Prothymosin-α Mediates Nuclear Import of the INrf2/Cul3·Rbx1 Complex to Degrade Nuclear Nrf2*

Nrf2-mediated coordinated induction of a battery of defensive genes is a critical mechanism in cellular protection and survival. INrf2 (Keap1), an inhibitor of Nrf2, functions as an adaptor for Cul3·Rbx1-mediated degradation of Nrf2. A majority of the INrf2/Cul3·Rbx1 complex is localized in the cytosol that degrades cytosolic Nrf2. However, 10-15% of INrf2 is also localized inside the nucleus. INrf2 does not contain a defined nuclear import signal, and the mechanism of nuclear import and its function inside the nucleus remain obscure. Present studies demonstrate that the DGR region of INrf2 is required for nuclear import of INrf2. Studies also demonstrate that Cul3 and Rbx1 are also imported inside the nucleus in complex with INrf2. Interestingly, Nrf2 and prothymosin-α both bind to the DGR region of INrf2. However, it is prothymosin-α and not Nrf2 that mediates nuclear import of INrf2/Cul3·Rbx1 complex. Antioxidant treatment increases nuclear import of INrf2/Cul3·Rbx1 complex. The INrf2/Cul3·Rbx1 complex inside the nucleus exchanges prothymosin-α with Nrf2, resulting in degradation of Nrf2. These results led to the conclusion that prothymosin-α-mediated nuclear import of INrf2/Cul3·Rbx1 complex leads to ubiquitination and degradation of Nrf2 inside the nucleus presumably to regulate nuclear level of Nrf2 and rapidly switch off the activation of Nrf2 downstream gene expression.

Nrf2:INrf2 2 serves as a sensor of chemical-and radiationinduced oxidative and electrophilic stress (1)(2)(3). Nrf2 is a nuclear transcription factor that regulates expression of several defensive genes including detoxifying enzymes and antioxidant genes (1)(2)(3). The induction of these enzymes is important for neutralizing the cellular stresses, providing cellular protection, and survival (1)(2)(3). Nrf2 resides predominantly in the cytoplasm, where it interacts with actin-associated cytosolic pro-tein, INrf2 (inhibitor of Nrf2) or Keap1 (Kelch-like ECH-associated protein 1) (4 -6). INrf2 functions as a substrate adaptor protein for a Cul3⅐Rbx1-dependent E3 ubiquitin ligase complex to ubiquitinate and degrade Nrf2, thus maintaining a steadystate level of Nrf2 (7). The exposure to oxidative/electrophilic stress leads to dissociation of Nrf2 from INrf2. Nrf2 is stabilized, translocates into the nucleus, and activates the transcription of several defensive genes. Recently, the mechanisms by which Nrf2 is released from INrf2 under stress have been actively investigated. One mechanism is that cysteine thiol groups of INrf2 were shown to function as sensors for oxidative stress, which are modified by the chemical inducers, causing formation of disulfide bonds between cysteines of two INrf2 peptides. This results in conformational change that renders INrf2 unable to bind to Nrf2 (8 -10). On the other hand, antioxidant-induced protein kinase C-mediated phosphorylation of serine 40 in Nrf2 leads to dissociation of Nrf2 from INrf2 (11)(12). It is possible that two mechanisms work in concert or independently of each other. However, the evidence to support either is lacking. Interestingly, we have recently demonstrated that an autoregulatory loop between Nrf2 and INrf2 controls their abundance inside the cells (13). In other words, Nrf2 induces INrf2 gene expression for its own degradation.
Several reports suggest that persistent accumulation of Nrf2 in the nucleus is harmful. INrf2-null mice demonstrated persistent accumulation of Nrf2 in the nucleus that led to postnatal death from malnutrition, resulting from hyperkeratosis in the esophagus and forestomach (14). Reversed phenotype of INrf2 deficiency by breeding to Nrf2-null mice suggested tightly regulated negative feedback might be essential for cell survival (15). The recent systemic analysis of INrf2 genomic locus in human lung cancer patients and cell lines showed that deletion, insertion, and missense mutations in functionally important domains of INrf2 results in reduction of INrf2 affinity for Nrf2 and elevated expression of cytoprotective genes (16,17). Taken together, unrestrained activation of Nrf2 in cells increases a risk of adverse effects including tumorigenesis. On the other hand, stress-induced activation of the Nrf2 pathway in normal cells is tightly regulated and confers cytoprotection against oxidative and electrophilic stress and carcinogens. Therefore, it appears that cells contain mechanisms that auto-regulate cellular abundance of Nrf2.
A majority of the INrf2 is present in the cytosol (7). However, 10 -15% of INrf2 is also localized inside the nucleus (7). Procite analysis revealed that INrf2 does not contain a defined nuclear import signal, and the mechanism of nuclear import of INrf2 and its function inside the nucleus remain unknown. In this report we investigated the mechanism of nuclear import of INrf2 and its role in Nrf2 degradation inside the nucleus. The studies demonstrated that the DGR region of INrf2 and prothymosin-␣ (PTM␣) are required for nuclear import of INrf2. Studies also demonstrated that INrf2 is not transported alone but in complex with Cul3 and Rbx1. Interestingly, PTM␣ containing a nuclear localization signal binds to the DGR region of INrf2 and transports the whole complex of INrf2/Cul3⅐Rbx1 inside the nucleus. The complex inside the nucleus releases PTM␣, binds to Nrf2, and triggers Nrf2 degradation. This is presumably to rapidly switch off the activation of Nrf2 downstream gene expression.

MATERIALS AND METHODS
Cell Cultures-Mouse hepatoma (Hepa-1) cells were obtained from the American Type Culture Collection (Manassas, VA). Cells were grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum, penicillin (40 units/ml), and streptomycin (40 g/ml). The cells were grown in monolayer in an incubator at 37°C in 95% air and 5% CO 2 .
Plasmid Constructs-Mammalian expression vector pEGFP-N1 was purchased from Clontech. This vector upon transfection in cells expresses green fluorescence protein (GFP). The vector was designated as 1XGFP. We modified this vector by cloning another coding sequence of GFP. The modified vector was designated as 2XGFP. A full-length mouse INrf2 coding sequence (1872 bp) was cloned into 1XGFP as well as 2XGFP vector. The primer sequences used for amplification of fulllength INrf2 and different deletion domains and mutations of INrf2 are shown in supplemental Table 1. Two leucine residues (Leu-308 and -310) present between functional nuclear export signals 2 (NES2) of INrf2 ware mutated to alanine by using specific mutant primers and the site-directed mutagenesis kit (Invitrogen). cDNA of PTM␣ was purchased from OriGene (Rockville, MD), and a full-length coding sequence was amplified by PCR and subcloned into pcDNA3.1/V5-His/Topo vector by TA cloning. The construction of pGL2B-NQO1-ARE, pcDNA-FLAG-INrf2, pcDNA-FLAG-Nrf2, pcDNA-INrf2-V5, and HA-ubiquitin has been described previously (18). All sitedirected mutations and deletions mutations were confirmed by DNA sequencing.
In Vitro Transcription/Translation-In vitro transcription/ translation of the plasmids encoding domains/deletions of INrf2-GFP and PTM␣-V5 constructs were performed using the TNT-coupled rabbit reticulocyte lysate system (Promega, Madison, WI). 0.2 g of plasmid DNA was incubated with 25 l of TNT-coupled rabbit reticulocyte lysate supplied with 40 M L-methionine at 30°C for 90 min. The plasmid encoding luciferase provided in the kit was used as a control for the transcription/translation reaction. After the coupled transcription/ translation, the proteins were checked for their correct size by SDS-PAGE followed by immunoblotting. All of the in vitro transcribed/translated proteins gave the expected size bands. Tetracycline-inducible FLAG INrf2, FLAG-⌬DGR-INrf2, and  FLAG-DGR-INrf2-Full-length INrf2, ⌬DGR-INrf2, and DGR-INrf2 regions were amplified by PCR using specific sets of prim-ers (supplemental Table 1) and cloned into modified pcDNA5/FRT/TO (FLAG) X2-(His) 8 vector. Flp-In T-REx HEK293 cells were purchased from Invitrogen and co-transfected separately with 0.1 g of FLAG-INrf2-pcDNA5/FRT/TO, FLAG-⌬DGR INrf2-pcDNA5/FRT/TO, and FLAG-DGRINrf2-pcDNA5/FRT/TO along with 0.9 g of pOG44 plasmids (Invitrogen) by the Effectene (Qiagen, Valencia, CA) method, as described by the manufacturer's instructions. Forty-eight hours after transfection, the cells were grown in medium containing 100 g/ml hygromycin B and 30 g/ml blasticidin (Invitrogen). The 293/FRT/TO cells stably expressing tetracycline-inducible N-terminal FLAG-tagged INrf2, FLAG-tagged ⌬DGR-INrf2, and FLAG-tagged DGR-INrf2 were selected. The stably selected cells were grown and treated with 0.5 g/ml tetracycline (Sigma) for varying periods of time to follow the overexpression of FLAG-tagged INrf2, FLAG-tagged ⌬DGR INrf2, and FLAG-tagged DGR INrf2 proteins.

Generation of Stable Flp-In T-REx HEK293 Cells Expressing
Subcellular Fractionation and Western Blotting-Hepa-1 cells, seeded in 100-mm plates and treated or transfected as displayed in the figures, were washed twice with ice-cold phosphate-buffered saline, trypsinized, and centrifuged at 1500 rpm for 5 min. For making whole cell lysates, the cells were lysed in radioimmune precipitation assay buffer (50 mM Tris, pH 8.0, 150 mM NaCl, 0.2 mM EDTA, 1% Nonidet P-40, 0.5% deoxycholic acid, 1 mM phenylmethylsulfonyl fluoride, and 1 mM sodium orthovanadate supplemented with protease inhibitor mixture (Roche Applied Science). Cytoplasmic and nuclear cell lysates were separated by using the Active Motif nuclear extract kit (Active Motif, Carlsbad, CA) by following the manufacturer's protocol. The protein concentration was determined using the protein assay reagent (Bio-Rad). 60 -80 g of proteins were separated by SDS-PAGE and transferred to nitrocellulose membranes. The membranes were blocked with 3% nonfat dry milk and incubated with anti-INrf2 (E-20) (1:1000), anti-Nrf2 (H-300) (1:500), anti-prothymosin ␣ (H-50) (1:1000), and antifetal alz-50 clone 1 (FAC1) (1:1000) antibodies, all purchased from Santa Cruz Biotechnology. Anti-Cul3 and anti-Rbx1 antibodies were purchased from Cell Signaling (Boston, MA). Anti-FLAG-HRP (1:10000), anti-HA-HRP (1:10000), and anti-actin antibodies were obtained from Sigma. Anti-GFP and anti-V5 HRP antibodies were obtained from Invitrogen. The membranes were washed three times with Tris-buffered saline-Tween, and immunoreactive bands were visualized using a chemiluminescence system ECL (Amersham Biosciences). The intensity of protein bands after Western blotting were quantitated by using QuantityOne 4.6.3 Image software (ChemiDoc XRS, Bio-Rad) and normalized against proper loading controls. To confirm the purity of nuclear-cytoplasmic fractionation, the membranes were reprobed with cytoplasm-specific, anti-lactate dehydrogenase (Chemicon) and nuclear specific, antilamin B antibodies (Santa Cruz Biotechnology). In related experiments, the cells were treated with 50 -100 M t-BHQ, 10 ng/ml leptomycin B (LMB), and 20 M genestein or DMSO as a vehicle for different time intervals.
Immunoprecipitation-1 mg of whole cell lysate or nuclear/ cytoplasmic extract was equilibrated in radioimmune precipitation assay buffer, pre-cleaned with protein-AGplus-agarose (Santa Cruz Biotechnology), and incubated with respective antibodies (1 g) at 4°C for overnight. Immune complexes were collected by the addition of protein AG-agarose again after centrifugation. The immune complexes were washed 3 times with radioimmune precipitation assay buffer containing 0.25% Nonidet P-40, and proteins were resolved by 10% reducing SDS-PAGE and transferred to nitrocellulose membrane. The membranes were blocked with 3% nonfat dry milk and incubated with their respective primary and secondary antibodies. Immunoreactive bands were visualized using a chemiluminescence system ECL (Amersham Biosciences).
Transient Transfection and Luciferase Assay-Hepa-1 cells were plated in 100-mm plates at a density of 1 ϫ 10 5 cells/plate 24 h before transfection. In the related experiments, the cells were transfected with 1 g of the indicated plasmids using Effectene transfection reagent (Qiagen) according to the manufacturer's instructions. After 36 h of transfection, cells were harvested, and cellular-specific protein regulation was examined by Western blotting. For luciferase reporter assay, Hepa-1 cells were grown in monolayer cultures in 12-well plates in Dulbecco's modified Eagle's medium-␣ supplemented with 10% fetal bovine serum. Then cells were co-transfected with 0.1 g of reporter construct human NQO1-ARE-Luc and 10 times less quantities of firefly Renilla luciferase encoded by plasmid pRL-TK. Renilla luciferase was used as the internal control in each transfections. After 24 h of transfection, the cells were washed with 1ϫ phosphate-buffered saline and lysed in 1ϫ Passive lysis buffer from the Dual-Luciferase reporter assay system kit (Promega). The luciferase activity was measured using the procedures described previously (18) and plotted.
siRNA Interference Assay-Control and Nrf2 and INrf2 siRNA purchased from Dharmacon were used to inhibit Nrf2 and INrf2 proteins, respectively, by a procedure described previously (10). Hepa-1 cells were transfected with 5, 25, and 50 nM concentrations of control or Nrf2 or INrf2 siRNA using Lipofectamine RNAiMAX reagent (Invitrogen) according to the manufacturer's instructions. Thirty-two hours after transfection, cells were harvested, and localization of INrf2, Cul3, and RBX1 was analyzed by Western blotting by probing the membranes with INrf2, Nrf2, and Cul3 or RBX1 antibodies. Fac1 and PTM␣ siRNA was also purchased from Ambion, and the effect of these siRNAs on localization of INrf2 was analyzed. The effect of PTM␣ siRNA on localization of INrf2 and NQO1 ARE luciferase activity was carried out after control siRNA or PTM␣ siRNA transfections.
Immunofluorescence-Hepa-1 cells were grown in Lab-Tek II chamber slides in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum. In related experiments cells were transfected with INrf2-GFP, INrf2-⌬DGR-GFP, or INrf2 DGR-2XGFP for 32 h. After fixing with 2% formaldehyde, cells were stained with DAPI and directly observed under fluorescence microscope. The effect of PTM␣ siRNA on localization of INrf2 was performed by transient transfection of PTM␣ siRNA using the procedures described above. Twenty-four hours later, cells were fixed with 2% formaldehyde at 4°C for 10 min and permeabilized by 0.25% Triton X-100. Cells were washed twice with PBS and incubated with a 1:1000 dilution of goat INrf2 primary antibodies in 2% bovine serum albumin at 4°C for 12 h. Then cells were washed twice with PBS and incu-bated with FITC-conjugated second antibody for 2 h at room temperature. Cells were washed and stained with DAPI and mounted. The effect of PTM␣ overexpression on localization of INrf2-V5 was also visualized by immunofluorescence. Hepa-1 cells were transfected with INrf2-V5 alone or in combination with pcDNA-PTM␣. Twenty-four hours later, cells were treated with 50 M t-BHQ for 2 h and fixed and permeabilized by 0.25% Triton X-100. After washing, cells were incubated with a 1:1000 dilution of FITC-conjugated anti-V5 tag antibodies for 12 h. Cells were washed twice with PBS, stained with DAPI, and mounted. After immunostaining, cells were observed under Nikon fluorescence microscope, and photographs were captured. The cytoplasmic and nuclear FITC fluorescence intensities were quantified from 10 different cells (n ϭ 10) by using NIS-Elements BR2.30 SP4 software (Nikon). The green fluorescence intensity of the images was quantified by using Nikon Elements Advanced Research Software (Melville, NY). The entire green region of nuclear and cytosolic compartment was first marked and delineated, and the average fluorescence intensity of the green channel was measured. The experiments were repeated three times.
Isolation and Purification of INrf2/Cul3⅐Rbx1/PTM␣/Nrf2 Complex-Human embryonic kidney FRT-HEK293 cells stably expressing FLAG-INrf2 were treated with 0.5 g/ml tetracycline for 24 h followed by treatment with DMSO or t-BHQ (50 M) for 1 and 2 h. Cells were harvested, and cytoplasmic and nuclear extracts were prepared by using an Active Motif kit. 5 mg of cytoplasmic and nuclear extracts were mixed with anti-FLAG-agarose beads (200 l) at 4°C for 8 h, beads washed 3 times with radioimmune precipitation assay buffer, and bound protein were eluted with 100 l of 1ϫ FLAG peptide (Sigma). The native protein complex was electrophoresed by 6% Native gel and stained with Coomassie Brilliant Blue. The same native complex was immunoblotted with anti INrf2, Nrf2, PTM␣, Cul3, and Rbx1 antibodies.
Ubiquitination Assay-Hepa-1 cells were transfected with control or PTM␣ siRNA (50 nM) for 10 h followed by co-transfection with FLAG-INrf2 (0.5 g), Nrf2-V5 (1.0 g), and HAubiquitin (0.2 g). Nuclear and cytoplasmic extracts were prepared using an Active Motif kit. To check the effect of PTM␣ siRNA on localization of INrf2 and PTM␣ protein, 60 g of the cytoplasmic and nuclear extracts were immunoblotted with anti-FLAG, anti-V5, and anti-PTM␣ antibodies. To check Nrf2 ubiquitination in cytoplasmic and nuclear compartments, 1 mg of cytoplasmic and nuclear lysates was immunoprecipitated with anti-V5 antibody (Invitrogen). Immune complexes were resolved by 10% SDS-PAGE followed by immunoblotting with anti-HA-HRP antibody.
Statistical Analyses-Statistical analysis of the data from luciferase assays and immunoblotting band intensities were performed by using the SPSS-16 software. Error bars indicate S.E. of triplicate samples and are presented in figures. Statistical analysis was performed by one-way analysis of variance followed by the Tukey-Kramer's post hoc test for multiple comparisons. Significance p values were also calculated and presented.

DGR Is Required for Nuclear Localization of INrf2-INrf2
and INrf2 domain deletion mutants were separately cloned into GFP vector (Fig. 1A). In vitro transcription/translation reactions confirmed these plasmids produced the expected size of GFP-tagged INrf2 protein (Fig. 1B). The various plasmids were transfected in Hepa-1 cells. The transfected cells were lysed to generate whole cell lysate (W) or fractionated to cytosolic (C) and nuclear (N) fractions. The purity of cytosolic and nuclear fractions was assessed by Western blotting and probing for cytosolic marker lactate dehydrogenase (LDH) and nuclear marker lamin B (Fig. 1C, panels 13 and 14). Western analysis of transfected cells revealed that deletion of the DGR region led to abrogation of nuclear localization of INrf2 (Fig. 1C, compare the top 10 panels). All other deletions showed mutant INrf2 protein in the nucleus. The mutation of NES2 region had no effect on localization of DGR-deficient mutant INrf2 (Fig. 1C, compare panels 7 and 8). In the same experiment DGR region attached to 2XGFP localized in the nucleus (Fig. 1C, panel 12). These results demonstrated that the DGR region is required for nuclear localization of INrf2. This interpretation was also supported by immunocytochemical analysis (Fig. 1D)  , and DGR-2XGFP on slides were fixed in 2% formaldehyde, washed with PBS, stained with Vectashield containing nuclear DAPI stain, and mounted. Cells were observed under Nikon fluorescence microscope, and photographs were captured. The cytoplasmic and nuclear GFP fluorescence were quantified from 10 different transfected cells (n ϭ 10) by using NIS-Elements BR2.30 SP4 software. All the experiments were repeated three times. Error bars indicate S.E. of fluorescence intensities. treatment led to nuclear accumulation of INrf2. This was similar to LMB-mediated nuclear accumulation of endogenous Nrf2, a positive control used in the experiment. However, LMB had no effect on the nuclear localization of DGR and INrf2⌬DGR (Fig. 2C). Similar results were also obtained in Hepa-1 cells transfected with GFP-tagged constructs (Fig. 2D). FRT 293 cells expressing FLAG-tagged INrf2 were treated with t-BHQ or LMB, and nuclear and cytoplasmic INrf2 levels were quantified from three independent experiments. As shown in Fig. 2E, 15-20% more INrf2 accumulates in the nucleus upon treatment with either t-BHQ or LMB compared with DMSO treated cells (Fig. 2, B and E).
INrf2 is known to contain a functional nuclear export signal (NES2) (19). The absence of an effect on deletion of NES2 ( These results suggested that INrf2, Cul3, and Rbx1 existed in a complex. This finding is similar, as reported earlier that Cul3 in complex with Rbx1 binds to INrf2 to ubiquitinates and degrade Nrf2 (7). It might be noteworthy that Rbx1 binds to Cul3 and not INrf2 (7). Rbx1 is immunoprecipitated with INrf2 because of its binding with Cul3 that binds to INrf2. Interestingly, treatment with t-BHQ led to significant loss in immunoprecipitation of Cul3 and Rbx1 with INrf2 antibody (Fig. 3A, upper panel, compare the t-BHQ lane with the DMSO lane). Similar results were also observed in immunoprecipitation with Cul3 and Rbx1 antibodies (Fig. 3A, middle and lower panels). These results indicated that t-BHQ destabilized the INrf2/Cul3⅐Rbx1 interaction to stabilize Nrf2. The replacement of whole cell lysate with cytosolic and nuclear fractions showed similar results (data not shown). We used INrf2 siRNA to demonstrate that inhibition of INrf2 led to nuclear translocation of not only INrf2 but also Cul3 and Rbx1 (Fig. 3, B and C). Hepa-1 cells were transfected with increasing concentrations of either control siRNA or INrf2 siRNA. INrf2 siRNA and not control siRNA showed a concentration-dependent decrease in INrf2 protein levels (Fig. 3B). In related experiments, cytoplasmic and nuclear proteins were isolated from mock, control, or INrf2 siRNAtransfected Hepa-1 cells and immunoblotted with INrf2, Cul3, and Rbx1 antibodies. Immunoblots were also probed for cytosolic marker LDH and nuclear marker lamin B. The results revealed that INrf2 siRNA significantly reduced whole cell lysate (W), cytosol (C), and nuclear (N) INrf2, as compared with mock and control siRNA-transfected Hepa-1 cells (Fig. 3C, top  panel). Interestingly, a decrease in siRNA-mediated INrf2 in the nucleus also decreased nuclear Cul3 and Rbx1 and correspondingly increased cytosolic Cul3 and Rbx1 (Fig. 3C, compare INrf2 siRNA lane with Mock and Control siRNA lanes in Cul3 and Rbx1 panels). These results suggested that loss of INrf2 reduced nuclear import of Cul3⅐Rbx1 or also suggested Cul3⅐Rbx1 translocate to nucleus in complex with INrf2. The effect of INrf2 siRNA on the overall localization of Cul3 and Rbx1 in cytoplasmic and nuclear compartments was quantified from three different experiments and presented (Fig. 3D). Further immunoprecipitation experiments confirmed that knocking down INrf2 by siRNA significantly decreased INrf2/ Cul3⅐Rbx1 complex in the nucleus, as compared with control siRNA-transfected cells. These data indicate that INrf2 interaction with Cul3⅐Rbx1 complex was essential for nuclear import of Cul3 and Rbx1 (Fig. 3E, upper, (Fig. 4B, upper two left panels). Reverse IP with anti-FLAG immunoprecipitated FLAG-Nrf2 (Fig. 4B, lowest left panel). In reverse IP, INrf2-V5 but not INrf2⌬DGR-V5 was pulled down (Fig. 4B, third left panel from the top). Interestingly, only the DGR domain of INrf2 interacts with FLAG Nrf2 (Fig. 4B, right  upper and lower panels). The interaction of endogenous Nrf2 with INrf2GFP and INrf2⌬DGRGFP was examined by GFP immune precipitation followed by Nrf2 immunoblotting. Endogenous Nrf2 interacts with INrf2-GFP but not with INrf2⌬DGR-GFP or GFP vector (Fig. 4C). These results com-bined demonstrated that Nrf2 interacts with the DGR domain of INrf2.
Next, we determined if Nrf2 interaction with the DGR domain of INrf2 is required for nuclear localization of INrf2. Hepa-1 cells were transfected with different concentrations of control or Nrf2 siRNA and analyzed by immunoblotting with Nrf2. The results showed an Nrf2 siRNA concentration-dependent decrease in Nrf2 (Fig. 4D). Control siRNA in the same experiment had no effect on expression of Nrf2 protein. siRNA inhibition of Nrf2 was used to determine whether the decrease in Nrf2 had an effect on nuclear localization of INrf2. The results demonstrated that siRNA-mediated decrease in Nrf2 did not alter nuclear localization of INrf2 (Fig. 4E). This indicated that Nrf2 is not essential for nuclear import of INrf2.
PTM␣ Is Required for Nuclear Import of INrf2-Human 293 cells permanently expressing tetracycline-inducible FLAG-INrf2, FLAG-DGR, and FLAG-INrf2⌬DGR were grown in culture, treated with tetracycline to induce respective proteins,  (Fig. 5A, left panel). Interestingly, immunoprecipitation of FLAG-INrf2⌬DGR with anti-FLAG antibody failed to immunoprecipitate either PTM␣ or Fac1 (Fig. 5A, left panel). Reverse immunoprecipitation with anti-PTM␣ and anti-Fac1 immunoprecipitated FLAG-INrf2 and FLAG-DGR but failed to immunoprecipitate FLAG-INrf2⌬DGR (Fig. 5A, right panels). These results showed that PTM␣ and Fac1 both interact with the DGR domain of INrf2. A similar experiment was performed in cytosolic and nuclear fractions of Hepa-1 cells to determine the interaction of endogenous PTM␣ and INrf2 and associated Cul3⅐Rbx1 proteins (Fig.   5B). INrf2 antibody immunoprecipitated INrf2 from both cytosolic and nucleus fractions (Fig. 5B, second blot from the top). PTM␣ was pulled down along with INrf2 from both the fractions (Fig. 5B, top blot). In reverse IP, PTM␣ antibody immunoprecipitated PTM␣ (Fig. 5B, second blot from bottom). INrf2, Cul3, and Rbx1 but not Nrf2 were pulled down along with PTM␣ (Fig. 5B, third though sixth blots from the top). These results clearly suggested that PTM␣ and INrf2/Cul3⅐Rbx1 exist in a complex in cytosolic and nuclear fractions.
Procite search revealed that both PTM␣ and Fac1 contain strong nuclear localization signals. We used PTM␣and Fac1specific siRNA to decrease the respective proteins in separate experiments with the aim of determining whether PTM␣ or Fac1 interaction with the DGR region leads to nuclear localization of INrf2. PTM␣ and Fac1 siRNA showed concentrationdependent inhibition of PTM␣ and Fac1 in Hepa-1 cells (Fig.  5C). In a related experiment, Hepa-1 cells were transfected with either PTM␣ siRNA or Fac1 siRNA, and cytosolic and nuclear extracts were immunoblotted for endogenous INrf2. As shown in Fig. 5D, siRNA-mediated inhibition of PTM␣ and not Fac1 led to abrogation of nuclear localization of INrf2. Furthermore, we used three different concentrations (25, 50, and 75 nM) of PTM␣ siRNA that led to an siRNA concentration-dependent decrease in PTM␣ (Fig. 5C, top panel), and significant loss of nuclear localization of INrf2 was also observed (Fig. 5E). These results clearly indicate that PTM␣ interaction with the DGR domain of INrf2 leads to nuclear localization of INrf2 in the nucleus. We reasoned that because Cul3⅐Rbx1 move to the nucleus through Cul3 interaction with INrf2, then inhibition of PTM␣-mediated nuclear localization should also lead to inhibition of nuclear localization of Cul3 and Rbx1.
Immunocytochemical analysis of INrf2 in PTM␣ siRNAtransfected Hepa-1 cells supported the role of PTM␣ in nuclear translocation of INrf2 (Fig. 6A). 75 nM PTM␣ siRNA caused almost complete loss of PTM␣ (Fig. 5C, top panel) and abrogation of nuclear import of INrf2 (Fig. 6A, immunocytochemistry left panels and intensity measurement right panels). Moreover, siRNA-mediated inhibition of PTM␣ led to siRNA concentration-dependent stabilization of Nrf2 and an increase in endogenous Nrf2 downstream NQO1 gene expression (Fig. 6B). This was presumably due to a decrease in import of INrf2 in the nucleus and stabilization of Nrf2. Additionally, we measured NQO1 ARE-luciferase activity from Hepa-1 cells, which were co-transfected with either control siRNA or PTM␣ siRNA. The results demonstrated that a PTM␣ siRNA-dependent increase in ARE-luciferase gene expression was significant compared with control siRNA (Fig. 6C, p Ͼ 0.004). This was due to stabilization of Nrf2 in PTM␣ siRNAtransfected Hepa-1 cells.
Overexpression of PTM␣-V5 Increases Nuclear Localization of INrf2-Hepa-1 cells were either transfected with vector control or PTM-␣V5, lysed, and immunoblotted with V5 antibody. Significant overexpression of PTM␣-V5 was observed in cells transfected with PTM␣-V5 (Fig. 7A). To examine the effect of overexpression of PTM␣ on nuclear localization INrf2, we overexpressed PTM␣-V5 in Hepa-1 cells, and cytosolic and nuclear extracts were immunoblotted with anti-V5 antibody for PTM␣-V5, anti-PTM␣ for endogenous PTM␣, and anti-V5 for INrf2-V5-transfected protein. As shown in Fig. 7B, overexpression of PTM␣-V5 led to increased nuclear localization of INrf2-V5 in the presence and absence of t-BHQ (Fig. 7B, Western analysis upper panels and relative band intensity lower panels, p Ͼ 0.005). Immunocytochemical analysis (Fig. 7C, upper panels) and intensity measurement (Fig. 7C, lower panels) showed significantly increased nuclear localization of INrf2 in cells overexpressing PTM␣-V5 (p Ͼ 0.0001). This again supported a role of PTM␣ in nuclear import of INrf2.
Antioxidant Treatment Induces Exchange of PTM␣ with Nrf2 in the Nucleus-Hepa-1 cells were treated with either DMSO or t-BHQ for the indicated periods as shown in Fig. 8A. Endogenous INrf2 protein was immunoprecipitated from equal amounts (1 mg) of cytosolic and nuclear extracts and immuno- Error bars indicate the S.E. Statistical analysis was performed as described under "Materials and Methods." *, p Ͻ 0.0005 (compared with mock or control siRNA-transfected nuclear INrf2 samples). B, Hepa-1 cells were transfected with control or PTM␣ siRNA and immunoblotted antibodies as shown. C, Hepa-1 cells transfected with control or PTM␣ siRNA. Ten hours after cells were transfected with 0.1 g of pGL2B-hARE-NQO1 plasmid DNA and 10 ng of pRL-TK plasmid encoding Renilla luciferase plasmid DNA (Promega). Twenty-four hours after the second transfection, the cells were lysed in 1X passive lysis buffer (Promega), and relative luciferase activity was measured and plotted. Error bars indicate S.E. of triplicate samples. Statistical analysis was performed as described under "Materials and Methods." *, p Ͻ 0.0004; **, p Ͻ 0.004 (compared with control siRNA or mock-transfected samples).
blotted for endogenous INrf2, PTM␣, and Nrf2 proteins for analyzing their interactions. INrf2 antibody immunoprecipitated INrf2 from nuclear and cytosolic fractions (Fig. 8A, Western analysis left panels and relative binding as determined by band intensity measurements, right panels). It showed an increase in immunoprecipitated INrf2 in cytosol but not in the nucleus in response to t-BHQ treatment. This was expected as t-BHQ is known to increase INrf2 gene expression (13). Nrf2 and PTM␣ were pulled down along with INrf2 because of their interaction with INrf2 (Fig. 8A). Interestingly, immunoprecipitation with INrf2 antibody demonstrated that in the nucleus Nrf2 interaction with INrf2 was more or less unchanged, whereas PTM␣ interaction with INrf2 was increased at 1 h of t-BHQ treatment as compared with DMSO-treated control. The magnitude of Nrf2 and PTM␣ interaction with INrf2 reversed in the nucleus at 2 h of t-BHQ treatment. The Nrf2 interaction increased and PTM␣ interaction decreased with INrf2. These results suggested that t-BHQ treatment in the first hour enhanced PTM␣-INrf2 interaction and nuclear import of INrf2. The results also suggest that in the nucleus Nrf2 replaces PTM␣ to bind to INrf2 presumably for degradation in the second hour of t-BHQ treatment of Hepa-1 cells. The Nrf2 interaction with INrf2 in the cytosol during the first hour of t-BHQ treatment decreased due to t-BHQ-mediated release of Nrf2 from INrf2 and nuclear translocation of Nrf2 to activate AREmediated defensive gene expression. The Nrf2 interaction with INrf2 showed an increase during the second hour of t-BHQ treatment to start degrading Nrf2. The PTM␣ interaction with INrf2 in the cytosol during the first and second hour of t-BHQ treatment showed only a marginal increase presumably to enhance nuclear translocation of INrf2 to capture and degrade nuclear Nrf2. The PTM␣ exchange with Nrf2 in the nucleus was further investigated by analyzing the immunoprecipitated INrf2/ Cul3⅐Rbx1/PTM␣/Nrf2 complex from the cytosol and nucleus on native blue gel (Fig. 8B) and immunoblotting (Fig. 8C). The INrf2/Cul3⅐Rbx1/PTM␣/Nrf2 complex increased in the nucleus at 1 h of t-BHQ treatment and then decreased at 2 h of t-BHQ treatment (Fig. 8B). Western analysis of the cytosolic and nuclear complex revealed interaction of both PTM␣ and Nrf2 with INrf2 in varying amounts (Fig. 8C). As expected, the cytosol from cells treated with t-BHQ for 1 h showed increased binding of PTM␣ with INrf2 and almost complete loss of interaction of Nrf2 with INrf2. This presumably is due to t-BHQmediated release of Nrf2 to activate gene expression and to increase in PTM␣-mediated import of INrf2 in the nucleus. The pretreatment of Hepa-1 cells with tyrosine kinase inhibitor genestein inhibited Nrf2 exchange with PTM␣ for binding to INrf2 in the nucleus (Fig. 8D). Immunoprecipitation with INrf2 antibody pull down PTM␣ but not Nrf2 at both 1 and 2 h of t-BHQ treatment in the nucleus. (Fig. 8D, Western analysis upper panel and band intensities in the lower panel). PTM␣-mediated Nuclear Import of INrf2 Leads to Degradation of Nuclear Nrf2-Hepa-1 cells co-transfected with HAubiquitin and Nrf2-V5 demonstrated very low levels of ubiquitination of Nrf2 (Fig. 9A). The inclusion of FLAG-INrf2 increased ubiquitination of Nrf2 (Fig. 9A) and slight degradation of Nrf2 (Fig. 9B). Prior transfection of cells with control siRNA had no effect on PTM␣ and ubiquitination/degradation of Nrf2 (Fig. 9, A and B). Prior transfection with PTM␣ siRNA inhibited PTM␣ in the cytosol as well as nucleus (Fig.  9B). The inhibition of PTM␣ in the nucleus but not in the cytosol led to decreased ubiquitination and degradation of Nrf2 (Fig. 9, A and B).

DISCUSSION
INrf2 is a negative regulator of Nrf2 (2). INrf2 functions as an adapter for bringing Nrf2 closer to Cul3⅐Rbx1 complex for ubiquitination and degradation of Nrf2 (7). This mechanism controls the Nrf2 level inside the cells. INrf2/Cul3⅐Rbx1 complex is predominately localized in the cytosol. Therefore, cytosol is the prime site for ubiquitination and degradation of Nrf2. However, 10 -15% of INrf2 is reported in the nucleus. The nucleus also contains the Cul3⅐Rbx1 complex, and our data confirmed that the INrf2/Cul3⅐Rbx1 complex is also present in the nucleus. Antioxidant t-BHQ is known to induce INrf2 through a proximal promoter ARE on the reverse strand (13). Interestingly, t-BHQ induces INrf2 in the cytosol as well as in the nucleus. In the current investigation we demonstrated the unique mechanism of nuclear import of INrf2 and Cul3⅐Rbx1 complex and its function in the nucleus. An analysis of INrf2 primary sequence did not reveal the presence of a putative nuclear import signal or nuclear localization signal. Therefore, we reasoned that the nuclear localization of INrf2 has to be mediated through an associate protein with a defined nuclear import signal or nuclear localization signal.
Domain deletion analysis clearly demonstrated that DGR region of INrf2 is required for nuclear translocation of INrf2. Deletion of DGR region from INrf2 abrogated nuclear localization of INrf2. In addition, attachment of DGR region with GFP led to nuclear translocation of GFP protein. The studies also revealed that PTM␣ bound to the DGR region and mediated Error bars indicate S.E. of triplicate samples. Statistical analysis was performed by one way analysis of variance followed by the Tukey-Kramer's post hoc test for multiple comparisons. *, p Ͻ 0.003 (compared with ubiquitinated nuclear Nrf2 of cells transfected with mock or control siRNA). B, 60 g of cytosolic and nuclear extracts in the above experiment were immunoblotted with anti-FLAG-HRP and anti-V5-HRP antibodies. The same extracts were also probed with anti-PTM␣, anti-LDH, and anti-lamin B antibodies for detection of endogenous proteins.
nuclear translocation of INrf2. The INrf2⌬DGR mutant deficient in DGR failed to bind to PTM␣ and also failed to localize in the nucleus. In addition, siRNA inhibition of PTM␣ resulted in a significant loss of nuclear localization of INrf2. Furthermore, the overexpression of PTM␣ enhanced nuclear localization of INrf2. These results concluded that PTM␣ interaction with DGR region led to nuclear localization of INrf2. A Procite search revealed that PTM␣ contains a strong nuclear localization signal at the C terminus of the protein that presumably helps PTM␣ to mediate nuclear localization of INrf2. The role of PTM␣ in nuclear translocation of INrf2 is also supported by an earlier observation that PTM␣ interaction with INrf2 increases nuclear INrf2 (20). Our data suggested that Nrf2 and Fac1 also bound to the DGR domain of INrf2 but were not required for nuclear localization of INrf2.
Cul3 is known to bind to the BTB domain of INrf2 (7). Rbx1 is known to bind to Cul3 but not INrf2 (7). The present studies demonstrated that the Cul3⅐Rbx1 complex is present in the nucleus. INrf2 knock down studies and immune precipitation assays revealed that Cul3⅐Rbx1 is imported in the nucleus along with INrf2, presumably due to Cul3 interaction with INrf2. Therefore, it is reasonable to conclude that PTM␣ through its binding with the DGR region of INrf2 mediates the nuclear import of INrf2/Cul3⅐Rbx1 complex.
The results also showed that the treatment of cells with antioxidant t-BHQ resulted in the release of Nrf2 from INrf2 and that Nrf2 translocates to the nucleus. This coincided with an increase in INrf2 in the cytosol and nucleus. The PTM␣ interaction with INrf2 led to an increase in nuclear translocation of INrf2 followed by exchange of PTM␣ with Nrf2 on INrf2. Interestingly, tyrosine kinase inhibitor genestein inhibited the exchange of PTM␣ with Nrf2 on INrf2. This was expected as it was reported earlier that Nrf2 is phosphorylated by tyrosine kinase Fyn at Tyr-568. The Nrf2Y568 phosphorylation is required for binding with INrf2 (21). Genestein inhibited tyrosine kinase-mediated phosphorylation of Nrf2 and exchange of PTM␣ with Nrf2. Therefore, it is possible that Nrf2Y568 has to be phosphorylated to bind to INrf2 in exchange for PTM␣. What happens to PTM␣ after exchange with Nrf2 inside nucleus remains unknown.
The exchange of PTM␣ with Nrf2 inside the nucleus on INrf2 leads to ubiquitination and degradation of Nrf2. The Cul3⅐Rbx1 complex is present in the nucleus which ubiquitinates Nrf2, leading to degradation of Nrf2. Therefore, PTM␣mediated nuclear import of INrf2/Cul3⅐Rbx1 complex is required for rapid degradation of used Nrf2 inside the nucleus. The nuclear degradation of Nrf2 along with cytosolic degradation of Nrf2 provides a complex but fine mechanism to regulate Nrf2 in the nucleus and cytosol. Nrf2 is known to promote cell survival by reducing apoptosis (22). This means that the damaged cells could escape apoptosis and survive if Nrf2 in the nucleus is not checked. Indeed, mutations in INrf2 result in loss of INrf2 and accumulation of nuclear Nrf2 in lung cancer (16,17). Nrf2 has to be rapidly degraded after its function is over to give opportunity to other cellular pathways of surveillance and the decision to survive, die, or undergo senescence.
In conclusion, we demonstrated that PTM␣ interaction with INrf2 leads to nuclear translocation of INrf2/Cul3⅐Rbx1 complex that exchanges PTM␣ with Nrf2, leading to ubiquitination and degradation of Nrf2. This process is a late response to oxidative/electrophilic stress after the release and nuclear translocation of Nrf2. The function of nuclear import of INrf2/ Cul3⅐Rbx1 complex is to rapidly degrade Nrf2 inside the nucleus and to switch off the activation of Nrf2 downstream genes to promote normal growth and survival of cells.