Construction of a prognostic model for gastric cancer patients based on NAD+ metabolism-related genes and tumor microenvironment analysis

doi:10.21203/rs.3.rs-2711008/v1

Download PDF

Article

Construction of a prognostic model for gastric cancer patients based on NAD+ metabolism-related genes and tumor microenvironment analysis

https://doi.org/10.21203/rs.3.rs-2711008/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background

Nicotinamide adenine dinucleotide (NAD+) metabolism is important in the regulation of tumor immune escape. This study endeavored to develop a NAD + metabolism-related signature in gastric cancer (GC), which could provide a theoretical foundation for prognosis and therapy of GC patients.

Methods

First, differentially expressed genes (DEGs) between GC and paraneoplastic tissues were intersected with NAD + metabolism-related genes (NMRGs) to obtain differentially expressed NMRGs (DE NMRGs). Then, based on the transcript levels of NMRGs, GC patients were classified into high and low scoring groups using the Gene set variation analysis (GSVA) algorithm. Next, the DEGs between the high and low scoring groups were intersected with DEGs between GC and paraneoplastic tissues to obtain the GC-NM DEGs. Additionally, univariate Cox analysis and Least absolute shrinkage and selection operator (LASSO) regression analysis of GC-NM DEGs were performed to obtain prognostic biomarkers, which were used to construct a risk model. In addition, independent prognostic factors were obtained by Cox analysis based on risk scores and clinicopathological factors. Gene set enrichment analysis (GSEA) enrichment analysis and immune infiltration analysis were performed for the high- and low-risk groups. Finally, the mRNA expression of prognostic related genes was verified by experiment.

Results

10 DE NMRGs were obtained and they were involved in the biological process of NAD biosynthetic process, nicotinamide nucleotide, and biosynthetic process. Further 7 biomarkers, including DNAJB13, CST2, THPO, CIDEA, ONECUT1, UPK1B, and SNCG, were obtained through univariate Cox and LASSO analyses of 1001 GC-NM DEGs. In addition, risk score and gender were demonstrated as credible independent prognostic factors for GC. Moreover, GSEA showed that the high-risk group was associated with bile secretion, intrinsic component of synaptic membrane and other pathways, while the low-risk group was associated with CMG complex. In addition, T cells, B cells, and natural killer cells were positively correlated with risk scores, and plasmacytoid dendritic cells were negatively correlated with risk scores. By QRT-PCR, the expression of prognostic genes in GC tissues was significantly up-regulated compared with paraneoplastic tissues.

Conclusion

This study established a NAD + metabolism-related signature based on DNAJB13, CST2, THPO, CIDEA, ONECUT1, UPK1B, and SNCG, which is of great significance in developing prognostic molecular biomarkers, clinical prognosis prediction, and treatment strategy decision for GC patients.

Health sciences/Oncology/Surgical oncology

Biological sciences/Immunology/Tumour immunology

Health sciences/Medical research/Biomarkers

Health sciences/Gastroenterology/Gastrointestinal diseases

Gastric cancer

Nicotinamide adenine dinucleotide (NAD+) metabolism

prognosis

risk signature

Gastric cancer (GC) is still one of the most common malignant tumors in the world, and its morbidity and mortality are at the forefront of all kinds of tumors. There were 1,089,103 new cancer cases and 768,793 new deaths in 2020. [1]. GC has an insidious onset, and early patients often have no apparent symptoms. Therefore, patients are easily missed in the early stage. At the time of diagnosis, they are often in an advanced stage, or distant organ metastasis has occurred. In addition, the limitations of current treatment methods often lead to poor prognoses[2]. Therefore, identifying and developing more robust prognostic markers and models of GC have important clinical significance in predicting the prognosis of patients with GC, guiding the clinical diagnosis and treatment of GC, and improving the survival rate of patients.

Nicotinamide adenine dinucleotide (NAD +) is an essential auxiliary regulatory factor in cells, which is involved in cellular energy metabolism, protein post-translational modification, DNA repair, cell stress adaptation, and other biological processes. NAD + metabolic disorders are closely related to various physiological changes and pathological damage, and it is closely linked to the aging process of the body, metabolic and neurodegenerative diseases, and the development and progression of various cancers [3, 4]. Recovery of intracellular NAD + levels demonstrated beneficial effects on cell function and related diseases. NAD + controls hundreds of key processes from energy metabolism to cell survival. NAD + levels decline steadily with age, leading to metabolic changes and increased disease susceptibility. Restoring NAD + levels in elderly or diseased animals can promote health and prolong life[5]. In the classic remedial pathway, Nicotinamide phosphoribosyl transferase (NAMPT) As a rate-limiting enzyme, NAM is converted into nicotinamide mononucleotide (NMN), which is then converted into NAD by nicotinamide mononucleotide adenylyl transferase (NMNAT)[6]. It has been reported that NAMPT is highly expressed in various tumors such as GC, breast cancer, pancreatic cancer, and prostate cancer to promote glycolysis, proliferation, survival, invasion, metastasis, and chemotherapy resistance of tumor cells[7–10]. Therefore, it is necessary to explore the correlation between NAD + metabolism-related genes and the prognosis of gastric cancer patients, which can provide new prognostic biomarkers and therapeutic targets for the treatment of gastric cancer.

In this study, GC-NM DEGs were obtained by crossing the DEGs between GC and paraneoplastic samples and the DEGs between NAD + high and low scoring groups. Based on this, the risk model was constructed by cox analysis and lasso analysis. After that, independent prognostic factors were obtained by COX analysis and a histogram was constructed to predict the survival rate of patients. Finally, immune infiltration analysis explored the relationship between the model genes and the differential immune cells between the high and low risk groups. These results can provide the theoretical basis for prognosis prediction and treatment strategies for patients with GC.

2.1 Data source

GC related RNA-seq data were downloaded from The Cancer Genome Atlas database (TCGA, https://portal.gdc.cancer.gov), and clinical data were downloaded from the cbioportal (https://www.cbioportal.org/) database. The TCGA database included sequencing data of 375 GC tissues and 32 paraneoplastic tissues, of which 354 GC samples had survival information. In addition, the GSE84437 dataset was downloaded from the Gene Expression Omnibus (GEO, http://www.ncbi.nlm.nih.gov/geo/) database, which included sequencing data of 433 GC samples. 51 NAD + metabolism-related genes (NMRGs) were sourced from the Kyoto Encyclopedia of Genes and Genomes (KEGG, https://www.kegg.jp/ or https://www.genome.jp/kegg/) pathway database (hsa00760) and the Reactome (https://reactome.org) database (R-HSA-196807)[11].

2.2 Differential expression and enrichment analysis

This study excavated the differentially expressed genes (DEGs) between GC tissues and paraneoplastic tissues through the “limma” R package (version 3.52.1) with adjust.P < 0.05 and |logFC| > 1[12]. Then heat map and volcano map of DEGs were plotted using the R package “pheatmap” (version 1.0.12) and R package “ggplot” (version 3.3.6). Moreover, NMRGs were crossed with DEGs to obtain differentially expressed NMRGs (DE NMRGs), and we applied the R package “clusterProfiler” (version 4.0.2)[13] for Gene Ontology (GO) and KEGG enrichment analysis of DE NMRGs. The threshold value was P < 0.05.

2.3 Correlation analysis of high and low score groups

Gene set variation analysis (GSVA) was performed on GC samples with survival information in the TCGA database using NMRGs as the background gene set, and GC patients were divided into high and low scoring groups according to the median value of the scores. Then the DEGs were obtained by differential analysis of GC patients in the high and low scoring groups, and the DEGs were intersected with DEGs between GC and paraneoplastic tissues to gain the GC-NM DEGs. In addition, the survival analysis of GC patients in the high and low scoring groups was performed using progression-free survival (PFS), and the genes were ranked according to the fold change of the genes, and the gene set enrichment analysis (GSEA) analysis was performed for the high and low scoring groups. This study also obtained the proportion of 22 immune cells in the high and low scoring groups by the Cell type Identification By Estimating Relative Subsets Of RNA Transcripts (CIBERSORT) algorithm and compared between high and low scoring groups by the Wilcox test.

2.4 Construction and validation of a prognostic model

The GC patients with survival information in the TCGA dataset were used as the training set. The univariate Cox (“survival” package, version 3.2–13) and Least absolute shrinkage and selection operator (LASSO) regression (“glmnet” package, version 4.1-3)[14] analysis based on GC-NM DEGs were deployed to select prognostic genes in the training set. Then we calculated the risk score according to the formula (Riskscore = h0(t) * exp (β1X1 + β2X2 + … + βnXn), β represented the regression coefficient obtained by LASSO, h0(t) was the baseline hazard function, and patients were separated into two risk subgroups (high risk and low risk) with the median of risk score. Kaplan-Meier (K-M) curves (“survminer” package, version 0.4.8), receiver operating characteristic (ROC) analysis (“timeROC”, version 0.4), and risk curves (“ggplot2” package, version 3.3.5) were generated employed to demonstrate the predictive efficiency of signature. In addition, the GSE84437 dataset was served as an external validation set.

2.5 Independent prognostic analysis and construction of a nomogram

The risk score and clinical pathological factors were enrolled into Cox analysis (univariate Cox and multivariate Cox) to authenticate independent prognostic predictors. The nomogram comprising the independent prognostic predictors was drawn using R package “Rms” (version 6.1-0)[15] to predict survival at 1, 3, and 5 years in GC patients. The calibration curves were plotted by “regplot” package (version 1.1) to appraise the precision and reliability of the nomogram predictions.

2.6 Functional enrichment analysis and immune infiltration analysis in high- and low-risk groups

To further investigate the signaling pathways and potential biological mechanisms in the high- and low-risk groups, GO and KEGG enrichment analyses were performed by GSEA on both high- and low-risk groups. The threshold was set at P < 0.05 and |NES| > 1. In addition, the proportion of 28 immune cells in the high- and low-risk groups were analyzed by single sample gene set enrichment analysis (ssGSEA) algorithm using the R package “GSVA”[16], and the immune cells with significant differences were obtained by rank sum test. Then, the relationships between immune cells and risk score were calculated by R package “ggplot”.

2.6 The mRNA expression of prognostic related genes was verified by experiment

In this study, tissue samples were collected from three patients with GC. All GC patients were from Tianjin Third Central Hospital and signed informed consent. This study was approved by the Ethics Committee of Tianjin Third Central Hospital. The control group was paraneoplastic tissue samples, and the tumor group was GC tissue samples. Real-time fluorescence quantitative QRT-PCR was carried out according to the basic operation steps, and all consumables and solutions were sterilized to prevent degradation and contamination of RNA during extraction. At low temperature, the tissue was ground into homogenate, and the total RNA of the tissue was extracted by the Trizol method. The total RNA of the tissue was obtained by reverse transcription reaction according to the operating guide of the reverse transcription kit. Related gene primers were diluted to 10uM with DEPC water, then the DNA amplification reaction system was constructed according to the instructions of the gene amplification kit: Reaction substrate, cDNA, and primers were added to each group to prepare a reaction system of 20 µl. Four complex Wells were made in each group. The system was mixed and put into the PCR instrument, and the reaction procedure was set: pre-denaturation at 94℃ for 30 s, denaturation at 94℃ for 10min, annealing at 65℃ for 15 s, extension at 72℃ for 10 s, and amplification for 40 cycles. Quantitative analysis was performed by comparison threshold method: quantitative copy number of target gene = 2-ΔΔCt, ΔCt = Ct value of target gene - Ct value of internal reference gene, ΔΔCt = ΔCt experimental group - ΔCt control group.

3.1 Enrichment analysis of 10 DE NMRGs

A total of 5018 DEGs between GC and paraneoplastic tissues were obtained, with 3964 up-regulated genes and 1054 down-regulated genes (Fig. 1A-B, Table S1). The 10 DE NMRGs were obtained (Fig. 1C), and they were enriched to 152 GO Biological Process (BPs), 38 GO Molecular Function (MFs), and 13 KEGG pathways, including NAD biosynthetic process, nicotinamide nucleotide biosynthetic process, positive regulation of immune response to tumor cell, positive regulation of immune effector process, JAK-STAT signaling pathway, nicotinate and nicotinamide metabolism and nucleotide metabolism, etc. (Fig. 1D-E, Table S2-3).

3.2 Correlation analysis of GSVA

The 354 GC patients were divided into the high scoring group (n = 177) and low scoring group (n = 177). There were 2956 DEGs between high and low scoring groups (Table S4), and a total of 1001 GC-NM DEGs were obtained (Fig. 2A). Figure 2B showed that the patients in the high scoring group had a lower survival rate compared to the low scoring group. GSEA showed 1501 GO items and 80 KEGG pathways were enriched (Table S5-6). The top5 GO and KEGG items enriched by high-score (NES > 0) and low-score (NES < 0) patients were showcased in Fig. 2C-D, including arrhythmogenic right ventricular cardiomyopathy, cell adhesion molecules, circadian entrainment, cell–cell adhesion via plasma-membraneadhesion molecules, intrinsic component of synaptic membrane etc. The CIBERSORT algorithm showed that there were native B cells, memory B cells, regulatory T cells (Tregs), marcophage M0, resting mast cells, activated mast cells, neutrophils with differences, in which B cell memory was close to zero in most samples.

3.3 Construction and validation of a prognostic model

To mine the NMRGs relevant to the overall survival (OS) of GC patients, we incorporated the 1001 GC-NM DEGs obtained above into a univariate Cox analysis in the training set, resulting in 387 genes associated with prognosis (Fig. 3A, Table S7). The above 387 genes were further integrated into the LASSO analysis, and 16 characteristic genes were screened out, namely NPY1R, NPR3, CCNA1, AFF2, DNAJB13, SNCG, OLFM3, CFHR3, CST2, CLDN10, DAO, THPO, CIDEA, NCCRP1, ONECUT1, UPK1B (Fig. 3B). Multivariate COX analysis was performed on the 16 prognostic-related genes obtained from the previous LASSO regression analysis, and 7 genes were finally determined as prognostic biomarkers, namely DNAJB13, CST2, THPO, CIDEA, ONECUT1, UPK1B, SNCG (Fig. 3C). Next, we computed the risk score for each GC patient in the training set and classified them into two risk subgroups (high risk and low risk) relying on the median value. Figure 3D revealed the distribution of ranked risk score and survival status for each patient in the training set. The expression heat map showed that DNAJB13, CST2, THPO, CIDEA, ONECUT1, UPK1B, and SNCG were highly expressed in patients with higher risk scores (Fig. 3E). The K-M curve illustrated that patients with higher risk had noteworthy poorer survival than those with lower risk (Fig. 3F). We graphed ROC curves to check the predictive efficiency of the signature. The area under the curve (AUC) values for OS in the training set were 0.699 (1 year), 0.728 (3 year), and 0.712 (5 year), indicating a decent accuracy (Fig. 3G). In addition, a comparable trend was verified in the external validation set (Fig. 4A-D). Above results revealed that the risk signature was a valid survival predictor for GC patients.

3.4 The risk score and gender were independent prognostic predictors

Via Cox analysis, we detected that risk score and gender were independent predictors of prognosis for patients with GC (P < 0.05) (Fig. 5A-B). The nomogram containing independent prognostic predictors was generated (Fig. 5C). The calibration curves proved that the capability of the nomogram in predicting survival at 1, 3, and 5 years in GC patients was satisfactory (Fig. 5D).

3.5 Functional enrichment analysis of high and low risk groups

872 GO items and 53 KEGG pathways were enriched (TableS8-9). The top5 GO and KEGG items enriched by high risk (NES > 0) and low risk (NES < 0) patients were shown in Fig. 6A-B. Homophilic cell adhesion via plasma membrane adhesion molecules, multicellular organismal signaling, and postsynaptic membrane et were activated in the patients with higher NMRGs-relevant risk scores (NES > 0). CMG complex, DNA strand elongation involved in DNA replication, PPAR signaling pathway, and complement and proteasome assembly et were activated in the patients with lower NMRGs-relevant risk scores.

3.6 Relationships between risk score and immune cell

There were differences between the high and low risk groups for 11 immune cells, with activated CD4 T cells, activated CD8 T cells, type 17 T helper cells highly expressed in the high risk group, and central memory CD4 T cells, T follicular helper cells, activated B cells, immature B cells, natural killer cells, plasmacytoid dendritic cells, immature dendritic cells, and mast cell were low expressed in the high risk group (Fig. 7A). In addition, 7 immune cells (activated CD4 T cell, activited CD8 T cell, activated dendritic cell, CD56 bright natural killer cell, gamma delta T cell, memory B cell, plasmacytoid dendritic cell) were significantly correlated with the risk score (Fig. 7B), suggesting that the risk score may be involved in the development of GC either directly or indirectly through the influence of immune cells.

3.7 Expression of prognostic genes

In order to further verify the expression of the prognostic gene, we studied the relative expression of its transcripts in paraneoplastic tissues and GC tissues by QRT-PCR. The results showed that the expressions of DNAJB13, CST2, THPO, CIDEA, ONECUT1, UPK1B, and SNCG were significantly up-regulated compared with those of paraneoplastic tissues (Fig. 8).

Studies have shown that the level of glycolysis in cancer cells is high and that cancer cells produce large amounts of NAD+. NAD + is essential in energy metabolism and can be involved in important cellular signaling processes such as mono-ADF and poly-ADP ribosylation, NAD+-dependent protein deacetylation [17]. These signaling events are closely related to the occurrence and development of GC. Therefore, in this study, we constructed the molecular characteristics of gastric cancer with NAD + metabolism-related genes based on TCGA and GEO data, and explored the correlation with tumor microenvironment.

We performed a univariate COX analysis of NAD + metabolism-related genes (NMRGs) to obtain 387 prognostic genes. Further LASSO and multivariate analysis were used to determine DNAJB13, CST2, THPO, CIDEA, ONECUT1, UPK1B, and SNCG as predictive biomarkers for constructing the risk model. CST2 expression was significantly up-regulated in GC samples, enhancing the growth, migration, and invasion of GC cells by regulating epithelial-mesenchymal transition (EMT) and TGF-β1 signaling pathways, leading to poor prognosis in patients[18]. Thrombopoietin (THPO) is involved in megakaryocyte proliferation and maturation of megakaryocyte growth and development factors. It is overexpressed in GC and causes poor prognosis by affecting cell viability, invasion, and migration[19]. In recurrent GC, elevated UPK1B expression has a worse prognosis[20]. The mechanism for poor prognosis may be that UPK1B significantly affects the expression of critical genes in the Wnt/β-catenin signaling pathway in cancer cells[21]. SNCG expression in gastric adenocarcinoma was considerably higher than in non-tumor adjacent gastric tissues and correlated with the depth of invasion and lymph node metastasis[22]. These genes are highly expressed in GC. The expression of these genes is higher, and the prognosis of GC patients is worse. This is consistent with the results of our study and further improves the reliability of our findings. Among these seven genes, we first found that DNAJB13, CIDEA, and ONECUT1 were associated with the prognosis of GC patients. Although these genes have not been reported in GC, we were surprised to find that these three genes have been reported in colon cancer. Studies have shown that compared with normal colonic mucosa, the expression of brown fat-associated protein (CIDEA) in colorectal tumors increased significantly. The brown fat-like phenotype characterized by CIDEA expression may be due to the infinite plasticity of cancer stem cells, which can differentiate into multiple cell lines with various phenotypes according to gene mutations and microenvironment factors. These cells undergo carcinogenesis under the influence of multiple factors[23]. ONECUT1 is highly expressed in colorectal cancer and has a poor prognosis. The mechanism may be that ONECUT1 promotes the proliferation and tumor growth of colorectal cancer cells and promotes liver metastasis by endogenic resistance to anoikis[24]. The high expression of DNAJB13 in colon cancer tissues also confirmed that DNAJB13 is an independent prognostic risk factor for colon cancer patients[25]. The results of the above three genes are consistent with our experiments. Therefore, we preliminarily speculate that these genes may be independent risk factors in GC.

This study is the first to construct a prognostic prediction model of NMRGs nomogram in GC patients. Compared with the traditional TNM staging system, the nomogram model based on multivariate analysis has been proven to obtain more accurate prognostic prediction in various cancer patients[26, 27]. RiskScore and gender were identified as independent prognostic factors for GC by univariate and multivariate COX analysis based on risk score and clinicopathological characteristics. We established a nomogram prognostic prediction model for GC patients. The internal validation C-index was 0.713, suggesting that it has an excellent predictive effect on the prognosis of patients with GC. The external validation set also further validates this model. Studies have shown that the NMRGs risk score prediction model is closely related to the prognosis of liver cancer, and the high-risk group was significantly worse than the low-risk group [28]. This is consistent with our experimental results. The simple and easy-to-use nomogram prediction model constructed in this study can help clinicians make more accurate survival predictions, identify high-risk patients, and conduct early individual intervention and treatment.

Tumor-associated lymphocytes have a positive effect on the prognosis of GC. Dendritic cells (DCs) and natural killer (NK) cells initiate specific immune responses to tumor cells. The decrease of subsets of DC and NK and cytokines may be one of the reasons for the decrease of DC and NK function in tissues and peripheral blood of patients with gastric cancer[29]. Human NK cells can be divided into two subsets according to the expression of CD56 and CD16. The H2O2 produced in the tumor microenvironment is inversely proportional to the infiltration of CD56, which may be due to its preferential induction of cell death. It further explains one of the mechanisms behind NK cell dysfunction often observed in the tumor microenvironment[30]. T cells are an important part of tumor infiltrating lymphocytes (TILs), and T cell-mediated adaptive immunity is considered to play an important role in anti-tumor immunity. T cell subsets are composed of CD8, CD4 and other cells. These lymphocytes can infiltrate the matrix and tumor cells to regulate the host 's immune response to tumor cells. Up-regulated PD-L1 or CTLA-4 expression inhibits these anti-tumor immune responses in some tumor microenvironments[31]. However, the above experiments did not include GC subtypes in humans, which may play different immune functions. In order to further understand the precise mechanism of TIL-mediated tumor immunity in GC, more experiments are needed in the future.

We first explored the predictive value of NMRGs in GC and created NMRG-related prognostic gene signatures. The shortcoming of this study is that more sample size, experimental mechanism research, and clinical application research are needed for further verification. Still, our current results lay a reliable foundation for future clinical research. We will continue to study and pay attention to the role of these genes.

Conflicts of interest

The authors declare that there are no conflicts of interest.

Acknowledgements

We thank the TCGA and KEGG databases for providing meaningful data sets. This work was sponsored by Tianjin Key Medical Discipline (Specialty) Construction Project (TJYXZDXK-035A) and Tianjin Science and Technology Project (21JCYBJC01590).

Ethics Approval and Consent to Participate

Ethical approval was obtained. Informed consent from the patient was obtained.

Declare

We used TCGA database retrieval, limma screening, K-M curve analysis, univariate and multivariate Cox regression analysis, pCR and other research analysis methods in line with relevant regulations, and strictly in accordance with relevant requirements.

The datasets and analyzed during the current study are available in the TCGA (https://portal.gdc.cancer.gov), cbioportal (https://www.cbioportal.org/) and GEO repository (https://www.ncbi.nlm.nih.gov/geo/) (GSE84437). All data generated during this study are included in this published article and Supplementary Information files.

Statement of author contributions

Yu Xing and Zili Zhang designed the research plan, compiled and analyzed the data, and drafted the manuscript. Wenqing Gao and Weiliang Song analyzed the data and completed the manuscript，Tong li designed and supervised the research plan. All authors participated in the writing of the paper and finally approved the submitted and published version.

Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, Bray F: Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J Clin 2021, 71(3):209-249.
Wagner AD, Syn NL, Moehler M, Grothe W, Yong WP, Tai BC, Ho J, Unverzagt S: Chemotherapy for advanced gastric cancer. Cochrane Database Syst Rev 2017, 8:CD004064.
Okabe K, Yaku K, Tobe K, Nakagawa T: Implications of altered NAD metabolism in metabolic disorders. J Biomed Sci 2019, 26(1):34.
Xie N, Zhang L, Gao W, Huang C, Huber PE, Zhou X, Li C, Shen G, Zou B: NAD(+) metabolism: pathophysiologic mechanisms and therapeutic potential. Signal Transduct Target Ther 2020, 5(1):227.
Rajman L, Chwalek K, Sinclair DA: Therapeutic Potential of NAD-Boosting Molecules: The In Vivo Evidence. Cell Metab 2018, 27(3):529-547.
Sharif T, Martell E, Dai C, Ghassemi-Rad MS, Kennedy BE, Lee PWK, Gujar S: Regulation of Cancer and Cancer-Related Genes via NAD(.). Antioxid Redox Signal 2019, 30(6):906-923.
Zhang H, Zhang N, Liu Y, Su P, Liang Y, Li Y, Wang X, Chen T, Song X, Sang Y et al: Epigenetic Regulation of NAMPT by NAMPT-AS Drives Metastatic Progression in Triple-Negative Breast Cancer. Cancer Res 2019, 79(13):3347-3359.
Lee J, Kim H, Lee JE, Shin SJ, Oh S, Kwon G, Kim H, Choi YY, White MA, Paik S et al: Selective Cytotoxicity of the NAMPT Inhibitor FK866 Toward Gastric Cancer Cells With Markers of the Epithelial-Mesenchymal Transition, Due to Loss of NAPRT. Gastroenterology 2018, 155(3):799-814 e713.
Qiu L, Tang Q, Li G, Chen K: Long non-coding RNAs as biomarkers and therapeutic targets: Recent insights into hepatocellular carcinoma. Life Sci 2017, 191:273-282.
Sampath D, Zabka TS, Misner DL, O'Brien T, Dragovich PS: Inhibition of nicotinamide phosphoribosyltransferase (NAMPT) as a therapeutic strategy in cancer. Pharmacol Ther 2015, 151:16-31.
Li C, Zhu Y, Chen W, Li M, Yang M, Shen Z, Zhou Y, Wang L, Wang H, Li S et al: Circulating NAD+ Metabolism-Derived Genes Unveils Prognostic and Peripheral Immune Infiltration in Amyotrophic Lateral Sclerosis. Front Cell Dev Biol 2022, 10:831273.
Ritchie ME, Phipson B, Wu D, Hu Y, Law CW, Shi W, Smyth GK: limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res 2015, 43(7):e47.
Wu T, Hu E, Xu S, Chen M, Guo P, Dai Z, Feng T, Zhou L, Tang W, Zhan L et al: clusterProfiler 4.0: A universal enrichment tool for interpreting omics data. Innovation (Camb) 2021, 2(3):100141.
Engebretsen S, Bohlin J: Statistical predictions with glmnet. Clin Epigenetics 2019, 11(1):123.
Ramadan F, Fahs A, Ghayad SE, Saab R: Signaling pathways in Rhabdomyosarcoma invasion and metastasis. Cancer Metastasis Rev 2020, 39(1):287-301.
Hanzelmann S, Castelo R, Guinney J: GSVA: gene set variation analysis for microarray and RNA-seq data. BMC Bioinformatics 2013, 14:7.
Diefenbach J, Burkle A: Introduction to poly(ADP-ribose) metabolism. Cell Mol Life Sci 2005, 62(7-8):721-730.
Zhang WP, Wang Y, Tan D, Xing CG: Cystatin 2 leads to a worse prognosis in patients with gastric cancer. J Biol Regul Homeost Agents 2020, 34(6):2059-2067.
Zhou CL, Su HL, Dai HW: Thrombopoietin is associated with a prognosis of gastric adenocarcinoma. Rev Assoc Med Bras (1992) 2020, 66(5):590-595.
Zhang Y, Yuan Z, Shen R, Jiang Y, Xu W, Gu M, Gu X: Identification of biomarkers predicting the chemotherapeutic outcomes of capecitabine and oxaliplatin in patients with gastric cancer. Oncol Lett 2020, 20(6):290.
Wang FH, Ma XJ, Xu D, Luo J: UPK1B promotes the invasion and metastasis of bladder cancer via regulating the Wnt/beta-catenin pathway. Eur Rev Med Pharmacol Sci 2018, 22(17):5471-5480.
Zheng S, Shi L, Zhang Y, He T: Expression of SNCG, MAP2, SDF-1 and CXCR4 in gastric adenocarcinoma and their clinical significance. Int J Clin Exp Pathol 2014, 7(10):6606-6615.
Alnabulsi A, Cash B, Hu Y, Silina L, Alnabulsi A, Murray GI: The expression of brown fat-associated proteins in colorectal cancer and the relationship of uncoupling protein 1 with prognosis. Int J Cancer 2019, 145(4):1138-1147.
Jiang K, Jiao Y, Liu Y, Fu D, Geng H, Chen L, Chen H, Shen X, Sun L, Ding K: HNF6 promotes tumor growth in colorectal cancer and enhances liver metastasis in mouse model. J Cell Physiol 2019, 234(4):3675-3684.
Lu Y, Wu S, Cui C, Yu M, Wang S, Yue Y, Liu M, Sun Z: Gene Expression Along with Genomic Copy Number Variation and Mutational Analysis Were Used to Develop a 9-Gene Signature for Estimating Prognosis of COAD. Onco Targets Ther 2020, 13:10393-10408.
Ruan R, Chen S, Tao Y, Yu J, Zhou D, Cui Z, Shen Q, Wang S: Retrospective analysis of predictive factors for lymph node metastasis in superficial esophageal squamous cell carcinoma. Sci Rep 2021, 11(1):16544.
Tonello AS, Capelli G, Bao QR, Marchet A, Farinati F, Pawlik TM, Gregori D, Pucciarelli S, Spolverato G: A nomogram to predict overall survival and disease-free survival after curative-intent gastrectomy for gastric cancer. Updates Surg 2021, 73(5):1879-1890.
Wang Y, Song F, Zhang X, Yang C: Mitochondrial-Related Transcriptome Feature Correlates with Prognosis, Vascular Invasion, Tumor Microenvironment, and Treatment Response in Hepatocellular Carcinoma. Oxid Med Cell Longev 2022, 2022:1592905.
Chen J, Yang J, Jiang J, Zhuang Y, He W: Function and subsets of dendritic cells and natural killer cells were decreased in gastric cancer. Int J Clin Exp Pathol 2014, 7(11):8304-8311.
Izawa S, Kono K, Mimura K, Kawaguchi Y, Watanabe M, Maruyama T, Fujii H: H(2)O(2) production within tumor microenvironment inversely correlated with infiltration of CD56(dim) NK cells in gastric and esophageal cancer: possible mechanisms of NK cell dysfunction. Cancer Immunol Immunother 2011, 60(12):1801-1810.
Oya Y, Hayakawa Y, Koike K: Tumor microenvironment in gastric cancers. Cancer Sci 2020, 111(8):2696-2707.

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Construction of a prognostic model for gastric cancer patients based on NAD+ metabolism-related genes and tumor microenvironment analysis

Status:

Version 1

Abstract

Background

Methods

Results

Conclusion

Figures

1. Introduction

2. Materials And Methods

2.1 Data source

2.2 Differential expression and enrichment analysis

2.3 Correlation analysis of high and low score groups

2.4 Construction and validation of a prognostic model

2.5 Independent prognostic analysis and construction of a nomogram

2.6 Functional enrichment analysis and immune infiltration analysis in high- and low-risk groups

2.6 The mRNA expression of prognostic related genes was verified by experiment

3. Results

3.1 Enrichment analysis of 10 DE NMRGs

3.2 Correlation analysis of GSVA

3.3 Construction and validation of a prognostic model

3.4 The risk score and gender were independent prognostic predictors

3.5 Functional enrichment analysis of high and low risk groups

3.6 Relationships between risk score and immune cell

3.7 Expression of prognostic genes

4. Discussion

5. Conclusion

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1