were significantly longer (HR, 1.565; 95% CI, 1.185–2.066; P < 0.001) (Figure 2e). A similar result was obtained with the validation cohort (HR, 1.626; 95% CI, 1.262–2.095; P < 0.001) (Figure 2f).

Next, Cox regression analysis of sex, age, pTNM stage, and OS risk score was performed on data from the training cohort. Our analysis revealed that age, pTNM stage, and the OS risk score were independent prognostic factors in patients (P < 0.001, P < 0.001, and P = 0.003, respectively) (Table 1). In the validation cohort, pTNM stage and OS risk score were again identified as independent prognostic factors in patients (P < 0.001 and P = 0.002). A nomogram was subsequently constructed using data from the training cohort (Figure 3a). The survival curve of the nomogram revealed that patients in the high-risk group had worse survival (HR, 2.977; 95% CI, 2.327–3.808; P < 0.001) (Figure 3b). The areas under the curves for evaluating postoperative survival at 1, 3, and 5 years were 0.681 (0.631–0.730), 0.718 (0.668–0.768), and 0.682 (0.561–0.824), respectively (Figure 3c). DCA, calibration curve analysis, and a C-index of 0.673 (0.655–0.690) provided further evidence of the good clinical application potential of the nomogram (Figure 3d and e).

Table 1

Univariate and multivariate analyses based on Cox regression in the training cohort

Characteristics	Total (N)	Univariate analysis		Multivariate analysis
Characteristics	Total (N)	Hazard ratio (95% CI)	P value	Hazard ratio (95% CI)	P value
Age (years)	619	1.019 (1.009–1.029)	<0.001	1.025 (1.014–1.036)	<0.001
Sex	619
Male	398	Reference
Female	221	0.891 (0.690–1.149)	0.373
pTNM stage	605
Stage I	83	Reference		Reference
Stage II	151	2.782 (1.436–5.391)	0.002	2.625 (1.354–5.090)	0.004
Stage III	327	5.119 (2.781–9.422)	<0.001	5.043 (2.735–9.296)	<0.001
Stage IV	44	9.519 (4.746–19.089)	<0.001	11.170 (5.552–22.471)	<0.001
Oxidative stress risk score	619	8.520 (2.568–28.273)	<0.001	7.621 (2.010–28.901)	0.003

The pTNM stage is classified according to the AJCC 8th edition of the American Joint Committee on Cancer Staging Manual. Bold values represent the category headings.

Figure 3

(a) Nomograms constructed by age, OS risk score, and pTNM stage in the training cohort. (b) K‒M survival curve for the nomogram. (c) timeROC curve for the nomogram. (d) DCA curve for the nomogram. (e) Calibration curve for the nomogram.

3.2 Relationship between mRNA MOXD1 expression and the OS risk score and its clinical significance

MOXD1 was present among the DEGs between the high- and low-OS risk score groups and was positively correlated with the OS risk score (R = 0.427, P < 0.001) (Figure 4a) (Additional file 1) (Figure S3a and b). In the training and validation cohorts, patients with high MOXD1 mRNA expression all had poor survival (HR, 1.525; 95% CI, 1.166–1.996; P < 0.001) (HR, 1.591; 95% CI, 1.198–2.113; P < 0.001) (Figure 4b and c). In the training cohort, the mRNA expression level of MOXD1 was independent of age and sex (Figure S4a and b). The mRNA expression of MOXD1 in tumor tissue was significantly greater than that in adjacent normal tissue (Figure 4d). The mRNA expression of MOXD1 in T3 and T4 patients was significantly greater than that in T1 and T2 patients (Figure 4e), the mRNA expression of MOXD1 in N2 and N3 patients was significantly higher than that in N0 and N1 (Figure 4f), and the mRNA expression of MOXD1 in stage III patients was the highest (Figure 4g), which indicated that MOXD1 was closely related to the progression of GC. High mRNA expression of MOXD1 in N0 and N1-2, stage II and III patients was associated with poor prognosis (all P < 0.05) (Figure S4c–k) (Table 2).

Figure 4

(a) Pearson correlation between MOXD1 expression and the OS risk score in the training cohort. (b) K‒M survival curve for MOXD1 expression in the training cohort. (c) K‒M survival curve for MOXD1 expression in the validation cohort. (d) Differential expression of MOXD1 between tumor and paratumorous normal tissue. (e–g) Differential expression of MOXD1 in the T, N, and pTNM stages. *P < 0.05, **P < 0.01, ***P < 0.001.

Table 2

Univariate and multivariate analyses based on Cox regression in the validation cohort

Characteristics	Total (N)	Univariate analysis		Multivariate analysis
Characteristics	Total (N)	Hazard ratio (95% CI)	P value	Hazard ratio (95% CI)	P value
Age (years)	492	1.007 (0.996–1.018)	0.230
Sex	492
Male	324	Reference
Female	168	0.947 (0.726–1.234)	0.686
pTNM stage	492
Stage I	61	Reference		Reference
Stage II	126	2.147 (1.041–4.427)	0.038	2.275 (1.101–4.697)	0.026
Stage III	168	5.257 (2.649–10.433)	<0.001	5.206 (2.622–10.335)	<0.001
Stage IV	137	12.832 (6.470–25.452)	<0.001	12.746 (6.419–25.307)	<0.001
Oxidative stress risk score	492	1.103 (1.056–1.153)	<0.001	1.074 (1.025–1.124)	0.002

The pTNM stage is classified according to the AJCC 8th edition of the American Joint Committee on Cancer Staging Manual. Bold values represent the category headings.

3.3 MOXD1 and the GC tumor immune microenvironment

Analysis was subsequently conducted on the basis of the median MOXD1 expression level. Patients with high MOXD1 expression presented higher ESTIMATE, stromal, and immune scores. Therefore, high expression of MOXD1 is often accompanied by a strong immune response. Patients with high MOXD1 mRNA expression also presented lower tumor purity and higher TIDE scores. Hence, patients with low MOXD1 mRNA expression might be a potential population for immunotherapy. Additionally, high MOXD1 mRNA expression was associated with a high epithelial–mesenchymal transition (EMT) score, providing further evidence that MOXD1 is involved in the progression of GC (Figure 5a). According to the CIBERSORT scores, high mRNA expression of MOXD1 was also associated with high tissue infiltration by M2 macrophages, activated NK cells, and resting dendritic cells. Conversely, low mRNA expression of MOXD1 was associated with tissue infiltration by plasma cells, activated mast cells, and resting NK cells (Figure 5c). Our ssGSEA also revealed that MOXD1 mRNA expression was largely related to immune cell type (Figure 5b). According to the single-cell database, MOXD1 mRNA expression was especially associated with plasma cells and fibroblasts (Figure 5d and e).

Figure 5

(a) Heatmap showing the correlations between MOXD1 expression and the ESTIMATE score, stromal score, immune score, tumor purity, TIDE, and EMT score. (b) Immune infiltration in the high- and low-MOXD1 expression groups in the training cohort (using the ssGSEA algorithm). (c) Immune infiltration in the high- and low-MOXD1 expression groups in the training cohort (using the CIBERSORT algorithm). (d) and (e) Location of MOXD1 expression in the STAD-GSE167297 and STAD-GSE134520 datasets.

3.4 Biological analysis of MOXD1

A DEG analysis was performed on the median MOXD1 expression level via Limma, which yielded 6749 mRNAs (Figure S3c and d) (Additional file 1). GO analysis of biological process (BP) terms revealed that the mRNA MOXD1 is involved mainly in extracellular matrix (ECM) organization, extracellular structure organization, and epithelial cell migration. GO analysis of cellular component terms revealed that MOXD1 is involved mainly in the following processes: cell–substrate junction; basement film; ER lumen; and cell–cell junction. GO analysis of molecular function terms revealed that MOXD1 was involved mainly in growth factor binding, fibronectin binding, and transforming growth factor beta binding (Figure 6a) (Additional file 1). KEGG analysis revealed that MOXD1 was involved mainly in ECM receptor interactions, the PI3K/Akt signaling pathway, the MAPK signaling pathway, and the cGMP/PKG signaling pathway (Figure 6b) (Additional file 1). Our PPI network analysis revealed that the MOXD1 protein interacted with STX7, LUM, TAAR5, TAAR6, SLC18B1, TMEM163, PF4, SRXN1, CPLX3, COL1A2, and other proteins (Figure 6c). Finally, GSEA revealed that MOXD1 was associated with EMT, angiogenesis, the interleukin (IL)-6/JAK/STAT3 signaling pathway, the inflammatory response, and hypoxia (Figure 6d) (Additional file 1).

Figure 6

(a) GO analysis of MOXD1. (b) KEGG analysis of MOXD1. (c) PPI network analysis of MOXD1. (d) GSEA of MOXD1.

3.5 Molecular characterization and chemosensitivity of MOXD1 mRNA expression

Considering the molecular classification standards of GC, correlations between mRNA MOXD1 and four categories of GC in TCGA-STAD were analyzed. The results indicate that patients with high MOXD1 mRNA expression were more likely to have CDH1 mutations (Figure 7a). The difference in chemosensitivity between the high- and low-MOXD1 mRNA expression groups was also analyzed. Although patients with low MOXD1 mRNA expression were more likely to respond to cisplatin, patients with high MOXD1 mRNA expression were more likely to respond to capecitabine and oxaliplatin (Figure 7b).

Figure 7

(a) Association between MOXD1 mRNA expression levels and molecular classification in the TCGA-STAD cohort. (b) Association between MOXD1 mRNA expression levels and chemotherapy sensitivity in the training cohort.

3.6 Clinical significance of the MOXD1 protein

The immunohistochemical results for the TMAs are shown in Figure 8(a)–(c). The expression of MOXD1 is not related to histological type. The survival curve indicated that patients with high MOXD1 protein expression had poor survival (HR, 2.576; 95% CI, 1.199–5.537; P = 0.011) (Figure 8d). Cox regression analysis revealed that MOXD1 protein expression and the metastatic lymph node ratio were independent risk factors related to prognosis (P = 0.020 and P = 0.047, respectively) (Table 3). Chi-square analysis revealed that MOXD1 protein expression was closely related to N stage, pTNM stage, and the metastatic lymph node ratio (all P < 0.001) (Table 4). Next, the independent prognostic factors were combined to construct a nomogram (Figure 8e). According to the nomogram, the survival time of patients in the high-risk group was significantly shorter (HR: 6.533; 95% CI: 3.112–13.713; P < 0.001) (Figure 8f). The areas under the curve for evaluating 1- and 3-year survival were 0.581 (0.336–0.826) and 0.741 (0.629–0.854), respectively (Figure 8g). DCA and calibration curve analysis revealed that the nomogram exhibited good potential for evaluating patient prognosis, with a C-index of 0.711 (0.661–0.761) (Figure 8h and i).

Figure 8

(a)–(c) IHC staining of TMAs for MOXD1 (well-differentiated, moderately differentiated, and poorly differentiated). (d) K‒M survival curve for MOXD1 expression in TMAs. (e) Nomogram constructed using MOXD1 expression and the metastatic lymph node ratio in TMAs. (f) K‒M survival curve for the nomogram. (g) timeROC curve for the nomogram. (h) DCA curve for the nomogram. (i) Calibration curve for the nomogram.

Table 3

Univariate and multivariate analyses of TMAs based on Cox regression

Characteristics	Total (N)	Univariate analysis		Multivariate analysis
Characteristics	Total (N)	Hazard ratio (95% CI)	P value	Hazard ratio (95% CI)	P value
MOXD1 expression	100
Low	60	Reference		Reference
High	40	2.590 (1.212–5.532)	0.014	2.608 (1.159–5.869)	0.020
Sex	100
Male	72	Reference
Female	28	0.851 (0.362–2.002)	0.712
Age (years)	100	0.992 (0.957–1.028)	0.646
BMI (kg/m ²)	100	0.944 (0.845–1.054)	0.303
Tumor infiltration pattern	100
INFb	16	Reference
INFa	20	0.674 (0.168–2.695)	0.577
INFc	48	1.186 (0.394–3.576)	0.761
N/A	16	1.228 (0.330–4.573)	0.760
Lymphatic infiltration	100
Negative	55	Reference
Positive	45	0.940 (0.445–1.988)	0.872
Venous infiltration	100
Negative	70	Reference
Positive	30	0.592 (0.240–1.460)	0.255
Nerve infiltration	100
Negative	25	Reference
Positive	75	2.243 (0.778–6.471)	0.135
Tumor location	100
Lower third	54	Reference		Reference
Middle and Upper third	42	1.866 (0.847–4.113)	0.122	1.324 (0.577–3.040)	0.508
Entire stomach	4	7.426 (2.017–27.337)	0.003	3.750 (0.710–19.814)	0.120
Histological type	100
Mucinous	8	Reference
Well to moderately differentiated	46	3.095 (0.409–23.430)	0.274
Poorly differentiated	26	1.832 (0.214–15.686)	0.581
Signet ring cell	20	3.485 (0.428–28.348)	0.243
HER2 expression	100
Negative	82	Reference
Positive	18	1.660 (0.705–3.906)	0.246
CEA	100
≤5 ng/mL	86	Reference
>5 ng/mL	14	0.679 (0.205–2.250)	0.526
CA-199	100
≤37 U/mL	88	Reference
>37 U/mL	12	1.745 (0.663–4.593)	0.260
pTNM stage	100
I	10	Reference
II	32	2.155 (0.259–17.902)	0.477
III	58	4.558 (0.613–33.907)	0.138
Metastatic lymph node ratio	100	14.056 (3.348–59.004)	<0.001	6.767 (1.023–44.780)	0.047

BMI: body mass index; CEA: carcinoembryonic antigen; HER2: human epidermal growth factor receptor 2. The histological type and pTNM stage were determined according to the 8th AJCC system. Bold values represent the category headings.

Table 4

Relationships between the expression of MOXD1 and clinicopathological characteristics

Characteristics	Low MOXD1 expression	High MOXD1 expression	P value
n	103	77
Sex, n (%)			0.815
Male	76 (42.2%)	58 (32.2%)
Female	27 (15%)	19 (10.6%)
Age (years), mean ± SD	59.913 ± 10.051	60.74 ± 8.6121	0.562
BMI (kg/m ² ), median (IQR)	23.14 (20.73, 25.43)	22.66 (20.58, 24.57)	0.455
Tumor infiltration pattern, n (%)			0.278
INFb	30 (16.7%)	14 (7.8%)
INFa	17 (9.4%)	19 (10.6%)
INFc	39 (21.7%)	29 (16.1%)
N/A	17 (9.4%)	15 (8.3%)
Lymphatic infiltration, n (%)			0.911
Negative	58 (32.2%)	44 (24.4%)
Positive	45 (25%)	33 (18.3%)
Venous infiltration, n (%)			0.238
Negative	79 (43.9%)	53 (29.4%)
Positive	24 (13.3%)	24 (13.3%)
Nerve infiltration, n (%)			0.159
Negative	31 (17.2%)	16 (8.9%)
Positive	72 (40%)	61 (33.9%)
Tumor location, n (%)			0.566
Lower third	59 (32.8%)	38 (21.1%)
Middle and upper third	41 (22.8%)	36 (20%)
Entire stomach	3 (1.7%)	3 (1.7%)
Histological type, n (%)			0.579
Well to moderately differentiated	46 (25.6%)	35 (19.4%)
Poorly differentiated	22 (12.2%)	22 (12.2%)
Signet ring cell	23 (12.8%)	14 (7.8%)
Mucinous	12 (6.7%)	6 (3.3%)
HER2 expression, n (%)			0.150
Negative	92 (51.1%)	63 (35%)
Positive	11 (6.1%)	14 (7.8%)
CEA, n (%)			0.600
≤5 ng/mL	91 (50.6%)	66 (36.7%)
>5 ng/mL	12 (6.7%)	11 (6.1%)
CA-199, n (%)			0.234
≤37 U/mL	93 (51.7%)	65 (36.1%)
>37 U/mL	10 (5.6%)	12 (6.7%)
T stage, n (%)			0.178
T1	7 (3.9%)	3 (1.7%)
T2	20 (11.1%)	7 (3.9%)
T3	35 (19.4%)	33 (18.3%)
T4	41 (22.8%)	34 (18.9%)
N stage, n (%)			<0.001
N0	39 (21.7%)	11 (6.1%)
N1	23 (12.8%)	13 (7.2%)
N2	19 (10.6%)	22 (12.2%)
N3	22 (12.2%)	31 (17.2%)
pTNM stage, n (%)			<0.001
I	19 (10.6%)	4 (2.2%)
II	41 (22.8%)	15 (8.3%)
III	43 (23.9%)	58 (32.2%)
Metastatic lymph node ratio, median (IQR)	0.056 (0, 0.156)	0.156 (0.061, 0.322)	<0.001

BMI: body mass index; CEA: carcinoembryonic antigen; HER2: Human epidermal growth factor receptor 2; IQR: interquartile range. The histological type and pTNM stage were determined according to the 8th AJCC system. Bold values represent the category headings.

3.7 MOXD1 knockdown significantly inhibited the proliferation and invasion of GC cells

Our qPCR and Western blot results provide strong evidence that MOXD1 was markedly knocked down (at both the transcriptome and proteome levels) in HGC-27 cells infected with MOXD1-interfering viruses (Figure 9a and b). According to our CCK-8 assay results, the proliferation rates of HGC-27 cells after MOXD1 knockdown (HGC-MOXD1sh1 group, HGC-MOXD1sh2 group) were significantly lower than those of control cells (HGC-NC group) (Figure 9c). Therefore, MOXD1 protein knockdown affected the proliferation of GC cells. Moreover, our Transwell assay results revealed that the invasion capacities of HGC-MOXD1sh1 and HGC-MOXD1sh2 cells were significantly diminished (Figure 9d). Therefore, MOXD1 protein knockdown also affected the invasion capacity of GC cells. Asterisk denotes statistically significant differences: *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001.

Figure 9

(a) qPCR analysis of HGC-NC, HGC-MOXD1sh1, and HGC-MOXD1sh2 cells. (b) Western blot analysis of HGC-NC, HGC-MOXD1sh1, and HGC-MOXD1sh2 cells. (c) CCK-8 assay results for HGC-NC, HGC-MOXD1sh1, and HGC-MOXD1sh2 cells. (d) Cell migration assay results for HGC-NC, HGC-MOXD1sh1, and HGC-MOXD1sh2 cells. All experiments were repeated at least three times.

4 Discussion

OS is an imbalance between oxidation reactions and antioxidation reactions in cells that results in excessive ROS production. ROS can originate from either internal sources, including mitochondria, neutrophils, and macrophages (by nicotinamide adenine dinucleotide phosphate oxidase), or from external sources, including ionizing radiation and external chemical stimulation. ROS generation may involve the production of 8-OH deoxyguanosine, which is involved in the transformation of GC pairs into TA pairs during DNA replication, resulting in chromosome instability, cell mutation, and potentially tumorigenesis. Although oxidative cleavage of RNA is more difficult than oxidative cleavage of DNA, 8-oxoguanosine, 8-hydroxyadenine, and other substances can affect the stability of RNA through base modification and excision, leading to abnormal protein translation processes. OS can also affect protein stability by altering amino acid chemical groups, including those of tyrosine and phenylalanine. Therefore, an overall reduction in ROS-induced mutations delays cancer development and is generally beneficial to patients [8,38,39].

In colorectal cancer patients, the score obtained from a 9 OS-related lncRNA construction could be used to predict prognosis [40]. IL (IL-6, IL-8, and IL-10) levels and tumor necrosis factor-α levels are highly correlated with the levels of malondialdehyde, a marker of OS, indicating that OS may be reflected by the immune response [41]. Additional processes, such as lipid metabolism and ferroptosis, are also closely related to OS [38,42]. Furthermore, OS is known to be associated with inflammation, neurodegenerative diseases, and cancer. In tumor cells treated with chemotherapy drugs, OS was also shown to affect the production of reactive species, ultimately modulating the therapeutic effect [43,44,45].

Although Wu et al. previously demonstrated the importance of OS in GC [11], GC is highly heterogeneous, necessitating elucidation of the role of OS in different GC patients. To identify potential molecules involved in OS, we first constructed an OS risk score related to prognosis using our training cohort. This method yielded two genes of interest, namely, CYP1B1 and APOD. Cytochrome P450 family 1 subfamily B member 1 (CYP1B1) is a member of the cytochrome P450 family that participates in the metabolism of carcinogenic compounds, such as 7,12-dimethylbenz[a]anthracene, and plays a key role in estrogen metabolism [46]. Apolipoprotein D (APOD) is a member of the lipoprotein superfamily, and this transporter modulates lipid metabolism during several important BPs, including inflammatory and antioxidant reactions [47]. CYP1B1 and APOD have already been associated with tumor pathobiology in the literature, and their roles in various cancers are explored below.

CYP1B1 has been shown to play an important regulatory role in estrogen-related malignant tumors, such as breast, ovarian, and uterine cancers [46,48]. Although CYP1B1 is considered to have an inhibitory effect on many malignant tumors, HIF-1α can affect the expression of CYP1B1 by activating estrogen receptor α in breast cancer. CYP1B1 polymorphisms have been shown to modulate estrogen regulatory enzyme activity, promoting the occurrence of breast cancer [49,50]. Moreover, CYP1B1 polymorphisms are known predisposing factors in male patients with breast cancer. In bladder cancer, CYP1B1 is associated with lymph node metastasis, and its expression increases with the progression of this disease [51]. In colorectal cancer cells, CYP1B1 modulates the cell cycle by affecting the expression of the PCNA and FEN1 genes, ultimately promoting proliferation [52]. As a potential therapeutic target, the knockdown of CYP1B1 could increase the sensitivity of breast cancer cells to paclitaxel, 5-fluorouracil, and cisplatin. CYP1B1 is also a potential target for anti-PD-1 treatment of colorectal cancer [53,54]. In GC, CYP1B1 expression was observed to increase with tumor stage and grade, and it was highly correlated with tumor-related fibroblast numbers, immunological checkpoints, microsatellite instability, tumor mutation burden (TMB), and neoantigens. Moreover, CYP1B1 polymorphisms affect the progression of GC [55,56].

APOD is known to modulate important processes such as inflammation and antioxidation, and its expression is closely associated with numerous malignant tumors [47]. A risk model constructed using APOD, APOC1, and SQLE was used to evaluate the prognosis of patients with cervical cancer and their corresponding immune microenvironment status [57]. APOD was also found to be highly expressed in high-grade prostatic intraepithelial neoplasia and was identified as a potential biomarker [58]. In colorectal cancer, APOD was identified as a cancer stem cell-related gene [59]. APOD was also used as a potential marker for dexamethasone treatment of lymphoma [60]. In GC, APOD was identified as a gene affecting the TMB score and basement membrane, and high APOD expression was often associated with poor prognosis. Moreover, APOD was identified as a potential target for evaluating immunotherapy [61,62,63].

As discussed above, the biological functions of both CYP1B1 and APOD are closely related to the OS process. Importantly, the OS risk score constructed with CYP1B1 and APOD was an independent risk factor related to the prognosis of GC patients. Moreover, constructing an operational nomogram from the OS risk score, age, and pTNM stage was possible. This nomogram demonstrated good potential for exploring and evaluating GC prognoses. Together, the above results provide further evidence that OS is an important biological pathway affecting the progression of GC.

MOXD1 encodes a member of the copper monooxygenase protein family (which includes dopamine β monooxygenase and the peptide glycine α hydroxylated monooxygenase), and it is involved in numerous biological functions, including copper ion binding and redox reactions. MOXD1 is highly expressed in pituitary gland, salivary gland, and GBM cells, whereas MOXD1 mRNA expression in gastric cells is a biomarker of early GC [28]. Lai et al. [64] reported that fat mass and the expression of obesity-associated protein (FTO), a demethylase, were positively correlated with MOXD1 expression and that FTO affected both the methylation of m6A MOXD1 mRNA and the prognosis of GC patients.

In the present study, we found that high MOXD1 mRNA expression was positively correlated with the OS risk score and that MOXD1 might also be involved in OS in patients with GC. GO analysis revealed that MOXD1 can participate in the translation of a multifactor binding protein that binds to growth factor and fibronectin (among other proteins) and is involved in ECM organization. KEGG analysis revealed that MOXD1 was associated with the PI3K/Akt signaling, MAPK signaling, and cGMP/PKG signaling pathways. Two of these pathways have previously been implicated in OS. For example, coptisine inhibits OS during the treatment of hyperuricemia by inhibiting the PI3K/Akt signaling pathway [65]. OS was also shown to be regulated by cryptotanshinone treatment of polycystic ovary syndrome via the MAPK/ERK signaling pathway [66]. Hence, KEGG analysis provides direct evidence that MOXD1 activity is closely associated with these important signaling pathways and indirect evidence that MOXD1 participates in the regulation of OS via these pathways. MOXD1 is also considered a potential molecular target for multiple drug therapies.

In this study, the associations between MOXD1 mRNA expression levels and treatment outcomes were further evaluated. Our results revealed that cisplatin therapy was more suitable for patients with low MOXD1 mRNA expression levels, whereas capecitabine and oxaliplatin therapies were more suitable for patients with high MOXD1 expression levels. Therefore, MOXD1 mRNA expression levels provide guidance for clinical treatment. In addition, GSEA revealed associations between MOXD1 and several important BP pathways, specifically hypoxia and angiogenesis. These BP pathways are closely related not only to OS but also to the progression of malignant tumors [67,68]. Although studies suggest that the occurrence of OS may affect EMT, further exploration is warranted to investigate the potential role of EMT in promoting GC progression by MOXD1 [69]. According to the molecular typing of GC, an association also exists between high MOXD1 expression levels in patients and the incidence of mutations in cadherin 1 (CDH1). CDH1 mutations are typically hereditary, and MOXD1 changes may also have a genetic basis [70].

The effects of MOXD1 mRNA expression on the immune microenvironment were also evaluated. The ESTIMATE, stromal, and immune scores of patients with high MOXD1 expression levels were higher than those of patients with low MOXD1 mRNA expression levels. Indeed, the immune response in patients with high MOXD1 mRNA expression levels was significantly greater than that in patients with low MOXD1 mRNA expression levels, and these patients also presented lower tumor purity. Gong et al. [71] reported that low tumor purity was associated with poor prognosis, the EMT pathway, infiltration of specific immune cells (e.g., M2 macrophages), and immune cell inhibition by chemokines. In the present study, patients with high MOXD1 mRNA expression levels were found to exhibit increased EMT and TIDE. Therefore, the effectiveness of immunotherapy should be evaluated in patients with low MOXD1 expression levels.

ssGSEA and CIBERSORT analyses also revealed a close association between MOXD1 and various immune cells. High mRNA expression of MOXD1 was accompanied by high infiltration of M2 macrophages, activated NK cells, and resting dendritic cells. Conversely, low mRNA expression of MOXD1 was accompanied by the infiltration of plasma cells, activated mast cells, and resting NK cells. The single-cell database provides evidence for links between MOXD1 mRNA expression levels, plasma cells, and fibroblasts. In the tumor microenvironment, M2 macrophages are postulated to produce an immunosuppressive response that promotes the metastasis of cancer cells [72]. Zhu et al. reported that M2 macrophage infiltration in the tumor microenvironment of endometrial cancer was related to an increase in ROS. A similar relationship was reported for GBMs [15,73]. These results suggest that OS may cause M2 macrophage infiltration and that MOXD1 mRNA expression reflects the degree of infiltration.

In addition, patients with high MOXD1 mRNA expression exhibited high infiltration of NK cells, which are known to play a key role in tumor cell killing. Klopotowska et al. reported that the antitumor activity of NK cells was weakened by elevated OS levels [74], potentially leading to immune escape and the promotion of distant metastasis [75]. High OS levels may also promote an increase in tumor-related fibroblast numbers via a growth factor present in the tumor microenvironment, possibly fibroblast growth factor 2 or tumor growth factor β. Notably, MOXD1 mRNA expression was also associated with fibroblasts in the GSE134520 dataset [76]. Finally, the observed differences in plasma cell infiltration may be due to different sequencing methods and patient populations.

To date, only a few studies on MOXD1 expression in malignant tumors have been published. In bladder cancer [24], MOXD1 was found to be associated with copper metabolism. In HGSOC [25], bioinformatics analysis revealed that MOXD1 was a potential biomarker related to prognosis. In GBM [27], MOXD1 knockdown inhibited the proliferation, migration, and invasion of GBM cells and triggered apoptosis in ER-associated mitochondria, ultimately impacting tumor progression. In the present study, MOXD1 mRNA expression levels in tumor tissue samples from the training cohort were significantly increased (compared with those in the corresponding para-tumorous normal tissue samples). Additionally, patients with high MOXD1 expression levels presented shorter survival times, a result that was verified in the validation cohort. Moreover, high MOXD1 expression was found to promote lymph node metastasis in GC patients.

MOXD1 mRNA expression levels could also be used for the prognostic grouping of patients with N0, N1-2, and stages II and III of GC. Specifically, the MOXD1 expression levels in T3, N2, N3, and stage III GC patients were greater than those in the other groups. Therefore, MOXD1 expression (at the transcriptome level) increases with tumor progression and metastasis. Moreover, MOXD1 (at the protein level) should be considered a biomarker for, and may be involved in, the promotion of GC progression. A chi-square analysis of our immunohistochemistry results for TMAs revealed that MOXD1 expression was associated with N stage and pTNM stage. Moreover, the survival of patients with high MOXD1 expression was relatively short. High MOXD1 expression levels are associated with later N stages, higher metastatic lymph node ratios, and later pTNM stages; therefore, the MOXD1 protein is postulated to promote the progression of GC.

For clinical applications, we constructed a nomogram of MOXD1 expression levels and metastatic lymph node ratios, two independent risk factors for prognosis in patients with GC. In evaluating the prognosis of patients, the C-index of the nomogram was 0.711. The clinical application potential of this nomogram was further confirmed via ROC analysis, DCA, and K‒M survival curve analysis. Although size and time limitations may constrain the applicability of this pilot study, our results could easily be expanded by increasing both patient numbers and follow-up times in a future study. To further validate the proposed function of MOXD1, we constructed a stable knockdown cell line in HGC-27GC cells, which is different from the cell line used in the study of Lai et al. [61]. Our results revealed that the deletion of MOXD1 significantly inhibited the proliferation and invasion capabilities of HGC-27 cells. These results provide further confirmation of the biological function of MOXD1 and validation of its use as a biomarker of GC.

Several shortcomings were identified in this study. First, this was a single-center retrospective study with a limited patient number and a limited follow-up time. Although MOXD1 demonstrated good prognostic ability in our study, its clinical significance should be verified in a subsequent study involving more patients, more centers, and a longer follow-up. Second, the mechanisms by which MOXD1 affects OS are not known and should be elucidated by in vitro and in vivo functional experiments.

5 Conclusions

The OS risk score is an independent prognostic factor in GC, and GC patients with a high score have a poor prognosis. MOXD1 mRNA expression levels were positively correlated with the OS risk score, and GC patients with high MOXD1 expression (at both the mRNA and protein levels) had shorter survival times. MOXD1 knockdown inhibited both proliferation and invasion in HGC-27 GC cells. MOXD1 is a potential prognostic biomarker related to OS in patients with GC.

Abbreviations

APOD: Apolipoprotein D
AUC: Area under the receiver operating characteristic curve
BMI: Body mass index
BPs: Biological processes
CCK-8: Cell counting kit-8
CDH1: Cadherin 1
CEA: Carcinoembryonic antigen
CI: Confidence interval
CYP1B1: Cytochrome P450 family 1 subfamily B member 1
DAB: Diaminobenzidine
DCA: Decision curve analysis
DEGs: Differentially expressed genes
EMT: Epithelial–mesenchymal transition
ER: Endoplasmic reticulum
FDR: False discovery rate
FTO: Fat mass and obesity-associated protein
GBM: Glioblastoma
GC: Gastric cancer
GEO: Gene Expression Omnibus
GO: Gene Ontology
GSEA: Gene set enrichment analysis
HER2: Human epidermal growth factor receptor 2
HGSOC: High-grade serous ovarian cancer
HMU: Harbin Medical University
HR: Hazard ratio
IHC: Immunohistochemistry
IL: Interleukin
INFa: Expanding growth and a distinct border with the surrounding tissue
INFc: Infiltrating growth and an indistinct border with the surrounding tissue
INFb: In-between INFa and INFc
IQR: Interquartile range
KEGG: Kyoto Encyclopedia of Genes and Genomes
K‒M: Kaplan‒Meier
MOXD1: Monooxygenase DBH like 1
MSI: Microsatellite instability
OS: Oxidative stress
OSRGs: OS-related genes
PD-1: Programmed cell death protein 1
PPI: Protein‒protein interaction
ROS: Reactive oxygen species
STAD: Stomach adenocarcinoma
STRING: Search tool for retrieving interacting genes
TCGA: The Cancer Genome Atlas
TIDE: Tumor immune dysfunction and exclusion
timeROC: Time-dependent receiver operating characteristic
TMAs: Tissue microarrays
TMB: Tumor mutation burden
TNM: Tumor node metastasis

# These authors are contributed equally to this work.

tel: +86-15846565177

tel: +86-13069894000

Acknowledgments

We are grateful to the individuals and institutions who contributed to the establishment of a common database and the sharing of resources. Thank you to all the authors for their hard work on this article.

Funding information: The research was supported by the post-doctoral startup funding of Heilongjiang Provincial Department of Human Resources and Social Security, China, No. LBH-Q20157.
Author contributions: The research conception and design of this study were led by Youming Xiao and Xiqing Zhu. Data collection and organization were carried out by Cong Wang, while Hongyu Gao completed the data analysis and interpretation of the results. The initial draft of the manuscript was written jointly by Youming Xiao and Zenghui Hao. Haibin Song and Zhaozhu Li were responsible for revising and finalizing the manuscript. All authors take responsibility for the study’s findings and have approved the final version of the manuscript for publication.
Conflict of interest: All the authors declare that they have no conflicts of interest.
Data availability statement: The HMU-GC cohort was stored in the GEO repository (GSE184336 and GSE179252). Patients’ data are saved in the Gastric Cancer Information Management System v1.2 of Harbin Medical University Cancer Hospital (Copyright no. 2013SR087424, https://www.sgihmu.com/). The contact to the laboratory where the experiments have been conducted is +86-0451-86605743.

References

[1] Zhang X, Yang L, Liu S, Cao LL, Wang N, Li HC, et al. Interpretation on the report of global cancer statistics 2022. Zhonghua Zhong Liu Za Zhi [Chin J Oncol]. 2024;46(7):710–21.Suche in Google Scholar

[2] Yang S, Sun B, Li W, Yang H, Li N, Zhang X. Fatty acid metabolism is related to the immune microenvironment changes of gastric cancer and RGS2 is a new tumor biomarker. Front Immunol. 2022;13:1065927.10.3389/fimmu.2022.1065927Suche in Google Scholar PubMed PubMed Central

[3] Gu R, Xia Y, Li P, Zou D, Lu K, Ren L, et al. Ferroptosis and its role in gastric cancer. Front Cell Dev Biol. 2022;10:860344.10.3389/fcell.2022.860344Suche in Google Scholar PubMed PubMed Central

[4] Bai Z, Bai Y, Fang C, Chen W. Oxidative stress-related patterns determination for establishment of prognostic models, and characteristics of tumor microenvironment infiltration. Front Surg. 2022;9:1013794.10.3389/fsurg.2022.1013794Suche in Google Scholar PubMed PubMed Central

[5] Zha Y, Xue C, Liu Y, Ni J, De La Fuente JM, Cui D. Artificial intelligence in theranostics of gastric cancer, a review. Med Rev (2021). 2023 Jul;3(3):214–29.10.1515/mr-2022-0042Suche in Google Scholar PubMed PubMed Central

[6] Su Y, Jin G, Zhou H, Yang Z, Wang L, Mei Z, et al. Development of stimuli responsive polymeric nanomedicines modulating tumor microenvironment for improved cancer therapy. Med Rev (2021). 2023 Mar;3(1):4–30.10.1515/mr-2022-0048Suche in Google Scholar PubMed PubMed Central

[7] Reznik SE, Tiwari AK, Chavda V, Ashby Jr CR. The delivery of N-myc downstream-regulated gene 2 (NDRG2) self-amplifying mRNA via modified lipid nanoparticles as a potential treatment for drug-resistant and metastatic cancers. Med Rev (2021). 2024 Apr;4(3):235–8.10.1515/mr-2024-0004Suche in Google Scholar PubMed PubMed Central

[8] Jelic MD, Mandic AD, Maricic SM, Srdjenovic BU. Oxidative stress and its role in cancer. J Cancer Res Ther. 2021;17(1):22–8.10.4103/jcrt.JCRT_862_16Suche in Google Scholar PubMed

[9] Kuo CL, Ponneri Babuharisankar A, Lin YC, Lien HW, Lo YK, Chou HY, et al. Mitochondrial oxidative stress in the tumor microenvironment and cancer immunoescape: foe or friend? J Biomed Sci. 2022;29(1):74.10.1186/s12929-022-00859-2Suche in Google Scholar PubMed PubMed Central

[10] Dogan R, Dogan EE, Guler EM, Senturk E, Yenigun A, Celik I, et al. Oxidative stress values of tumor core, edge, and healthy thyroid tissue in thyroid masses. Eur Arch Otorhinolaryngol. 2021;278(8):2953–60.10.1007/s00405-020-06422-9Suche in Google Scholar PubMed

[11] Wu Z, Wang L, Wen Z, Yao J. Integrated analysis identifies oxidative stress genes associated with progression and prognosis in gastric cancer. Sci Rep. 2021;11(1):3292.10.1038/s41598-021-82976-wSuche in Google Scholar PubMed PubMed Central

[12] Forcados GE, James DB, Sallau AB, Muhammad A, Mabeta P. Oxidative stress and carcinogenesis: Potential of phytochemicals in breast cancer therapy. Nutr Cancer. 2017;69(3):365–74.10.1080/01635581.2017.1267777Suche in Google Scholar PubMed

[13] Zhang M, Du G, Li Z, Li D, Li W, Li H, et al. An oxidative stress-related genes signature for predicting survival in bladder cancer: Based on TCGA database and bioinformatics. Int J Gen Med. 2022;15:2645–67.10.2147/IJGM.S348945Suche in Google Scholar PubMed PubMed Central

[14] Wang X, Xu Y, Dai L, Yu Z, Wang M, Chan S, et al. A novel oxidative stress- and ferroptosis-related gene prognostic signature for distinguishing cold and hot tumors in colorectal cancer. Front Immunol. 2022;13:1043738.10.3389/fimmu.2022.1043738Suche in Google Scholar PubMed PubMed Central

[15] Zhu X, Xu Y, Wang J, Xue Z, Qiu T, Chen J. Loss of NLRP3 reduces oxidative stress and polarizes intratumor macrophages to attenuate immune attack on endometrial cancer. Front Immunol. 2023;14:1165602.10.3389/fimmu.2023.1165602Suche in Google Scholar PubMed PubMed Central

[16] Gambardella V, Castillo J, Tarazona N, Gimeno-Valiente F, Martinez-Ciarpaglini C, Cabeza-Segura M, et al. The role of tumor-associated macrophages in gastric cancer development and their potential as a therapeutic target. Cancer Treat Rev. 2020;86:102015.10.1016/j.ctrv.2020.102015Suche in Google Scholar PubMed

[17] Chan JS, Tan MJ, Sng MK, Teo Z, Phua T, Choo CC, et al. Cancer-associated fibroblasts enact field cancerization by promoting extratumoral oxidative stress. Cell Death Dis. 2017;8(1):e2562.10.1038/cddis.2016.492Suche in Google Scholar PubMed PubMed Central

[18] Lv C, Fu S, Dong Q, Yu Z, Zhang G, Kong C, et al. PAGE4 promotes prostate cancer cells survive under oxidative stress through modulating MAPK/JNK/ERK pathway. J Exp Clin Cancer Res. 2019;38(1):24.10.1186/s13046-019-1032-3Suche in Google Scholar PubMed PubMed Central

[19] Jee S, Hwang D, Seo S, Kim Y, Kim C, Kim B, et al. Microarray analysis of insulin-regulated gene expression in the liver: The use of transgenic mice co-expressing insulin-siRNA and human IDE as an animal model. Int J Mol Med. 2007;20(6):829–35.10.3892/ijmm.20.6.829Suche in Google Scholar

[20] Hoerder-Suabedissen A, Wang WZ, Lee S, Davies KE, Goffinet AM, Rakic S, et al. Novel markers reveal subpopulations of subplate neurons in the murine cerebral cortex. Cereb Cortex. 2009;19(8):1738–50.10.1093/cercor/bhn195Suche in Google Scholar PubMed

[21] Chen L, Zhu D, Huang J, Zhang H, Zhou G, Zhong X. Identification of Hub genes associated with COPD through integrated bioinformatics analysis. Int J Chron Obstruct Pulmon Dis. 2022;17:439–56.10.2147/COPD.S353765Suche in Google Scholar PubMed PubMed Central

[22] Chen J, Luo SF, Yuan X, Wang M, Yu HJ, Zhang Z, et al. Diabetic kidney disease-predisposing proinflammatory and profibrotic genes identified by weighted gene co-expression network analysis (WGCNA). J Cell Biochem. 2022;123(2):481–92.10.1002/jcb.30195Suche in Google Scholar PubMed

[23] Mao R, Wang Y, Wang F, Zhou L, Yan S, Lu S, et al. Identification of four biomarkers of human skin aging by comprehensive single cell transcriptome, transcriptome, and proteomics. Front Genet. 2022;13:881051.10.3389/fgene.2022.881051Suche in Google Scholar PubMed PubMed Central

[24] Wang W, Hua S, Li J, Zhao J, Zhang Y, Jiang J, et al. Tumour microenvironment landscape and immunotherapy response in bladder cancer decoded by stromal MOXD1 based on copper-related genes signature. Front Oncol. 2022;12:1081091.10.3389/fonc.2022.1081091Suche in Google Scholar PubMed PubMed Central

[25] Liu H, Zhou L, Cheng H, Wang S, Luan W, Cai E, et al. Characterization of candidate factors associated with the metastasis and progression of high-grade serous ovarian cancer. Chin Med J (Engl). 2023;136(24):2974–82.10.1097/CM9.0000000000002328Suche in Google Scholar PubMed PubMed Central

[26] Neuckermans J, Lequeue S, Claes P, Heymans A, Hughes JH, Colemonts-Vroninks H, et al. Hereditary tyrosinemia type 1 mice under continuous nitisinone treatment display remnants of an uncorrected liver disease phenotype. Genes (Basel). 2023;14(3):693.10.3390/genes14030693Suche in Google Scholar PubMed PubMed Central

[27] Shi P, Xu J, Xia F, Wang Y, Ren J, Liang P, et al. MOXD1 knockdown suppresses the proliferation and tumor growth of glioblastoma cells via ER stress-inducing apoptosis. Cell Death Discov. 2022;8(1):174.10.1038/s41420-022-00976-9Suche in Google Scholar PubMed PubMed Central

[28] Nation JB, Cabot-Miller J, Segal O, Lucito R, Adaricheva K. Combining algorithms to find signatures that predict risk in early-stage stomach cancer. J Comput Biol. 2021;28(10):985–1006.10.1089/cmb.2020.0568Suche in Google Scholar PubMed

[29] Ooi CH, Ivanova T, Wu J, Lee M, Tan IB, Tao J, et al. Oncogenic pathway combinations predict clinical prognosis in gastric cancer. PLoS Genet. 2009;5(10):e1000676.10.1371/journal.pgen.1000676Suche in Google Scholar PubMed PubMed Central

[30] Cristescu R, Lee J, Nebozhyn M, Kim KM, Ting JC, Wong SS, et al. Molecular analysis of gastric cancer identifies subtypes associated with distinct clinical outcomes. Nat Med. 2015;21(5):449–56.10.1038/nm.3850Suche in Google Scholar PubMed

[31] Leek JT, Johnson WE, Parker HS, Jaffe AE, Storey JD. The sva package for removing batch effects and other unwanted variation in high-throughput experiments. Bioinformatics. 2012;28(6):882–3.10.1093/bioinformatics/bts034Suche in Google Scholar PubMed PubMed Central

[32] Liu Q, Yang X, Yin Y, Zhang H, Yin F, Guo P, et al. Identifying the role of oxidative stress-related genes as prognostic biomarkers and predicting the response of immunotherapy and chemotherapy in ovarian cancer. Oxid Med Cell Longev. 2022;2022:6575534.10.1155/2022/6575534Suche in Google Scholar PubMed PubMed Central

[33] Wilkerson MD, Hayes DN. ConsensusClusterPlus: A class discovery tool with confidence assessments and item tracking. Bioinformatics. 2010;26(12):1572–3.10.1093/bioinformatics/btq170Suche in Google Scholar PubMed PubMed Central

[34] Wang T, Guo K, Zhang D, Wang H, Yin J, Cui H, et al. Disulfidptosis classification of hepatocellular carcinoma reveals correlation with clinical prognosis and immune profile. Int Immunopharmacol. 2023;120:110368.10.1016/j.intimp.2023.110368Suche in Google Scholar PubMed

[35] Colaprico A, Silva TC, Olsen C, Garofano L, Cava C, Garolini D, et al. TCGAbiolinks: an R/Bioconductor package for integrative analysis of TCGA data. Nucleic Acids Res. 2016;44(8):e71.10.1093/nar/gkv1507Suche in Google Scholar PubMed PubMed Central

[36] Cibulskis K, Lawrence MS, Carter SL, Sivachenko A, Jaffe D, Sougnez C, et al. Sensitive detection of somatic point mutations in impure and heterogeneous cancer samples. Nat Biotechnol. 2013;31(3):213–9.10.1038/nbt.2514Suche in Google Scholar PubMed PubMed Central

[37] Corsello SM, Nagari RT, Spangler RD, Rossen J, Kocak M, Bryan JG, et al. Discovering the anti-cancer potential of non-oncology drugs by systematic viability profiling. Nat Cancer. 2020;1(2):235–48.10.1038/s43018-019-0018-6Suche in Google Scholar PubMed PubMed Central

[38] Klaunig JE. Oxidative stress and cancer. Curr Pharm Des. 2018;24(40):4771–8.10.2174/1381612825666190215121712Suche in Google Scholar PubMed

[39] Hjelmeland A, Zhang J. Metabolic, autophagic, and mitophagic activities in cancer initiation and progression. Biomed J. 2016;39(2):98–106.10.1016/j.bj.2015.10.002Suche in Google Scholar PubMed PubMed Central

[40] Chen R, Wei JM. Integrated analysis identifies oxidative stress-related lncRNAs associated with progression and prognosis in colorectal cancer. BMC Bioinforma. 2023;24(1):76.10.1186/s12859-023-05203-5Suche in Google Scholar PubMed PubMed Central

[41] Ghafoor DD. Correlation between oxidative stress markers and cytokines in different stages of breast cancer. Cytokine. 2023;161:156082.10.1016/j.cyto.2022.156082Suche in Google Scholar PubMed

[42] Zeng W, Long X, Liu PS, Xie X. The interplay of oncogenic signaling, oxidative stress and ferroptosis in cancer. Int J Cancer. 2023;153(5):918–31.10.1002/ijc.34486Suche in Google Scholar

[43] Gill JG, Piskounova E, Morrison SJ. Cancer, oxidative stress, and metastasis. Cold Spring Harb Symp Quant Biol. 2016;81:163–75.10.1101/sqb.2016.81.030791Suche in Google Scholar

[44] Thanan R, Oikawa S, Hiraku Y, Ohnishi S, Ma N, Pinlaor S, et al. Oxidative stress and its significant roles in neurodegenerative diseases and cancer. Int J Mol Sci. 2014;16(1):193–217.10.3390/ijms16010193Suche in Google Scholar

[45] Victorino VJ, Pizzatti L, Michelletti P, Panis C. Oxidative stress, redox signaling and cancer chemoresistance: Putting together the pieces of the puzzle. Curr Med Chem. 2014;21(28):3211–26.10.2174/0929867321666140601164647Suche in Google Scholar

[46] Shou M, Korzekwa KR, Krausz KW, Buters JT, Grogan J, Goldfarb I, et al. Specificity of cDNA-expressed human and rodent cytochrome P450s in the oxidative metabolism of the potent carcinogen 7,12-dimethylbenz[a]anthracene. Mol Carcinog. 1996;17(4):241–9.10.1002/(SICI)1098-2744(199612)17:4<241::AID-MC8>3.3.CO;2-9Suche in Google Scholar

[47] Rassart E, Desmarais F, Najyb O, Bergeron KF, Mounier C, Apolipoprotein D. Gene. 2020;756:144874.10.1016/j.gene.2020.144874Suche in Google Scholar

[48] Ma X, Idle JR, Krausz KW, Gonzalez FJ. Metabolism of melatonin by human cytochromes p450. Drug Metab Dispos. 2005;33(4):489–94.10.1124/dmd.104.002410Suche in Google Scholar

[49] Rizzolo P, Silvestri V, Valentini V, Zelli V, Bucalo A, Zanna I, et al. Evaluation of CYP17A1 and CYP1B1 polymorphisms in male breast cancer risk. Endocr Connect. 2019;8(8):1224–9.10.1530/EC-19-0225Suche in Google Scholar

[50] Gajjar K, Martin-Hirsch PL, Martin FL. CYP1B1 and hormone-induced cancer. Cancer Lett. 2012;324(1):13–30.10.1016/j.canlet.2012.04.021Suche in Google Scholar

[51] Al-Saraireh YM, Alshammari F, Youssef AMM, Al-Sarayreh S, Almuhaisen GH, Alnawaiseh N, et al. Profiling of CYP4Z1 and CYP1B1 expression in bladder cancers. Sci Rep. 2021;11(1):5581.10.1038/s41598-021-85188-4Suche in Google Scholar

[52] Jin L, Huang J, Guo L, Zhang B, Li Q, Li H, et al. CYP1B1 promotes colorectal cancer liver metastasis by enhancing the growth of metastatic cancer cells via a fatty acids-dependent manner. J Gastrointest Oncol. 2023;14(6):2448–65.10.21037/jgo-23-895Suche in Google Scholar PubMed PubMed Central

[53] Hollis PR, Mobley RJ, Bhuju J, Abell AN, Sutter CH, Sutter TR. CYP1B1 augments the mesenchymal, claudin-low, and chemoresistant phenotypes of triple-negative breast cancer cells. Int J Mol Sci. 2022;23(17):9670.10.3390/ijms23179670Suche in Google Scholar PubMed PubMed Central

[54] Chen C, Yang Y, Guo Y, He J, Chen Z, Qiu S, et al. CYP1B1 inhibits ferroptosis and induces anti-PD-1 resistance by degrading ACSL4 in colorectal cancer. Cell Death Dis. 2023;14(4):271.10.1038/s41419-023-05803-2Suche in Google Scholar PubMed PubMed Central

[55] Yuan B, Liu G, Dai Z, Wang L, Lin B, Zhang J. CYP1B1: A novel molecular biomarker predicts molecular subtype, tumor microenvironment, and immune response in 33 cancers. Cancers (Basel). 2022;14(22):5641.10.3390/cancers14225641Suche in Google Scholar PubMed PubMed Central

[56] Freedman ND, Ahn J, Hou L, Lissowska J, Zatonski W, Yeager M, et al. Polymorphisms in estrogen- and androgen-metabolizing genes and the risk of gastric cancer. Carcinogenesis. 2009;30(1):71–7.10.1093/carcin/bgn258Suche in Google Scholar PubMed PubMed Central

[57] Zhang Y, Qin Y, Li D, Yang Y. A risk prediction model mediated by genes of APOD/APOC1/SQLE associates with prognosis in cervical cancer. BMC Womens Health. 2022;22(1):534.10.1186/s12905-022-02083-4Suche in Google Scholar PubMed PubMed Central

[58] Allina DO, Andreeva YY, Zavalishina LE, Moskvina LV, Frank GA. [Estimation of the diagnostic potential of APOD, PTOV1, and EPHA4 for prostatic neoplasms]. Arkh Patol. 2016;78(5):9–14.10.17116/patol20167859-14Suche in Google Scholar PubMed

[59] Liang Y, Su Q, Wu X. Identification and validation of a novel six-gene prognostic signature of stem cell characteristic in colon cancer. Front Oncol. 2020;10:571655.10.3389/fonc.2020.571655Suche in Google Scholar PubMed PubMed Central

[60] Jiang D, Jin H, Zuo J, Kong Y, Zhang X, Dong Q, et al. Potential biomarkers screening to predict side effects of dexamethasone in different cancers. Mol Genet Genomic Med. 2020;8(4):e1160.10.1002/mgg3.1160Suche in Google Scholar PubMed PubMed Central

[61] Wang Z, Chen H, Sun L, Wang X, Xu Y, Tian S, et al. Uncovering the potential of APOD as a biomarker in gastric cancer: A retrospective and multi-center study. Comput Struct Biotechnol J. 2024;23:1051–64.10.1016/j.csbj.2024.02.015Suche in Google Scholar PubMed PubMed Central

[62] Wang J, Liu D, Wang Q, Xie Y. Identification of basement membrane-related signatures in gastric cancer. Diagnostics (Basel). 2023;13(11):1844.10.3390/diagnostics13111844Suche in Google Scholar PubMed PubMed Central

[63] Guo X, Liang X, Wang Y, Cheng A, Zhang H, Qin C, et al. Significance of tumor mutation burden combined with immune infiltrates in the progression and prognosis of advanced gastric cancer. Front Genet. 2021;12:642608.10.3389/fgene.2021.642608Suche in Google Scholar PubMed PubMed Central

[64] Lai Y, Dong H, Xu P, Wang J, Feng W, Zhao Z, et al. RNA N6-methyladenosine demethylase FTO targets MOXD1 promoting the malignant phenotype of gastric cancer. BMC Gastroenterol. 2024;24(1):29.10.1186/s12876-023-03065-ySuche in Google Scholar PubMed PubMed Central

[65] Liu Y, Gong S, Li K, Wu G, Zheng X, Zheng J, et al. Coptisine protects against hyperuricemic nephropathy through alleviating inflammation, oxidative stress and mitochondrial apoptosis via PI3K/Akt signaling pathway. Biomed Pharmacother. 2022;156:113941.10.1016/j.biopha.2022.113941Suche in Google Scholar PubMed

[66] Khalid M, Petroianu G, Adem A. Advanced glycation end products and diabetes mellitus: Mechanisms and perspectives. Biomolecules. 2022;12(4):542.10.3390/biom12040542Suche in Google Scholar PubMed PubMed Central

[67] Wigner P, Grebowski R, Bijak M, Saluk-Bijak J, Szemraj J. The interplay between oxidative stress, inflammation and angiogenesis in bladder cancer development. Int J Mol Sci. 2021;22(9):4483.10.3390/ijms22094483Suche in Google Scholar PubMed PubMed Central

[68] Bae T, Hallis SP, Kwak MK. Hypoxia, oxidative stress, and the interplay of HIFs and NRF2 signaling in cancer. Exp Mol Med. 2024;56(3):501–14.10.1038/s12276-024-01180-8Suche in Google Scholar PubMed PubMed Central

[69] Farahzadi R, Valipour B, Fathi E, Pirmoradi S, Molavi O, Montazersaheb S, et al. Oxidative stress regulation and related metabolic pathways in epithelial-mesenchymal transition of breast cancer stem cells. Stem Cell Res Ther. 2023;14(1):342.10.1186/s13287-023-03571-6Suche in Google Scholar PubMed PubMed Central

[70] Hansford S, Kaurah P, Li-Chang H, Woo M, Senz J, Pinheiro H, et al. Hereditary diffuse gastric cancer syndrome: CDH1 mutations and beyond. JAMA Oncol. 2015;1(1):23–32.10.1001/jamaoncol.2014.168Suche in Google Scholar PubMed

[71] Gong Z, Zhang J, Guo W. Tumor purity as a prognosis and immunotherapy relevant feature in gastric cancer. Cancer Med. 2020;9(23):9052–63.10.1002/cam4.3505Suche in Google Scholar PubMed PubMed Central

[72] Pan Y, Yu Y, Wang X, Zhang T. Tumor-associated macrophages in tumor immunity. Front Immunol. 2020;11:583084.10.3389/fimmu.2020.583084Suche in Google Scholar PubMed PubMed Central

[73] Liang X, Wang Z, Dai Z, Liu J, Zhang H, Wen J, et al. Oxidative stress is involved in immunosuppression and macrophage regulation in glioblastoma. Clin Immunol. 2024;258:109802.10.1016/j.clim.2023.109802Suche in Google Scholar PubMed

[74] Klopotowska M, Bajor M, Graczyk-Jarzynka A, Kraft A, Pilch Z, Zhylko A, et al. PRDX-1 supports the survival and antitumor activity of primary and CAR-modified NK cells under oxidative stress. Cancer Immunol Res. 2022;10(2):228–44.10.1158/2326-6066.CIR-20-1023Suche in Google Scholar PubMed PubMed Central

[75] Guillerey C. NK cells in the tumor microenvironment. Adv Exp Med Biol. 2020;1273:69–90.10.1007/978-3-030-49270-0_4Suche in Google Scholar PubMed

[76] Eskandari-Malayeri F, Rezaei M. Immune checkpoint inhibitors as mediators for immunosuppression by cancer-associated fibroblasts: A comprehensive review. Front Immunol. 2022;13:996145.10.3389/fimmu.2022.996145Suche in Google Scholar PubMed PubMed Central

Received: 2025-03-30

Revised: 2025-08-14

Accepted: 2025-08-19

Published Online: 2025-09-19

This work is licensed under the Creative Commons Attribution 4.0 International License.

Supplementary material

Artikel in diesem Heft

https://doi.org/10.1515/med-2025-1271

Schlagwörter für diesen Artikel

gastric cancer; MOXD1; mRNA; oxidative stress; prognosis

Creative Commons

BY 4.0

The mRNA MOXD1: Link to oxidative stress and prognostic significance in gastric cancer

Artikel

Abstract

Purpose

Methods

Results

Conclusion

1 Introduction

2 Materials and methods

2.1 GC data

2.2 Data processing

2.3 Structure of the OS risk score

2.4 Immune microenvironment analysis

2.5 Bioinformatics analyses

2.6 Drug analysis

2.7 Immunohistochemistry

2.8 Cell line and transfection

2.9 RNA extraction and quantitative real-time PCR (qPCR)

2.10 Western blot analysis

2.11 Cell proliferation measurement

2.11.1 Cell migration assay

2.12 Statistical analyses and bioinformatics analyses

3 Results

3.1 The prognostic significance of OS in patients with GC

3.2 Relationship between mRNA MOXD1 expression and the OS risk score and its clinical significance

3.3 MOXD1 and the GC tumor immune microenvironment

3.4 Biological analysis of MOXD1

3.5 Molecular characterization and chemosensitivity of MOXD1 mRNA expression

3.6 Clinical significance of the MOXD1 protein

3.7 MOXD1 knockdown significantly inhibited the proliferation and invasion of GC cells

4 Discussion

5 Conclusions

Abbreviations

Acknowledgments

References

Zusatzmaterial

Artikel in diesem Heft

Artikel in diesem Heft

Artikel in diesem Heft