Background
Human papillomavirus (HPV) is the primary etiologic agent of cervical cancer [
1]. However, HPV alone is not sufficient for tumor progression; the clinical manifestation of HPV infection depends on the immune response of the host [
2]. Tumors are recognized by the immune system and their development can be stopped or controlled through a process known as immunosurveillance [
3]. The mucosal epithelium represents the first line of defense against virus invasion. An immature or weakened innate immunity of the uterine cervical epithelium may exacerbate viral infection. Therefore, despite the improvements in vaccines against HPV, more studies are needed to identify new therapeutic inducers for the reinforcement of the innate immune responses against HPV infection in cervical cancer patients.
The NADPH oxidase (NOX) family, the major family of enzymes that catalyze reactive oxygen species (ROS) production, comprises seven members: NOX1–5, dual oxidase (DUOX) 1, and DUOX2 [
4]. ROS induce oxidative stress and diverse inflammatory responses [
5]. Excessive ROS production by NOX homologs as a result of chronic inflammation can also promote proliferative and invasive malignancies [
6]. However, oxidative innate immune defense mechanism mediated by NADPH oxidase family members has been emerged, especially, DUOX plays an important role in host mucosal immunity by producing hydrogen peroxide [
7‐
9]. Host-defense properties of DUOX have also been identified in non-mammalian organisms [
10‐
13]. Homologs of DUOX are found in nearly all multicellular organisms, and DUOX enzymes seem to be evolved to fundamentally serve host immune defense [
14]. DUOX1 and DUOX2 may have unique roles in specific arms of the innate immune response. Nevertheless, the immunologic effect of DUOX in the uterine cervical mucosa, which provides the first line of defense to HPV invasion, especially in cervical cancer, has not yet been investigated.
The present study aimed to investigate whether NOX family members are involved in cervical cancer progression or host immunity in response to cervical cancer. We used data from 307 cervical cancer patients obtained from The Cancer Genome Atlas (TCGA). Indeed, we discovered a prognostic value of DUOX1 and NOX2 expression in cervical cancer patients, and we attempted to elucidate the underlying mechanisms by using bioinformatics analyses, including gene set enrichment analysis (GSEA) and cell-type identification by estimating relative subsets of known RNA transcript (CIBERSORT). Moreover, we analyzed the protein expression of NOX2, DUOX1, and DUOX2 using clinical tissue samples from cervical cancer patients.
Methods
Gene and protein expression profiles
Western blotting
Total protein samples were isolated from frozen liver tissue using RIPA lysis buffer, containing protease and phosphatase inhibitor cocktail (TransLab, #30-04CLI19SSH). Samples were separated in a 10% SDS-polyacrylamide gel electrophoresis and transferred onto nitrocellulose membrane (GE Healthcare Life Sciences, #10600023). After the membranes were blocked in 5% skim milk for 1 h at room temperature, they were incubated with primary antibodies overnight at 4 °C and then with the corresponding secondary antibodies for 1 h at room temperature. All of the primary antibodies gp91-phox antibody (Santa Cruz Biotechnology, #K0817) and β-actin (Cell Signaling, #4970 s) were used at a dilution of 1:1000 except DUOX1 (Santa Cruz Biotechnology, #B2817) (1:500) and DUOX2 (Santa Cruz Biotechnology, #D0317) (1:500). Secondary antibodies were used at 1:2500 dilution. Immunoreactive bands were detected using the enhanced chemiluminescence (ECL) detection system with a PhosphorImager (GE Healthcare). Protein expression levels were normalized to the levels of the β-actin, which was used as a loading control.
Patients samples
Frozen cervical cancer tissue samples were obtained from some of patients with cervical cancer and their controls were obtained from the cohort of the Department of Obstetrics and Gynecology, Chungnam National University Hospital (Daejeon, South Korea) and were analyzed by western blot. In this study, each three normal cervical cancer tissues, early-stage cervical squamous cell carcinoma, advanced-stage cervical squamous cell carcinoma, and endocervical adenocarcinoma tissues deposited with the Human Resources Bank of Korea in Chungnam National University Hospital were used for this study. Authorization for the use of these tissues for research purposes and ethical approval were obtained from the Institutional Review Board of Chungnam National University Hospital (IRB number: 2019-05-087). Written informed consents, which were approved by Institutional Review Board of Chungnam National University Hospital, were received from the entire patients who had provided the tissue.
Functional enrichment analysis
Gene Set Enrichment Analysis (GSEA) was used to assess enrichment of mRNAs associated with Hallmark and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways sets [
15]. GSEA was conducted using the 10% of CESC samples with the most strongly upregulated
DUOX1 and
NOX2 expression and the 10% of samples with the most strongly downregulated
DUOX1 and
NOX2 expression. Enrichment maps were visualized in Cytoscape v.3.5.1 (
www.cytoscape.org). A
p-value of less than 0.05 was considered significant.
Analysis of immune cell subsets from mRNA expression profiles
To quantify the relative abundances of 22 tumor-associated leukocyte subsets in samples from HPV-positive and -negative CESC patients, we utilized the Cell type Identification By Estimating Relative Subsets Of known RNA Transcript (CIBERSORT) method and the LM22 gene signature, which allow for highly sensitive and specific discrimination of hematopoietic cells and were well-designed and validated based on gene expression profiles from Affymetrix Human Genome U133A/Plus2 [
16]. CIBERSORT analysis was conducted using the 10% samples with the most strongly upregulated
DUOX1 and
NOX2 expression and the 10% of samples with the most strongly downregulated
DUOX1 and
NOX2 expression.
Survival analysis
Survival analysis of cervical cancer patients was performed using GEPIA. The cumulative event (death) rate was calculated by the Kaplan–Meier method, using the time from the date of operation to the date of death as the outcome variable. Survival curves stratified by risk factors were compared by log-rank test, with p-values less than 0.05 considered to indicate statistical significance. The median group cutoff was median.
Statistical analysis
Data were analyzed in Prism version 5.0 (GraphPad Prism Software, La Jolla, CA, USA) and Statistical Package for Social Sciences for Windows version 13.0 (SPSS, Chicago, IL, USA). Distributions between two groups were compared by t-test (or by Kolmogorov-Smirnov test when the expected frequency in any group was less than 5) for continuous variables, and by Chi-square test (or Fisher’s exact test when the expected frequency in any group was less than 5) for categorical variables. Three or more groups were compared by one-way analysis of variance. A p-value of less than 0.05 was considered significant.
Discussion
We tried to identify new therapeutic targets for the reinforcement of immune responses against HPV infection. This study was the first to examine the immunologic role and clinical significance of NADPH oxidase family members in cervical cancer patients. We initially evaluated
DUOX1 and
DUOX2 mRNA levels in the normal ectocervix, endocervix, and vagina (Additional file
4). Interestingly, we found that
DUOX1 and
DUOX2 mRNA levels were dramatically increased in cervical cancer patients infected with HPV 16 (Fig.
1d). DUOX1 and DUOX2 protein expression were also identified in cervical squamous cell carcinoma (Additional file
1). In line with our findings, a previous study reported that
DUOX and
DUOX-derived ROS were upregulated in the respiratory mucosa upon influenza virus infection [
17]. Moreover, in our study, high expression levels of
DUOX1 mRNA were significantly associated with favorable overall survival as well as disease-free survival in cervical cancer patients. Indeed, several studies were reported that the relationship between expression and prognostic effect of DUOX1 depend on the cancer tissue type. For example, DUOX enzymes were first identified in thyroid tissues and were found to be involved in thyroid hormone biosynthesis [
18‐
20]. In thyroid cancer, DUOX1 is upregulated upon radiation, and DUOX1-dependent H
2O
2 production promotes persistent DNA damage and genome instability, which might contribute to cancer development [
21,
22]. In contrast, in the respiratory tract, DUOX1 is mostly expressed in the tracheal and bronchial epithelium [
9], and DUOX1 mRNA and protein are suppressed in lung cancer as a consequence of hypermethylation in the promoter region, and this suppression is associated with poor prognosis [
23‐
25]. Moreover,
DUOX1 expression is low in the gastrointestinal tract and has been detected in the stomach lining [
24,
26]. In gastric cancer, mRNA expression of
DUOX1 was downregulated, whereas, high levels of
DUOX1 mRNA were correlated with poor prognosis, paradoxically [
27]. It is conceivable that the expression and prognostic effect of DUOX1 depend on the organ and cancer type.
The role of DUOX2 has been actively investigated in various malignancies [
6,
23]. DUOX2 is the main isoform within the gastrointestinal tract and is expressed most prominently within the colon epithelium and rectal glands [
9,
28]. It has been reported that strong DUOX2 expression accelerates the development of colorectal and pancreatic cancers in patients with inflammatory bowel disease and chronic pancreatitis, respectively [
6]. Overexpression of DUOX2/DUOX2A during ulcerative colitis is also thought to be responsible for oxidative DNA damage, which predisposes these patients to colon cancer development [
29]. However, in our study,
DUOX2 mRNA was detected in the vagina, and rarely detected on the cervix (Additional file
4). DUOX2 mRNA was also dramatically increased in cervical cancer patients; however, high
DUOX2 mRNA level was not associated with significant favorable prognosis. Moreover,
NOX2 mRNA was rarely detected on the cervix and vagina (Additional file
4). However,
NOX2 mRNA was significantly increased in cervical cancer patients with HPV, and high
NOX2 mRNA level was significantly associated with favorable overall survival. NOX2 protein expression were also identified in cervical squamous cell carcinoma and adenocarcinoma (Additional file
1). Indeed, it has been indicated that high levels of
NOX2 mRNA are implicated in promoting oncogenic characteristics in breast cancer, rectal cancer, and prostate cancer [
30‐
32].
We conducted GSEA to verify the effects of
DUOX1 and
NOX2 on survival in cervical cancer patients. Notably, expression of both
DUOX1 and
NOX2 was significantly associated with immune pathways related to IFN-alpha and IFN-gamma. IFN is well known to be important for tumor suppression because it not only directly kills tumor cells, but also activates immune cells in the tumor microenvironment [
33]. In addition, estrogen response and NK cell signaling pathways were closely related to
DUOX1 expression. Moreover, the pathways of TNF alpha and cytokine–cytokine receptor interaction were closely related to
NOX2 expression (Table
3). The effects of
DUOX1 and
NOX2 on survival in cervical cancer patients depend commonly on IFN-alpha and IFN-gamma, and differential pathways of
DUOX1 and
NOX2 were identified.
We investigated IHC staining of DUOX1 and NOX2 in cervical cancer tissues based on data from the Human Protein Atlas. Specifically, we discovered that DUOX1 and NOX2 staining in uterine cervical glands and intraepithelial infiltrating cells in cervical cancer tissues. These findings are supported by several recent reports on the presence of DUOX1 in non-epithelial cell types such as T-cells [
34], macrophages [
35], and innate lymphoid cells [
36], and the presence of NOX2 in phagocytes [
37]. To investigate the immune cell types regulated by
DUOX1 and
NOX2 mRNA expression in cervical cancer tissues more specifically, we utilized CIBERSORT analysis. Notably, high mRNA levels of
DUOX1 were closely related with increased innate immune cells, especially, NK cells, monocytes, dendritic cells, and mast cells, and also with a decreased fraction of adaptive immune cells, including B cells, CD8+ T, and CD4+ T cells. This indicates that
DUOX1 expression is highly associated with the innate immune cell response in cervical cancer. Recent evidence indicates that DUOX1 is expressed in innate lymphoid cells, where it has potential roles in innate lymphoid cell polarization, indicating broad host defense functions of DUOX1 [
36]. Moreover, the patients with high mRNA expression levels of
NOX2 were closely related with increased fractions of M1/M2 macrophages and neutrophils among innate immune cells. In addition, the patients with high mRNA expression levels of
NOX2 mRNA levels were related with increased percentages of CD8+ T cells and follicular helper T cells among adaptive immune cells. These findings indicate that NOX2 expression is not only associated with phagocytes, such as macrophages and neutrophils [
37], but also with adaptive immune cells, including CD8+ and follicular helper T cells. Based on GSEA and CIBERSORT analysis, it is suggested that
DUOX1 and
NOX2 have differential effects on the immune cell-mediated response in cervical cancer patients. In the tumor microenvironment, different types of infiltrating immune cells, including macrophages, dendritic cells, mast cells, NK cells, B cells, and effector T cells have diverse effects on cancer progression [
38]. Especially, NK cells collaborate with dendritic cells to induce an immune response against viral infections and tumors [
39]. Activated dendritic cells also play an important role in tumor therapy by acting as natural adjuvants, and tumor-specific follicular helper T cells act as potent antigen-presenting cells [
40,
41]. In addition, an increased population of mast cells was related with favorable prognosis [
42]. In this study, the increased mRNA levels of
DUOX1,
DUOX2, and
NOX2 in cervical cancer were identified in TCGA and GEO databases. Moreover, the protein expression and their localization of DUOX1 and NOX2 were also confirmed in our own patient samples and Human Protein Atlas database, respectively. However, analyses presented here are mainly suggested on the basis of different databases and there was still a challenge to experimentally validate the proposed underlying mechanism in a large cohort of cervical cancer patients.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.