Nicotinamide adenine dinucleotide phosphate (NADPH) oxidase complex-derived reactive oxygen species (ROS) promote chronic liver inflammation and remodeling that can drive hepatocellular carcinoma development. The role of NOX expression in hepatocellular carcinoma (HCC) has been partially investigated; however, the clinical relevance of collective or individual NOX family member expression for HCC survival remains unclear. Here, we obtained NOX mRNA expression data for 377 HCC samples and 21 normal liver controls from the TCGA data portal and performed Kaplan-Meier survival, gene ontology functional enrichment, and gene set enrichment analyses. Although most NOX genes exhibited little change, some were significantly induced in HCC compared to that in normal controls. In addition, HCC survival analyses indicated better overall survival in patients with high NOX4 and DUOX1 expression, whereas patients with high NOX1/2/5 expression showed poor prognoses. Gene-neighbour and gene set enrichment analyses revealed that NOX1/2/5 were strongly correlated with genes associated with cancer cell survival and metastasis, whereas increased NOX4 and DUOX1 expression was associated with genes that inhibit tumour progression. On the basis of these data, NOX family gene expression analysis could be a predictor of survival and identify putative therapeutic targets in HCC.
Persistent inflammation can initiate and accelerate tumorigenesis1. The liver functions to metabolise various endogenous and exogenous substances; however, this constant exposure results in sustained chronic inflammatory stimulation and damage that leads to hepatitis and subsequent remodeling. When left uncontrolled, these processes can cause portal fibrosis, liver cirrhosis, and ultimately liver cancer2, 3. Among the primary hepatic malignancies, hepatocellular carcinoma (HCC) is a clear example of inflammation-related cancer, as most cases result from liver damage incurred after prolonged inflammation in the cirrhotic liver4. Therefore, it is essential to identify the fundamental factors in the inflammatory cascade regulating the transition from chronic hepatic injury to dysplastic or regenerative nodules or HCC tumours, which may serve as therapeutic targets and prognostic markers in HCC5.
Several inflammatory factors may influence HCC prognosis and survival. In this regard, several studies have suggested a role for the NADPH oxidases (NOX) complex, which is mainly involved in the generation of reactive oxygen species (ROS). The mammalian NOX complex family comprises seven paralogues: NOX1-5 and dual oxidase 1/2 (DUOX1/2)6, 7. NOX-derived ROS are central factors in oxidative stress and related reduction-oxidase signalling incongruity involved in HCC initiation and development and therefore, considered to be oncogenic factors8, 9; however, the functional significance of NOX family members varies by tumour origin10. Several NOX studies have suggested close associations between the expression of specific NOX members and HCC prognosis, as well as angiogenesis and metastasis as in other malignancies10,11,12,13. For instance, high NOX1 expression has unfavourable effects on recurrence-free survival, whereas increased NOX4 expression has beneficial effects on both recurrence-free survival and overall survival on HCC12. Moreover, another study reported that patients with high DUOX1 expression had longer recurrence-free and overall survival on HCC13. Analysing the effects of the NOX family members on HCC prognosis requires assessing the entire family of NOX genes, using both tumour and adjacent non-tumour tissues and the clinicopathological information of the patients with HCC. However, no study has investigated the effects of the entire family of NOX family genes on HCC prognosis to date.
The present study analysed the expression of all NOX family members in tissue samples from 377 patients with HCC and 21 normal liver tissue samples to determine whether they could be used to predict prognosis. We also analysed the expression of related genes using The Cancer Genome Atlas (TCGA), Gene Set Enrichment Analysis (GSEA), and Database for Annotation, Visualization and Integrated Discovery (DAVID). Significantly, our results validated the relevance of the expression of each NOX family gene to overall survival in the clinical treatment of HCC.
NOX family mRNA expression in liver hepatocellular carcinoma (LIHC)
Figure 1 shows the mRNA expression levels of NOX family members in LIHC. Interestingly, NOX1/4/5 and DUOX1/2 expression was higher in LIHC as compared to that in normal controls, whereas NOX2 and NOX3 expression levels were unchanged. Moreover, tumour NOX4 and NOX2 expression was markedly higher and lower, respectively, compared to that in the normal controls, the NOX2 mRNA level was the highest among the NOX family genes. On the other hand, NOX3 expression was undetectable (Fig. 1A and B). NOX family gene expression showed no significant gene alterations overall (Table 1 and Supplementary Fig. 1). Although no marked differences were observed between sexes, NOX1 and NOX4 expression was significantly higher in male and younger patients, respectively (Supplementary Fig. 2). With respect to tumour stage, DUOX2 expression was significantly higher in advanced TNM stage tumours, however, no significant difference was observed between histological stages (Supplementary Fig. 3).
Effect of NOX family gene expression on LIHC patient survival
Table 2 summarises the clinicopathological data of the patients with LIHC enrolled in this study. To determine the prognostic significance of the NOX family genes in patients with LIHC, we examined the correlations between expression of NOX family members and overall survival of the patients. Initially, Kaplan-Meier curves were used to plot overall survival with mRNA expression, using Cutoff Finder (http://molpath.charite.de/cutoff) (Fig. 2). High expression levels of NOX4 and DUOX1 were significantly associated with a better prognosis (Hazard ratio [HR]: NOX4, 0.37 [95% CI, 0.16–0.84]; DUOX1, 0.69 [95% CI, 0.49–0.99]) (Fig. 2A). However, high NOX1/2/5 expression was significantly associated with a poor prognosis (NOX1, 2.91 [95% CI, 1.35–6.26]; NOX2, 2.59 [95% CI 1.45–4.62]; NOX5, 3.26 [95% CI 1.32–8.07]) (Fig. 2B).
Gene neighbours of the NOX family in LIHC
The 100 genes most correlated with NOX family members were identified using GeneNeighbors and classified according to DAVID-based analyses on biological processes, cellular components, and molecular functions (Supplementary Table 1), of which all three showed significant differences (P < 0.05).
NOX1 gene neighbours highly expressed in LIHC were primarily associated with metabolic processes (NADP, glucose, and glucose-6-phosphate) when analysed according to biological processes (Supplementary Table 1). For cellular components, NOX1 was associated with the cytosol and binding-related molecular functions (protein, NADP, and snRNP binding).
NOX2 gene neighbours highly expressed in LIHC were associated with the inflammatory response, cell adhesion, and signalling pathways (VEGF, B cell receptor, integrin-mediated, and cell surface receptor signalling pathways) when analysed according to biological processes (Supplementary Table 1). For cellular components, NOX2 was associated with the plasma membrane, receptor activity (phosphatidylinositol and tyrosine kinase), and binding-related molecular function (MHC class I and actin).
NOX3 gene neighbours highly expressed in LIHC were mainly associated with the immune response (defence response and innate response) when analysed according to biological processes (Supplementary Table 1). For cellular components, NOX3 was associated with keratin filaments and the extracellular region, and receptor activity (G-protein and transmembrane signalling) for molecular functions.
NOX4 gene neighbours highly expressed in LIHC were mainly associated with angiogenesis and cell adhesion when analysed according to biological processes (Supplementary Table 1). For cellular components, NOX4 was associated with focal adhesion and extracellular components, and binding-related molecular functions (integrin and extracellular matrix).
NOX4 gene neighbours highly expressed in LIHC were mainly associated with angiogenesis, cell adhesion, and various signalling molecules (VEGF, Rho protein, GTPase and MAPK) when analysed according to biological processes (Supplementary Table 1). For cellular components, NOX5 was associated with the extracellular matrix and plasma membrane, and binding-related molecular functions (growth factor, integrin, VEGF-activated receptor, and actin).
DUOX1 gene neighbours highly expressed in LIHC were mainly associated with GTPase activity and the mitotic cell cycle when analysed according to biological processes (Supplementary Table 1). For cellular components, DUOX1 was associated with intracellular cellular components, and for molecular functions, it was associated with binding-related functions (kinesin, GTPase, and NADPH oxidase).
DUOX2 gene neighbours highly expressed in LIHC were mainly associated with differentiation, hydrogen peroxide metabolism, and negative regulation of cell proliferation when analysed according to biological processes (Supplementary Table 1). For cellular components, DUOX2 was associated with the extracellular space and membrane, and binding-related molecular functions (calcium-dependent phospholipid binding, RNA polymerase DNA binding, and CXCR receptor binding).
GSEA of the NOX family in LIHC
GSEA was performed to identify significantly enriched pathways that differed between the high (top 10%) and low (bottom 10%) NOX-expressing groups based on pathways provided in curated gene set enrichment analysis, KEGG, and oncogenic signatures of gene sets that represent the signatures of cellular pathways often dis-regulated in cancer (Figs. 3 and 4, and Supplementary Table 2).
In the high-NOX1 group, the GO terms filament, STAT phosphorylation, mitochondrial translation, and type I interferon, and pathways involving the regulation of autophagy and oxidative phosphorylation, were significantly enriched when compared with those in the low-NOX1 group (Supplementary Table 2).
In the high-NOX2 group, the GO terms immune response, cell activation, and cell adhesion, and pathways involving chemokine, immune response and phagocytosis, were significantly enriched when compared with those in the low-NOX2 group. Moreover, high NOX2 expression was associated with oncologic signatures involving EGFR, RAF, and KRAS (Supplementary Table 2).
In the high-NOX3 group, the GO terms filament, mitochondrial translation, and ribosome, and pathways involving RNA polymerase, ribosome, and oxidative phosphorylation, were significantly enriched when compared with those in the low-NOX3 group (Supplementary Table 2).
In the high-NOX4 group, the GO terms chromatin and DNA replication, and pathways involving the cell cycle, DNA replication, mismatch repair, p53 signalling, and oxidative phosphorylation, were significantly enriched when compared with those in the low-NOX4 group (Supplementary Table 2).
In the high-NOX5 group, the GO terms type I interferon, filament, STAT phosphorylation, and drug metabolic process, and pathways involving autophagy regulation, fatty acid metabolism, peroxisome, and the RIG-1 signalling pathway, were significantly enriched when compared with those in the low-NOX5 group. In addition, increased NOX5 expression was associated with oncologic signatures involving PKCA (Supplementary Table 2).
In the high-DUOX1 group, the GO terms chromatid segregation and cell division, and pathways involving the cell cycle, DNA replication, mismatch repair, and the p53 signalling pathway, were significantly enriched when compared with those in the low-DUOX1 group (Supplementary Table 2).
In the high-DUOX2 group, the GO terms extracellular matrix, metabolic process, leukocyte migration, and Rho protein signal transduction, and pathways involving extracellular matrix receptor interaction, phagocytosis, and the cell cycle, were significantly enriched when compared with those in the low-DUOX2 group (Supplementary Table 2).
Several factors associated with hepatic inflammatory signalling are involved in HCC occurrence and progression, including ROS14. In mammalian cells, physiologically low levels of ROS play a central role in molecular communication by acting as secondary messengers; however, persistent ROS accumulation can induce inflammatory responses, resulting in genetic instability, chromosomal damage, and tumour development and metastasis14,15,16. Because NOX family members are involved in ROS production and carcinogenesis17,18,19, the present study sought to determine whether these factors were significantly correlated with HCC patient survival.
Previous studies have reported that several NOX subtype proteins are tumour-suppressive or serve as favourable prognostic factors in the liver. For example, high DUOX1 expression in HCC is associated with prolonged disease-free and overall survival after radical tumour resection; however, this gene is frequently silenced by promoter hyper-methylation in human HCC tumour tissues and cancer cell lines13. Furthermore, restoration of DUOX1 expression significantly inhibits cancer cell proliferation by inducing G2/M phase arrest20. As such, it was concluded that DUOX1 acts as a tumour suppressor in HCC development. In addition, another study has suggested that NOX1 and NOX4 are useful prognostic biomarkers after HCC resection12. In this study, HCC patients with high NOX1 or low NOX4 expression, as assessed by immunohistochemistry, had worse recurrence-free and overall survival rates12. Therefore, it was suggested that these patients would benefit from adjuvant treatment after surgical tumour resection, although the underlying molecular mechanisms remain unknown. Similarly, Isable et al. reported that NOX4 may act as a tumour suppressor in the liver, based on data from partial hepatectomy and xenograft mouse models and analyses data from human HCC cell lines and liver tumour tissues21. It was concluded that NOX4 in liver tumour cells acts as a growth inhibitor, consistent with its potential role in counteracting growth factor signals and/or inducing senescence21. However, these studies addressed only NOX1/4 and DUOX1. Moreover, only some of these studies discussed the underlying mechanisms involved, especially those pertaining to the cell cycle, and none presented a comprehensive analysis of the underlying mechanisms. Given the lack of information on the relationship between these genes and HCC survival, we investigated the potential of NOX family members as prognostic factors in HCC.
In our study, NOX1/4/5 and DUOX1/2 expression was higher in hepatocellular carcinoma tissues than in control normal liver tissues (Fig. 1). Paradoxically, higher mRNA expression levels of NOX4 and DUOX1 were significantly associated with prolonged overall survival (Fig. 2A), whereas increased NOX1/2/5 expression was significantly associated with a poor overall survival (Fig. 2B). These findings are supported by those of several previous studies in which NOX4 and DUOX1 expression was associated with a favourable prognosis and NOX1 expression was correlated with decreased survival12, 13, 21.
We performed bioinformatics analyses to verify the effect of NOX family expression on cumulative overall survival in HCC. Notably, NOX1 expression was significantly associated with metabolic processes, including those of NADP, glucose, and glucose-6-phosphate. Cancer cells often exhibit accelerated aerobic glycolysis to deal with unfavourable situations, primarily hypoxic conditions, and this can distinguish cancer cells from normal cells22, 23. In addition, NOX2 expression was significantly associated with the inflammatory response and VEGF signalling pathway in HCC, consistent with previous reports that VEGF functions as a direct link between chronic inflammation and tumour progression24. Especially, the oncologic signatures of NOX2 expression correlate with EGFR, RAF, and KRAS activity. Moreover, NOX5 expression was significantly associated with angiogenesis and VEGF signalling, as well as STAT phosphorylation, type I interferon, and autophagy regulation in HCC. The NOX5 oncologic signature was also significantly enriched for PKCA signalling. Our analyses revealed that expression of both NOX1 and NOX5 was associated with increased STAT protein phosphorylation, type I interferon receptor-related gene expression, and the induction of autophagy regulatory genes. Autophagy is a typical survival strategy utilized by tumour cells in unfavourable environments such as hypoxia or hypoperfusion25, 26. This process has been reported to induce JAK2/STAT3 phosphorylation and can contribute to cancer cell survival. Lastly, NOX2 expression was associated with the induction of oncogenic KRAS and RAF signalling, and upregulation of angiogenesis-related genes that contribute to tumour progression, metastasis, and reduced survival rate.
Next, we analysed the genes associated with prolonged survival of HCC patients. A significant increase in NOX4 expression was found in LIHC tissues compared to that in normal controls, paradoxically, however, NOX4 expression was correlated with a better prognosis. Indeed, NOX4 promotes angiogenesis and cell adhesion control27, 28; however, genes associated with pathways involving the cell cycle, DNA replication, mismatch repair, and p53 signalling were also significantly enriched in patients with high NOX4 expression. Previous reports indicate that NOX4 not only promotes angiogenesis via endothelial nitric oxide synthase activation, HIF-1α-mediated angiogenesis, and VEGF expression, but also inhibits the epithelial-to-mesenchymal transition to attenuate liver cancer progression21, 27,28,29. Thus, the functional significance of NOX4 expression in HCC requires further study.
With respect to the dual oxidase enzymes, DUOX1 expression was associated with NADPH oxidase activity and the cell cycle, DNA replication, mismatch repair, and p53 signalling pathways, which were responsible for favourable effects on HCC survival, whereas DUOX2 expression was associated with attenuated cell proliferation, immunological pathways involving Fc-gamma receptor-mediated phagocytosis, and cell cycle regulation, which are associated with a better prognosis. LIHC samples showed higher DUOX1 and DUOX2 expression as compared to non-malignant tissues, and this expression pattern was previously associated with diminished recurrence-free and overall survival30. Unlike our present study, NOX4 expression was not significantly correlated with survival in the previous study30. This discrepancy may be due to the relatively large number of patients enrolled in our study (n = 377 vs. n = 107). Moreover, genes associated with mismatch repair, G2M checkpoint, and the cell cycle were also induced in patients with high DUOX1 expression levels. Similarly, recent genome sequencing and cell cycle analyses demonstrated that DUOX1 overexpression was associated with cell accumulation in the G2/M phase and cell cycle arrest20. In addition, genes associated with leukocyte migration and Fc-gamma receptor-mediated phagocytosis were highly induced in patients with high DUOX2 expression. In the tumour environment, Fc gamma receptor activity elicits antibody-dependent cellular cytotoxicity or phagocytosis31,32,33. According to these data, NOX4 and DUOX1/2 exhibit a tumour suppressive function.
In summary, our data shed light on the clinicopathological mechanisms of NOX family members in HCC, including their diverse roles in inflammatory signalling, angiogenesis, tumour-suppressive and oncogenic gene expression, DNA repair, and the cell cycle control. Thus, NOX genes can positively or negatively affect HCC patient survival through various mechanisms. These findings support the use of a NOX gene expression panel to predict HCC patient survival after tumour resection and suggest that molecular targeting of NOX enzymes may be an effective strategy for HCC therapy.
Gene expression profiles
Level 3 mRNA expression and clinical data from 377 LIHC and 21 normal control samples were obtained from the TCGA data portal, and RSEM_genes_normalised RNA-Seq data were acquired from Firebrowse for analysis of gene expression. All of these datasets downloaded during and analysed during the current study are available in the TCGA data portal (http://tcga-data.nci.nih.gov/tcga) and Firebrowse (http://firebrowse.org/?cohort=LIHC&download_dialog=true).
Analysis of mRNA microarray data
The data from TCGA portal were analysed using R software (v.3.2.5; http://www.r-project.org). The Rank Normalize module in GenePattern (http://broadinstitute.org/cancer/software/genepattern) was used to normalise the chip data. The data for the fold change of NOX family genes in LIHC compared to normal controls were derived from Firebrowse (http://firebrowse.org/viewGene.html?gene=NOX1), where the gene name in the web address for Firebrowse could be substituted by the name of the gene of interest. Illuminahiseq_maseqv2-RSEM_genes_normalised RNA-Seq data were acquired from Firebrowse (http://firebrowse.org/?cohort=LIHC&download_dialog=true) and analysed using GeneNeighbors, a module programmed in GenePattern, to select genes closely related to NOX family genes34. cBioportal (http://www.cbioportal.org/) was used to analyse changes in the expression of NOX family genes in LIHC.
Functional enrichment analysis
First, 100 differentially expressed genes (DEGs) analysed using GeneNeighbors were imported into Database for Annotation, Visualization and Integrated Discovery (DAVID) (http://david.abcc.ncifcrf.gov/) for gene ontology (GO) functional enrichment analyses. After the data for DEGs were submitted, they were converted and analysed using the gene accession conversion tool and functional annotation tool in DAVID, respectively. Gene set enrichment analysis (GSEA) was performed to find enriched mRNAs predicted to have a correlation with pathway in C2, a curated gene set and the Kyoto Encyclopedia of Genes and Genomes (KEGG), and in C5, a gene set that contain genes annotated by the same GO term, and in C6, oncogenic signatures of gene sets that represent the signatures of cellular pathways that are often dis-regulated in cancer. Especially, GO analysis encompassed three domains: biological processes, cellular components, and molecular functions. P < 0.05 was considered to indicate a statistical significance.
Cutoff Finder (http://molpath.charite.de/cutoff) was used to determinate cutoff values for LIHC mRNA expression. Illuminahiseq_maseqv2-RSEM_genes_normalised RNA-seq data of NOX family genes were uploaded from a tab-separated files, and the rows represented patients and columns represented variables (http://molpath.charite.de/cutoff/load.jsp). The cutoff determination in Cutoff Finder was based on survival for significance based on the log-rank test for patient outcome (http://molpath.charite.de/cutoff/assign.jsp). The cumulative event (death) rate was calculated with the Kaplan-Meier method, using the time to the first event as the outcome variable. The criterion for the survival date for statistical analysis was the day from the date of operation to the date of death. Survival curves were compared by the log-rank test for high and low expression groups on each NOX gene family.
Statistical analyses were performed using Prism software (v.5.0; GraphPad Prism Software, La Jolla, CA, USA) and the Statistical Package for Social Sciences for Windows (SPSS 13.0, Inc., Chicago, IL, USA). Distributions were compared between two groups by the t-test (or the Kolmogorov-Smirnov test when the expected frequency within any cell was <5) for continuous variables, and the χ2 test (or Fisher’s exact test when the expected frequency within any cell was <5 for categorical variables) for categorical variables. Distributions of the characteristics among three or more groups were compared by ANOVA. P < 0.05 was considered statistically significant.
Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Bishayee, A. The role of inflammation and liver cancer. Adv. Exp. Med. Biol. 816, 401–435 (2014).
Tu, T. et al. Novel aspects of the liver microenvironment in hepatocellular carcinoma pathogenesis and development. Int J Mol Sci. 15, 9422–9458 (2014).
Shi, J. H. & Line, P. D. Effect of liver regeneration on malignant hepatic tumors. World J Gastroenterol. 20, 16167–16177 (2014).
Li, H. & Zhang, L. Liver regeneration microenvironment of hepatocellular carcinoma for prevention and therapy. Oncotarget. 8, 1805–1813 (2017).
Weber, A., Boege, Y., Reisinger, F. & Heikenwalder, M. Chronic liver inflammation and hepatocellular carcinoma: persistence matters. Swiss Med Wkly. 141, w13197 (2011).
Bedard, K. & Krause, K. H. The NOX family of ROS-generating NADPH oxidases: physiology and pathophysiology. Physiol. Rev. 87, 245–313 (2007).
Lambeth, J. D. NOX enzymes and the biology of reactive oxygen. Nat. Rev. Immunol. 4, 181–189 (2004).
Wondrak, G. T. Redox-Directed Cancer Therapeutics: Molecular Mechanisms and Opportunities. Antioxid Redox Signal. 11, 3013–3069 (2009).
Liou, G.-Y. & Storz, P. Reactive oxygen species in cancer. Free Radic Res. 44, 479–496 (2010).
Roy, K. et al. NADPH oxidases and cancer. Clin Sci. 128, 863–875 (2015).
Ushio-Fukai, M. VEGF signaling through NADPH oxidase-derived ROS. Antioxid Redox Signal. 9, 731–739 (2007).
Ha, S. Y. et al. NADPH Oxidase 1 and NADPH Oxidase 4 Have Opposite Prognostic Effects for Patients with Hepatocellular Carcinoma after Hepatectomy. Gut and Liver. 10, 826–835 (2016).
Chen, S. et al. Dual oxidase 1: A predictive tool for the prognosis of hepatocellular carcinoma patients. Oncol. Rep. 35, 3198–3208 (2016).
Chio, I. I. C. & Tuveson, D. A. ROS in Cancer: The Burning Question. Trends Mol. Med. 23, 411–429 (2017).
Circu, M. L. & Aw, T. Y. Reactive oxygen species, cellular redox systems and apoptosis. Free Radic Biol Med. 48, 749–762 (2010).
Reuter, S., Gupta, S. C., Chaturvedi, M. M. & Aggarwal, B. B. Oxidative stress, inflammation, and cancer: How are they linked? Free radical biology & medicine. 49, 1603–1616 (2010).
Teufelhofer, O. et al. Superoxide generation from Kupffer cells contributes to hepatocarcinogenesis: studies on NADPH oxidase knockout mice. Carcinogenesis. 26, 319–329 (2005).
Calvisi, D. F., Ladu, S., Hironaka, K., Factor, V. M. & Thorgeirsson, S. S. Vitamin E down-modulates iNOS and NADPH oxidase in c-Myc/TGF-alpha transgenic mouse model of liver cancer. J Hepatol. 41, 815–822 (2004).
Choi, J., Corder, N. L., Koduru, B. & Wang, Y. Oxidative stress and hepatic Nox proteins in chronic hepatitis C and hepatocellular carcinoma. Free Radic Biol Med. 72, 267–284 (2014).
Ling, Q. et al. Epigenetic silencing of dual oxidase 1 by promoter hypermethylation in human hepatocellular carcinoma. Am J Cancer Res. 4, 508–517 (2014).
Crosas-Molist, E. et al. The NADPH oxidase NOX4 inhibits hepatocyte proliferation and liver cancer progression. Free Radic Biol Med. 69, 338–347 (2014).
Alfarouk, K. O. et al. Glycolysis, tumor metabolism, cancer growth and dissemination. A new pH-based etiopathogenic perspective and therapeutic approach to an old cancer question. Oncoscience. 1, 777–802 (2014).
Hay, N. Reprogramming glucose metabolism in cancer: can it be exploited for cancer therapy? Nat. Rev. Cancer. 16, 635–649 (2016).
Waldner, M. J. et al. VEGF receptor signaling links inflammation and tumorigenesis in colitis-associated cancer. J. Exp. Med. 207, 2855–2868 (2010).
Yoon, S. et al. STAT3 transcriptional factor activated by reactive oxygen species induces IL6 in starvation-induced autophagy of cancer cells. Autophagy. 6, 1125–1138 (2010).
Yoon, S. et al. NF-kappaB and STAT3 cooperatively induce IL6 in starved cancer cells. Oncogene. 31, 3467–3481 (2012).
Helfinger, V. et al. The NADPH Oxidase Nox4 mediates tumour angiogenesis. Acta Physiol (Oxf). 216, 435–446 (2016).
Crosas-Molist, E. et al. The NADPH oxidase NOX4 represses epithelial to amoeboid transition and efficient tumour dissemination. Oncogene. (2016).
Craige, S. M. et al. NADPH oxidase 4 promotes endothelial angiogenesis through endothelial nitric oxide synthase activation. Circulation. 124, 731–740 (2011).
Lu, C. L. et al. NADPH oxidase DUOX1 and DUOX2 but not NOX4 are independent predictors in hepatocellular carcinoma after hepatectomy. Tumour Biol. 32, 1173–1182 (2011).
Stewart, R., Hammond, S. A., Oberst, M. & Wilkinson, R. W. The role of Fc gamma receptors in the activity of immunomodulatory antibodies for cancer. Journal for ImmunoTherapy of Cancer. 2, 29 (2014).
Wilson, N. S. et al. An Fcgamma receptor-dependent mechanism drives antibody-mediated target-receptor signaling in cancer cells. Cancer Cell. 19, 101–113 (2011).
Bulliard, Y. et al. Activating Fc gamma receptors contribute to the antitumor activities of immunoregulatory receptor-targeting antibodies. J. Exp. Med. 210, 1685–1693 (2013).
Golub, T. R. et al. Molecular classification of cancer: class discovery and class prediction by gene expression monitoring. Science. 286, 531–537 (1999).
This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government, Ministry of Science, ICT and Future Planning. (NRF-2017R1C1B1004924).
Electronic supplementary material
About this article
Journal of Experimental & Clinical Cancer Research (2019)
Clinical Significance of the Thioredoxin System and Thioredoxin-Domain-Containing Protein Family in Hepatocellular Carcinoma
Digestive Diseases and Sciences (2019)