Nuclear Export of Retinoid X Receptor α in Response to Interleukin-1β-mediated Cell Signaling

As the obligate heterodimer partner to class II nuclear receptors, the retinoid X receptor α (RXRα) plays a vital physiological role in the regulation of multiple hepatic functions, including bile formation, intermediary metabolism, and endobiotic/xenobiotic detoxification. Many RXRα-regulated genes are themselves suppressed in inflamed liver via unknown mechanisms, which constitute a substantial component of the negative hepatic acute phase response. In this study we show that RXRα, generally considered a stable nuclear resident protein, undergoes rapid nuclear export in response to signals initiated by the pro-inflammatory cytokine interleukin-1β (IL-1β), a central activator of the acute phase response. Within 30 min of exposure to IL-1β, nuclear levels of RXRα are markedly suppressed in human liver-derived HepG2 cells, temporally coinciding with its appearance in the cytoplasm. The nuclear residence of RXRα is maintained by inhibiting c-jun N-terminal kinase (JNK, curcumin or SP600125) or CRM-1-mediated nuclear export (Leptomycin B). Pretreatment with the proteasome inhibitor MG132 blocks IL-1β-mediated reductions in nuclear RXRα levels while increasing accumulation in the cytoplasm. Mutational studies identify one residue, serine 260, a JNK phosphoacceptor site whose phosphorylation status had an unknown role in RXRα function, as critical for IL-1β-mediated nuclear export of transfected human RXRα-green fluorescent fusion constructs. These findings indicate that inflammation-mediated cell signaling leads to rapid and profound reductions in nuclear RXRα levels, via a multistep, JNK-dependent mechanism involving Ser260, nuclear export, and proteasomal degradation. Thus, inflammation-meditated cell signaling targets RXRα for nuclear export and degradation; a potential mechanism that explains the broad suppression of RXRα-dependent gene expression in the inflamed liver.

After injury or infection, the liver participates in a program of modified gene expression known collectively as the acute phase response (APR) 2 (1). A wide variety of hepatic functions are altered during the APR, much of which occurs by cytokine-mediated activation or suppression of target gene transcription. Among the principal hepatic physiologic processes inhibited during the negative hepatic APR are genes involved in endobiotic/xenobiotic metabolism, glucose homeostasis, and bile formation, which then leads to cholestasis. Activation of the negative hepatic APR leads to cholestasis by decreasing bile salt synthesis (2), reducing canalicular bile salt export (3), and suppressing bile salt import, the latter of which occurs primarily by transcriptional down-regulation of the sodium-dependent taurocholate co-transporting polypeptide (Slc10A1) (4,5). Recent evidence suggests that cytokine-mediated activation of cell signaling pathways during the APR leads to this coordinated response (6). One possibility is targeted repression of the essential heterodimer partner for type II nuclear receptor, RXR␣ (NR2B1), which is an attractive mechanism to explain the suppression of many hepatic genes during the APR. Recent studies from multiple groups, including our own, support the involvement of inflammation-based cell signaling pathways as a suppressor of RXR␣-dependent gene expression, although the underlying mechanisms are unknown (7)(8)(9)(10)(11). One cytokine in particular, interleukin-1␤, (IL-1␤) appears to be a major player in mediating these effects, both in vivo and in vitro (10,12). How IL-1␤-activated pathways ultimately leads to reduced RXR␣ heterodimer DNA binding in the nucleus is not known.
In this study we sought to investigate the hypothesis that IL-1␤-mediated activation of c-jun N-terminal kinase (JNK) cell signaling reduces RXR␣ function by inducing its export from the nucleus and initiating proteasome-mediated degradation. We found that the subcellular localization of RXR␣ is responsive to IL-1␤ signaling, whereby it undergoes a rapid JNK-mediated, CRM-1-dependent nuclear export, leading to decreased nuclear RXR␣ levels and subsequent reduced nuclear DNA binding activity. Using green fluorescent protein (GFP) technology, we show that JNK-mediated RXR␣-GFP is exported out of the nucleus like native RXR␣ and that nuclear export involves the JNK phosphoacceptor site, serine 260. Finally, IL-1␤-induced cell signaling leads to rapid proteasome-mediated degradation of RXR␣, suggesting that JNK-mediated nuclear export is the first and critical step that results in reduced nuclear levels of RXR␣. Taken together, these studies reveal that RXR␣ is a target for pro-inflammatory cytokine cell signaling and provide a novel molecular mechanism for the broad and significant reduction of hepatic RXR␣-regulated gene expression in the inflamed liver.

EXPERIMENTAL PROCEDURES
Materials-HepG2 cells were obtained from the American Type Culture Collection (ATCC, Rockville, MD). Cell culture reagents were purchased from Invitrogen, human recombinant IL-1␤ was purchased from R&D Systems Inc. (Minneapolis, MN), Curcumin and SP600125 were from Calbiochem, MG132 was from Biomol (Plymouth Meeting, PA), and Leptomycin B was from Sigma-Aldrich. FuGENE 6 was purchased from Roche Applied Science. Subcloning and mutation reagents were obtained from Stratagene (La Jolla, CA). Antibodies for JNK (num-* A portion of this work was presented in abstract form at the 2004 annual meeting of the American Association for the Study of Liver Diseases (AASLD). This work was supported by Grants AI46773 (to A. R. B.), DK56239 (to S. J. K.), and DK56338 (supporting the Texas Gulf Coast Digestive Diseases Center) from the National Institutes of Health and the Texas Children's Hospital Foundation. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1  Cell Culture-HepG2 cells, a human hepatoblastoma-derived cell line, were maintained in minimum essential medium containing Earle's salts and supplemented with 10% certified fetal bovine serum, penicillin-streptomycin, and L-glutamine. For protein analysis, cells were plated at 2.5 ϫ 10 6 cells/10-cm dish and maintained in serum-containing medium for 48 h and then maintained in serum-free medium for 20 h prior to treatments. For transfections and immunofluorescent staining, HepG2 cells were plated at 5 ϫ 10 5 cells onto glass coverslips.
Cytokine and Inhibitor Treatments-After 20 h in serum-free medium, cells were treated with either 10 ng/ml IL-1␤ or vehicle control (0.0001% bovine serum albumin in phosphate-buffered saline (PBS)) for 0 -16 h. For inhibitor experiments cells were pretreated with 25 M curcumin or vehicle (0.05% Me 2 SO), or 30 M SP600125 or vehicle (0.3% Me 2 SO) for 30 min, or pretreated with 1 nM leptomycin B or vehicle (0.075% methanol) and then treated with 10 ng/ml IL-1␤ or vehicle for 30 min. For MG132 experiments, cells were pretreated for 1 h with 10 M MG132 or vehicle (0.1% Me 2 SO) and then treated for 30 min with 10 ng/ml IL-1␤ or vehicle.
Cell Fractionation and Immunoblotting-Total cell protein or nuclear and cytosolic cell fractions were extracted as previously reported by Itoh et al. (13). Protein concentrations were determined using the BCA Assay from Pierce. Ten g of protein/well was electrophoresed in 12% acrylamide-ready gels (Bio-Rad Laboratories, Inc.). After transfer to nitrocellulose membranes, blots were blocked with 5% nonfat milk in Tris-buffered saline with 0.1% Tween 20 (TBS-T) for either 1 h at room temperature or 20 h at 4°C. Incubation in a 1:1000 dilution of anti-JNK and anti-phospho-JNK antibodies in 5% bovine serum albumin/TBS-T (Sigma) was carried out at 4°C for 12-16 h. All other primary antibody incubations were done for 1 h at 25°C at a 1:1000 dilution in the blocking buffer. After washing, membranes were incubated in 1:1000 goat anti-rabbit IgG horseradish peroxidase-linked antibody in the blocking buffer for 1 h at 25°C. Membranes were then washed and analyzed using Amersham Biosciences's Western Lightening Chemiluminescence kit and exposed to either Biomax film or the Kodak Imaging Station 2000R for detection.
Electrophoretic Mobility Shift Assays-Nuclear extracts were prepared from treated HepG2 cell cultures according to published methods (12). Protein concentrations were determined using the Bradford reagent (Sigma-Aldrich). Double-stranded rat sodium-dependent taurocholate co-transporting polypeptide RXR␣:RAR␣ DR2 element (rat sodium-dependent taurocholate co-transporting polypeptide Ϫ53/Ϫ40) was endlabeled, purified, and incubated with 10 g of HepG2 nuclear extracts for 30 min as described (12,13). After binding, each reaction was electrophoresed through a non-denaturing 5% polyacrylamide gel, the gel was dried and exposed to BioMax film for varying time periods or developed with Cyclone Storage phosphor system screens and analyzed.
Immunoflourescent Staining-HepG2 cells grown on coverslips were treated for 30 min with either saline vehicle or 10 ng/ml IL-1␤ and then washed with cold PBS. Cells were fixed for 20 min using 4% paraformaldehyde washed and permeabilized with 0.1% Triton X-100. After 30 min of blocking with 10% goat serum, cells were incubated for 1 h in a 1:50 dilution of anti-RXR␣ antibody to 1.5% goat serum/PBS. After three 5-min washes, cells were incubated with 1:500 goat anti-rabbit IgG-fluorescein isothiocyanate (Catalog number D1504, Santa Cruz) in 1.5% serum/PBS for 45 min. Cells were washed and mounted in Vectashield (Vector Labs, Burlingame, CA) mounting medium contain-ing 1.5 g/ml nuclear DNA stain DAPI and imaged using an Olympus microscope. Images were processed using softWorx software (Applied Precision, Issaquah, WA).
Transient Transfections-Twenty-four hours after plating onto coverslips, HepG2 cells were transfected using a 2:3 l of FuGENE 6:g of plasmid DNA ratio/35-mm dish for 16 h. Cells on coverslips were placed in serum-free conditions for 20 h before treatment with either saline or IL-1␤ for 30 or 60 min, fixed with 4% paraformaldehyde in PBS for 20 min at 25°C. Coverslips were mounted on microscope slides using Vectashield.
Microscopy and Deconvolution Analysis-Treated and fixed HepG2 cells were examined using an Olympus microscope and acquired images were deconvolved using softWorx software (Applied Precision, Issaquah, WA). Intracellular location was quantified in a blinded fashion by counting 100 consecutive GFP-expressing cells/treatment group, and categorizing each cell as either exclusively nuclear (N), or both nuclear and cytoplasmic (NϩC), expression. Experiments were performed a minimum of three times.
Statistical Analyses-Data were expressed as the means Ϯ S.D. from at least three independent experiments. Differences between experimental groups were evaluated for statistical significance using Student's t test where p Ͻ 0.05 were considered to be statistically significant.

IL-1␤ Induces a Rapid Shift in RXR␣ Subcellular
Location-We first sought to determine the effects of IL-1␤ on RXR␣ protein levels and localization in the human liver-derived cell line, HepG2, a well studied model of IL-1␤ modulation of RXR␣-regulated hepatic gene expression (9,12). Within 30 min of exposure to IL-1␤, nuclear levels of RXR␣ fall significantly (Fig. 1, A and B) and return to baseline levels within 16 h after treatment. Correspondingly, within 30 min, RXR␣ protein is readily detectable in cytoplasmic fractions from IL-1␤-treated cells. Importantly, nuclear and cytoplasmic levels of two transcription factors, the heterodimer partner of RXR␣, RAR␣, and the non-nuclear receptor transcription factor Oct1, were not significantly affected by IL-1␤ treatment, (Fig. 1, A and B), demonstrating that the effects of IL-1␤ on RXR␣ protein subcellular localization are specific. Notably, slower migrating RXR␣ species were present in the nuclear fractions obtained 30 -60 min after cytokine treatment (Fig. 1A, *), consistent with rapid post-translational modification. Immunohistochemical detection of native RXR␣ protein in HepG2 cells mirrors the response to IL-1␤ seen in immunoblot analyses (Fig. 1, compare A with C-H). The typical native nuclear residence of RXR␣ is unchanged by saline ( Fig. 1, C, E, and G), whereas IL-1␤ treatment for 30 min clearly leads to a predominant cytoplasmic localization of RXR␣ (Fig. 1, D, F, and H). Thus, as evident from both immunoblot and immunocytochemical analyses, IL-1␤ treatment leads to a rapid and profound reduction in nuclear RXR␣, coincident with a relocalization of RXR␣ from the nucleus to cytoplasm.
Electrophoretic mobility shift assays (Fig. 1I ) performed on nuclear extracts from IL-1␤-treated HepG2 cells show decreased binding of nuclear proteins to a typical RXR␣ heterodimer binding site (DR2), with a timeline that corresponds to the rapid and sustained depletion of nuclear RXR␣ (Fig. 1, compare I with B). Because activation of JNK is involved in IL-1␤-mediated suppression of RXR␣:RAR␣ activation of the sodium-dependent taurocholate co-transporting polypeptide DR2 element, it was important to determine whether there is a temporal link between JNK activation and subcellular relocalization of RXR␣ (12). As seen in Fig. 1J, the timeline of IL-1␤-JNK activation (as phospho-JNK) is not only consistent with previous reports (15) but coincides with the peak subcellular relocalization of RXR␣ (Fig. 1J).
IL-1␤-mediated RXR␣ Nucleocytoplasmic Relocalization Is Critically Dependent upon JNK-JNK can phosphorylate RXR␣ at several sites, yet the physiological consequences are controversial (16). Nor is there a known role for IL-1␤-mediated effects on RXR␣ nucleocytoplasmic relocalization, although results in Fig. 1 support this possibility (14). To investigate the role of JNK in IL-1␤ modulation of RXR␣ subcellular localization, we utilized two potent inhibitors of JNK activity, curcumin, the yellow pigment derived from the spice turmeric (12,17), and SP600125 (18). As seen in Figs. 2, A-D, pretreatment with either agent potently inhibits the effects of IL-1␤ on the subcellular localization of RXR␣. These findings clearly support a critical role for IL-1␤-activated JNK in determining the subcellular localization of RXR␣.
Il-1␤-induced Rapid Subcellular Redistribution of RXR␣ Is due to Nuclear Export-Two potential explanations for the effects of IL-1␤ on RXR␣ subcellular localization were considered, either that there was an inhibition of nuclear import of recently translated RXR␣ or that IL-1␤ induced rapid nuclear export of pre-existing nuclear RXR␣. Given the rapid timeline and the limited evidence for significant residence of RXR␣ in the cytoplasm (see Fig. 1C (19, 20)) we considered the former possibility to be less likely than the latter. HepG2 cells were pretreated with either vehicle or leptomycin B (LMB), a potent inhibitor of CRM-1 (chromosome region maintenance-1)-dependent nuclear export, prior to IL-1␤ treatments (21). In extracts from IL-1␤-treated cells, nuclear RXR␣ levels were modestly, but insignificantly, reduced when pretreated with LMB (Fig. 3), whereas cytoplasmic RXR␣ levels remained essentially undetectable. Inhibition of newly translated RXR␣ import into the nucleus was unlikely to be a major component of the actions of IL-1␤, because pretreatment with the protein synthesis inhibitor cyclo-heximide had no discernible effect on either nuclear or cytoplasmic RXR␣ levels in response to IL-1␤ (data not shown). Taken together, these finding indicate that the rapid reduction in nuclear RXR␣, accompanied by its appearance in the cytoplasm in response to IL-␤-induced signaling, is because of nuclear export of resident nuclear RXR␣ protein and not via an impairment in nuclear import of newly synthesized RXR␣.
IL-1␤ Induces GFP-tagged RXR␣ Export from the Nucleus-To further elucidate the molecular mechanisms of RXR␣ nuclear export in response to IL-1␤, a RXR␣ C-terminally tagged GFP construct was made and expressed in transiently transfected HepG2 cells. The RXR␣-GFP fusion protein construct transactivated a reporter gene to a similar extent as transfected RXR␣, indicating that the GFP tag had no significant effect on the transcriptional activity of RXR␣ (data not shown). HepG2 cells were transiently transfected with RXR␣-GFP and analyzed in a blinded fashion by categorizing the subcellular location of GFP in individual transfected GFP-positive cells as either exclusively nuclear (N), or both nuclear and cytoplasmic (NϩC). Between 68 and 73% of transfected cells express the RXR␣-GFP protein exclusively in the nuclear compartment (vehicle treatments, Fig. 4 and 5), whereas the remaining 27-32% of transfected cells showed RXR␣-GFP expression throughout the cell (NϩC). When treated with IL-1␤ for 60 min, only 25-30% of transfected cells exhibited exclusive nuclear RXR␣-GFP localization (Ϸ70% reduction, Figs. 4 and 5), similar to the response of endogenous RXR␣ protein to IL-1␤ seen in immunoblot and immunocytochemical analyses (Fig. 1).
The Response of RXR␣-GFP to IL-1␤ Is Curcumin-and LMB-sensitive-To test the role of JNK phosphorylation and CRM-1-dependent nuclear export, we followed the response of RXR␣-GFP to IL-1␤ after a short pretreatment period with either curcumin or LMB. A 30-min pretreatment with the JNK inhibitor curcumin completely blocked the export of RXR␣ in response to IL-1␤ treatment (Fig. 4A), similar to that seen for native RXR␣ (Fig. 2, A and B). One hour of LMB pretreatment blocked the ability of IL-1␤ to cause a decrease in  nuclear levels of RXR␣-GFP (Fig. 4B). In concert with the immunoblot analyses (Figs. 2 and 3), these findings support critical and integral roles for JNK and CRM-1 in IL-1␤-induced nuclear export of both native and GFP-tagged, RXR␣, and that C-terminally tagged RXR␣-GFP faithfully models the response of native RXR␣ to IL-1␤.
Ser 260 Is Critically Involved in RXR␣-GFP Nuclear Export-Several amino acid residues in RXR␣, including Ser 260 , have been identified as JNK phosphorylation sites (14), although the functional consequences related to the phosphorylation status of these sites are uncertain (22). An analysis of the potential JNK sites in RXR␣ led us to first explore a role for Ser 260 in IL-1␤-mediated nuclear export by creating RXR␣S260A-GFP, where Ala is substituted for Ser 260 . Upon IL-1␤ treatment, nuclear export of RXR␣S260A-GFP was significantly attenuated in comparison to wild type RXR␣-GFP (Fig. 5, G-L compared with A-F ). After IL-1␤ treatment, 52% of cells transfected with RXR␣S260A-GFP still retained exclusive nuclear residence (Fig. 5M), significantly higher than 25% of cells transfected with wild type RXR␣-GFP (Fig. 5M). Transfected cells expressing either RXR␣-GFP or RXR␣S260A-GFP showed no significant difference in the predominantly nuclear localization of the fusion protein (73 and 71%, respectively; see Fig. 5, A, G, and M), nor in transactivating ability (data not shown). This indicates that Ser 260 is not required for either the native nuclear localization or function of RXR␣ but is required for full nuclear export in response to IL-1␤.
IL-1␤ Induces Proteasome-mediated Degradation of RXR␣-IL-1␤ induces prolonged suppression of nuclear RXR␣ levels, suggesting that the modified and exported RXR␣, once in the cytoplasm, may be highly susceptible to degradation, rather than acting as a substrate for remodification and reimportation back into the nucleus. Thirty minutes after exposure to IL-1␤, we isolated nuclear and cytoplasmic extracts from cultures pretreated for 1 h with vehicle (Me 2 SO) or the 26 S proteasome inhibitor, 1 MG132 (Fig. 6). Both nuclear and cytoplasmic concentrations of RXR␣ were increased by preincubation with MG132 at baseline (time 0), consistent with an inhibition of ongoing proteasomal degradation during the preincubation period by MG132. Importantly, there is clearly an accumulation of RXR␣ in both compartments in extracts prepared from cells pretreated with MG132 before exposure to IL-1␤ (30-min samples). These data support critical roles for proteasome-mediated degradation of RXR␣ in FIGURE 4. Curcumin and leptomycin B block IL-1␤-induced nuclear export of RXR␣-GFP. A, HepG2 cells were plated onto coverslips and transfected with RXR␣-GFP then pretreated with Me 2 SO vehicle or 25 mM curcumin for 30 min followed by a 60 min treatment with saline or 10 ng/ml IL-1␤. Cells were formalin-fixed, followed by blinded counts of 100 consecutive GFP-positive cells, scored as either nuclear (N, more than 90% of GFP protein was exclusively nuclear) or a mixed pattern of nuclear and cytosolic (NϩC) (n ϭ 3). B, HepG2 cells were transfected with RXR␣-GFP and pretreated with either methanol vehicle (MetOH) or 1 nM leptomycin B for 1 h followed by 60 min of saline or 10 ng/ml IL-1␤ exposure. Cells were fixed and counted as in A for N or NϩC GFP-expressing cells (n ϭ 3). Bar graphs depict mean Ϯ S.D.  Fig. 4 (n ϭ 3). Bar graphs depict mean Ϯ S.D.
response to IL-1␤ signaling, while still allowing for the processes of modification and export to continue.

DISCUSSION
In previous studies, we have identified inflammation-mediated suppression of hepatobiliary transporter gene expression via involvement of IL-1␤-induced JNK activation and subsequent reduction of nuclear RXR␣-containing heterodimer binding capacity and function (12). However, the consequences and roles of JNK on RXR␣ function or subcellular relocalization were unknown. In this report, we sought to determine the underlying molecular mechanisms and provide evidence to support the hypothesis that the reduction in RXR␣ nuclear binding activity is because of the rapid and specific JNK-dependent modification and nuclear export of RXR␣. In addition, RXR␣ nucleocytoplasmic translocation is LMB-sensitive and therefore exported from the nucleus via a CRM-1/exportin-1 dependent mechanism. We identify one residue, Ser 260 , a known JNK phosphorylation site whose functional importance was previously unknown, as critical to IL-1␤-mediated RXR␣ nuclear export. Finally, exported RXR␣ undergoes proteasome-mediated degradation, linking IL-1␤ cell signaling pathways to reductions in nuclear RXR␣ levels. These studies provide an explanation for the suppression of RXR␣-dependent gene expression by IL-1␤ (12). Taken together, this is the first description of a direct molecular link between activation of cell signaling pathways and alteration of the nuclear residence of the central type II nuclear receptor heterodimer partner, RXR␣.
Understanding the mechanisms of the hepatic response to inflammation is clinically relevant, because inflammation is a major component of both acute and chronic liver diseases, and RXR␣-regulated genes are essential to multiple physiological processes in liver (23)(24)(25). Moreover, Geier et al. (10) recently showed that IL-1␤ is a critical mediator of LPS-mediated suppression of transporter RNA expression, thereby placing IL-1␤-induced signaling in hepatocytes as central to the effects of inflammation on hepatic gene expression. If long term inflammatory signaling leads to continued impairment of RXR␣-regulated hepatic gene expression, the consequences would likely lead to further exacerbation of hepatic damage and disruption of homeostatic mechanisms. For example, genes responsible for the hydroxylation and basolateral export of toxic bile acids such as lithocholic acid, are positively regulated by RXR␣ and its heterodimer partner PXR (pregnane X receptor), but down-regulated in inflammation (26,27). Prolonged depletion of RXR␣ would lead to the impairment of this hepatoprotective mechanism, leading to the accumulation of toxic bile acids, with resultant increased hepatotoxicity. Similar impairments in other RXR␣-regulated hepatic pathways (glucose metabolism, xenobiotic detoxification, cholesterol and lipid metabolism, and others) are known to occur in hepatic inflammation and chronic liver disease (7). This makes it attractive to speculate that therapies targeted toward ameliorating this particular cell signaling pathway on RXR␣ function may provide a novel means of restoring vital hepatic functions in these disease states.
Phosphorylation of RXR␣ has been studied in several systems; however the results do not provide a straightforward interpretation, likely because of multiple variables including differing experimental cells and conditions, use of a variety of JNK activators, and perhaps most importantly, a reliance on following the responses of transfected and overexpressed RXR␣, and not native RXR␣. Depending upon the experimental model and JNK inducer, researchers have reported no, increased, or decreased response of RXR␣ function to JNK activation (14, 16, 19, 22, 28 -31, 34). Adam-Stitah et al. (14) provided a detailed exploration of JNK-mediated phosphorylation sites in RXR␣, which included three sites in the N-terminal region, as well as Ser 260 in the ligand binding domain, but did not describe any role for phosphorylation of these sites in directing subcellular location. Combining the present and previous  studies on the effects of IL-1␤, it is clear that IL-1␤-induced cell signaling rapidly and reproducibly reduces nuclear RXR␣ levels and function via a JNK-dependent pathway (12); however, a direct role for JNK-dependent phosphorylation of Ser 260 as a critical component of IL-1␤mediated nuclear export of RXR␣ remains to be proven.
Linking IL-1␤-induced Signaling to Phosphorylation, Nuclear Export, and Proteasome-mediated Degradation of RXR␣-Several lines of evidence support this novel pathway of regulating the nuclear residence of RXR␣: 1) IL-1␤-induced JNK activation is central to RXR␣ nuclear export, 2) IL-1␤-activated JNK directly phosphorylates RXR␣ (12), 3) Ser 260 is necessary for IL-1␤ induced nuclear export, 4) inhibition of CRM-1-dependent nuclear export maintains nuclear RXR␣ levels, 5) proteasome inhibition maintains whole cell RXR␣ levels, and 6) immunoblot analyses suggest a transient higher molecular weight species of RXR␣, (Fig. 1A, *) consistent with phosphorylated RXR␣ (14,32,34). It is unlikely that a significant contribution is made from JNK-mediated modification of one or more RXR␣ partners and that RXR␣ is "dragged" out of the nucleus because of a partner effect, although this indirect route to nuclear export cannot be completely excluded (19,33). Further experimentation will help clarify any possible contributing role for an indirect route of IL-1␤ on RXR␣ nuclear export.
Preincubation of cells with the 26 S proteasome inhibitor MG132 led to a block in the IL-1␤-mediated suppression of RXR␣ in the nucleus and accumulation of the protein in the cytoplasm. Whether or not there is a direct inhibition of both nuclear and cytoplasmic proteasomes by MG132 or that there is some re-entry of cytoplasmic RXR␣ during this short time point remains to be determined in future experiments. Combined with the data in Fig. 3, where there is a small decrease in nuclear RXR␣ in the presence of LMB (whereas the appearance of cytoplasmic RXR␣ is blocked), it is intriguing to speculate that there may be a proportion of proteasome-mediated degradation that takes place in the nucleus. Perhaps in response to IL-1␤-induced signaling, the modified RXR␣ molecules that remain in the nucleus may still be targets for nuclear proteasomes. Future experiments are needed to help determine the relative contributions of proteasome-mediated degradation that takes place in each subcellular compartment.
RXR␣ Is a Central Target for Modulation of Gene Expression during the Negative Hepatic APR-One of the hallmarks of the hepatic APR is the broad suppression of multiple hepatic physiological processes (endobiotic/xenobiotic metabolism, bile formation, lipid metabolism, etc.). Many of the genes involved in these processes are critically regulated by RXR␣-containing heterodimers (22). Any impairment in RXR␣ activity or nuclear concentration will have wide ranging consequences, given the multiple target gene families regulated by RXR␣ and its partners (32,33). In the studies presented here, we indicate that RXR␣ is a direct target for cell signaling pathways, with rapid post-translational modification that results in nuclear export and subsequent proteasomal degradation.
In summary, the rapid nucleocytoplasmic translocation and degradation of RXR␣ in response to IL-1␤-induced cell signaling is a novel and potentially powerful component of the broad changes in hepatic gene expression seen in the negative hepatic APR (Fig. 7). It is likely that prolonged suppression of RXR␣-regulated pathways plays a prominent role in the progression of liver injury in chronic liver diseases, most all of which invoke inflammation as a major pathophysiological factor or cofactor. The pathophysiological mechanisms provided in this study points to several new therapeutic targets for treating altered hepatic gene function in acute and chronic liver diseases by addressing cell signaling effects on RXR␣.