UBE2S Activates NF-κB Signaling By Binding With IκBα and Promotes Metastasis of Lung Adenocarcinoma Cell

Nuclear factor (NF)-κB signaling in cancer cells was reported to be involved in tumorigenesis. Phosphorylation and degradation of inhibitor of NF-κBα (IκBα) is a canonical pathway of NF-κB signaling. Herein, we report non-canonical activation of NF-κB signaling without phosphorylation of IκB but by directly binding by ubiquitin-conjugating enzyme E2S (UBE2S) for degradation in adenocarcinoma cells.


Abstract Background
Nuclear factor (NF)-κB signaling in cancer cells was reported to be involved in tumorigenesis.
Phosphorylation and degradation of inhibitor of NF-κBα (IκBα) is a canonical pathway of NF-κB signaling. Herein, we report non-canonical activation of NF-κB signaling without phosphorylation of IκB but by directly binding by ubiquitin-conjugating enzyme E2S (UBE2S) for degradation in adenocarcinoma cells.

Methods
TCGA and the Human Atlas Protein Database were used to analyze the survival rate and expression of UBE2S. PC9, H460, H441 and A549 cells were used in this study. PC9 and H460 cells were used for further analysis because of different protein levels of UBE2S. Speci c IKK inhibitors, PS1145 and SC514, were used to analyze the phosphorylation of IκBα. Western blotting experiment was used to analyze the protein levels PC9 and H460 cells. Wound-healing experiment was used to analyze the migrative ability of PC9 and H460 cells. Overexpression and knockdown of UBE2S in H460 and PC9 cells were used to analyze the downstream proteins levels. Immunoprecipitation, immuno uorescent staining, a glutathione S transferase (GST) pull-down assay, and an in vitro binding assay were used to analyze the interaction of UBE2S and IκBα. Luciferase assay was used to analyze the activation of NF-κB signaling regulated by UBE2S. Zebra sh xenograft model was used analyzed the metastasis of PC9 cells regulated by UBE2S.

Results
UBE2S in lung adenocarcinoma patients was negatively related to the survival rate. The protein levels of UBE2S and IκBα were shown opposite change in PC9 and H460 cells. PC9 cells showed higher UBE2S expression and migrative ability than H460 cells. Phosphorylation of IκBα was not changed by treatment with IKK speci c inhibitors, PS1145 and SC514, in PC9 and H460 cells. Overexpression and knockdown of UBE2S in H460 and PC9 cells showed the protein levels of IκBα were regulated. Immunoprecipitation, immuno uorescent staining, a glutathione S transferase (GST) pull-down assay, and an in vitro binding assay showed the direct binding of UBE2S with IκBα. Protein levels of nuclear p65 and luciferase assay showed the NF-κB signaling was regulated UBE2S. EMT markers and migrative ability of cancer cells were also regulated by UBE2S. Zebra sh xenograft tumor model showed the reduction of migrative PC9 cells by knockdown of UBE2S.

Conclusion
Higher UBE2S expressed in lung adenocarcinomas could bind with IκBα for activation of NF-κB signaling to promote metastasis of cancer cells. UBE2S might be a potential therapeutic target for lung adenocarcinomas.

Page 3/22
Background The ubiquitin-conjugating enzyme E2S (UBE2S) ubiquitin carrier protein is an E2 ubiquitination ligase that helps the E1, E2, and E3 ligases link ubiquitin with target proteins, which then targets them towards proteasome degradation [1][2][3]. The ubiquitin-proteasome pathway was documented to play critical roles in tumor formation and progression [4]. The extent of UBE2S, which is highly expressed in several types of tumors, was correlated with the progression of tumors in esophageal cancer patients [5], breast tumors tissues [6], lung tumor cells [7], papillary renal cell carcinoma [8], endometrial tumors [9], glioblastoma multiforme [10], and colon cancer development [11]. UBE2S was suggested to be a potential biomarker for diagnostic, prognostic, and therapeutic targets [12]. UBE2S is involved in stabilization of hypoxiainducible factor (HIF)-1α-induced tumorigenesis [13]. UBE2S was also reported to be overexpressed in lung cancer [7,14]. Nuclear factor (NF)-κB expression was related to the tumor stage, lymph node metastasis, and 5-year overall survival rate for non-small cell lung cancer patients [15].
NF-κB is constitutively activated in some types of cancer cells and is able to induce several aspects of tumorigenesis, including proliferation, anti-apoptosis, angiogenesis, and metastasis [16]. Activation of NF-κB is generally involved in phosphorylation of inhibitor of NF-κBα (IκBα) by IκB kinase (IKK) for degradation. NF-κB is then translocated from the cytoplasm into nuclei. It binds to the promoter region of target genes with speci c site-activated transcription of target genes related to cell proliferation [17], antiapoptosis [18,19], drug resistance [20], angiogenesis, and metastasis [21,22]. Low IκBα and high constitutive NF-κB activities were reported in adenocarcinoma cells and induced only weak NF-κB activity by tumor necrosis factor (TNF)-α treatment [23]. Stabilization of free IκB is not regulated by IKK but is related to the proline (P), glutamic acid (E), serine (S), threonine (T) (PEST) domain and proteasome degradation [24]. Degradation of IκBα to activate NF-κB signaling in cancer cells without IKK activation might play an important role in tumorigenesis.
In our earlier reports, metastasis of tumor cells was activated by matrix metalloproteinase (MMP)-9 through epithelial-to-mesenchymal transition (EMT) signaling [25][26][27][28][29][30]. Snail and MMP-9 are downstream target genes of NF-κB [31]. We also found that acceleration of nuclear entry of NF-κB activated expression of EMT markers in A431-III cells [32]. Treatment with the dietary avonoids, luteolin and quercetin, was able to inhibit stemness, reactive oxygen species (ROS), and the invasive capacity of A431-III cells by reducing the EMT and inhibiting the phosphorylation of Akt and glycogen synthase kinase-3β (GSK3β) [32][33][34]. UBE2S is highly expressed and promotes the migratory and invasive abilities of cancer cells [28]. UBE2S is stabilized by Akt phosphorylation and is involved in DNA repair mechanisms by degradation of the Ku70 complex [10]. Akt might stabilize UBE2S by phosphorylation and mediate NF-κB signaling to promote metastasis of cancer cells.
In this study, we investigated the interaction and effects of UBE2S with IκBα on lung cancer cells. The binding ability of UBE2S with IκBα was analyzed by immunoprecipitation (IP) and in vitro pull-down assays. Protein levels of UBE2S and IκBα were analyzed in lung cancer cells. Knockdown and overexpression of UBE2S in lung cancer cells were performed to analyze protein levels of IκBα and EMT markers.

Database
The survival rate of lung adenocarcinoma patients correlated with UBE2S mRNA was analyzed by GEPIA2 web site (http://gepia2.cancer-pku.cn) [35]. The protein level of UBE2S in lung cancer patients was obtained from staining with CAB015228 antibody in the Human Atlas Protein Database (https://www.proteinatlas.org) [36].

Cell culture
The H441, A549, PC9, and H460 human lung cancer cell lines were obtained as described previously [52].

Plasmid construction
The pcDNA3-UBE2S-ag plasmid was obtained as described in our earlier report [28]. pCMV4-3 HA/IκBα was purchased from addgene (Cambridge, MA, USA). The coding region of IκBα was ampli ed by a KAPA HiFi PCR Kit (Roche) and cloned to the pcDNA3-HA vector as the pcDNA3-IkBa-HA plasmid. The pcDNA3-IkBa-ag plasmid was digested with BamHI and SalI and inserted into the PQE30 vector as the PQE30-IkBa plasmid. The pcDNA3-UBE2S-ag plasmid was digested with BamHI and XhoI and inserted into the pGEX-4T-2 vector as the pGEX-UBE2S plasmid. The following primer pairs were used for the PCR: IκBαforward (5'-ATG TTC CAG GCG GCC G-3') and IκBα-reverse (5'-TAA CGT CAG ACG CTG G-3').
Cell viability assay H460 and PC-9 cells (10 5 cells/well) were seeded in 48-well plates overnight. PC-9 cells were treated with DMSO, SC514, PS1145, and Ly294002 for 24 h. Cells were refreshed with culture medium with 5 mg/mL of MTT (Merck), and incubation continued at 37 °C for 3 h. The culture medium was removed, and 200 μl DMSO was added to resolve precipitation. The absorbance was measured at 570 nm using a Spark multimode microplate reader (TECAN, Männedorf, Switzerland).

Western blotting
Protein samples were heated to 100 °C with sample buffer (250 mM Tris-HCl at pH 6.8, 10% sodium dodecylsulfate (SDS), 30% glycerol, 5% β-mercaptoethanol, and 0.02% bromophenol blue) for 5 min. SDSpolyacrylamide gel electrophoresis (PAGE) was used to separate proteins, which were then transferred to polyvinylidene di uoride (PVDF) membranes (Millipore, Billerica, MA, USA). After blocking with 3% bovine serum albumin (BSA) for 30 min at room temperature, a speci c primary antibody diluted in 3% BSA was incubated with the membrane on a shaker at 4 °C overnight. The membrane was washed with PBST (0.25% Tween-20 in PBS) three times, and the secondary antibody conjugated with horseradish peroxide (Millipore) was incubated for 1 h at room temperature. After washing with PBST three times, proteins were detected using an enhanced chemiluminescence (ECL) reagent kit (Millipore) followed by exposure to Amersham Imager 600 imagers (GE). Protein images were quanti ed using ImageJ software (http://rsb.info.nih.gov/ij/index.html, National Institutes of Health, Bethesda, MA, USA).
Immunostaining COS-1 cells (10 5 cells/well) were seeded onto 12-well plates overnight. The pcDNA3-UBE2S-HA and pcDNA3-IkBa-ag plasmids were transfected into COS-1 cells by PolyJet (SignaGen) for 24 h. PFA at 4% was used to x COS-1 cells for 20 min, and PBST (0.1% Triton X-100 in PBS) was used to permeabilize them for 10 min. After being washed with PBS three times and blocking with 3% BSA for 1 h at room temperature, anti-HA and anti-ag monoclonal antibodies were incubated at 4 °C overnight. Cy3 A niPure goat anti-mouse immunoglobulin G (IgG) was added, incubated for 30 min, and then stained with 0.1% DAPI for 5 min. Images were captured with an Olympus IX70-FLA inverted uorescence microscope (Olympus, Tokyo, Japan) and SPOT system (Diagnostic Instruments, Sterling Heights, MI, USA).
Knockdown of UBE2S by small interfering (si)RNA UBE2S and control siRNAs were purchased from ThermoFisher Scienti c and prepared following the manufacturer's instructions. In total, 10 6 cells were seeded onto six-well plates overnight and then transfected with 40 nM of siRNA using the GenMute siRNA transfection reagent (SignaGen Laboratories, Rockville, MD, USA) following the manufacturer's instructions. All assays were performed 24 h after transfection.
Expression and puri cation of GST-UBE2S and His-IκBα protein in Escherichia coli PQE30-IκBα and pET4T2-UBE2S were transformed into E. coli DH5a strands and expressed by 0.2 M IPTG induction. Cells grown overnight were re-grown in 1/10 dilution of lysogeny broth (LB) for 1 h and inducted by 0.2 M IPTG for 3 h. Cells were collected by centrifugation and resuspended in cold GST buffer (50 mM Tris-HCl at pH 7.5, 150 mM NaCl, 1 mM EDTA, and a proteinase inhibitor cocktail) and His binding/wash buffer (0.5 M NaCl, 100 mM HEPES, 10 mM Imidazole at pH 8.0, 0.5% NP-40, 1 mM PMSF, and a proteinase inhibitor cocktail). Cells were processed by an Untrasonic Processors (Chrom Tech, Apple Valley, MN, USA) on ice. The lysate was centrifuged at 13,000 g for 10 min at 4 °C. The supernatant was applied to a glutathione Sepharose 4B GST-tagged protein puri cation resin (GE) and Ni Sepharose high-performance histidine-tagged protein puri cation resin (GE). After being washed with GST wash buffer (0.5% Triton X-100 and 1 mM EDTAU in PBS) and His binding/wash buffer, GST-UBE2S was resuspended in PBS. His-IκBα proteins were eluted with elution buffer (100 mM HEPES and 0.5 M imidazole at pH 8.0), followed by changing the buffer to PBS by Amicon Ultra Centrifugal Filters (GE).

Pull-down and in vitro binding assays
For the pull-down assay, GST-UBE2S proteins on beads were added to the total lysate of H460 cancer cells and incubated at 4 °C overnight on a rolling shaker. After centrifugation at 500 g for 2 min and washing three times with PBS, the beads were resuspended in PBS and analyzed by Western blotting. The in vitro binding assay was performed by mixing equal volumes of GST-UBE2S and the IκBα protein to interact at 4 °C overnight on a rolling shaker. After centrifugation at 500 g for 2 min and washing three times with PBS, the beads were resuspended in PBS and analyzed by Western blotting.

Immunoprecipitation (IP)
Cultured cells in 10-cm plates were collected and treated with IP-lysis buffer (150 mM NaCl, 20 mM NaCl, 20 mM HEPES at pH 7.2, 10 mM NaF, 1 mM EDTA, 1% NP-40, 1 mM Na3V04, 1 mM PMSF, 1 DTT, and a proteinase inhibitor cocktail) for 10 min. Total cell lysate was collected by centrifugation at 13,000 rpm for 10 min. Pretreatment with protein G Mag Sepharose magnetic beads (GE) for 1 h was used to collect total proteins. An anti-UBE2S (GeneTex) or anti-IκBα (Abcam) monoclonal antibody was coated onto protein G Mag Sepharose magnetic beads (GE) for 1 h and washed with PBS three times. Coated protein G Mag Sepharose magnetic beads (GE) were added to total cell lysates, and the complex was pulleddown overnight. After washing with PBS three times, the protein G Mag Sepharose complex was immunoblotted with an anti-UBE2S (GeneTex) or anti-IκBα (Abcam) monoclonal antibody.
Cell migration assay PC9 cells (5x10 5 cells/well) were plated onto six-well culture plates overnight in RPMI-1640 containing 10% FBS. Monolayer cells were scratched with a pipette tip. After washing with PBS, the monolayers were incubated at 37 °C for 24 h. Images of the monolayers were captured at 0 and 24 h using a phasecontrast Zeiss Axio Vert.A1 inverted microscope (Zeiss, Jena, Germany) and a Leadview 2800AM-FL camera (Leadview, Taipei, Taiwan). Migrating cells were calculated from triplicate determinations for each treatment group. Zebra sh metastasis model Zebra sh (Danio rerio) and embryos were maintained at 28 °C. All animal procedures were approved by the Institutional Animal Care and Use Committee or Panel (IACUC/IACUP) (protocol # LAC-2019-0355). The methods were carried out in accordance with the approved guidelines. The migration assay of cancer cells in zebra sh was analyzed following our earlier report [30,44]. Brie y, GFP expressing PC9 cells were transfected with control or UBE2S siRNA for 24h. The 0.25 % trypsin (ThermoFisher) was used to detach cancer cells and following stained with 1 μM CM-DiI (ThermoFisher) for 20 min. After 3 times wash with PBS, tumor cells were diluted as 2x10 6 cells/mL with PBS. Tumor cells were microinjected into the yolk of 3 days post-fertilization (dpf) zebra sh larvae by IM300 Microinjector (Narishige, Tokyo, Japan) at 30 psi. Zebra sh larvae were then incubated at 28 °C for 1 h and then transferred to 32 °C for further incubation. The migrative tumor cells in zebra sh larvae were analyzed at 7 dpf zebra sh larvae using a phase-contrast Olympus IX70 microscope (Olympus, Tokyo, Japan) and a SPOT camera (Sterling Heights, MI, USA).

Statistical analysis
Three independent experiments were used to analyze the mean ± standard deviation (SD). Statistical signi cance was analyzed by a one-way analysis of variation (AVOVA) followed by Tukey's test. Statistical signi cance was indicated by * p < 0.05, ** p < 0.01, and *** p < 0.001.

UBE2S has an important role in clinical lung adenocarcinoma patients
To investigate the role of UBE2S in lung adenocarcinoma patients, The Cancer Genome Atlas (TCGA) database [35] and Human Protein Atlas database [36] were used to analyze messenger (m)RNA and protein levels of UBE2S. High levels of UBE2S mRNA were correlated with low survival rates in lung adenocarcinoma patients from TCGA database ( Figure 1A). Strong and medium (MS) levels of the UBE2S protein were seen in 81.8% of lung adenocarcinoma patients compared to weak and normal (NW) levels by immunohistochemical (IHC) staining from the Human Protein Atlas database ( Figure 1B). These results show that UBE2S might play an important role in tumorigenesis of lung adenocarcinomas.

UBE2S promotes tumorigenesis related to NF-κB signaling
To analyze the role of UBE2S in tumorigenesis of lung adenocarcinomas, we analyzed UBE2S expression levels in H441, A549, PC9, and H460 lung cancer cells. UBE2S was more highly expressed in PC9 cells than in H441, A549, and H460 cells (Figure 2A). In earlier reports, NF-κB signaling was suggested to be an important tumorigenesis and drug-resistance pathway in lung cancer [37][38][39]. UBE2S is an E2 ligase involved in K11-linked polyubiquitination of component proteins to promote NF-κB signaling in tumor cells [1,40,41]. UBE2S promotes proliferation and survival of lung adenocarcinoma cells [14]. To investigate the role of UBE2S in NF-κB signaling, we analyzed protein levels of UBE2S, IκBα, and phosphorylated (p)-IκBα in H460 and PC9 cancer cells which expressed different levels of UBE2S ( Figure  2A). Protein levels of IκBα were higher in H460 than PC9 cells, which were opposite results than with UBE2S. Protein levels of p-IκBα were similar in these two cancer cells lines ( Figure 2B). In a further analysis of p65 protein levels in the cytosol and nuclei, PC9 cells expressed higher p65 than H460 cells in the cytosol and nuclei ( Figure 2C). A wound-healing experiment showed higher migratory abilities of PC9 than H460 cells ( Figure 2D, E). Collectively, these results suggested that UBE2S might be involved in regulating tumorigenesis through NF-κB signaling without IKK activation in lung cancer cells.

UBE2S activates NF-κB signaling by targeting IκBα
In our present data, we found that expression levels of UBE2S and IκBα, but not p-IκBα, were opposite in H460 and PC9 cells. UBE2S is an E2 ligase and regulates degradation of proteins through polyubiquitination. To elucidate whether UBE2S regulates the degradation of IκBα and activates NF-κB signaling to promote tumorigenesis, we treated PC9 cells with the IKK inhibitors, SC514 and PS1145, to reduce IκBα phosphorylation. Protein levels of p-IκBα did not signi cantly differ between the control and treatment groups ( Figure 3A, B). In an earlier report, UBE2S was activated by Akt and could be inhibited by Ly294002 [10]. We found that UBE2S protein levels had decreased but IκBα had increased after treatment with 5 and 10 μM Ly294002 ( Figure 3C). These data suggested that UBE2S might regulate activation of NF-κB signaling in PC9 cells without IKK-induced activation. To further analyze the interaction of UBE2S and IκBα, we knocked-down and overexpressed UBE2S in PC9 and H460 cells to analyze IκBα expression. We found that IκBα protein levels increased by knockdown with two UBE2S small interfering (si)RNAs in PC9 cells ( Figure 3D). UBE2S overexpression in H460 cells showed that IκBα protein levels decreased in a dose-dependent manner ( Figure 3E). Overexpression of UBE2S and treatment with MG132 in H460 cells retained IκBα protein levels ( Figure 3F). Collectively, these results showed that UBE2S might target IκBα for degradation and promoted activation of NF-κB signaling.

Binding of UBE2S and IκBα
To identify whether UBE2S interacts with IκBα, we prepared an immunostaining experiment in COS-1 cells. The pcDNA3-UBE2S-ag and pcDNA3-IκBα -HA plasmids were transfected into COS-1 cells, and we analyzed localization of UBE2S (green) and IκBα (red) using anti-ag and anti-hemoagglutinin (HA) monoclonal antibodies. Expressions of UBE2S and IκBα in COS-1 cells were analyzed by Western blotting ( Figure 4A). Immunostaining experiments showed that UBE2S and IκBα were co-localized in the cytosol ( Figure 4B). In a further analysis of the interaction of UBE2S with IκBα, an IP experiment was carried out in H460 and PC9 cells. UBE2S was precipitated with IκBα in H460 and PC9 cells ( Figure 4C). To investigate whether UBE2S directly binds IκBα, GST-tagged UBE2S (GST-UBE2S) and His-tagged IκBα (His-IκBα) were individually expressed and puri ed in an Escherichia coli expression system. Coomassie blue (CBB) staining and Western blot analysis using anti-GST and anti-UBE2S antibodies revealed the expression of GST-UBE2S ( Figure 4D). His-IκBα expression was also detected by CBB staining and Western blotting using anti-His and anti-IκBα antibodies ( Figure 4E). An in vitro pull-down assay was conducted to analyze the binding of UBE2S with IκBα in PC9 cells. GST-UBE2S in agarose beads was shown to pull-down IκBα in PC9 total cell lysate ( Figure 4F). An in vitro binding assay was also conducted to elucidate the binding of UBE2 and IκBα. GST-UBE2S in agarose beads was shown to bind with the puri ed IκBα protein ( Figure 4G). These data showed the direct binding of UBE2S with IκBα.

UBE2S activates NF-κB signaling and EMT signaling in lung adenocarcinoma cells
Earlier studies reported that EMT signaling was activated by NF-κB signaling [42]. In our present data, UBE2S directly bound to IκBα to degrade and promote activation of NF-κB signaling. To analyze UBE2S' regulation of activation downstream genes of NF-κB signaling, we prepared overexpression and knockdown experiments to analyze p65 expression in PC9 and H460 cells. UBE2S overexpression in H460 cells produced increased protein levels of p65 in the cytosol and nuclei ( Figure 5A). UBE2Sknockdown in PC9 cells produced decreases in the p65 protein in the cytosol and nuclei ( Figure 5B). We further prepared a luciferase report assay to analyze activation of NF-κB signaling by UBE2S in H460 and PC9 cells. pGL3-5xκB-Luc plasmid DNA contains ve repeats of κB-binding sites which are activated by NF-κB signaling. The transactivation activity increased 13-, 23-, and 35-fold after transfection with 0.25, 0.5, and 1 μg of pGL3-5xκB-Luc compared to pGL3-basic in H460 cells ( Figure 5C). With the combined transfection with 0.5, 1, and 2 μg pcDNA3-UBE2S-ag and 0.5 μg of pGL3-5xκB-Luc plasmid DNA, the transactivation activity of pGL3-5xκB-Luc increased 1.4-, 1.8-, and 2.9-fold compared to the control group (pGL3-5xκB-Luc only) in H460 cells ( Figure 5D). In PC9 cells, the transactivation activity of pGL3-5xκB-Luc increased 40-, 72.2-, and 81-fold after transfection with 0.25, 0.5, and 1 μg pGL3-5xκB-Luc compared to pGL3-basic ( Figure 5E). UBE2S-knockdown with UBE2S siRNA (siUBE2S#1 and -#2) decreased the transactivation activity to 42% and 53% of that of the control group (control siRNA) ( Figure 5F). These results show that UBE2S activated NF-κB signaling in lung adenocarcinoma cells. In our earlier report, we found that UBE2S promoted migration and invasion through EMT signaling in A431-III cells [28]. Collectively, UBE2S might activate NF-κB and EMT signaling in lung adenocarcinoma cells.

UBE2S promotes EMT signaling and metastasis of lung adenocarcinomas in zebra sh xenograft model
In our and others' reports, UBE2S activated EMT signaling in cancer cells [28,43]. To investigate the function of UBE2S in metastasis of lung adenocarcinoma cells, we prepared a wound-healing experiment to analyze the migratory abilities of PC9 cells. UBE2S-knockdown by UBE2S siRNA (siUBE2S#1 and -#2) decreased the migratory abilities to 42% and 23% that of control siRNA (Con) in PC9 cells ( Figure 6A, B). To investigate the function of UBE2S in activating EMT signaling in lung cancer cells, we prepared overexpression and knockdown of UBE2S in H460 and PC9 cells. UBE2S overexpression in H460 cells increased MMP-9 and TWIST protein levels. E-Cadherin (E-Cad) protein levels also decreased ( Figure 6C).
Knockdown of UBE2S by UBE2S siRNA (siUBE2S#1 and -#2) decreased MMP-9 and TWIST protein levels. In contrast, E-Cad protein levels increased ( Figure 6D). These data suggested that UBE2S might promote EMT signaling in lung adenocarcinoma cells. In our earlier reports, zebra sh was shown as an e cient in vivo tumor cell migration model [30,44]. To further investigate metastasis of lung adenocarcinoma cell in vivo, we prepared in vivo xenograft zebra sh model. PC9 cells was transfected with control (Con) and UBE2S siRNA (siUBE2S#1, #2) for 24 h. Tumor cells were collected and stained with CM-DiI (Figure 6Cab) and microinjected into yolk of 2 days post-fertilization (dpf) zebra sh larvae. The migrative tumor cells were identi ed at 3-dpf zebra sh larvae (Figure 6Cc-f). The metastatic tumor cells were signi cantly decreased to 32% and 33% compared to control group (Con) (Figure 6Cg). These results suggested that UBE2S might activate the metastasis of lung adenocarcinoma cells.

Discussion
This study reveals novel non-canonical activation of NF-κB signaling by UBE2S. NF-κB signaling was activated by UBE2S in lung adenocarcinoma cells without IKK-induced phosphorylation or degradation of IκBα. NF-κB and EMT signaling were activated by UBE2S. UBE2S in lung cancer patients was shown to be correlated with a poor survival rate in TCGA database. Contrasting simultaneous expressions of IκBα with UBE2S were observed and affected p65 accumulation in nuclei of lung adenocarcinoma cells. IκBα and p-IκBα protein levels were not affected by treatment with the speci c IκB kinase (IKK) inhibitors, SC514 and PS1145, in PC9 cells, but they were reduced by treatment with Ly294002, which was reported to inhibit phosphorylation of UBE2S by Akt [10]. Knockdown and overexpression of UBE2S in PC9 and H460 cells also affected IκBα protein levels, accumulation of p65 in nuclei, and transactivation activity of the 5x-NF-κB-binding motif containing the promoter. EMT marker expressions were also regulated by UBE2S.
Immunoprecipitation, immunostaining, in vitro pull-down, and in vitro binding assays showed the direct binding of UBE2S with IκBα. Collectively, we hypothesized novel non-canonical activation of NF-κB signaling in lung adenocarcinomas (Figure 7). In low-malignancy lung adenocarcinoma cells, low UBE2S expression failed to bind with IκBα for degradation. IκBα bound with p65 and p50. Weak IKK signaling was activated to process basal NF-κB signaling. In malignant lung adenocarcinoma cells, high UBE2S expression was bound to IκBα for degradation. p65 and p50 entered nuclei to activate NF-κB signaling.
UBE2S is an E2 ubiquitination ligase and is involved in different signaling pathways during tumorigenesis in many tumor types. UBE2S mediated tumor progression via Sox6/β-catenin signaling in endometrial cancer [45], was associated with malignant properties of breast cancer [46], enhanced ubiquitination of p53 to exert oncogenic activities in hepatocellular carcinoma [47], was mediated by VHL/Hif1α/Stat3 signaling to promote EMT signaling [43], and promoted the proliferation and survival of lung adenocarcinoma cells by activating multiple gene expressions [14]. In our earlier report, UBE2S activated the migration and invasion of squamous cell carcinoma cells through EMT signaling [28]. Although UBE2S is involved in tumorigenesis of multiple tumors, the detailed mechanism is still unclear. In this study, we showed novel non-canonical activation of NF-κB signaling by UBE2S which promoted the metastasis of adenocarcinoma cells. This nding provides an important function of UBE2S in regulating and activating NF-κB signaling in lung adenocarcinomas. Therefore, we did not provide the distinct binding position of UBE2S and IκBα; the PEST domain of the C-terminal region of IκBα, which was suggested to be a binding region of free IκBα for degradation [24], might be a potential site for UBE2S binding for degradation. This suggestion requires further investigation.
NF-κB signaling was reported to regulate cancer progression in multiple cancers. High constitutive NF-κB activity and low IκBα expression were found in some cancer cell lines [48,49]. Phosphorylation and degradation of IκBα by IKK to activate NF-κB signaling constitute the canonical pathway. The noncanonical NF-κB pathway might not involve degradation of IκBα and may be regulated by p100 [50]. However, a report showed that NF-κB signaling was activated in IKK1/2-devoid mouse embryo broblasts. Degradation of IκBα was induced by doxorubicin-induced DNA damage and was inhibited by the phosphatidylinositol 3-kinase inhibitor, LY294002 [51]. In a previous report, UBE2S was regulated by Akt and was associated with the DNA damage response by the Ku70 complex [10]. In our present data, Ly294002 reduced protein levels of UBE2S and increased IκBα in lung adenocarcinoma cells. Knockdown and overexpression of UBE2S in adenocarcinoma cells increased nuclear entry of p65 and activated the transactivation activity of the NF-κB reporter. This activation further induced EMT signaling and metastasis of lung adenocarcinoma cells in zebra sh xenograft tumor model. These results showed a new tumor progression mechanism, and UBE2S might be a new therapeutic target in lung adenocarcinoma.

Conclusions
UBE2S in lung adenocarcinoma patients was negatively related to the survival rate. Higher UBE2S expressed in lung adenocarcinomas was directly binding with IκBα to activate NF-κB signaling. The downstream EMT signaling was activated to promote metastasis of cancer cells. UBE2S might be a potential therapeutic target for lung adenocarcinomas.

Declarations
Ethics approval and consent to participate Not applicable.

Consent for publication
Not applicable.

Availability of data and materials
Not applicable.

Con ict of interests
The authors declare that they have no con icts of interest.     of PC9 lung cancer cells in zebra sh larvae. The PC9 cells were stained with CM-DiI (a-b). PC9 cells knockdown with siUBE2S (e-h) or control siRNA (c-d) were microinjected into 3 dpf zebra sh larvae and then analyzed the migrative PC9 cells at 7 dpf zebra sh larvae. (i) Statistical analysis of metastatic PC9 cells in zebra sh larvae by knockdown with control (control) and UBE2S siRNA (siUBE2S#1, #2).

Figure 7
UBE2S activated nuclear factor (NF)-κB signaling in lung adenocarcinomas. In low-malignancy lung adenocarcinoma cells, low ubiquitin-conjugating enzyme E2S (UBE2S) expression failed to bind with inhibitor of nuclear factor-κBα (IκBα) for degradation. IκBα was bound with p65 and p50. Weak inhibitor of κB kinase (IKK) signaling was activated to process basal NF-κB signaling. In malignant lung adenocarcinomas, high UBE2S expression was bound with IκBα for degradation, then p65 and p50 entered nuclei to activate NF-κB signaling.