The adaptor protein TNF receptor-associated factor 6 (TRAF6) is a key mediator in inflammation. However, the molecular mechanisms controlling its activity and stability in cancer progression remain unclear. Here we show that death-associated protein kinase-related apoptosis-inducing kinase 1 (DRAK1) inhibits the proinflammatory signaling pathway by targeting TRAF6 for degradation, thereby suppressing inflammatory signaling-mediated tumor growth and metastasis in advanced cervical cancer cells. DRAK1 bound directly to the TRAF domain of TRAF6, preventing its autoubiquitination by interfering with homo-oligomerization, eventually leading to autophagy-mediated degradation of TRAF6. Depletion of DRAK1 in cervical cancer cells resulted in markedly increased levels of TRAF6 protein, promoting activation of the IL1β signaling-associated pathway and proinflammatory cytokine production. DRAK1 was specifically underexpressed in metastatic cervical cancers and inversely correlated with TRAF6 expression in mouse xenograft model tumor tissues and human cervical tumor tissues. Collectively, our findings highlight DRAK1 as a novel antagonist of inflammation targeting TRAF6 for degradation that limits inflammatory signaling-mediated progression of advanced cervical cancer.

Significance:

Serine/threonine kinase DRAK1 serves a unique role as a novel negative regulator of the inflammatory signaling mediator TRAF6 in cervical cancer progression.

Cervical cancer is a major cause of mortality in women worldwide and strongly associated with persistent infection with high-risk (HR) human papillomavirus (HPV), mainly HPV16 and HPV18 (1, 2). Most HPV viruses can be cleared by the immune system, and a preventive and effective vaccine against HPV infection decreases the development of cervical cancer (3–5). However, approximately 1% of HPV-positive patients infected with HR-HPV eventually develop advanced cervical cancer from premalignant lesions (6), indicating that not all HPV infections progress into cervical cancer. In this regard, the different molecular features of cervical cancer progression have not been extensively investigated, and the identification of new therapeutic strategies and biomarkers for advanced cervical cancer is urgently needed.

Inflammation is a hallmark of cancer progression. Notably, the excessive or prolonged production of proinflammatory cytokines serves a crucial role in advanced cancer progression (7–10). Furthermore, inflammatory responses are highly associated with high-grade tumor lesions, and the hyperactivation of NF-κB, a proinflammatory transcription factor, is also observed in many types of cancer, which is correlated with tumor grade and aggressive behavior (11–14). Many studies have established that TNF receptor-associated factor 6 (TRAF6) is central to the activation of the NF-κB–mediated signaling pathway and that its overexpression or dysregulation is closely associated with inflammation-induced tumorigenesis (15–18). TRAF6, which functions as an E3 ubiquitin ligase, is composed of an N-terminal ring finger domain that confers ubiquitin ligase activity and a coiled-coil domain and a C-terminal TRAF domain, which are needed for oligomerization and adaptor function (19). The autoubiquitination of TRAF6 by its homo-oligomerization is a critical regulatory event that activates the downstream NF-κB signaling pathway, which in turn induces the expression of proinflammatory cytokines. Although the role of TRAF6 in inflammation and cancer progression has been extensively studied, regulators that control the function of TRAF6 remain to be identified.

Death-associated protein (DAP) kinase-related apoptosis-inducing protein kinase 1 (DRAK1), which is also known as STK17A, is a member of the DAP kinase (DAPK) family and a serine/threonine protein kinase and proapoptotic protein kinase that induces morphologic changes associated with apoptosis (20). However, the functional role of DRAK1 in cancer has not been well characterized, and its role in regulating apoptosis is still controversial. DRAK1, a p53 target gene, enhances cisplatin-induced cell death in testicular cancer cells (21) and was identified as a regulator of chemotherapeutic resistance in melanoma, pancreatic cancer, and ovarian cancer cells (22–24). Furthermore, considerable attention has recently been paid to the relevance of DRAK1 expression in colorectal cancer progression and metastasis. The loss of DRAK1 increases the invasive potential of colorectal cancer cells by inducing partial epithelial–mesenchymal transition (EMT), whereas its overexpression maintains the epithelial phenotype (25). Although it was reported that dysregulation of DRAK1 contributes to tumor growth in head and neck cancers and gliomas (26, 27), most studies suggest DRAK1 as a tumor suppressor and regulator of chemotherapeutic resistance. However, despite these findings, the precise role of DRAK1 in cancer progression and metastasis, especially in the context of cervical cancers, has not been extensively investigated.

In this study, we show that DRAK1 specifically decreases the stability of the TRAF6 protein via an autophagy-mediated degradation pathway by interfering with the homo-oligomerization of TRAF6, eventually suppressing tumor growth and metastasis in inflammation-associated advanced cervical cancer cells.

Cell culture and reagents

The human cervical cancer cell lines (SiHa, CaSki), breast cancer cell lines (MCF7, MDA-MB-231), and colon cancer cell lines (SW620, HCT116) were purchased from the ATCC, and C33A, HeLa, HeLa229, and ME180 cells were obtained from the Department of Otorhinolaryngology (Seoul National University College of Medicine, Seoul, South Korea). HeLa, MCF7, MDA-MB-231, SW620, and HCT116 cells were grown in DMEM (WelGENE) supplemented with 10% FBS and 1% penicillin/streptomycin, and C33A, CaSki, SiHa, HeLa229, and ME180 cells were grown in RPMI1640 medium (WelGENE) containing 10% FBS and 1% penicillin/streptomycin. The cell lines in this study were routinely tested for Mycoplasma contamination by PCR. Cycloheximide (CHX; C7698), MG-132 (C2211), 3-MA (M9281), and N-ethylmaleimide (NEM; E3876) were purchased from Sigma-Aldrich. IL1β was obtained from PeproTech. Primer sequences, shRNA sequences, and primary antibodies used in this study are listed in Supplementary Tables S1 to S3.

Yeast two-hybrid screening

The DRAK1 sequence was cloned into the EcoRI and BamHI sites of the pGBKL vector. The bait DRAK1 plasmid (pGBKL-DRAK1) was transformed into yeast strain PBN204. Transformants containing the DRAK1 bait plasmid were “mated” with a pretransformed HeLa human cervical cancer cell cDNA library (Clontech). Two-hybrid screening was performed according to the manufacturer's protocol. After positive clones were grown in minimal medium lacking tryptophan, leucine, adenosine, histidine, and β-galactosidase (β-gal) to induce expression, plasmids were “rescued,” and the junction between the GAL4 DNA-binding domain and the DRAK1 gene was confirmed by DNA sequencing.

Human cervical cancer tissue microarray and IHC

Human cervical cancer tissue microarrays (TMA) were purchased from US Biomax (CR208) and collected at the Seoul National University College of Medicine in compliance with all relevant ethical regulations regarding research involving human participants after approval by the Institutional Review Board (IRB approval no. 1707-063-869). All volunteers gave their written informed consent to participate in this study, according to the Declaration of Helsinki. For IHC analysis, TMA sections were fixed and stained using anti-DRAK1 (Abcam, ab111963) and anti-TRAF6 (Santa Cruz Biotechnology, SC-8409) antibodies and counterstained with hematoxylin. After staining, DRAK1 and TRAF6 levels were scored by microscope and analyzed.

In vivo tumor formation and lung metastasis

All animal experiments procedures were approved by the Woo Jung Bio animal facility (Suwon, Korea). For the tumor formation assay, DRAK1-overexpressing SiHa and DRAK1-depleted CaSki cells (1 × 107) were resuspended in a 1:3 PBS/hydrogel (The Well Bioscience) solution and subcutaneously injected into 6-week-old female NOD/ShiLtJ-Rag2em1AMC Il2rgem1AMC (NRGA) mice. Tumor dimensions were measured once per week. The mice were sacrificed 6 weeks after injection, and the tumors were surgically isolated. The tumor volume (V) was calculated using the formula V = (S × S × L) × 0.5, where S and L were the short and long dimensions of the tumor, respectively. To analyze lung metastasis, DRAK1-overexpressing SiHa and DRAK1-depleted CaSki cells (1 × 106) were injected into the lateral tail veins of NRGA mice. After 12 weeks, the mice were sacrificed, and their lungs were fixed with 10% formalin. All tumor and lung tissues were embedded in paraffin for hematoxylin and eosin staining and IHC. All of the animals were approved and maintained according to the Woo Jung Bio Facility (Suwon, Korea) and Use Committee guidelines under protocol number IACUC110004.

RNA sequencing

Total RNA from each cell for RNA sequencing was isolated using TRIzol reagent following the manufacturer's instructions. The total RNA samples were treated with DNase I, purified with an miRNeasy Mini Kit (Qiagen) and subsequently examined for quality using an Agilent 2100 bioanalyzer (Agilent). An Illumina platform (Illumina) was used to analyze transcriptomes with a 90 bp paired-end library. Samples were pair-end sequenced with the Illumina HiSeq 2000 platform using HiSeq Sequencing kits.

Bioinformatics analysis

After sequencing, low-quality reads were filtered out according to the following criteria: reads containing more than 10% skipped bases (marked as “N”s), reads with more than 40% bases with quality scores less than 20, and reads in which the average quality score of each read is less than 20. The whole filtering process was performed using scripts developed in-house. Filtered reads were mapped to the reference genome related to the species using STAR v.2.4.0b alignment software (28). Gene expression levels were measured with Cufflinks v2.1.1 (29) using the ensemble database and quantified as the ratio of reads mapped to a gene to the gene length in kilobases and expressed as the fragments per kilobase of transcript per million fragments mapped. Noncoding gene regions were excluded from gene expression measurement. To improve the accuracy of the measurements, the multiread correction and frag-bias-correct options were applied. All other options were set to their default values. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses were performed by using the Database for Annotation, Visualization, and Integrated Discovery (DAVID) tool (http://david.abcc.ncifcrf.gov) and the KEGG orthology-based annotation system online tool (http://geneontology.org) with cut-off values of P < 0.01. String (www.string-db.org) was used to generate protein–protein interaction (PPI) networks.

Quantification and statistical analyses

Statistical significance was calculated by using GraphPad Prism 5. The significance of predicting overall survival in patients with cervical cancer was analyzed by using the log-rank Mantel–Cox test. For all other comparisons, the two-tailed unpaired Student t test was used, and P < 0.05 indicated statistical significance.

DRAK1 decreases the stability of TRAF6 through its interaction with the TRAF domain

To identify novel DRAK1-binding proteins, we performed a yeast two-hybrid screening assay. Through sequence analysis of positive transformants, we obtained eight positive clones and selected to TRAF6. To confirm the interaction between DRAK1 and TRAF6 in mammalian cells, 293T cells were transiently cotransfected with Myc-DRAK1 and Flag-TRAF6 plasmids. Indeed, immunoprecipitation assays showed an interaction between these two proteins (Fig. 1A). Most interestingly, the expression level of TRAF6 in the whole cell lysate (WCL) was markedly decreased in the presence of DRAK1. This finding was supported by a dose-dependent decrease in TRAF6 expression with the ectopic expression of DRAK1 (Fig. 1B). Furthermore, we examined the interaction between DRAK1 and TRAF family members. Interestingly, the expression level of TRAF6 was specifically decreased in the presence of DRAK1 among TRAF family members (Supplementary Fig. S1). To determine which degradation pathway is involved in the decreased expression of the TRAF6 protein with the ectopic expression of DRAK1, 293T cells were transiently cotransfected with Flag-TRAF6 in the presence or absence of Myc-DRAK1 upon treatment with MG-132, a proteasome degradation inhibitor, or 3-methyladenine (3-MA), an autophagy inhibitor. Interestingly, the decreased expression of TRAF6 byDRAK1 was markedly restored by 3-MA treatment but not by MG-132 treatment, suggesting that overexpression of DRAK1 induces the autophagy-mediated degradation of TRAF6 (Fig. 1C). Next, to identify which domain of TRAF6 interacts with DRAK1, we generated five human TRAF6 deletion constructs (Fig. 1D). Immunoprecipitation assays indicated that DRAK1 specifically binds to the TRAF domain of TRAF6, and residues 358 to 364 in its domain are a critical region for the interaction of TRAF6 with DRAK1 (Fig. 1EG). Considering that the TRAF domain of TRAF6 is important for the autoubiquitination of TRAF6 through homo-oligomerization-mediated its stabilization, we examined whether DRAK1 inhibits the autoubiquitination of TRAF6. Overexpression of DRAK1 significantly decreased both the autoubiquitination and total expression of TRAF6 without 3-MA treatment. Although the expression of TRAF6 with DRAK1 in the WCL was strongly rescued by 3-MA treatment, pull-down experiments revealed that the overexpression of DRAK1 reduced the autoubiquitination of TRAF6 regardless of 3-MA treatment, suggesting that DRAK1 inhibits the autoubiquitination of TRAF6 by interfering with the interaction of TRAF domains in TRAF6 (Fig. 1H). Indeed, an immunoprecipitation assay showed that the ectopic expression of DRAK1 strongly inhibited the interaction of TRAF domains (Fig. 1I). This was further corroborated by a dose-dependent decrease in the homodimerization of TRAF6 with ectopic DRAK1 expression upon 3-MA treatment (Fig. 1J).

Figure 1.

DRAK1 induces the autophagy-mediated degradation of TRAF6 through interaction with the TRAF domain of TRAF6. A, Immunoprecipitation assay showing the interaction between Myc-DRAK1 and Flag-TRAF6 in 293T cells. B, Immunoblot analysis showing the dose-dependent changes in the expression levels of TRAF6 with the ectopic expression of DRAK1. C, Immunoblot analysis of 293T cells cotransfected with Myc-DRAK1 and Flag-TRAF6 plasmids upon treatment with 10 μmol/L MG-132 for 4 hours or 10 mmol/L 3-MA for 9 hours. D, Schematic representations of different truncated forms of TRAF6. E and F, Immunoprecipitation assays of 293T cells cotransfected with GST-DRAK1 and different truncated TRAF6 mutants. G, Immunoprecipitation assays of 293T cells cotransfected with GST-DRAK1 and various TRAF6 deletion mutants. H, Ni-NTA-mediated pull-down assays in 293T cells cotransfected with plasmids encoding Flag-TRAF6, Myc-DRAK1, and His-ubiquitin with or without 3-MA. I, Immunoprecipitation assay showing the dimerization of the TRAF domain of TRAF6 with the ectopic expression of DRAK1. J, Immunoprecipitation assay showing the dimerization of TRAF6 with the ectopic expression of DRAK1 upon 3-MA treatment. These data are representative of at least three independent experiments.

Figure 1.

DRAK1 induces the autophagy-mediated degradation of TRAF6 through interaction with the TRAF domain of TRAF6. A, Immunoprecipitation assay showing the interaction between Myc-DRAK1 and Flag-TRAF6 in 293T cells. B, Immunoblot analysis showing the dose-dependent changes in the expression levels of TRAF6 with the ectopic expression of DRAK1. C, Immunoblot analysis of 293T cells cotransfected with Myc-DRAK1 and Flag-TRAF6 plasmids upon treatment with 10 μmol/L MG-132 for 4 hours or 10 mmol/L 3-MA for 9 hours. D, Schematic representations of different truncated forms of TRAF6. E and F, Immunoprecipitation assays of 293T cells cotransfected with GST-DRAK1 and different truncated TRAF6 mutants. G, Immunoprecipitation assays of 293T cells cotransfected with GST-DRAK1 and various TRAF6 deletion mutants. H, Ni-NTA-mediated pull-down assays in 293T cells cotransfected with plasmids encoding Flag-TRAF6, Myc-DRAK1, and His-ubiquitin with or without 3-MA. I, Immunoprecipitation assay showing the dimerization of the TRAF domain of TRAF6 with the ectopic expression of DRAK1. J, Immunoprecipitation assay showing the dimerization of TRAF6 with the ectopic expression of DRAK1 upon 3-MA treatment. These data are representative of at least three independent experiments.

Close modal

DRAK1, a serine/threonine kinase, is activated by autophosphorylation (20). To confirm whether the kinase activity of DRAK1 is required to inhibit the autoubiquitination of TRAF6, we generated a kinase-dead form of DRAK1 (DRAK1 KD) by replacing Lys90, an ATP-binding site in the kinase domain, with an alanine residue. Interestingly, DRAK1 KD interacted with TRAF6 (Supplementary Fig. S2A) and decreased the autoubiquitination of TRAF6 (Supplementary Fig. S2B) in a manner similar to wild-type DRAK1. Moreover, the expression levels of TRAF6 were markedly reduced in the WCL regardless of the kinase activity of DRAK1, implying that DRAK1 may function as an adaptor protein in the regulation of TRAF6. Taken together, our results suggest that DRAK1 decreases the homo-oligomerization-mediated stabilization of TRAF6 by binding to its TRAF domain.

DRAK1 induces the autophagy-mediated degradation of TRAF6 in cervical cancer cells

Because IL1R/TLR signaling-activated TRAF6 plays an important role in the development and invasion of cervical cancer, we hypothesized that DRAK1 regulates the function of TRAF6 in cervical cancer cells. We initially investigated the expression of DRAK1 and TRAF6 in various cervical cancer cells. Interestingly, the expression levels of DRAK1 and TRAF6 were inversely correlated in cervical cancer cells (Fig. 2A). We next examined the physiologic significance of the interaction between DRAK1 and TRAF6 in cervical cancer cells. As expected, an immunoprecipitation assay showed an endogenous interaction between these two proteins (Fig. 2B). To assess the expression of endogenous TRAF6 according to the expression status of DRAK1, we generated DRAK1-depleted (HeLa and CaSki) or DRAK1-overexpressed (SiHa) cervical cancer cells. Interestingly, DRAK1 knockdown markedly increased the expression of TRAF6 protein, whereas DRAK1 overexpression strongly decreased its endogenous expression without affecting the expression level of TRAF6 mRNA (Fig. 2C and D; Supplementary Figs. S3A and S3B). In addition, ectopic expression of DRAK1 reduced the expression of endogenous TRAF6 in various cancer cells (Supplementary Fig. S3C). To determine whether the reduced expression of TRAF6 in DRAK1-overexpressing SiHa cells was due to decreased stability of TRAF6 protein, we measured the half-life of endogenous TRAF6 protein after treatment with CHX, a protein synthesis inhibitor. DRAK1 overexpression dramatically increased the rate of TRAF6 degradation compared with that observed in control cells (Fig. 2E). Moreover, the attenuated stability of endogenous TRAF6 with DRAK1 overexpression was markedly restored by treatment with 3-MA but not MG-132 (Fig. 2F). We next investigated whether degradation of TRAF6 by DRAK1 is dependent on autophagy. Interestingly, depletion or overexpression of DRAK1 markedly led to altered expression levels of LC3II, an autophagic marker, and TRAF6 in cervical cancer cells (Fig. 2G and H). Furthermore, knockdown of Beclin-1 and ATG5 significantly rescued decreased expression of TRAF6 in DRAK1-overexprssing SiHa cell (Fig. 2I and J). Taken together, these results suggest that DRAK1 induces the autophagy-mediated degradation of TRAF6 in cervical cancer cells.

Figure 2.

Overexpression of DRAK1 facilitates the autophagy-mediated degradation of TRAF6 in cervical cancer cells. A, Immunoblot analysis showing DRAK1 and TRAF6 expression in various cervical cancer cells. B, Immunoprecipitation assay showing the endogenous interaction between DRAK1 and TRAF6. C and D, Immunoblot analysis showing the expression levels of endogenous TRAF6 in DRAK1-depleted (C) and DRAK1-overexpressing (D) cervical cancer cells. E, Immunoblot analysis showing the stability of the TRAF6 protein in control and DRAK1-overexpressing SiHa cells in the presence of CHX (50 μg/mL) for the indicated times (left). The data were quantified using ImageJ software (right). F, Immunoblot analysis of DRAK1-overexpressing SiHa cells treated with 10 μmol/L MG-132 for 4 hours or 10 mmol/L 3-MA for 9 hours. G and H, Immunoblot analysis showing the expression levels of endogenous LC3B and TRAF6 in DRAK1-depleted (G) and DRAK1-overexpressing (H) cervical cancer cells. I and J, Immunoblot analysis showing the expression levels of endogenous TRAF6 by knockdown of Beclin-1 (I) and ATG5 (J) in DRAK1-overexpressing SiHa cells. Cells were transiently transfected with BECN1 and ATG5 siRNA and then immunoblotted with the indicated antibodies.

Figure 2.

Overexpression of DRAK1 facilitates the autophagy-mediated degradation of TRAF6 in cervical cancer cells. A, Immunoblot analysis showing DRAK1 and TRAF6 expression in various cervical cancer cells. B, Immunoprecipitation assay showing the endogenous interaction between DRAK1 and TRAF6. C and D, Immunoblot analysis showing the expression levels of endogenous TRAF6 in DRAK1-depleted (C) and DRAK1-overexpressing (D) cervical cancer cells. E, Immunoblot analysis showing the stability of the TRAF6 protein in control and DRAK1-overexpressing SiHa cells in the presence of CHX (50 μg/mL) for the indicated times (left). The data were quantified using ImageJ software (right). F, Immunoblot analysis of DRAK1-overexpressing SiHa cells treated with 10 μmol/L MG-132 for 4 hours or 10 mmol/L 3-MA for 9 hours. G and H, Immunoblot analysis showing the expression levels of endogenous LC3B and TRAF6 in DRAK1-depleted (G) and DRAK1-overexpressing (H) cervical cancer cells. I and J, Immunoblot analysis showing the expression levels of endogenous TRAF6 by knockdown of Beclin-1 (I) and ATG5 (J) in DRAK1-overexpressing SiHa cells. Cells were transiently transfected with BECN1 and ATG5 siRNA and then immunoblotted with the indicated antibodies.

Close modal

DRAK1 inhibits the TRAF6-mediated activation of NF-κB in cervical cancer cells

Considering that TRAF6 is a signal transducer that activates the NF-κB signaling pathway in response to proinflammatory cytokines, we investigated whether DRAK1 negatively regulates proinflammatory cytokine-induced TRAF6 signaling cascades in cervical cancer cells. Interestingly, DRAK1 overexpression specifically decreased the IL1β-activated phosphorylation of TAK1 and p38 MAPK (Fig. 3A). Conversely, the IL1β-induced activation of TAK1 and p38 MAPK was enhanced in DRAK1-depleted cells (Fig. 3B). In addition, DRAK1 overexpression significantly reduced the IL1β-induced translocation of p65, a subunit of the NF-κB transcription complex, into the nucleus and the activity of the NF-κB promoter (Fig. 3CE), whereas DRAK1-knockdown CaSki cells exhibited the opposite effects (Supplementary Figs. S4A–S4C). Next, we examined whether activation of TAK1/p38 MAPK by DRAK1 knockdown is dependent on the expression of TRAF6. siRNA-induced TRAF6 depletion clearly attenuated the activation of TAK1 and p38 MAPK in both control cells and DRAK1-depleted cells, suggesting that DRAK1 specifically suppresses TRAF6-dependent TAK1/p38 MAPK signaling to activate NF-κB in cervical cancer cells (Fig. 3F; Supplementary Fig. S4D). Considering that autoubiquitination of TRAF6 is critical for the activation of its downstream signaling pathway in inflammatory responses, we hypothesized that DRAK1 regulates the IL1β-induced autoubiquitination of TRAF6 in cervical cancer cells. Indeed, an immunoprecipitation assay showed that DRAK1 overexpression strongly decreased the IL1β-induced autoubiquitination of TRAF6 by reducing its stabilization, whereas DRAK1 knockdown inversely increased TRAF6 autoubiquitination and stabilization (Fig. 3G; Supplementary Fig. S4E). Moreover, ectopic expression of DRAK1 strongly attenuated the TRAF6-induced polyubiquitination of TAK1, which is served as its substrate (Supplementary Fig. S4F). Previously, it was reported that p62/SQSTM1 promoted oligomerization of TRAF6 through interaction with the TRAF domain, eventually activating NF-κB signaling (30). We assumed that DRAK1 might interfere the complex formation between TRAF6 and p62 to block NF-κB activation. Indeed, an immunoprecipitation assay showed that DRAK1 inhibited the binding of p62 to TRAF6 independent of 3-MA treatment (Fig. 3H). Interestingly, p62 interacted with DRAK1 and did not influence reduced expression of TRAF6 by DRAK1 in the absence of 3-MA. Furthermore, DRAK1 significantly decreased IL1β-induced NF-κB promoter activity in p62 and TRAF6-cotransfected SiHa cells (Fig. 3I). Collectively, these results suggest that DRAK1 specifically decreases proinflammatory cytokine-activated TRAF6-mediated NF-κB signaling cascade by inhibiting interaction between p62 and TRAF6, thereby acting as a novel negative regulator of TRAF6 in cervical cancer cells.

Figure 3.

DRAK1 strongly decreases the TRAF6-activated NF-κB signaling pathway by interfering interaction of between TRAF6 and p62. A and B, Immunoblot analysis of DRAK1-overexpressing SiHa cells (A) and -knockdown CaSki cells (B) upon IL1β (10 ng/mL) treatment for the indicated times. C, Subcellular distribution of DRAK1 and p65 in control and DRAK1-overexpressing SiHa cells upon treatment with IL1β (10 ng/mL) for 1 hour. The cells were fractionated into cytoplasmic and nuclear extracts, and then the fractionated extracts were immunoblotted with the indicated antibodies. α-Tubulin and lamin B were used as cytoplasmic and nuclear markers, respectively, and loading controls. D, Immunofluorescence analysis showing p65 localization in control or DRAK1-overexpressing SiHa cells upon IL1β (10 ng/mL) treatment for 1 hour. Original magnification, ×400. E, Luciferase activities in DRAK1-overexpressing SiHa cells were measured after 1 hour with or without IL1β (10 ng/mL). F, Immunoblot analysis showing levels of DRAK1, TRAF6, p-TAK1, and p-p38 MAPK in control or DRAK1 knockdown HeLa cells transiently transfected with TRAF6 siRNA. G, Immunoprecipitation assay showing endogenous TRAF6 ubiquitination in control or DRAK1-overexpressing SiHa cells upon treatment with IL1β (10 ng/mL) for the indicated times. H, Immunoprecipitation assay in 293T cells cotransfected with plasmids encoding Flag-TRAF6, HA-p62, and Myc-DRAK1 with or without 3-MA. I, Luciferase activities of NF-κB promoter in SiHa cells transiently cotransfected with Flag-TRAF6, HA-p62, and Myc-DRAK1 plasmids with or without IL1β (10 ng/mL). All P values were calculated by unpaired two-tailed Student t tests and error bars indicate the mean ± SD of three independent experiments in E and I.

Figure 3.

DRAK1 strongly decreases the TRAF6-activated NF-κB signaling pathway by interfering interaction of between TRAF6 and p62. A and B, Immunoblot analysis of DRAK1-overexpressing SiHa cells (A) and -knockdown CaSki cells (B) upon IL1β (10 ng/mL) treatment for the indicated times. C, Subcellular distribution of DRAK1 and p65 in control and DRAK1-overexpressing SiHa cells upon treatment with IL1β (10 ng/mL) for 1 hour. The cells were fractionated into cytoplasmic and nuclear extracts, and then the fractionated extracts were immunoblotted with the indicated antibodies. α-Tubulin and lamin B were used as cytoplasmic and nuclear markers, respectively, and loading controls. D, Immunofluorescence analysis showing p65 localization in control or DRAK1-overexpressing SiHa cells upon IL1β (10 ng/mL) treatment for 1 hour. Original magnification, ×400. E, Luciferase activities in DRAK1-overexpressing SiHa cells were measured after 1 hour with or without IL1β (10 ng/mL). F, Immunoblot analysis showing levels of DRAK1, TRAF6, p-TAK1, and p-p38 MAPK in control or DRAK1 knockdown HeLa cells transiently transfected with TRAF6 siRNA. G, Immunoprecipitation assay showing endogenous TRAF6 ubiquitination in control or DRAK1-overexpressing SiHa cells upon treatment with IL1β (10 ng/mL) for the indicated times. H, Immunoprecipitation assay in 293T cells cotransfected with plasmids encoding Flag-TRAF6, HA-p62, and Myc-DRAK1 with or without 3-MA. I, Luciferase activities of NF-κB promoter in SiHa cells transiently cotransfected with Flag-TRAF6, HA-p62, and Myc-DRAK1 plasmids with or without IL1β (10 ng/mL). All P values were calculated by unpaired two-tailed Student t tests and error bars indicate the mean ± SD of three independent experiments in E and I.

Close modal

Cell-penetrating DRAK1 peptide suppresses the TRAF6-mediated signaling pathway by inhibiting homo-oligomerization-mediated stabilization

Given that DRAK1 negatively regulates the TRAF6-mediated signaling cascade by reducing TRAF6 stabilization, we next determined which region of DRAK1 is responsible for its interaction with TRAF6. GST pull-down assay revealed that residues 112 to 120 in the DRAK1 kinase domain are a critical motif for its interaction with TRAF6 (Fig. 4A and B). In parallel with this observation, we designed a specific cell-penetrating peptide (11R-DRAK1) with the potential DRAK1-binding residues (residue 110-AVLELAQDNPWVINL-residue 124) and a membrane-permeable polyarginine residue (11R) at its N-terminus to mimic the function of DRAK1 as a dominant negative inhibitor, which may promote degradation of TRAF6 (Fig. 4C and D). To test whether the 11R-DRAK1 peptide regulates the stability of TRAF6 proteins, we measured the half-life of endogenous TRAF6 protein after its treatment with CHX with or without the 11R-DRAK1 peptide. Interestingly, the stability of the TRAF6 protein was markedly decreased in the presence of the 11R-DRAK1 peptide compared with that in the presence of a scrambled peptide (Fig. 4E). Next, we examined the effects of 11R-DRAK1 on the TRAF6-mediated downstream signaling cascades. Treatment with 11R-DRAK1 significantly decreased the expression level of TRAF6 and the phosphorylation of TAK1 and p38 MAPK in both control cells and DRAK1-knockdown cells (Fig. 4F). Furthermore, the enhanced activity of the NF-κB promoter following DRAK1 depletion was markedly attenuated by 11R-DRAK1 treatment (Fig. 4G). We further tested whether 11R-DRAK1 regulates the autoubiquitination of TRAF6 in cervical cancer cells. As expected, 11R-DRAK1 clearly decreased the IL1β-induced autoubiquitination of TRAF6 in a dose-dependent manner (Fig. 4H). In addition, consistent with previous results (Fig. 1H), an immunoprecipitation assay showed that 11R-DRAK1-induced reduction in the expression of endogenous TRAF6 was markedly restored by 3-MA treatment, whereas endogenous autoubiquitination of TRAF6 was decreased by 11R-DRAK1 regardless of 3-MA treatment (Fig. 4I), indicating that 11R-DRAK1 may effectively attenuate the autoubiquitination of TRAF6 by interfering the homo-oligomerization of TRAF6. Collectively, these results suggest that 11R-DRAK1 exerts a specific inhibitory effect on the TRAF6-mediated NF-κB signaling pathway via reduction of homo-oligomerization-mediated stabilization of TRAF6.

Figure 4.

The 11R-DRAK1 peptide decreases inflammatory signaling activation through destabilization of the TRAF6 protein. A and B, GST pull-down assay of Flag-TRAF6 bound to the kinase domain of GST-DRAK1. 293T cells were transiently cotransfected with Flag-TRAF6 and GST-DRAK1 deletion constructs. C, The predicted amino acid sequence of DRAK1 that interacts with TRAF6. D, Peptide sequences of polyarginine residues (11R)-conjugated scramble and DRAK1. E, Immunoblot analysis showing the stability of the TRAF6 protein in 11R-scramble- or 11R-DRAK1 peptide-treated SiHa cells in the presence of CHX (50 μg/mL) for the indicated times (left). SiHa cells were pretreated with 5 μmol/L 11R-scramble or 11R-DRAK1 peptides for 8 hours. The data were quantified using ImageJ software (right). F, Immunoblot analysis of the inhibitory effect of 11R-DRAK1 on TRAF6-mediated signaling increased by DRAK1 knockdown. Control or DRAK1-depleted CaSki cells were treated with 5 μmol/L 11R-scramble or 11R-DRAK1 peptides for 14 hours. G, Luciferase activities of the NF-κB in control or DRAK1-depleted CaSki cells treated with 5 μmol/L 11R-scramble or 11R-DRAK1 peptides for 14 hours. P values were calculated by unpaired two-tailed Student t tests. Error bars, mean ± SD of three independent experiments. H, Immunoprecipitation assays of SiHa cells pretreated with 11R-Scramble or 11R-DRAK1 peptide in the absence or presence of IL1β. SiHa cells were pretreated with 5 and 10 μmol/L 11R-DRAK peptide for 14 hours, followed by treatment with IL1β for 15 minutes. Endogenous TRAF6 ubiquitination was detected by immunoprecipitation with an anti-TRAF6 antibody, followed by immunoblotting with anti-ubiquitin antibody. I, Immunoprecipitation assays of SiHa cells pretreated with 11R-Scramble or 11R-DRAK1 peptide in the absence or presence of 3-MA. All data are representative of at least three independent experiments.

Figure 4.

The 11R-DRAK1 peptide decreases inflammatory signaling activation through destabilization of the TRAF6 protein. A and B, GST pull-down assay of Flag-TRAF6 bound to the kinase domain of GST-DRAK1. 293T cells were transiently cotransfected with Flag-TRAF6 and GST-DRAK1 deletion constructs. C, The predicted amino acid sequence of DRAK1 that interacts with TRAF6. D, Peptide sequences of polyarginine residues (11R)-conjugated scramble and DRAK1. E, Immunoblot analysis showing the stability of the TRAF6 protein in 11R-scramble- or 11R-DRAK1 peptide-treated SiHa cells in the presence of CHX (50 μg/mL) for the indicated times (left). SiHa cells were pretreated with 5 μmol/L 11R-scramble or 11R-DRAK1 peptides for 8 hours. The data were quantified using ImageJ software (right). F, Immunoblot analysis of the inhibitory effect of 11R-DRAK1 on TRAF6-mediated signaling increased by DRAK1 knockdown. Control or DRAK1-depleted CaSki cells were treated with 5 μmol/L 11R-scramble or 11R-DRAK1 peptides for 14 hours. G, Luciferase activities of the NF-κB in control or DRAK1-depleted CaSki cells treated with 5 μmol/L 11R-scramble or 11R-DRAK1 peptides for 14 hours. P values were calculated by unpaired two-tailed Student t tests. Error bars, mean ± SD of three independent experiments. H, Immunoprecipitation assays of SiHa cells pretreated with 11R-Scramble or 11R-DRAK1 peptide in the absence or presence of IL1β. SiHa cells were pretreated with 5 and 10 μmol/L 11R-DRAK peptide for 14 hours, followed by treatment with IL1β for 15 minutes. Endogenous TRAF6 ubiquitination was detected by immunoprecipitation with an anti-TRAF6 antibody, followed by immunoblotting with anti-ubiquitin antibody. I, Immunoprecipitation assays of SiHa cells pretreated with 11R-Scramble or 11R-DRAK1 peptide in the absence or presence of 3-MA. All data are representative of at least three independent experiments.

Close modal

DRAK1 negatively regulates the TRAF6-mediated inflammatory gene network in cervical cancer cells

Given that TRAF6 activation induces inflammation-associated genes through the NF-κB-mediated signaling pathway, eventually promoting cancer progression, our findings led us to verify the functions of DRAK1 in regulating the TRAF6-mediated gene network in cervical cancer cells. We initially performed transcriptome analysis using the RNA sequencing of DRAK1-knockdown HeLa cells. Notably, DRAK-depleted cells exhibited the significant induction of genes enriched in inflammatory responses, as shown in a heatmap of gene expression (with a 2-fold cutoff, P < 0.05; Fig. 5A). To gain further insights into the genes upregulated in cervical cancer cells with the loss of DRAK1, we conducted GO and KEGG pathway analyses using the DAVID functional annotation tool. The genes in cervical cancer cells upregulated by the loss of DRAK1 were highly enriched in GO terms involved in the inflammatory response and in KEGG pathways including cytokine–cytokine receptor interactions, the NF-κB signaling pathway, the TNF signaling pathway, and ABC transporters (Fig. 5B). Furthermore, a network of genes enriched in the inflammatory response was part of a substantially interconnected regulatory pathway with TRAF6 (Fig. 5C). In accordance with these observations, qRT-PCR showed that the depletion of DRAK1 significantly enhanced the expression of the well-known NF-κB target genes, such as IL1B, IL8, CXCL1, CXCL2, and ABCB1 (Fig. 5D), which have been implicated in cancer progression. We next examined whether the enhanced expression of proinflammatory cytokines in DRAK1-depleted cells is dependent on the expression of TRAF6. The siRNA-induced depletion of TRAF6 significantly decreased the expression of IL1B and IL8 in DRAK1-depleted cells compared with its expression in control cells (Fig. 5E). Taken together, these findings indicate the negative role of DRAK1 in regulating the TRAF6-mediated induction of the inflammatory response in cervical cancer cells.

Figure 5.

DRAK1 negatively regulates the inflammatory signaling-associated gene network in cervical cancer cells. A, Heatmap showing the downregulated genes in DRAK1-depleted HeLa cells. Threshold values are as follows: corrected value P < 0.05 and absolute log2-fold change >1.0. B, KEGG pathways and GO terms enriched in differentially expressed genes upregulated by DRAK1 knockdown from A. C, STRING analysis showing protein interactome from B. Red, NF-κB signaling pathway; blue, TNF signaling pathway; green, cytokine-cytokine receptor interaction; yellow, TRAF6-mediated induction of NF-κB and MAP kinase. D, Real-time qRT-PCR showing the expression of altered target genes in DRAK1-depleted HeLa cells. E, Real-time qRT-PCR to analyze the expression of IL1B and IL8 mRNAs following TRAF6 knockdown in DRAK1-depleted HeLa cells. Total RNA was isolated from control and DRAK1-depleted cells transiently transfected with TRAF6 siRNA. All P values were calculated by unpaired two-tailed Student t tests. Data in D and E are representative of at least three independent experiments.

Figure 5.

DRAK1 negatively regulates the inflammatory signaling-associated gene network in cervical cancer cells. A, Heatmap showing the downregulated genes in DRAK1-depleted HeLa cells. Threshold values are as follows: corrected value P < 0.05 and absolute log2-fold change >1.0. B, KEGG pathways and GO terms enriched in differentially expressed genes upregulated by DRAK1 knockdown from A. C, STRING analysis showing protein interactome from B. Red, NF-κB signaling pathway; blue, TNF signaling pathway; green, cytokine-cytokine receptor interaction; yellow, TRAF6-mediated induction of NF-κB and MAP kinase. D, Real-time qRT-PCR showing the expression of altered target genes in DRAK1-depleted HeLa cells. E, Real-time qRT-PCR to analyze the expression of IL1B and IL8 mRNAs following TRAF6 knockdown in DRAK1-depleted HeLa cells. Total RNA was isolated from control and DRAK1-depleted cells transiently transfected with TRAF6 siRNA. All P values were calculated by unpaired two-tailed Student t tests. Data in D and E are representative of at least three independent experiments.

Close modal

DRAK1 suppresses tumor growth and metastasis in cervical cancer cells

Given that the dysregulation of TRAF6 is highly associated with cancer progression, our findings led us to verify the physiologic function of DRAK1 in the TRAF6-mediated cancer progression of cervical cancer cells in vitro and in vivo. To this end, we first investigated whether the expression level of DRAK1 affects tumorigenesis in cervical cancer cells. The depletion of DRAK1 markedly increased cell growth as well as clonogenic potential in CaSki and HeLa cells, whereas its overexpression had an antiproliferative effect in SiHa cells (Fig. 6A and B; Supplementary Figs. S5A and S5B). On the basis of these in vitro results, to further examine the effect of DRAK1 on the tumorigenic capacity of cervical cancer cells in vivo, we subcutaneously injected DRAK1-depleted CaSki cells or DRAK1-overexpressing SiHa cells into the flanks of immunodeficient mice. DRAK1 knockdown significantly increased the ability of CaSki cells to form tumors and the expression of Ki-67, a marker of cell proliferation, compared with that in the control cells, whereas DRAK1-overexpressing SiHa cells exhibited an attenuated tumor volume and the decreased expression of Ki-67 (Fig. 6CF). More strikingly, the markedly increased expression of TRAF6 was observed in DRAK1-depleted primary tumor tissues compared with that in control tissues (Fig. 6D). In contrast, DRAK1-overexpressing primary tumor tissues showed a significant reduction in TRAF6 expression (Fig. 6F).

Figure 6.

DRAK1 significantly suppresses the tumorigenicity and lung metastasis of cervical cancer cells. A and B, Cell doubling times of DRAK1-depleted HeLa and CaSki cells and DRAK1-overexpressed SiHa cells. The data represent the mean ± SD of three independent experiments. C and E, Tumor formation and growth of control and DRAK1-depleted CaSki cells (C) or DRAK1-overexpressed SiHa cells (E) subcutaneously injected into the flanks of NOD/SCID mice (n = 8). Primary tumor volumes were measured weekly, and mice were sacrificed at 6 weeks. Representative primary tumor images (top) and tumor volumes (bottom) are shown. The data represent the mean ± SD of independent experiments. D and F, Representative IHC images showing DRAK1, TRAF6, and Ki-67 expression in primary tumor tissues from C and E. Original magnification, ×100. Scale bar, 50 μm. G and H, Matrigel invasion assays of control and DRAK1-depleted CaSki and HeLa cells (G) or DRAK1-overexpressed SiHa cells (H). Invaded cells were counted following staining with crystal violet. I and J, Scatter plot (left) and representative images of hematoxylin and eosin staining (right) showing lung metastatic nodules 8 weeks after the lateral tail vein injection of control and DRAK1-depleted CaSki cells (I) or DRAK1-overexpressed SiHa cells (J). Original magnification, ×100. Scale bar, 50 μm. K, Representative IHC images showing DRAK1 and TRAF6 expression from J. Original magnification, ×100. Scale bar, 50 μm. L, Kaplan–Meier survival curves for mice from J. P values were calculated by log-rank Mantel–Cox test. The data in G and H were statistically analyzed by unpaired two-tailed Student t tests and represent the mean ± SD.

Figure 6.

DRAK1 significantly suppresses the tumorigenicity and lung metastasis of cervical cancer cells. A and B, Cell doubling times of DRAK1-depleted HeLa and CaSki cells and DRAK1-overexpressed SiHa cells. The data represent the mean ± SD of three independent experiments. C and E, Tumor formation and growth of control and DRAK1-depleted CaSki cells (C) or DRAK1-overexpressed SiHa cells (E) subcutaneously injected into the flanks of NOD/SCID mice (n = 8). Primary tumor volumes were measured weekly, and mice were sacrificed at 6 weeks. Representative primary tumor images (top) and tumor volumes (bottom) are shown. The data represent the mean ± SD of independent experiments. D and F, Representative IHC images showing DRAK1, TRAF6, and Ki-67 expression in primary tumor tissues from C and E. Original magnification, ×100. Scale bar, 50 μm. G and H, Matrigel invasion assays of control and DRAK1-depleted CaSki and HeLa cells (G) or DRAK1-overexpressed SiHa cells (H). Invaded cells were counted following staining with crystal violet. I and J, Scatter plot (left) and representative images of hematoxylin and eosin staining (right) showing lung metastatic nodules 8 weeks after the lateral tail vein injection of control and DRAK1-depleted CaSki cells (I) or DRAK1-overexpressed SiHa cells (J). Original magnification, ×100. Scale bar, 50 μm. K, Representative IHC images showing DRAK1 and TRAF6 expression from J. Original magnification, ×100. Scale bar, 50 μm. L, Kaplan–Meier survival curves for mice from J. P values were calculated by log-rank Mantel–Cox test. The data in G and H were statistically analyzed by unpaired two-tailed Student t tests and represent the mean ± SD.

Close modal

Considering that tumorigenic potential is often associated with invasion ability for the metastasis of tumor cells, we next investigated whether DRAK1 affects invasiveness in cervical cancer cells using Matrigel invasion assays. Cell invasion was significantly increased by DRAK1 knockdown compared with that in control cells, whereas DRAK1 overexpression had the opposite effect (Fig. 6G and H). Consistent with these observations, the tail vein injection of immunodeficient mice with DRAK1-depleted CaSki cells markedly increased metastatic pulmonary nodules in vivo (Fig. 6I). Notably, DRAK1 overexpression in SiHa cells significantly reduced metastatic pulmonary nodules and TRAF6 expression, prolonging survival (Fig. 6JL). To examine the possibility that DRAK1 directly influences the TRAF6-mediated invasion of cervical cancer cells, DRAK1-depleted HeLa and CaSki cells were transiently transfected with TRAF6 siRNA. Interestingly, siRNA-transfected, TRAF6-depleted cells dramatically exhibited attenuation of the enhanced invasion capability following DRAK1 knockdown, indicating that DRAK1 may suppress TRAF6-dependent metastasis in cervical cancer cells (Supplementary Fig. S6A). Furthermore, given that DRAK1 overexpression suppresses the metastasis of cervical cancer cells, we also examined the ability of the 11R-DRAK1 peptide to inhibit the invasiveness of cervical cancer cells. 11R-DRAK1 treatment significantly reduced cell invasion enhanced by DRAK1 knockdown compared with that following 11R-scramble treatment (Supplementary Fig. S6B). We then examined whether the loss of DRAK1 induces aggressive potential in cervical cancer cells. Interestingly, although DRAK1 overexpression in SiHa cells did not change the expression of the EMT markers E-cadherin, N-cadherin, and Vimentin, DRAK1-depleted CaSki cells showed upregulated N-cadherin and Vimentin expression without altered E-cadherin expression, implying that the loss of DRAK1 may induce partial EMT (Supplementary Figs. S7A and S7B). Furthermore, we examined whether loss of DRAK1 increases the stemness capacity of cervical cancer cells. Notably, DRAK1 depletion significantly increased the formation of spheres and the expression of cancer stemness-related markers, including CD44, OCT4, NANOG, and KLF4, indicating that loss of DRAK1 might increase the stem-like properties by inducing the EMT in cervical cancer cells (Supplementary Figs. S7C–S7E). Taken together, these results strongly suggest that DRAK1 acts as a negative regulator of the TRAF6-mediated aggressive cancer progression of cervical cancer cells.

Expression levels of DRAK1 and TRAF6 are inversely correlated in patients with cervical cancer

To evaluate the clinical significance of DRAK1 expression in patients with cervical cancer, we initially performed immunohistochemical staining using commercial cervical cancer TMAs. Interestingly, the expression of DRAK1 was remarkably lower in distant metastasized tumors compared with that in the primary tumor compartments of patients with cervical cancer (Supplementary Figs. S8A and S8B). This observation prompted an investigation of the relevance between DRAK1 and TRAF6 expression in patients with cervical cancer. To this end, we used cervical cancer TMAs obtained from the Seoul National University College of Medicine in South Korea. Notably, TRAF6 and DRAK1 were inversely expressed in primary tumor tissues as well as the metastatic tumor tissues of patients with cervical cancers (Fig. 7A and B). To further support this observation, we examined matched tumor samples in patients with cervical cancers. Consistent with the results in unmatched samples, a significant inverse relationship between DRAK1 and TRAF6 expression was observed in matched primary tumors as well as the metastatic tumor tissues of patients with cervical cancer (Fig. 7C and D). Taken together, these data suggest that DRAK1 may serve as a predictive factor for the risk of metastasis in patients with cervical cancer.

Figure 7.

DRAK1 is significantly underexpressed and inversely correlated with TRAF6 expression in metastatic cervical cancers. A and B, Representative IHC images showing the expression of DRAK1 and TRAF6 (A) and scatter plots showing the Z-score of DRAK1 and TRAF6 IHC intensity (B) in unmatched cases from TMAs of cervical cancer patients. Original magnification, ×100. Scale bar, 50 μm. C and D, Representative IHC images showing the expression of DRAK1 and TRAF6 (C) and graphs showing the Z-score of DRAK1 and TRAF6 IHC intensity (D) in matched cases from TMAs of patients with cervical cancer. Original magnification, ×100. Scale bar, 50 μm. All P values were calculated by unpaired two-tailed Student t tests and all data represent the mean ± SD in B and D.

Figure 7.

DRAK1 is significantly underexpressed and inversely correlated with TRAF6 expression in metastatic cervical cancers. A and B, Representative IHC images showing the expression of DRAK1 and TRAF6 (A) and scatter plots showing the Z-score of DRAK1 and TRAF6 IHC intensity (B) in unmatched cases from TMAs of cervical cancer patients. Original magnification, ×100. Scale bar, 50 μm. C and D, Representative IHC images showing the expression of DRAK1 and TRAF6 (C) and graphs showing the Z-score of DRAK1 and TRAF6 IHC intensity (D) in matched cases from TMAs of patients with cervical cancer. Original magnification, ×100. Scale bar, 50 μm. All P values were calculated by unpaired two-tailed Student t tests and all data represent the mean ± SD in B and D.

Close modal

Activation of inappropriate inflammatory signaling is linked to cervical cancer progression. Here, we report a unique role of DRAK1 as a crucial negative regulator of TRAF6, which is central activator of inflammatory signaling pathway. DRAK1 specifically decreases the stability of the TRAF6 protein via an autophagy-mediated degradation pathway by interfering with the homo-oligomerization of TRAF6, eventually suppressing tumor growth and metastatic potential of advanced cervical cancer cells. In addition, given that DRAK1 and TRAF6 are inversely expressed in patients with cervical cancers, our findings highlight DRAK1 as a potential biomarker and novel therapeutic target for treatment of TRAF6-associated advanced cervical cancers.

In addition to the major function of TRAF6 in the maintenance of immunological homeostasis, several studies have investigated TRAF6 as a crucial regulator of innate immune signaling in cancer progression (31, 32). The amplification/overexpression of TRAF6 is frequently observed in several cancer cells and enhances tumor formation by activating inappropriate inflammatory signaling. For example, TRAF6 was shown to promote RAS-driven NF-κB activation and tumor-promoting responses in lung cancer (16). Furthermore, TRAF6 is known to mediate Helicobacter pylori-induced NF-κB activation and IL8 secretion in gastric cancers, which are associated with inflammatory signaling in tumorigenesis (33, 34). Notably, patients with HR-HPV-infected cervical cancers exhibit increased levels of inducible nitric oxygen synthase (iNOS), which is induced by the TLRs/TRAF6/NF-κB signaling pathway, compared with those in HPV-negative cervical cancer patients, indicating that TRAF6 is an important activator in inflammatory signaling-induced cancer progression (35–37). Consistent with the important role of TRAF6 in activating NF-κB signaling in the inflammatory signaling pathway, the overexpression of DRAK1 destabilized TRAF6 through its autophagy-mediated degradation, decreasing inflammatory signaling cascades. Further supporting this observation, transcriptome analysis showed that DRAK1 negatively regulates the inflammatory response-associated gene network and decreases the expression of the TRAF6-induced proinflammatory cytokines IL1B and IL8, suggesting that DRAK1 negatively regulates TRAF6-mediated advanced cervical progression. In addition, given that TRAF6 is able to stimulate an inflammatory response leading to tumor development in tumor-infiltrating immune cells, further comprehensive work is needed to understand the functional roles between DRAK1 and TRAF6 in the tumor microenvironment.

Our findings in this study raise questions regarding the mechanism by which DRAK1 destabilizes TRAF6. Previous studies have identified various negative regulators targeting stabilization and autoubiquitination of TRAF6, which is a key regulatory event in its activation (38–41). In response to TLR signaling, the germinal center kinase MST4 directly induces the phosphorylation of TRAF6 and prevents its autoubiquitination by blocking its oligomerization, which in turn limits the activation of downstream NF-κB signaling (38). Our results showed that DRAK1 directly interacts with the TRAF domains of TRAF6 to destabilize TRAF6 and prevent its autoubiquitination, which consequently suppresses the IL1β-activated TRAF6-dependent downstream TAK1/p38 MAPK/NF-κB signaling cascade to inhibit inflammatory signaling-mediated tumorigenesis. In addition, considering that DRAK1 is a serine/threonine kinase, it is feasible that activity of DRAK1 kinase is required to decrease the stabilization of TRAF6. Unexpectedly, like wild-type DRAK1, kinase-inactive DRAK1 reduced the stability and autoubiquitination of TRAF6, suggesting that DRAK1 functions as a negative adaptor protein without inducing the phosphorylation of TRAF6. This observation was further supported by experiments with the cell-penetrating 11R-DRAK1 peptide, which interacted with the TRAF domain of TRAF6, corroborating the function of DRAK1 as an adaptor that limits inflammatory signaling in tumor progression through TRAF6. As the inhibitory effect of DRAK1 as an antagonist targeting TRAF6 was tested in only TRAF6-overexpressing metastatic cervical cancer cells, it is necessary to gain deeper insight into the association of TRAF6 with other malignant tumors as well as inflammatory diseases.

Our findings also demonstrated that DRAK1 is downregulated in patients with metastatic cervical cancers and that its expression is inversely correlated with TRAF6 expression in primary tumors as well as the metastatic tumor tissues of patients with cervical cancer, resulting in poor clinical outcomes. Immune responses in cervical lesions with HPV infection are enhanced through TLR4-activated TRAF6 signaling, leading to advanced cervical cancer progression (6). Thus, considering that HPV is detected in more than 90% of invasive cervical cancers and persistent infection of HR-HPV, such HPV16 and HPV18, can lead to advanced cervical cancers, DRAK1 expression in cervical cancer may be regulated by HPV infection. However, although we could not analyze the correlation between DRAK1 and HPV in patients with cervical cancer, the DRAK1 protein was mostly expressed in HPV-positive cervical cancer cells rather than HPV-negative C33A cells (Fig. 2A), and its expression was not affected by HPV16 E6 and E7 oncoproteins, which were closely related to the occurrence of cervical cancer development in C33A (HPV-negative) and CaSki (HPV-positive) cells. Moreover, overexpression of TLR4 is highly associated with the development of cervical cancer but is not mediated by HPV infection (6). Therefore, we reasoned that altered DRAK1 expression may be indirectly involved in advanced cervical cancer progression through persistent HPV infection. Further comprehensive work is required to be clarified the alteration of DRAK1 expression in cervical cancer progression.

In conclusion, our results suggest a mechanism by which DRAK1 regulates the stability of the TRAF6 protein by blocking its autoubiquitination and homo-oligomerization, thereby indicating a unique function of DRAK1 in inflammatory signaling-mediated metastatic cervical cancer progression. In addition, given that TRAF6 is strongly linked to inflammatory signaling, the modulation of DRAK1 as an antagonist of TRAF6 may be a therapeutic intervention against advanced cervical cancers as well as inflammation diseases.

S.-J. Kim is a CSO at TheragenEtex Co. No potential conflicts of interest were disclosed by the other authors.

Conception and design: Y. Park, K.-M. Yang, S.-J. Kim

Development of methodology: Y. Park, Y.S. Song, K.-M. Yang, S.-J. Kim

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): J. Park, E. Hong, A. Ooshima, H.-S. Kim, J.H. Cho, Y.S. Song

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): A. Ooshima, H.-S. Kim, J.H. Cho, Y. Han, C. Lee, Y.S. Song, K.-M. Yang, S.-J. Kim

Writing, review, and/or revision of the manuscript: Y. Park, K. Pang, K.-M. Yang, S.-J. Kim

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): J. Lee, H.-S. Kim, J.H. Cho, Y. Han, Y.S. Song

Study supervision: A. Ooshima, K.-S. Park, K.-M. Yang, S.-J. Kim

This work was supported by a grant of the Korea Health Technology R&D Project through the National Cancer Center, Ministry for Health and Welfare, Republic of Korea (HA17C0037).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
zur Hausen
H
. 
Papillomavirus infections—a major cause of human cancers
.
Biochim Biophys Acta
1996
;
1288
:
F55
78
.
2.
Fernandes
JV
,
DE Medeiros Fernandes
TA
,
DE Azevedo
JC
,
Cobucci
RN
,
DE Carvalho
MG
,
Andrade
VS
, et al
Link between chronic inflammation and human papillomavirus-induced carcinogenesis (Review)
.
Oncol Lett
2015
;
9
:
1015
26
.
3.
Trimble
CL
,
Peng
S
,
Thoburn
C
,
Kos
F
,
Wu
TC
. 
Naturally occurring systemic immune responses to HPV antigens do not predict regression of CIN2/3
.
Cancer Immunol Immunother
2010
;
59
:
799
803
.
4.
Kim
TJ
,
Jin
HT
,
Hur
SY
,
Yang
HG
,
Seo
YB
,
Hong
SR
, et al
Clearance of persistent HPV infection and cervical lesion by therapeutic DNA vaccine in CIN3 patients
.
Nat Commun
2014
;
5
:
5317
.
5.
Deivendran
S
,
Marzook
KH
,
Radhakrishna Pillai
M
. 
The role of inflammation in cervical cancer
.
Adv Exp Med Biol
2014
;
816
:
377
99
.
6.
Yang
X
,
Cheng
Y
,
Li
C
. 
The role of TLRs in cervical cancer with HPV infection: a review
.
Signal Transduct Target Ther
2017
;
2
:
17055
.
7.
Wee
ZN
,
Yatim
SM
,
Kohlbauer
VK
,
Feng
M
,
Goh
JY
,
Bao
Y
, et al
IRAK1 is a therapeutic target that drives breast cancer metastasis and resistance to paclitaxel
.
Nat Commun
2015
;
6
:
8746
.
8.
Kelly
MG
,
Alvero
AB
,
Chen
R
,
Silasi
DA
,
Abrahams
VM
,
Chan
S
, et al
TLR-4 signaling promotes tumor growth and paclitaxel chemoresistance in ovarian cancer
.
Cancer Res
2006
;
66
:
3859
68
.
9.
Hartman
ZC
,
Poage
GM
,
den Hollander
P
,
Tsimelzon
A
,
Hill
J
,
Panupinthu
N
, et al
Growth of triple-negative breast cancer cells relies upon coordinate autocrine expression of the proinflammatory cytokines IL-6 and IL-8
.
Cancer Res
2013
;
73
:
3470
80
.
10.
Coussens
LM
,
Werb
Z
. 
Inflammation and cancer
.
Nature
2002
;
420
:
860
7
.
11.
Jost
PJ
,
Ruland
J
. 
Aberrant NF-kappaB signaling in lymphoma: mechanisms, consequences, and therapeutic implications
.
Blood
2007
;
109
:
2700
7
.
12.
Perkins
ND
. 
The diverse and complex roles of NF-kappaB subunits in cancer
.
Nat Rev Cancer
2012
;
12
:
121
32
.
13.
Dolcet
X
,
Llobet
D
,
Pallares
J
,
Matias-Guiu
X
. 
NF-kB in development and progression of human cancer
.
Virchows Arch
2005
;
446
:
475
82
.
14.
Prusty
BK
,
Husain
SA
,
Das
BC
. 
Constitutive activation of nuclear factor -kB: preferential homodimerization of p50 subunits in cervical carcinoma
.
Front Biosci
2005
;
10
:
1510
9
.
15.
Fang
J
,
Muto
T
,
Kleppe
M
,
Bolanos
LC
,
Hueneman
KM
,
Walker
CS
, et al
TRAF6 mediates basal activation of NF-kappaB necessary for hematopoietic stem cell homeostasis
.
Cell Rep
2018
;
22
:
1250
62
.
16.
Starczynowski
DT
,
Lockwood
WW
,
Delehouzee
S
,
Chari
R
,
Wegrzyn
J
,
Fuller
M
, et al
TRAF6 is an amplified oncogene bridging the RAS and NF-kappaB pathways in human lung cancer
.
J Clin Invest
2011
;
121
:
4095
105
.
17.
Grivennikov
SI
,
Greten
FR
,
Karin
M
. 
Immunity, inflammation, and cancer
.
Cell
2010
;
140
:
883
99
.
18.
Schimmack
G
,
Schorpp
K
,
Kutzner
K
,
Gehring
T
,
Brenke
JK
,
Hadian
K
, et al
YOD1/TRAF6 association balances p62-dependent IL-1 signaling to NF-κB
.
Elife
2017
;
6
:
e22416
.
DOI: 10.7554/eLife.22416.
19.
Yin
Q
,
Lin
SC
,
Lamothe
B
,
Lu
M
,
Lo
YC
,
Hura
G
, et al
E2 interaction and dimerization in the crystal structure of TRAF6
.
Nat Struct Mol Biol
2009
;
16
:
658
66
.
20.
Sanjo
H
,
Kawai
T
,
Akira
S
. 
DRAKs, novel serine/threonine kinases related to death-associated protein kinase that trigger apoptosis
.
J Biol Chem
1998
;
273
:
29066
71
.
21.
Mao
P
,
Hever
MP
,
Niemaszyk
LM
,
Haghkerdar
JM
,
Yanco
EG
,
Desai
D
, et al
Serine/threonine kinase 17A is a novel p53 target gene and modulator of cisplatin toxicity and reactive oxygen species in testicular cancer cells
.
J Biol Chem
2011
;
286
:
19381
91
.
22.
Wittig
R
,
Nessling
M
,
Will
RD
,
Mollenhauer
J
,
Salowsky
R
,
Munstermann
E
, et al
Candidate genes for cross-resistance against DNA-damaging drugs
.
Cancer Res
2002
;
62
:
6698
705
.
23.
Guo
Q
,
Chen
Y
,
Wu
Y
. 
Enhancing apoptosis and overcoming resistance of gemcitabine in pancreatic cancer with bortezomib: a role of death-associated protein kinase-related apoptosis-inducing protein kinase 1
.
Tumori
2009
;
95
:
796
803
.
24.
Gao
J
,
Liu
D
,
Li
J
,
Song
Q
,
Wang
Q
. 
Effect of STK17A on the sensitivity of ovarian cancer cells to paclitaxel and carboplatin
.
Oncol Lett
2016
;
12
:
1107
12
.
25.
Short
SP
,
Thompson
JJ
,
Bilotta
AJ
,
Chen
X
,
Revetta
FL
,
Washington
MK
, et al
Serine threonine kinase 17A maintains the epithelial state in colorectal cancer cells
.
Mol Cancer Res
2019
;
17
:
882
94
.
26.
Park
Y
,
Kim
W
,
Lee
JM
,
Park
J
,
Cho
JK
,
Pang
K
, et al
Cytoplasmic DRAK1 overexpressed in head and neck cancers inhibits TGF-beta1 tumor suppressor activity by binding to Smad3 to interrupt its complex formation with Smad4
.
Oncogene
2015
;
34
:
5037
45
.
27.
Mao
P
,
Hever-Jardine
MP
,
Rahme
GJ
,
Yang
E
,
Tam
J
,
Kodali
A
, et al
Serine/threonine kinase 17A is a novel candidate for therapeutic targeting in glioblastoma
.
PLoS One
2013
;
8
:
e81803
.
28.
Metsalu
T
,
Vilo
J
. 
ClustVis: a web tool for visualizing clustering of multivariate data using principal component analysis and heatmap
.
Nucleic Acids Res
2015
;
43
:
W566
70
.
29.
Szklarczyk
D
,
Franceschini
A
,
Wyder
S
,
Forslund
K
,
Heller
D
,
Huerta-Cepas
J
, et al
STRING v10: protein-protein interaction networks, integrated over the tree of life
.
Nucleic Acids Res
2015
;
43
:
D447
52
.
30.
Sanz
L
,
Diaz-Meco
MT
,
Nakano
H
,
Moscat
J
. 
The atypical PKC-interacting protein p62 channels NF-kappaB activation by the IL-1-TRAF6 pathway
.
EMBO J
2000
;
19
:
1576
86
.
31.
Varney
ME
,
Niederkorn
M
,
Konno
H
,
Matsumura
T
,
Gohda
J
,
Yoshida
N
, et al
Loss of Tifab, a del(5q) MDS gene, alters hematopoiesis through derepression of Toll-like receptor-TRAF6 signaling
.
J Exp Med
2015
;
212
:
1967
85
.
32.
Fang
J
,
Bolanos
LC
,
Choi
K
,
Liu
X
,
Christie
S
,
Akunuru
S
, et al
Ubiquitination of hnRNPA1 by TRAF6 links chronic innate immune signaling with myelodysplasia
.
Nat Immunol
2017
;
18
:
236
45
.
33.
Lamb
A
,
Chen
LF
. 
Role of the Helicobacter pylori-induced inflammatory response in the development of gastric cancer
.
J Cell Biochem
2013
;
114
:
491
7
.
34.
Maeda
S
,
Yoshida
H
,
Ogura
K
,
Mitsuno
Y
,
Hirata
Y
,
Yamaji
Y
, et al
H. pylori activates NF-kappaB through a signaling pathway involving IkappaB kinases, NF-kappaB-inducing kinase, TRAF2, and TRAF6 in gastric cancer cells
.
Gastroenterology
2000
;
119
:
97
108
.
35.
De Marco
F
. 
Oxidative stress and HPV carcinogenesis
.
Viruses
2013
;
5
:
708
31
.
36.
De Marco
F
,
Bucaj
E
,
Foppoli
C
,
Fiorini
A
,
Blarzino
C
,
Filipi
K
, et al
Oxidative stress in HPV-driven viral carcinogenesis: redox proteomics analysis of HPV-16 dysplastic and neoplastic tissues
.
PLoS One
2012
;
7
:
e34366
.
37.
Hiraku
Y
,
Tabata
T
,
Ma
N
,
Murata
M
,
Ding
X
,
Kawanishi
S
. 
Nitrative and oxidative DNA damage in cervical intraepithelial neoplasia associated with human papilloma virus infection
.
Cancer Sci
2007
;
98
:
964
72
.
38.
Jiao
S
,
Zhang
Z
,
Li
C
,
Huang
M
,
Shi
Z
,
Wang
Y
, et al
The kinase MST4 limits inflammatory responses through direct phosphorylation of the adaptor TRAF6
.
Nat Immunol
2015
;
16
:
246
57
.
39.
Wertz
IE
,
O'Rourke
KM
,
Zhou
H
,
Eby
M
,
Aravind
L
,
Seshagiri
S
, et al
De-ubiquitination and ubiquitin ligase domains of A20 downregulate NF-kappaB signalling
.
Nature
2004
;
430
:
694
9
.
40.
Allen
IC
,
Moore
CB
,
Schneider
M
,
Lei
Y
,
Davis
BK
,
Scull
MA
, et al
NLRX1 protein attenuates inflammatory responses to infection by interfering with the RIG-I-MAVS and TRAF6-NF-kappaB signaling pathways
.
Immunity
2011
;
34
:
854
65
.
41.
Schneider
M
,
Zimmermann
AG
,
Roberts
RA
,
Zhang
L
,
Swanson
KV
,
Wen
H
, et al
The innate immune sensor NLRC3 attenuates Toll-like receptor signaling via modification of the signaling adaptor TRAF6 and transcription factor NF-kappaB
.
Nat Immunol
2012
;
13
:
823
31
.

Supplementary data