Tip110 is identified as an interacting partner with HIV-1 Tat protein that has been reported to enhance NF-κB activity (28, 29). USP15, another Tip110-interacting protein, that has been reported to regulate NF-κB activity (27, 30–32). In addition, Tip110 interacts with and/or regulates other proteins which are known to play a role on NF-κB by targeting NF-κB signaling components (9, 10, 21, 24, 26, 27). All those findings prompted us to investigate the possible role of Tip110 in the NF-κB signaling pathway. Transfection of luciferase-reporter vector pGL3-NF-κB(3)-Luc containing NF-κB responsive element with an increasing amount of Tip110 in 293T enhanced NF-κB transcriptional activity in a dose-dependent manner ( Figure 1A ). Treatment of Tip110 expressing cells with TNF-α further augmented NF-κB transcriptional activity compared to untreated cells ( Figure 1B ). On the contrary, Tip110 knockdown by siRNA in 293T led to inhibition in the NF-κB transcriptional activity in response to TNF-α treatment ( Figure 1C ). Interestingly, deletion of Tip110 nuclear localization signal (NLS) ( Figure 2A ) abolished the NF-κB transcriptional activity ( Figure 1D ).

Even though NF-κB and AP-1 transcription factors are regulated by different mechanisms, they appear to be activated by the same multitude of stimuli (33–35). Indeed, the activation of JNK by inflammatory cytokines or by stress is often accompanied by the nuclear translocation of NF-κB, and many genes require the concomitant activation of AP-1 and NF-κB, suggesting that these transcription factors work cooperatively (36, 37). To determine whether Tip110 expression would also regulate AP-1 transcriptional activity, 293T were transfected with a luciferase-reporter vector containing the AP-1 responsive element with an increasing amount of Tip110. The results showed that Tip110 expression suppressed AP-1 transcriptional activity in a dose-dependent manner ( Figure 1E ). A similar result was also found under phorbol 12-myristate 13-acetate (PMA) treatment ( Figure 1F ). Under our experimental settings, TNF-α showed low activation of AP-1 luciferase activity in 293T, and the effect was not significant; therefore, we used PMA, a known activator for AP-1 transcriptional activity. These data suggest that Tip110 activates NF-κB transcriptional activity through Tip110 nuclear localization but inhibits AP-1 transcriptional activity.

Tip110 contains 12 half-a-tetratricopeptide repeats (HAT) and an NLS at the N terminus and two RNA recognition motifs (RRM) at the C terminus. To identify the Tip110 domains involved in the regulation of NF-κB activity, a series of Tip110 mutant proteins ( Figure 2A ( 25, 28 ), were analyzed. Similarly, 293T were transfected with wild-type Tip110 and its mutants together with the NF-κB reporter. Compared to wild-type Tip110, deletion of Tip110 N-terminal domain (Tip110ΔN), including the NLS, resulted in a decrease in the NF-κB activity ( Figure 2B ). Deletion up to 557 aa from the C-terminal, including the NLS, completely abolished NF-κB activity similar to the cDNA3-transfected control. While deletion of the C-terminal domain on Tip110 (Tip110ΔC and Tip110Δ785-963 mutants) led to higher NF-κB activity than the wild-type ( Figure 2B ). Western blotting was performed to confirm the expression of Tip110 mutant proteins ( Figure 2C ). These results demonstrated that an approximately 300 aa at the N-terminal domain, which includes the NLS region on Tip110, 556-785 aa, is required for the NF-κB activation.

We have previously shown that Tip110 is ubiquitinated and this ubiquitination is regulated by its associated-protein USP15 (25). Therefore, we then determined whether Tip110 ubiquitination would be affected by its intracellular localization. 293T were transfected with UB.HA alone or plus Tip110.His or Tip110ΔNLS.His on the presence or absence of pUSP15.Myc. Cells were treated with MG132, an inhibitor of 26S proteasome degradation, then collected for cell lysates. Tip110 expression was confirmed by Western blotting ( Figure 2D , Input). Immunoprecipitation using anti-His antibody followed by Western blotting using anti-HA antibody led to the detection of a highly ubiquitinated Tip110ΔNLS mutant protein compared to wild-type Tip110 and the expression of USP15 deubiquitinated both Tip110 and its cytoplasmic mutant Tip110ΔNLS ( Figure 2D , IP). Then we used the same strategy to determine whether TNF-α stimulation would alter Tip110 ubiquitination. 293T were transfected with Tip110.His or its control plasmid then either treated with TNF-α or left untreated. Immunoprecipitation assay data showed little effect in ubiquitinated Tip110 upon TNF-α stimulation compared to unstimulated cells with reference to the amount of Tip110 input ( Figure 2E ). Those results indicate that Tip110 ubiquitination level affects its intracellular localization and might be involved in NF-κB activity.

We and others have shown that a number of Tip110-associated proteins translocated to the nucleus when they co-expressed with Tip110 such as USP15, USP4, and YB-1 ((12, 25, 38), data not shown for YB-1). In addition, NF-κB activation is mainly associated with nuclear translocation of endogenous NF-κB p65 (2). Thus we analyzed the distribution of NF-κB p65 between the cytoplasm and nucleus in relation to Tip110 expression and TNF-α treatment by using an immunofluorescence assay. Expression of endogenous NF-κB p65 localized mainly in the cytoplasm and weak expression was also detected in the nucleus due to its continuous shuttling between both compartments (39) ( Figure 3A–I ). Treatment of cells with TNF-α for 30 min caused accumulation of NF-κB p65 in the nucleus ( Figure 3A–II ). Tip110 overexpression showed no significant effect on p65 nuclear accumulation ( Figure 3A–III ). However, treatment of Tip110 expressing cells with TNF-α greatly enhanced NF-κB p65 nuclear accumulation compared to control expressing cells ( Figure 3A–IV ). On the other hand, Tip110 knockdown by siRNA in the presence or absence of TNF-α treatment has no significant effect on NF-κB p65 protein expression and localization pattern compared to si-RNA control cells ( Figures 3B–III, IV ). Then we determined whether Tip110 would be associated with NF-κB p65. 293T were transfected with Tip110 or vector control then treated with TNF-α for 20 hr. Cells were lysed, and an immunoprecipitation assay was performed. Tip110 and p65 were expressed ( Figure S1A ), but there was no complex formation between Tip110 and p65 ( Figure S1B ). These data together demonstrated that Tip110 expression altered NF-κB p65 nuclear accumulation only in the presence of TNF-α stimulation, suggesting that other mechanisms may also contribute to NF-κB activation by Tip110.

The canonical NF-κB transactivation pathway involves IκBα phosphorylation and degradation (2). Thus, we next investigated whether Tip110 expression would alter IκBα expression and stability in the presence TNF-α stimulation. 293T were transfected with Tip110 expression vector or its control and treated with TNF-α for various lengths of time. Western blotting showed that under TNF-α stimulation, phosphorylated IκBα increased under the context of Tip110 expression ( Figure 4A ), which is paralleled with changes in the IκBα protein stability. To further investigate whether this effect could also be observed in normal primary cells which retain a native signaling pathways, we used primary mouse embryonic fibroblasts (MEF) under the context of limited amount of Tip110. Cells were transfected with siRNA-Tip110 or control siRNA, treated with TNF-α at various lengths of time up to 120 min, and harvested for Western blotting. Similarly, Tip110 knockdown led to decreases in IκBα phosphorylation and subsequent stabilization of the total amount of IκBα in the presence of TNF-α stimulation ( Figure 4B ). These results also showed that Tip110 had no significant effects on the amount of NF-κB p65 protein. These data suggest that Tip110 alters the phosphorylation status and stability of IκBα and subsequently regulates NF-κB activity under TNF-α stimulation.

Unexpectedly, no significant effect was detected on the basal level of IκBα in the absence of TNF-α treatment ( Figures 4A, B ). Thus to further examine whether Tip110 expression regulates IκBα protein turnover rate, a cycloheximide (CHX) chase assay was performed. 293T were transfected with IκBα.HA alone or plus Tip110.His or Tip110ΔNLS expression vectors, treated with CHX, to inhibit new protein synthesis, for various lengths of time, and harvested to determine Tip110 and IκBα protein expression by Western blotting. Tip110 overexpression accelerated IκBα degradation compared to its transfection control ( Figures 4C, D ), while expression of Tip110ΔNLS that previously showed an inhibitory effect on NF-κB activity ( Figure 1D ) showed no apparent effects on IκBα protein level compared to its transfection control ( Figures 4D, E ). Further, we observed that Tip110ΔNLS expression caused more accumulation of IκBα protein before the beginning of CHX treatment compared to Tip110 and the control ( Figures 4C, E ). Then we examined whether Tip110 or Tip110ΔNLS expressions affect the turnover rate of the nuclear targeted IκBα by using IκBα-Nuc.Myc expression plasmid. Interestingly, CHX chase assay showed that Tip110 expression stabilized the nuclear IκBα compared to its transfection control while the expression of Tip110ΔNLS accelerated nuclear IκBα degradation compared to Tip110 but still higher than the transfection control ( Figures 4E, F ). In addition, qRT-PCR results showed that Tip110 expression only had a little effect on the IκBα mRNA level at 2 hr, but had no effect at 24 hr following TNF-α treatment compared to the control transfection ( Figure 4G ). These data showed that the Tip110 affects IκBα phosphorylation under TNF-α treatment and the regulation of IκBα protein stability.

We showed that nuclear localization was required for potentiation of TNF-α-induced NF-κB activation by Tip110 and different effect of Tip110 on the IκBα protein stability is dependent on its localization ( Figures 1 , 4 ). The TNF-α-induced NF-κB signaling cascade is initiated within the cytoplasm (37). Furthermore, Tip110 interacts with USP15 and regulates its localization between the cytoplasm and nucleus (25, 38), while USP15 is shown to regulate NF-κB activity (27, 40, 41). Those findings together raised the possibility that USP15 might be the link between Tip110 and NF-κB activation. To test this possibility, 293T were transfected with pGL3-NF-κB(3)-Luc, Tip110, USP15, or USP15 mutants ( Figure 5A ). The data showed that the expression of USP15 inhibited NF-κB transcriptional activity ( Figure 5B ). Interestingly, increasing the amount of USP15 expression plasmid in the presence of Tip110 led to gradual decreases of Tip110-induced NF-κB activity in a dose-dependent manner ( Figure 5B ). Then we determined which USP15 domain was involved in the inhibition of NF-κB activity. The results showed that the N-terminal domain of USP15, which contains the domain present in ubiquitin-specific proteases (DUSP) and ubiquitin-like domain (UBL) ( Figure 5A ), was responsible for this inhibition as deletion of this domain (USP15ΔN) enhanced the NF-κB activity ( Figure 5C ). While deletion of the C-terminal domain, USP15ΔC, led to inhibition of NF-κB activity. Furthermore, co-expression of Tip110 with the USP15ΔC mutant led to reduction on the Tip110-induced NF-κB activity. In contrast, co-expression with the USP15ΔN mutant had no such effects ( Figure 5C ), indicating that the N-terminal domain of USP15, which physically associates with Tip110 (25), is responsible for USP15 inhibitory function on Tip110-induced NF-κB activity.

Next, we examined the role of Tip110 mutants on NF-κB activity in the context of the USP15 expression. Expression of Tip110 mutant proteins in the presence of USP15 attenuated their NF-κB basal transcriptional activity ( Figure 5D ). Interestingly, co-expression of Tip110ΔNLS or Tip110Δ274-963 mutants with USP15 led to decreases in the NF-κB activity compared to the mutants alone, indicating that USP15 expression may also affect cytoplasmic expressed Tip110 mutants. To further verify the opposing effects of Tip110 and USP15 on the NF-κB activity, we utilized NF-κB p65 protein, which is known to activate NF-κB transcriptional activity in a reporter gene assay. The results showed an enhancement in the transcriptional activation of NF-κB by the co-expression of p65 and Tip110. Expression of USP15 abrogated this enhancement effect, yet the level was still higher than Tip110 alone due to the effect of p65, indicating that p65 expression reversed the inhibitory effect of USP15 on Tip110-induced NF-κB activity ( Figure 5E ). Then, we investigated whether the abrogation of Tip110-enhanced NF-κB activity by USP15 was due to modulation of IκBα protein stability. 293T were transfected with Tip110 or USP15 alone, or in combination, then treated with TNF-α for 15 or 30 min. Similar to our previous observation, Tip110 expression led to increases in p-IκBα paralleled with decreases in IκBα protein level. While co-expression of USP15 and Tip110 reduced the IκBα protein level compared to the USP15 expressing cells only ( Figure 5F ), yet this reduction was not paralleled with increased in the p-IκBα, which is more likely due to the effect of Tip110 on the IκBα phosphorylation. Interestingly, expression of USP15 on the presence of TNF-α led to increase on IκBα phosphorylation compared to untreated cells which is not parallel with decrease on IκBα level. It has been reported that overexpression of USP15 potentiated TNF-α-induced NF-κB activation (42). In addition, deletion of USP15 N-terminal domain enhanced NF-κB activity ( Figure 5C ). Thus it’s possible that USP15 regulates NF-κB at multiple mechanisms. These data further confirming the opposing effect of Tip110 and USP15 on IκBα stability.

As we previously alluded to, Tip110 expression promoted the nuclear translocation of USP15 where both are associated (25) and the above observations prompted us to examine whether Tip110 also targeted nuclear IκBα to regulate NF-κB activity. 293T were transfected same as above then treated with leptomycin B (LMB), a nuclear export inhibitor found to sequester NF-κB/IκBα complexes in the nucleus (43), alone or together with TNF-α. Then the cells were prepared for an NF-κB reporter gene assay. As reported previously, treatment of cells with LMB inhibited NF-κB activity while further inhibition was observed by combined treatment with TNF-α ( Figure S2A ) (39, 44). However, LMB treatment had no significant effect on NF-κB activity under the context of Tip110 expression compared to untreated cells, but inhibition of NF-κB activity was observed with combined treatment with TNF-α. Furthermore, LMB treatment led to decreases in p65-induced NF-κB transcriptional activity while Tip110 co-expression resulted in a marginal increase compared to p65 expression alone ( Figure S2B ). These data indicate that USP15 expression abrogated the effect of Tip110-induced NF-κB activity and that Tip110 could regulate both nuclear and cytoplasmic IκBα protein stability.

To gain further insight into the consequence of Tip110 effect in the TNF-α-induced NF-κB signaling pathway, we assessed the NF-κB dependent gene expression of various inflammatory cytokines when Tip110 was knocked down. 293T were transfected with control or Tip110 siRNA then stimulated with TNF-α for short (2 and 4 hr) and long (24 hr) lengths of time due to different stimulation kinetics exhibited by various cytokines in response to TNF-α treatment. The cells were harvested for the determination of mRNA level by qRT-PCR. In line with the previous findings, the overall expression of NF-κB target genes such as TNF-α and IL-6 were substantially decreased in Tip110 knockdown cells as compared to their respective control ( Figure 6 ). While IL-8 expression showed no significant reduction, indicating that these differences could reflect the quantitative requirement for NF-κB by different target genes under a limited amount of Tip110 or that the regulation of IL-8 by NF-κB under the context of Tip110 acted differently than other cytokines. Recently, we reported that knockdown of Tip110 in melanoma cell line resulted in the up-regulation of IL-8 expression while no such effect was observed in the cell lines from other cancer types (22).

The possible role of Tip110 on NF-κB signaling pathway was further illustrated by an analysis of whole-transcriptome array data using Affymetrix Expression Console (TAC) and pathway analysis software that was generated previously during our investigation of the embryonic lethality instigated by knockout of Tip110 in mice (45). The microarray analysis data from the Tip110-knockout mESC revealed that several genes associated with NF-κB regulatory pathways were highly downregulated. Specifically, FGF4 (-17.8 fold) and GSTA4 (-46.7 fold) related to MAPK signaling cascades ( Figure 7A ), PMI (-4.97 fold) and FLNA (-7.26 fold) related to the TNF-α-NF-κB pathway ( Figure 7B ), CDK1 (-9.99 fold) and CYCT (-40.0 fold) related to p53 signaling pathway ( Figure 7C ), SPRY2 (-8.22 fold) related EGFR1 signaling pathway ( Figure S3A ), CDK5 (-8.4 fold) related to IL-6 signaling pathway ( Figure S3B ), and SSP1 (-39.2 fold) related to TGF-β1 receptor signaling pathway ( Figure S3C ). Interestingly, analysis of TNF-α-NF-κB, EGF, and TGF-β signaling cascades demonstrated that several genes within the cascade were also upregulated in the knockout Tip110 cells ( Figures 7D and S3D ), indicating there is a compensatory mechanism that was impacted these signaling pathways in the absence of Tip110, which may also explain the little effect of Tip110 on the basal level of IκBα and the gene expression of cytokines ( Figure 6 ).

HEK 293T cell line was purchased from American Tissue Culture Collection (ATCC) and cultured in Dulbecco’s modified Eagle’s medium (DMEM) containing 10% fetal bovine serum (FBS). Mouse embryonic fibroblasts (MEF) isolation was reported previously (25). Briefly, MEF cells were isolated from embryos of 13.5-14.5 days post coitus pregnant mice and cultured in DMEM with 10% FBS and 50 μM 2-β mercaptoethanol. All media contained 100 IU/ml penicillin, 100 μg/ml streptomycin. Plasmid DNA transfection was performed using the standard calcium phosphate precipitation method for 293T. MEF cells were transfected using Lipofectamine LTX (Invitrogen) according to the manufacturer’s instructions. Si-RNA transfection was performed using Lipofectamine 2000 according to the manufacturer’s instructions (Invitrogen).

Tip110.His, Tip110ΔN.His, Tip110ΔC.His, GFP-Tip110ΔNLS, Tip110ΔNLS.His, and GFP-Tip110 plasmids were described elsewhere (28, 74). Tip110Δ274-963.His, Tip110Δ387-963.His, Tip110Δ557-963.His, and Tip110Δ786-963.His plasmids were constructed in the backbone of pcDNA3 (Invitrogen) using standard PCR techniques with Tip110.His as a template and using EcoRI and XhoI cloning sites. USP15.Myc (1–952 aa), USP15ΔC.Myc (1–384 aa) and USP15ΔN.Myc (385-952 aa) were described elsewhere (25). NF-κB p65 was a kindly provided by Dr. Michael Klemsz of Indiana University School of Medicine. pNF-κB-luc and pAP-1-Luc were purchased from Clontech Laboratories Inc., CA. UB.HA plasmid was kindly provided by Dr. Mark Hannink of the University of Missouri. IκBα.HA plasmid was kindly provided by Dr. Michael Karin of the University of California (75). The IκBα-Nuc.Myc plasmid was constructed using IκBα.HA as a template, backbone of pCMV-Nuc-Myc (Addgene) and using SalI and NotI cloning sites. On-TARGETplus Tip110 siRNA (L-013447-01) and SiRNA control (D-001810-01) were purchased from Dharmacon.

The firefly luciferase activity was measured using the luciferase assay substrate (Promega, Madison, WI) according to the manufacturer’s instructions. Briefly, cells were washed with ice-cold PBS and lysed with 120 μl 1× firefly luciferase lysis buffer (Promega) at room temperature for 15 min. The lysates were centrifuged at 12,000 × g for 2 min to remove cell debris. The cleared lysates (5 μl) were then mixed with 20 μl firefly luciferase substrate (Promega), and the luciferase activity was measured using an Opticomp luminometer (MGM Instruments, Hamden, CT). The same cell lysates were then used for immunoblotting to confirm the protein expression.

Cells were washed in cold PBS and lysed in lysis buffer (50 mM Tris.HCl, pH 8.0, 280 mM NaCl, 0.5% NP-40, 0.2 mM EDTA, 2 mM EGTA, 10% glycerol, 2 mM PMSF and protease inhibitor cocktail). Lysates were cleared of cell debris by centrifugation at 12,000 x g and fractionated on SDS-PAGE, followed by Western blotting. For detection of Tip110 and its mutants, we used polyclonal and monoclonal Tip110 antibodies (9, 28) because not all of the Tip110 mutants were able to be recognized by the same antibody. Also, not all the constructed mutants in our study were tagged. Anti-His (abm), anti-HA (Santa Cruz Biotechnology; sc-7392), anti-p65 (Abcam; ab-7970), anti-Myc (Santa Cruz Biotechnology; sc-9E10), anti-USP15 (Santa Cruz Biotechnology; sc-100629), and anti-actin (Sigma A1978). For immunoprecipitation, 1 mg of cell lysates were incubated with 1 μg of primary antibody at 4°C on a rotating device overnight, and protein A agarose beads (Upstate) were then added for 45 min. Agarose beads were recovered and washed thoroughly three times with the above lysis buffer and processed for SDS-PAGE and Western blotting analysis.

Cells were transfected on 24-well culture plates, fixed with 4% paraformaldehyde for 20 min, and then permeabilized with 0.5% Triton X-100 for 10 min. After extensive washing with PBS, the cells were blocked with 3% BSA and incubated with anti-p65 and anti-Tip110 antibodies (1:500) for 1 hr at room temperature in a humidified chamber. Then the cells were washed and incubated with either anti-Alexa Fluor 555 or anti-Alexa Fluor 488 secondary antibodies (Invitrogen; 1:500) for 1 hr and then 1 ng/ml 4,6-diamidino-2-phenylindole (DAPI) for 5 min to stain the nuclei. The coverslips were washed with PBS and mounted with fluoromount-G (SouthernBiotech) on glass slides. Fluorescence micrographs were taken using a Zeiss Model Axiovert 200 M microscope.

RNA was extracted from cells using TRizol (Invitrogen) according to the manufacturer’s instructions. RNA (1 μg) was converted to cDNA using the iScript cDNA synthesis kit (Bio-Rad, Hercules, CA) and used as the template for PCR using SsoAdvanced SYBR green Supermix and the CFX96 real-time PCR detection system (Bio-Rad). The quantitative reverse transcription-PCR (qRT-PCR) primers used and their sequences are as follows: for β-actin, 5′-AAA CTG GAA CGG TGA AGG TG-3′ and 5′-AGA GAA GTG GGG TGG CTT TT-3′; for TNF-α, 5′-TCT TCT CGA ACC CCG AGT GA-3′ and 5-CCT CTG ATG GCA CCA CCA G-3′; for IL-6, 5′-ACA ACA AAT TCG GTA GAT CCT CG-3′ and 5′-AGC CAT CTT TGG AAG GTT CAG G-3′; for IL-8, 5′-TGC CAA GGA GTG CTA AAG-3′ and 5′-TCT CAG CCC TCT TCA AAA-3′; and, for IκBα, 5′- ACC AAC TAC AAT GGC CAC AC-3′ and 5′- ATT ACA GGG CTC CTG AGC AT-3′. Threshold cycle (C_T ) values were calculated using Bio-Rad CFX manager software. The 2^{−ΔΔC T} value was calculated to represent the fold change of the target gene mRNA compared to untreated siRNA control and normalized using β-actin as the reference.

Cells were treated with 20 µg/ml cycloheximide (Sigma), and then the whole-cell lysates were prepared after different durations. The lysates were subjected to Western blotting analysis to identify the IκBα turnover rate alone or in the presence of Tip110 and Tip110ΔNLS. The amount of the remaining proteins was quantified by using densitometry and normalized to the β-actin.

Where appropriate, values are expressed as means ± standard deviations (SD) from triplicate samples. All comparisons were made based on the control using a two-tailed Student’s t-test and two-way analysis of variance (ANOVA) as appropriate. P values of <0.05 were considered statistically significant, <0.01 highly significant, and <0.001 strongly significant. If statistical significance (P <0.05) was determined by ANOVA, the data were further analyzed by Turkey’s post hoc test for multigroup Comparisons.