DDX19A Promotes Metastasis of Cervical Squamous Cell Carcinoma by Inducing NOX1-Mediated ROS Production

The major obstacle to treat cervical squamous cell carcinoma (CSCC) is the high prevalence of metastasis, which severely affects 5-year survival rate and quality of life for cancer patients. The DEAD-box helicase family has been reported to be a critical mediator in the development and metastasis of various cancers. DEAD-box helicase 19A (DDX19A) is a member of the DEAD-box helicase family; however, its functional role in CSCC is unclear. In this study, bioinformatics analysis of clinical samples from public databases demonstrated that the expression of DDX19A was elevated in CSCC tissues and that high expression of DDX19A was positively correlated with metastasis and poor clinical outcome. Functionally, we found that DDX19A promoted CSCC cell migration and invasion in vitro and lung metastasis in vivo. Mechanistically, overexpression of DDX19A increased NADPH oxidase 1 (NOX1) expression, enhanced reactive oxygen species (ROS) production, and induced the migration and invasion of CSCC cells. Rescue experiments revealed that DDX19A-induced CSCC functional alterations were dependent on NOX1 and that DDX19A-promoted CSCC metastasis was abrogated upon the inhibition of ROS. Our results demonstrated that DDX19A could promote CSCC metastasis by inducing NOX1-mediated ROS production and that blockage of the NOX1/ROS axis might serve as a potential therapeutic target for patients with DDX19A-overexpressed CSCC.


INTRODUCTION
Cervical cancer is one of the most common gynecological malignancies and the second leading cause of cancer-related deaths among women worldwide (1). Thanks to advances in surgery, chemotherapy, radiotherapy, and immunotherapy, the 5-year survival rate for patients with cervical cancer has increased to 80%. However, once local metastasis or distant metastasis occurs, the 5-year survival rate slumps to ∼50% (2). Metastasis is the primary cause of death in patients with cervical cancer (3,4), and cervical squamous cell carcinoma (CSCC) is the most common pathologic type of cervical cancer, accounting for ∼80% of all cases (5). Thus, it is imperative to understand the underlying mechanism regarding the metastasis of CSCC and to identify novel targets and therapies.
The DEAD-box family, characterized by the conserved Asp-Glu-Ala-Asp (DEAD) motif, represents a large group of RNA helicases consisting of 37 members (6,7). DEAD-box proteins are involved in various RNA metabolic processes including transcription, RNA transport, and RNA degradation (8)(9)(10). Recent studies have revealed that many DEAD-box proteins were abnormally expressed and play pivotal roles in cancer metastasis (11)(12)(13)(14). For example, DEAD-box helicase 39 (DDX39) could promote hepatocellular carcinoma progression and metastasis by activating the Wnt/β-catenin pathway (15), and DDX3 activated CBC-eIF3-mediated translation of uORF-containing oncogenic mRNAs to promote metastasis in head and neck squamous cell carcinoma (16). On the contrary, DDX1 could inhibit ovarian tumor metastasis through regulating primary microRNA maturation (17). However, the expression and function of DEAD-box proteins in CSCC remain unknown.
DDX19A, a member of the DEAD-box family, was identified as a novel cytosolic RNA sensor that could activate the NLRP3 inflammasome during virus infection (18). DDX19A was proven to be associated with NADPH oxidase 1 (NOX1)mediated oxidative stress in tumor necrosis factor (TNF)-αinduced A549 cells (19). The expression and function of DDX19A in tumor development have not been defined. Interestingly, immunohistochemical (IHC) staining data from the Human Protein Atlas website (https://www.proteinatlas.org/) indicate that the expression of DDX19A protein was highest in cervical cancer among all the common cancer types. According to data from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) websites, DDX19A was found highly expressed in CSCC samples compared with normal cervical tissues. High expression of DDX19A was positively correlated with metastasis and poor clinical outcome of patients. Therefore, we attempt to verify the function and investigate the underlying mechanism of DDX19A in the metastasis of CSCC. Our findings may provide a novel therapeutic target of CSCC.

Clinical Tissues
A tissue microarray (TMA) containing 76 pairs of CSCC tissues, non-adjacent normal tissues, and 10 non-paired CSCC tissues was constructed with clinical specimens obtained between January 2008 and February 2016 from the Fifth Affiliated Hospital of Sun Yat-sen University. The diagnosis of CSCC was confirmed by pathologists, and the study was approved by

Immunohistochemical Staining
Seventy-six pairs of CSCC and adjacent normal tissues and 10 non-paired CSCC tissues were used to prepare a TMA. After incubating at 60 • C for 2 h, the tissues were deparaffinized with dimethylbenzene and rehydrated in different concentrations of alcohol. To retrieve antigen, the slides were heated at 95 • C in 0.01 M citrate buffer (pH 6.0), and 3% hydrogen peroxide was used to quench peroxidase activity for 20 min. To avoid non-specific staining, the tissues were incubated with normal goat serum, followed by incubation overnight with anti-human DDX19A (HPA045252, 1:250, Sigma) or anti-human NOX1 (#GTX103888-S, 1:100, GeneTex) at 4 • C. After rinsing with PBS, the tissues were incubated with a secondary antibody for 1 h and stained with 3,3 ′ -diaminobenzidine (DAB; Zhongshan Biotech, Beijing, China). After hematoxylin counterstaining, the sections were dehydrated and sealed. Two experienced pathologists independently evaluated the percentage of positive tumor cells and the staining intensity. The values for DDX19A and NOX1 staining intensity were assigned as follows: 0 (negative), 1 (weak), 2 (moderate), and 3 (strong). The values for the percentage of positive tumor cells were scored as follows: (1) (0-25%), (2) (26-50%), (3) (51-75%), and (4) (76-100%). The immunoreactive score (IRS) for each section was calculated by the product of the staining intensity and the percentage of tumor cells.

Cell Migration and Invasion Assays
Cell migration and invasion assays were performed as previously described (20). Transwell chambers (Corning, Corning, NY, USA) equipped with 8-µm pore insets were used for the migration and invasion assays.

Measurement of Cellular Reactive Oxygen Species
ROS levels in the human CSCC cell lines were measured using the Reactive Oxygen Species (ROS) Assay Kit (Beyotime, China) following the manufacturer's protocol. CSCC cells were incubated with 2 ′ ,7 ′ -dichlorofluorescin diacetate (DCFH-DA) probe at 37 • C for 20 min and washed three times with serum-free DMEM medium. DCFH-DA fluorescence was measured using a Micro Fluorescence Reader with an excitation wavelength of 488 nm and an emission wavelength of 525 nm.

In vivo Metastasis Experiment
BALB/c nude mice (4-5 weeks of age, female, 20-24 g) were purchased from Vital River Lab Animal Technology Co., Ltd., housed under standard conditions at the animal care facility at Guangdong Provincial Key Laboratory of Biomedical Imaging. BALB/c nude mice were used for tail vein injection experiments to evaluate the metastatic ability of CSCC cells in vivo. SiHa cells with DDX19A stable knockdown (1 × 10 6 ) or negative control SiHa cells (1 × 10 6 ), and SiHa cells overexpressing DDX19A or control SiHa cells, were injected into the tail veins of nude mice. After 10 weeks of caudal intravenous injection, the lungs were harvested and weighed, and the number of metastatic nodules was counted on the surface of the lungs. Then, lungs were fixed in 4% formalin and embedded in paraffin, and 5-µm sections were stained with H&E. Animal protocols were approved by the Institutional Animal Care and Use Committee of the Fifth Affiliated Hospital of Sun Yat-sen University.

Bioinformatics and Data Analyses
DDX19A mRNA expression in normal cervical tissues and CSCC tissues were examined according to TCGA (https://gdc. cancer.gov/) and the GEO dataset (GSE7803) (see URL https:// www.ncbi.nlm.nih.gov/geo). Gene expression was presented as the mean value of multiple probes for each gene after log2 transformation. Comparisons (normal cervical tissues vs. CSCC tissues) were analyzed by the Mann-Whitney U-test. p < 0.05 was considered statistically significant.

Statistical Analysis
Results are presented as the mean ± SEM. All statistical analyses were performed using GraphPad Prism 5 software (GraphPad software, Inc., La Jolla, CA). Data from three independent experiments were expressed as the mean ± SD. A value of p < 0.05 was considered to be significant, p < 0.05, p < 0.01, and p < 0.001 are designated as * , * * , and * * * , respectively.   (Figure 1A), suggesting a pivotal role for DDX19A in this gynecological malignancy. Furthermore, microarray analysis [GEO dataset (GSE7803) and TCGA (CSCC data)] revealed that the expression of DDX19A mRNA was upregulated in CSCC samples compared with that in normal cervical tissues (Figures 1B,C). In-depth analysis of TCGA database showed that DDX19A expression was higher in CSCC with distant metastasis (M1) than that in specimens without metastasis (M0) (Figure 1D).
In order to validate these findings, the protein expression of DDX19A was examined in 76 pairs of CSCC and adjacent normal tissues and 10 non-paired CSCC tissues by TMA. Our results showed that CSCC tissues had significantly higher staining intensity of DDX19A compared with the adjacent non-tumor tissues (Figures 1E,F, Supplementary Table 2). To further assess the clinical significance of DDX19A expression, we divided 86 CSCC cases into low-expression (n = 31) group and high-expression (n = 55) group, with the cutoff value defined as the median of the IHC immunoreactive score (IRS = 6). Intriguingly, high expression of DDX19A was associated with lymph node metastasis and larger tumor size ( Figure 1G, Supplementary Table 1). Furthermore, Kaplan-Meier survival plot revealed that patients with high DDX19A expression levels had an unfavorable prognosis relative to those with low DDX19A expression ( Figure 1H). Moreover, TCGA data echoed the above survival correlation (Figure 1I). In summary, these data indicate that DDX19A may play a tumor-promoting role in CSCC.

DDX19A Promotes the Metastasis of Cervical Squamous Cell Carcinoma Cells in vitro and in vivo
Using bioinformatics analysis and confirmatory experiments in clinical specimens, we found that the expression of DDX19A was positively correlated with tumor metastasis. Moreover, the occurrence of metastasis is an important reason for the poor prognosis of CSCC patients. To validate the function of DDX19A in CSCC metastasis, we conducted functional studies both in vitro and in vivo. Firstly, DDX19A expression was identified in various CSCC lines by Western blot (Supplementary Figure 1A). Subsequently, we knocked down DDX19A in two CSCC cell lines, SiHa (Figure 2A) and MS751 (Figure 2B), and overexpressed DDX19A in two cell lines, SiHa ( Figure 2C) and HCC94 (Figure 2D). Transwell migration and cell invasion assays showed that knockdown of DDX19A suppressed the migration and invasion abilities of CSCC cells (Figures 2E,F), whereas, DDX19A upregulation markedly enhanced these functions (Figures 2G,H).
We further investigated whether DDX19A could affect tumor metastasis in vivo. Stable SiHa cell lines with DDX19A knockdown or overexpression were established and verified (Supplementary Figures 1B,C). The pulmonary metastasis nude mouse model was established by caudal intravenous injection of the above cell lines into nude mice. As shown in Figure 2I, compared with the control groups, DDX19A knockdown cells exhibited fewer lung metastases in nude mice, as measured by both the number of nodules and tumor weight. In contrast, overexpression of DDX19A significantly promoted lung metastasis ( Figure 2J). Collectively, these results indicate that DDX19A may promote the metastasis of CSCC cells both in vitro and in vivo.
In addition, epithelial-mesenchymal transition (EMT) has been widely recognized as a critical mechanism of cancer metastasis (21)(22)(23). Thus, we further explored whether DDX19A affected EMT of CSCC cells. As shown in Supplementary Figure 1D, DDX19A knockdown in SiHa cells led to a significantly increased protein level of zonula occludens 1 (ZO-1) and E-cadherin and decreased expression of N-cadherin, Snail, and β-catenin. Therefore, these data indicate that DDX19A promotes the metastatic ability of CSCC cells by inducing EMT.

DDX19A Regulates NOX1 Expression and Enhances Reactive Oxygen Species Production in Cervical Squamous Cell Carcinoma
It has been previously reported that DDX19A participates in the activation of the NOX1 promoter in TNF-α-induced A549 cells (19). NOX1, a member of the nicotinamide adenine dinucleotide phosphate (NADPH) oxidase family, is a key enzyme that regulates redox reactions in the body (24). It has been widely reported that the biological function of NOX1 is to produce ROS (25)(26)(27), and ROS is found to regulate the metastasis through extracellular signal-regulated kinase (ERK) signaling pathway and EMT in a variety of tumors (28)(29)(30). In cervical cancer, ROS can promote cervical cancer metastasis through  the β-catenin-WNT signaling pathway (31) and EMT (32). We assumed that DDX19A promotes the expression of NOX1 increases the production of ROS. Therefore, we measured the expression of NOX1 in DDX19A-knockdown or -overexpressing cells. Both NOX1 mRNA and protein were decreased as the expression of DDX19A was inhibited in SiHa and MS751 (Figures 3A,B) and vice versa (Figures 3C,D).
Next, we examined the expression of NOX1 in CSCC tissues. As shown in Figures 3E,F, CSCC tissues with a higher DDX19A expression exhibited significantly increased staining intensity of NOX1 protein compared with those showing a lower DDX19A expression. A correlation analysis showed that the IHC score of DDX19A was positively correlated with that of NOX1 ( Figure 3G). DDX19A knockdown resulted in reduced  (M) Correlation between lymph node metastasis and DDX19A expression in CSCC patients. Chi-square test was used. (N) Kaplan-Meier analysis was performed for our CSCC patients' cohort to evaluate the association between DDX19A protein level and 86 CSCC patients' overall survival. Results represent three independent experiments (A-J). The results were shown as mean ± SD, ***p < 0.001 by two-tailed Student's t-test.
ROS production in SiHa and MS751 (Figure 3H), and DDX19A overexpression stimulated ROS production in SiHa and HCC94 ( Figure 3I). Overall, these results confirmed that DDX19A regulated NOX1 expression and enhanced ROS production in CSCC.

NOX1 Promotes Metastasis and Reactive Oxygen Species Production in Cervical Squamous Cell Carcinoma Cells and May Serve as a Prognostic Marker in Cervical Squamous Cell Carcinoma Patients
We further explored the effect of NOX1 on the function of CSCC. SiHa and MS751 were generated by transfection with NOX1 siRNA (Figures 4A,B). Inhibition of NOX1 expression arrested the migration and invasion of SiHa and MS751 cells (Figures 4E,F), whereas enhanced NOX1 expression (Figures 4C,D) promoted cell migration and invasion (Figures 4G,H). ROS production followed the same trend as NOX1 (Figures 4I,J). Taken together, these data demonstrated that NOX1 could promote cell migration and invasion and enhanced ROS production in CSCC cells.
To explore the clinical significance of NOX1, we performed IHC staining of NOX1 in CSCC TMA (Figure 4K) and demonstrated that NOX1 expression level was significantly elevated in CSCC tissues compared to adjacent non-tumor tissues ( Figure 4L, Supplementary Table 3). Furthermore, NOX1 expression was associated with tumor size and lymph node metastasis ( Figure 4M, Supplementary Table 4). Additionally, high expression of NOX1 significantly decreased overall survival ( Figure 4N). Collectively, these data show that NOX1 may function as a prognostic marker for CSCC.

The NOX1/Reactive Oxygen Species Axis Exerts a Pro-metastasis Effect Downstream of DDX19A
To further investigate whether NOX1/ROS signaling acts downstream of DDX19A in regulating CSCC metastasis, we performed a series of rescue experiments. Western blot analysis was used to evaluate the efficacy of NOX1 overexpression in DDX19A-knockdown SiHa and MS751 (Figures 5A,B). Overexpression of NOX1 reversed the reduction of cell migration and invasion (Figures 5C,D) as well as ROS production induced by DDX19A knockdown (Figure 5E). In addition, we treated HCC94 and SiHa cells with the ROS inhibitor, N-acetylcysteine (NAC), which effectively abolished ROS production induced by DDX19A overexpression ( Figure 5F). Consequently, NAC treatment reversed the decrease in migration and invasion induced by DDX19A overexpression in HCC94 and SiHa cells (Figures 5G,H). In addition, treating SiHa cells with the ROS inhibitor NAC led to a significantly increased protein level of ZO-1 and E-cadherin and decreased the expression of N-cadherin, Snail, and βcatenin (Supplementary Figures 2A,B), while elevating ROS in DDX19A-knockdown SiHa can decrease the protein level of ZO-1 and E-cadherin and increase the expression of N-cadherin, Snail, and β-catenin (Supplementary Figures 2C,D). These data suggest that the NOX1/ROS axis exerts a metastasis-promoting effect downstream of DDX19A.

DISCUSSION
Tumor metastasis is considered the major cause for treatment failure and death in CSCC. However, the mechanisms underlying CSCC metastasis have not been fully understood. In this study, we identified DDX19A as a novel regulator of CSCC metastasis, which exerts pro-metastasis function via the NOX1/ROS pathway.
DDX19A was identified as a novel cytosolic RNA sensor that bridged porcine reproductive and respiration syndrome virus RNA and NLRP3 to activate the NLRP3 inflammasome (18). Moreover, a meta-analysis showed that a downsized fourgene signature, consisting of DDX19A, FOXM1, KPNA4, and H2AFV, represents a highly significant finding for the biology underlying histological grades in breast cancer, in particular, regarding cell proliferation, and DNA stability (33). However, the expression and function of DDX19A in cancer have not been reported. In this study, data from public databases and the results obtained from our clinical specimens consistently showed increased expression of DDX19A in CSCC, suggesting a tumorpromoting role in CSCC. Functional assays further confirmed that DDX19A exerted tumor-promoting roles in CSCC by stimulating cell metastasis.
A recent study has revealed that DDX19A participates in the activation of the NOX1 promoter in TNF-α-induced A549 cells (19). NOX1, a member of the NADPH oxidase family, is the main source of ROS production (34). Accumulating evidence suggests that NOX1 is involved in the occurrence and development of a variety of cancers (35). For example, NOX1 promoted cell metastasis by mediating ROS production in hepatocellular carcinoma (36). NOX1 regulated colorectal cancer metastasis by modulating the stability of a disintegrin and metalloprotease domain 17 (37). NOX1 could stimulate gastric carcinogenesis by enhancing inflammation or oxygen radical activity (38). However, the expression and function of NOX1 in CSCC remain undefined. We assume that DDX19A promotes the expression of NOX1 and thus enhances the production of ROS. Our results showed that DDX19A affected NOX1 expression in CSCC cell lines and that DDX19A expression was found positively correlated with NOX1 expression in CSCC specimens. DDX19A could promote migration and invasion of CSCC cells using both gain-and loss-of-function assays. Furthermore, overexpression of NOX1 reversed the influence of DDX19A downregulation on metastasis in vivo. Taken together, our data demonstrated that DDX19A could accelerate migration and invasion of CSCC cells in vitro and promote cancer cell metastasis in vivo via enhancing NOX1 expression. However, the underlying molecular mechanism of interaction between DDX19A and NOX1 in CSCC remains unknown, which is worthy of a future study.
As a product of NOX1, ROS levels were essential regulators in cancer growth, metastasis, and other malignant behaviors (39). Numerous researches have shown that ROS can affect the invasion and metastasis of different tumors through different ways. For example, in colon cancer, chloride intracellular channel 1 can regulate cancer cell migration and invasion through ROS-ERK-matrix metalloproteinase 2 (MMP2) pathway (40). In breast cancer, benzo[a]pyrene exposure leads to cancer cell migration and invasion through ROS-ERK-MMP9 axis signaling (30). In addition, ROS can influence the invasion, and migration of tumor cells in lung cancer (41), gastric cancer (28), and liver cancer (29) through the EMT process. In cervical cancer, ROS has been confirmed to regulate cervical cancer cell invasion and metastasis through induction of EMT, activation of Nrf2, and Wnt/β-catenin signaling (42)(43)(44). In this study, a novel regulator of ROS production, DDX19A, was identified. DDX19A knockdown reduced ROS production in SiHa and MS751 cells, while DDX19A overexpression resulted in the opposite effects in HCC94 and SiHa cells. Moreover, NAC treatment blocked the promoting effects of DDX19A on cell migration and invasion. Therefore, the regulatory effect of DDX19A on ROS production is responsible for the capabilities of cellular migration and invasion in CSCC cells with DDX19A overexpression.
In summary, the current study demonstrated that DDX19A contributed to metastasis of CSCC by inducing NOX1-mediated ROS production. Moreover, DDX19A/NOX1 may represent biomarkers of metastasis and novel therapeutic targets in CSCC patients.

DATA AVAILABILITY STATEMENT
The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/Supplementary Material.

ETHICS STATEMENT
The studies involving human participants were reviewed and approved by the Institutional Review Board of the Fifth Affiliated Hospital of Sun Yat-sen University (2020K78-1). The patients/participants provided their written informed consent to participate in this study. The animal study was reviewed and approved by Animal Experimentation Ethics Committee of the Fifth Affiliated Hospital of Sun Yat-sen University (Permission Number: 00083).

AUTHOR CONTRIBUTIONS
YJ, YL, BW, SC, HY, and JS performed the experiments. YW and FZ analyzed the data. CL and FW contributed reagents/materials/analysis tools. YJ wrote the paper. JL, YC, and HH conceived and designed the experiments. All authors read and approved the final manuscript.