An RNA Interference Screen Identifies the Deubiquitinase STAMBPL1 as a Critical Regulator of Human T-Cell Leukemia Virus Type 1 Tax Nuclear Export and NF-κB Activation

Biological reagents and antibodies.

Human embryonic kidney cells (HEK 293T) and Jurkat T cells were purchased from ATCC. The HTLV-1-transformed cell line MT-2 was described previously (20). 293T cells were cultured in Dulbecco's modified Eagle's medium (DMEM); Jurkat and MT-2 cells were cultured in RPMI medium. Medium was supplemented with fetal bovine serum (10%) and penicillin-streptomycin (1%). Expression vectors encoding pCMV4-Tax, green fluorescent protein (GFP)-Tax and Flag-Tax, hemagglutinin (HA)-Ub, HA-Ub K48 only, κB-luciferase (Luc), pU3R-Luc, pRL-TK (thymidine kinase), and pCMV4-p100 have all been described previously (2, 20, 41). pCDNA-Myc ELKS expression plasmid was a gift from I. Verma (8). Flag/HA-STAMBPL1 was from J. Wade Harper and obtained from Addgene (plasmid 22559) (45). Flag-Tax1 was a gift from U. Bertazzoni (48). The following antibodies were used in this study: anti-STAMBPL1 (ab99172; Abcam), anti-p100/p52 (Santa Cruz Biotechnology), anti-β-actin (AC15; Abcam), anti-IκBα (C-21; Santa Cruz Biotechnology), anti-phospho-IκBα (14D4; Cell Signaling), and anti-TATA-binding protein (anti-TATA-BP) (MAB3658; Millipore). The monoclonal anti-Tax antibody was prepared from a Tax hybridoma (168B17-46-34) that was obtained from the AIDS Research and Reference Program, NIAID, National Institutes of Health. Alexa Fluor 555-conjugated donkey anti-mouse IgG, Alexa Fluor 488-conjugated donkey anti-rabbit IgG, and Alexa Fluor 647-conjugated donkey anti-rabbit IgG were purchased from Molecular Probes/Invitrogen. DAPI(4′,6-diamidino-2-phenylindole) and MG-132 were purchased from EMD Biosciences. Cycloheximide (CHX) was purchased from Sigma. Recombinant TNF-α and interleukin-1β (IL-1β) were purchased from R&D Systems. The SMARTpool siRNA human deubiquitinating enzyme library, control scrambled siRNA, and ELKS and STAMBPL1 SMARTpool siRNAs were purchased from Dharmacon/Thermo Scientific.

Transfections and luciferase assays.

Transient transfections in 293T cells were performed with GenJet (SignaGen) according to the manufacturer's instructions. Jurkat and MT-2 cells were transfected with FuGene HD (Roche) according to the manufacturer's instructions. For siRNA transfections, cells were transfected with 20 or 40 pmol of control or STAMBPL1 siRNA using Lipofectamine 2000 (Invitrogen). For luciferase assays, cells were lysed 24 h after transfection using 1× passive lysis buffer (Promega). Luciferase activity was measured with the dual-luciferase assay system according to the manufacturer's instructions (Promega). Firefly luciferase values were normalized based on the Renilla luciferase internal control values. Luciferase values are presented as “fold induction” relative to the untreated control transfected with empty vector or control siRNA.

qRT-PCR.

Quantitative real-time PCR (qRT-PCR) was performed essentially as described previously (5, 19). Total RNA was isolated from cells using the RNeasy minikit (Qiagen). RNA was converted to cDNA using the First Strand cDNA synthesis kit for RT-PCR (avian myeloblastosis virus [AMV]; Roche). Real-time PCR was performed with a LightCycler instrument (Roche) using LightCycler FastStart DNA MasterPLUS HybProbe (Roche) and gene-specific TaqMan probes (Applied Biosystems). Gene expression for IκBα was normalized to the internal control 18S rRNA.

CHX chase assay.

CHX chase assays were performed as described previously (40). Cells were treated with CHX (100 μg/ml) for various times 2 days after transfection. Cells were lysed in radioimmunoprecipitation assay (RIPA) buffer, and immunoblotting was conducted with anti-Tax.

Immunoblotting.

Immunoblotting was performed essentially as described previously (20, 42). Whole-cell lysates were resolved by SDS-PAGE, transferred to nitrocellulose membranes, blocked in 5% milk or bovine serum albumin (BSA) (for phospho-specific antibodies), incubated with the indicated primary and secondary antibodies, and detected with Western Lightning enhanced chemiluminescence reagent (Perkin Elmer).

Immunostaining and confocal microscopy.

Cells were seeded in 6-well plates onto poly-l-lysine-coated coverslips (BD Biosciences) and transfected with either siRNAs or plasmids. After 24 h, the plates were briefly centrifuged, and cells were washed with (phosphate-buffered saline (PBS) and fixed in 1% paraformaldehyde for 10 min at room temperature. The fixed cells were permeabilized with PBS containing 0.2% Triton X-100 (Sigma) and incubated in PBS containing 10% normal donkey serum (Jackson ImmunoResearch) and 0.1% Triton X-100 to prevent nonspecific binding. The primary antibodies were diluted in PBS containing 0.1% Triton X-100 and added to the cells for 1 h, followed by washes in PBS containing 0.1% Triton X-100. Secondary antibodies were added and incubated for 1 h, followed by several washes with PBS containing 0.1% Triton X-100. The coverslips were stained with DAPI to visualize nuclei. Finally, the coverslips were mounted onto slides using AquaPolymount (Polysciences) and analyzed with a Zeiss confocal laser scanning microscope.

Subcellular fractionation.

Cytoplasmic and nuclear extracts were prepared using hypotonic buffer (20 mM HEPES, pH 8.0, 10 mM KCl, 1 mM MgCl₂, 0.1% [vol/vol] Triton X-100, 400 mM NaCl) and insoluble buffer (20 mM Tris [pH 8.0], 150 mM NaCl, 1% [vol/vol] NP-40, 10 mM iodacetamide), respectively. The purity of the obtained fractions was confirmed by examining β-actin for cytoplasmic expression and TATA-binding protein (TATA-BP) for nuclear expression. Cells were lysed in hypotonic buffer to obtain the cytoplasmic fractions, and then the insoluble pellets were washed five times with the same buffer. The pellets were then suspended in nuclear buffer to obtain the nuclear fractions.

Ubiquitination assay.

The ubiquitination assay was performed essentially as described previously (41). Briefly, 293T cells were transfected with Flag-Tax, HA-Ub, and HA-Ub K48-only plasmids. Lysates were subjected to immunoprecipitation (IP) with anti-Flag and immunoblotting (IB) with anti-HA to examine ubiquitination of Tax.

Deubiquitination assay.

The deubiquitination assay was performed as described previously (29). 293T cells were transfected with the indicated plasmids and lysed in RIPA buffer after 48 h, and lysates were subjected to immunoprecipitations (IP) with anti-Flag antibody and protein A-agarose beads (Roche). Flag immunoprecipitates were eluted with 3× Flag peptide (Sigma) and incubated with 1 μg immunoaffinity-purified human STAMBPL1 (Origene) at 37°C for 18 h and immediately subjected to SDS-PAGE and immunoblotting. As a positive control for STAMBPL1 DUB activity, STAMBPL1 was incubated with recombinant K63-linked penta-ubiquitin (Ub⁵) chains (Boston Biochem) for 18 h at 37°C.

Statistical analysis.

All error bars represent the standard deviation (SD) of triplicate samples. Statistical significance was determined by Student's t test. On the figures, * indicates a P value of <0.05 and ** indicates a P value of <0.01.

STAMBPL1 is a positive regulator of Tax-mediated NF-κB signaling.

Polyubiquitination of Tax plays a critical role in IKKγ binding, Tax trafficking, and NF-κB activation (20, 27). However, the regulation of Tax polyubiquitination and NF-κB activation by host proteins is poorly understood. Thus, we used an unbiased approach to identify DUBs that regulate Tax activation of NF-κB signaling. We screened an siRNA library containing 100 siRNAs specific for all known human deubiquitinating enzymes (DUBs). 293T cells were transfected with the siRNA library to evaluate the role of DUBs in Tax-mediated NF-κB activation using a luciferase reporter assay. As expected, a number of negative regulators of Tax-induced NF-κB activation were identified, including the cylindromatosis (CYLD) gene (data not shown) (50). Interestingly, a small number of DUBs were found to be required for Tax activation of NF-κB, including AMSH-LP/STAMBPL1 (STAM binding protein-like 1). STAMBPL1 is a member of the JAMM deubiquitinating enzyme family, which cleaves Lys-63-linked polyubiquitin chains (37). AMSH proteins regulate the recycling of receptors conjugated with K63-linked polyubiquitination chains by opposing ubiquitin-dependent sorting to the lysosome (31). First, we confirmed that siRNA-mediated knockdown of STAMBPL1 indeed impaired Tax-mediated activation of an NF-κB reporter in 293T cells (Fig. 1A).

To further confirm the effect of STAMBPL1 on Tax-induced NF-κB activation, additional NF-κB signaling functional studies were performed. A catalytically inactive mutant of STAMBPL1 was generated in which glutamic acid at position 292 was replaced with alanine (STAMBPL1 E292A) (37). The effects of wild-type STAMBPL1 and the E292A mutant on Tax-induced activation of an NF-κB target gene were evaluated by qRT-PCR (Fig. 1B). Wild-type STAMBPL1, but not the catalytically inactive mutant, enhanced the expression of IκBα mRNA induced by Tax (Fig. 1B). However, overexpression of either STAMBPL1 alone or the STAMBPL1 mutant had no effect on the induction of IκBα mRNA (Fig. 1B). In support of these findings, using an NF-κB reporter assay, wild-type STAMBPL1, but not the catalytically inactive mutant, enhanced Tax-induced NF-κB activation in Jurkat T cells (Fig. 1C). The effects of STAMBPL1 on NF-κB appear to be specific for Tax since siRNA-mediated knockdown of STAMBPL1 had no effect on TNF-α- or IL-1β-induced NF-κB activation (Fig. 1D and E). We next examined the effect of STAMBPL1 overexpression on NF-κB signaling in MT-2 cells, a T-cell line chronically infected with HTLV-1. Wild-type STAMBPL1, but not the E292A mutant, potentiated NF-κB activation in MT-2 cells (Fig. 1F). As a negative control, we overexpressed ELKS (a protein rich in glutamate, leucine, lysine, and serine), an adaptor molecule involved in NF-κB signaling by genotoxic stress, which had no effect on NF-κB activation in MT-2 cells (Fig. 1F). Both wild-type STAMBPL1 and mutant STAMBPL1 were expressed similarly, as detected by Western blotting (Fig. 1F). Taken together, these data suggest that STAMBPL1 is a novel positive regulator of Tax-induced NF-κB activation.

Tax is also a potent activator of the noncanonical NF-κB pathway, which is persistently activated in HTLV-1-transformed cell lines (51), thus prompting us to next examine if STAMBPL1 regulates Tax-induced p100 processing. Silencing of STAMBPL1 with siRNA resulted in an inhibition of Tax-induced p100 processing to p52 (Fig. 1G). Overexpression of STAMBPL1 did not further promote the processing of p100 elicited by Tax (Fig. 1G), perhaps suggesting that Tax activation of the noncanonical pathway had already reached maximal levels. We next examined if STAMBPL1 modulated the phosphorylation and degradation of IκBα in HTLV-1-transformed MT-2 cells. Indeed, knockdown of STAMBPL1 in MT-2 cells reduced IκBα phosphorylation with a concomitant increase in total IκBα (Fig. 1H). Interestingly, knockdown of STAMBPL1 also decreased Tax expression, but not the Env-Tax fusion previously reported in MT-2 cells (Fig. 1H) (47). Collectively, our results strongly suggest that STAMBPL1 positively regulates Tax activation of the canonical and noncanonical NF-κB pathways.

STAMBPL1 stabilizes Tax and inhibits K48-linked polyubiquitination of Tax.

A previous study has demonstrated that Tax undergoes proteasomal degradation in the nucleus, specifically in the nuclear matrix (54). Since STAMBPL1 functions as a deubiquitinating enzyme, we hypothesized that STAMBPL1 may enhance the protein stability of Tax, possibly by removing K48-linked degradative polyubiquitin chains. To test this hypothesis, we conducted cycloheximide (CHX) chase assays with 293T and MT-2 cells. Although ectopic Tax in 293T cells was fairly stable in the presence of control siRNA, STAMBPL1 siRNA greatly accelerated the turnover of Tax (Fig. 2A). STAMBPL1 expression was efficiently silenced by siRNA, as determined by Western blotting (Fig. 2A). The loss of Tax was rescued by the proteasome inhibitor MG-132, indicating that the shorter half-life of Tax in the absence of STAMBPL1 was due to accelerated proteasomal degradation (Fig. 2A). Consistent with these results, knockdown of STAMBPL1 with siRNA in MT-2 cells triggered Tax degradation after 4 h of CHX treatment that was rescued with MG-132 treatment (Fig. 2B). Furthermore, an siRNA specific for an unrelated gene (coding for ELKS) had no effect on Tax stability, thus indicating specificity (Fig. 2B). Therefore, STAMBPL1 plays an important role in the stabilization of Tax protein.

We next examined the effect of STAMBPL1 on Tax ubiquitination using HA-tagged ubiquitin (HA-Ub) and a ubiquitin mutant with all lysines mutated to arginines, except for lysine 48 (HA-Ub K48 only). This ubiquitin mutant only has K48 available for the linkage of polyubiquitin chains and thus only supports K48-linked polyubiquitin instead of the other lysine residues in ubiquitin. 293T cells were transfected with Flag-Tax, wild-type HA-Ub, and K48-only HA-Ub constructs and control or STAMBPL1 siRNA for ubiquitination assays. Tax ubiquitination was observed when Tax and HA-Ub plasmids were coexpressed, but MG-132 treatment greatly increased total as well as K48-linked Tax polyubiquitination (Fig. 2C). STAMBPL1 knockdown resulted in greater Tax polyubiquitination and, in particular, K48-linked polyubiquitination, despite lower levels of Tax expression, as observed by Western blotting (Fig. 2C). These results confirm that STAMBPL1 regulates Tax K48-linked polyubiquitination and Tax stability.

STAMBPL1 enhances the stability of Tax by promoting its nuclear/cytoplasmic shuttling.

Tax undergoes a number of posttranslational modifications, including ubiquitination and sumoylation, that regulate Tax trafficking in cells (27). Whereas ubiquitinated Tax is predominantly found in the cytoplasm, sumoylated Tax is located primarily in the nucleus. However, nuclear Tax is unstable since Tax is degraded by the proteasome in the nuclear matrix upon K48-linked polyubiquitination (54). Thus, we reasoned that STAMBPL1 might indirectly stabilize Tax by promoting its nuclear export. To examine this hypothesis, we took advantage of previous findings that suggested that nuclear Tax is relocalized to the cytoplasm in a CRM1-dependent manner in response to genotoxic stress and DNA damage (11, 12). Thus, we performed immunofluoresence and confocal microscopy using a Tax expression plasmid and subjected cells to UV irradiation. Consistent with previous studies, Tax localization was heterogeneous with staining in both the cytoplasm and nucleus in 293T cells (Fig. 3A) (48). However, upon siRNA-induced silencing of STAMBPL1, a strongly nuclear distribution of Tax in the Tax speckled structure (TSS) was readily apparent (Fig. 3A). As expected, Tax nuclear staining in the speckles colocalized with the splicing factor SC-35 (data not shown). Upon UV stimulation of cells, Tax was efficiently exported from the nucleus to the cytoplasm, as previously described (Fig. 3A) (12). However, when STAMBPL1 was knocked down with siRNA, UV-induced Tax nuclear export was impaired and Tax was mainly confined to nuclear speckles (Fig. 3A). Similar results were obtained in HTLV-1-transformed MT-2 cells (data not shown). Western blotting of lysates confirmed the knockdown of STAMBPL1 and the destabilization of Tax by UV together with STAMBPL1 siRNA (data not shown). Quantitation of the confocal results revealed that Tax expression in the cytoplasm (either completely cytoplasmic or cytoplasmic and nuclear) decreased from 84% to 39% upon silencing of STAMBPL1 (Fig. 3B). Tax expression in the cytoplasm decreased from 93% to 25% in UV-stimulated cells transfected with control or STAMBPL1 siRNA, respectively (Fig. 3B). The effect of STAMBPL1 appears to be specific for Tax trafficking since knockdown or overexpression of STAMBPL1 did not influence p53 localization (data not shown). Together, these studies suggest that STAMBPL1 is essential for the shuttling of Tax from the nucleus to the cytoplasm, thus allowing Tax to elude proteasome-dependent degradation in the nucleus.

Given that STAMBPL1 plays an essential role in the shuttling of Tax from the nucleus to the cytoplasm, we next examined Tax activation of the HTLV-1 LTR, an event that occurs in the nucleus. In contrast to its effects on NF-κB activation, knockdown of STAMBPL1 enhanced Tax activation of the HTLV-1 LTR reporter (Fig. 3C). The effects of STAMBPL1 loss were even more significant when cells were exposed to UV irradiation (Fig. 3C). These results are consistent with our hypothesis that STAMBPL1 regulates Tax nuclear export; thus, when STAMBPL1 is knocked down, there is more abundant Tax in the nucleus, which correlates with greater LTR activation.

We next examined the localization of STAMBPL1 in cells by confocal microscopy to determine where STAMBPL1 may potentially be regulating Tax. STAMBPL1 was predominantly expressed in the nucleus, and Tax had no effect on the localization of STAMBPL1 (Fig. 3D). Western blotting confirmed the expression of both STAMBPL1 and Tax (Fig. 3D). Therefore, STAMBPL1 is a nuclear protein that is essential for Tax nuclear export.

STAMBPL1 promotes Tax relocalization to the cis-Golgi.

We have previously demonstrated that Tax promotes the relocalization of IKKγ to perinuclear “hot spots” in the proximity of the Golgi apparatus (20). The relocalization of IKK to the cis-Golgi by Tax is an important mechanism of Tax-induced activation of NF-κB. Since STAMBPL1 enhanced Tax-induced NF-κB activation, we next performed confocal studies to determine if STAMBPL1 promoted Tax colocalization with GM-130 (a cis-Golgi matrix protein). 293T cells were transfected with GFP-Tax and either control or STAMBPL1 siRNA, and cells were irradiated with UV light. As seen in Fig. 4, Tax colocalized with GM-130 in the presence of control siRNA, but not STAMBPL1 siRNA, which resulted in mostly nuclear Tax expression. Upon UV stimulation, Tax colocalization with GM-130 was also readily observed, although Tax expression was reduced and was predominantly nuclear when STAMBPL1 was silenced with siRNA (Fig. 4). Thus, STAMBPL1 facilitates Tax nuclear export and trafficking to the cis-Golgi, where it activates IKK and NF-κB. In the absence of STAMBPL1, Tax is trapped in the nucleus and undergoes proteasomal degradation.

Tax is degraded by the proteasome in the nucleus in the absence of STAMBPL1.

In order to further test the hypothesis that STAMBPL1 stabilizes Tax expression by promoting nuclear export, we performed additional immunostaining and confocal microscopy studies to determine if a proteasome inhibitor rescues Tax nuclear expression in the absence of STAMBPL1. 293T cells were transfected with Tax and either control or STAMBPL1 siRNA, and cells were treated with MG-132 and UV light (Fig. 5A). As demonstrated earlier, Tax was localized to perinuclear “hot spots” when transfected with control siRNA, but was predominantly nuclear in the presence of STAMBPL1 siRNA (Fig. 5A). Furthermore, MG-132 stabilized nuclear Tax that otherwise would have been degraded in cells expressing STAMBPL1 siRNA (Fig. 5A). Western blotting confirmed the diminished Tax expression by STAMBPL1 knockdown, and the stabilization of Tax by MG-132 treatment (Fig. 5A). Again, the effects of STAMBPL1 siRNA on Tax trafficking were more pronounced upon UV stimulation of cells. Strong nuclear focal Tax staining was observed in UV- and MG-132-treated cells knocked down for STAMBPL1 (Fig. 5A).

To confirm the confocal results using an independent assay, we also examined Tax localization by biochemical fractionation. We performed subcellular fractionation to separate cytoplasmic and nuclear fractions and examined Tax localization by Western blotting. Detection of TATA-binding protein and β-actin in the nuclear and cytoplasmic compartments, respectively, confirmed the purity of the fractions (Fig. 5B). Consistent with the confocal studies, Tax was expressed in both the cytoplasm and nucleus, but after UV irradiation, Tax was almost exclusively expressed in the cytoplasm (Fig. 5B). However, STAMBPL1 knockdown impaired Tax shuttling to the cytoplasm and Tax was more evenly distributed between the cytoplasmic and nuclear compartments (Fig. 5B). This effect on Tax was also observed upon treatment of cells with both UV and MG-132 (Fig. 5B). Thus, both biochemical fractionation and confocal analysis confirm that STAMBPL1 is essential for the nuclear export of Tax in response to UV stimulation.

Together, these results suggest that STAMBPL1 indirectly stabilizes Tax by promoting its nuclear export and preventing its proteasomal degradation in the nucleus. To further evaluate the role of STAMBPL1 in regulating Tax nuclear export, we used leptomycin B (LMB), a drug originally discovered as an antibiotic that blocks the nuclear export of proteins dependent on the nuclear export receptor CRM-1 (34). Wild-type STAMBPL1 and the catalytically inactive mutant E292A, together with Tax, were transfected into cells and treated with MG-132 and/or LMB. When STAMBPL1 was transfected together with Tax, the localization of Tax was almost completely cytoplasmic, forming a perinuclear aggregate (Fig. 6). However, this effect was dependent on STAMBPL1 catalytic activity since Tax was mainly confined to the nucleus in the presence of the STAMBPL1 mutant (Fig. 6). Upon treatment of cells with LMB, Tax expression was diminished and was localized predominantly in the nucleus; however, MG-132 stabilized nuclear Tax that was prevented from shuttling by LMB and thus was strongly expressed in the TSS (Fig. 6). These results were also corroborated by Western blotting, which demonstrated that LMB destabilized Tax but was rescued with MG-132 treatment (Fig. 6). Collectively, these results indicate that STAMBPL1 inhibits Tax proteasomal degradation by controlling Tax nuclear/cytoplasmic transport.

STAMBPL1 does not directly deubiquitinate Tax.

The previous experiments using the STAMBPL1 DUB mutant strongly suggest that the positive regulation of STAMBPL1 on Tax-induced NF-κB activation is driven by the deubiquitinating activity of this protein. Furthermore, STAMBPL1 knockdown enhances the K48-linked polyubiquitination of Tax (Fig. 2C). However, it is unclear if STAMBPL1 directly disassembles polyubiquitin chains from Tax or another target that regulates Tax trafficking. Thus far, we have been unable to detect an interaction between Tax and STAMBPL1 by co-IP (data not shown). Nevertheless, we performed an in vitro DUB assay using immunoaffinity-purified STAMBPL1. 293T cells were transfected with Flag-Tax together with wild-type HA-Ub, K63-only HA-Ub, or K48-only HA-Ub. Tax was immunoprecipitated from lysates using Flag antibody and eluted with Flag peptide. The ubiquitinated Tax proteins (either wild type, K48 only, or K63 only) were incubated with STAMBPL1 for a DUB assay. STAMBPL1 had little, if any, deubiquitination activity on either wild-type, K48-, or K63-linked polyubiquitin chains on Tax in vitro (Fig. 7A). However, STAMBPL1 effectively hydrolyzed recombinant K63-linked penta-ubiquitin (Ub⁵) chains in a DUB assay (Fig. 7B). Therefore, STAMBPL1 does not directly deubiquitinate Tax in vitro, suggesting that it likely regulates Tax nuclear export via an indirect mechanism.