IκBβ Interacts with the Retinoid X Receptor and Inhibits Retinoid-dependent Transactivation in Lipopolysaccharide-treated Cells

To elucidate the molecular action of the NFκB inhibitor IκBβ, we isolated a number of IκBβ interactors using the yeast two-hybrid system. These include the retinoid X receptor (RXR), whose interaction with IκBβ is significantly stimulated by the RXR ligand 9-cis-retinoic acid, as shown in the yeast system as well as the glutathione S-transferase pull down assays. RXR is a nuclear protein, whereas IκBβ accumulates in the nucleus only in cells stimulated with lipopolysaccharide or other inducers that result in prolonged activation of NFκB. Consistent with this, cotransfection with IκBβ specifically repressed the 9-cis-RA-induced transcriptional activities of RXR in an lipopolysaccharide-dependent manner. These results suggest a novel IκBβ-mediated antagonism between the signaling pathways of NFκB and RXR.

The transcription factor NFB is important for the inducible expression of a wide variety of cellular and viral genes (1,2). NFB is composed of homo-and heterodimeric complexes of members of the Rel/NFB family of polypeptides. In vertebrates, this family comprises p50, p65 (RelA), c-Rel, p52, and RelB. These proteins share a 300-amino acid region, known as the Rel homology domain, which binds to DNA and mediates homo-and heterodimerization. This domain also is the target of the IB␤ inhibitors, which include IB␣, IB␤, IB␥, Bcl-3, p105, and p100 (3). In the majority of cells, NFB exists in an inactive form in the cytoplasm, bound to the inhibitory IB proteins. Treatment of cells with various inducers results in the degradation of IB proteins. The bound NFB is released and translocates to the nucleus, where it activates appropriate target genes. IB␣ is degraded in response to all of the known inducers of NFB, whereas IB␤ is degraded only when cells are stimulated with inducers such as lipopolysaccharide (LPS) 1 and interleukin-1 that cause persistent activation of NFB (4). Following degradation of the initial pool of IB␤ in response to LPS or interleukin-1, newly synthesized IB␤ accumulates in the nucleus as an unphosphorylated protein that forms a stable complex with NFB and prevents it from binding to newly synthesized IB␣ (5)(6)(7).
To understand the molecular action of IB␤, we exploited the yeast two-hybrid system (8) to isolate a series of cDNAs encoding proteins that specifically interact with IB␤. Interestingly, these include retinoid X receptor (RXR), a member of the nuclear hormone receptors that comprise a large family of liganddependent transcription factors, bind as homodimers or heterodimers to their cognate DNA elements, and regulate genes involved in critical aspects of cell proliferation, differentiation, and homeostasis (9). Herein, we show that the RXR-IB␤ interactions are stimulated by the RXR ligand 9-cis-RA and that cotransfection with IB␤ specifically represses the 9-cis-RAinduced transcriptional activities of RXR in an LPS-dependent manner. These results are consistent with a novel IB␤-mediated antagonism between the signaling pathways of NFB and RXR.
Yeast Two-hybrid Screening of cDNA Library-Candidate IB␤ interacting clones were isolated from a mouse liver cDNA library (13) using the yeast two-hybrid system as described (8), with slight modifications. Approximately 10 6 primary yeast transformants of a derivative of EGY48 expressing the LexA-IB␤⌬1 were generated from an initial transformation with the cDNA library with selection for Trp auxotrophy. Library transformed cells were pooled and selected for Leu auxotrophy and expression of the LexA-␤-galactosidase construct. A number of leucine-independent colonies harboring cDNAs encoding candidate LexA-IB␤⌬1 interactors were obtained. The cDNA library plasmids were recovered from appropriate yeast strains, propagated in Escherichia coli, and reintroduced into EGY48 derivatives expressing LexA alone, LexA-IB␤⌬1, or other LexA chimeras to confirm specific interaction. Finally, identities of isolated cDNAs were determined by DNA sequencing. ʈʈ To whom correspondence should be addressed. Tel.: 82-62-520-7423; Fax: 82-62-522-5654; E-mail: jlee@chonnam.chonnam.ac.kr. 1 The abbreviations used are: LPS, lipopolysaccharide; RXR, retinoid X receptor; TR, thyroid hormone receptor; GST, glutathione S-transferase; ␤-gal, ␤-galactosidase; CAT, chloramphenical acetyltransferase; RA, retinoic acid.
Yeast Two-hybrid Test-For the yeast two-hybrid tests, plasmids encoding LexA fusions and B42 fusions were cotransformed into Saccharomyces cerevisiae EGY48 strain containing the ␤-gal reporter plasmid, SH/18 -34 (8). Plate and liquid assays of ␤-gal expression were carried out as described (10,13). Similar results were obtained in more than two similar experiments.
GST Pull Down Assays-The GST fusions or GST alone was expressed in E. coli, bound to glutathione-Sepharose-4B beads (Pharmacia), and incubated with labeled receptors or luciferase expressed by in vitro translation by using the TNT-coupled transcription-translation system, with conditions as described by the manufacturer (Promega, Madison, WI). Specifically bound proteins were eluted from beads with 40 mM reduced glutathione in 50 mM Tris (pH 8.0) and analyzed by SDS-polyacrylamide gel electorphoresis and autoradiography as described (12).
Cell Culture and Transfections-CV1 cells were grown in 24-well plates with medium supplemented with 10% charcoal-stripped serum for 24 h and transfected with 150 ng of ␤-galactosidase expression vector pRSV-␤-gal and 100 ng of a reporter gene TREpal-CAT along with 10 ng of RXR␣ and increasing amounts (10 -200 ng) of IB␤ expression vectors. For control experiments, Gal4-TKLuc and Gal4-VP16 (14) replaced TREpal-CAT and RXR, respectively. Total amounts of expression vectors were kept constant by adding decreasing amounts of the CDM8 expression vector to transfections containing increasing amounts of the IB␤ vector. After 12 h, cells were washed and refed with Dulbecco's modified Eagle's medium containing 10% charcoalstripped fetal bovine serum. After 12 h, cells were left unstimulated or stimulated with 2 g/ml LPS either in the presence or the absence of 10 Ϫ7 M 9-cis-RA. Cells were harvested 24 h later, and CAT or luciferase activity was assayed as described (15), and the results were normalized to the ␤-galactosidase expression. Similar results were obtained in more than two similar experiments.

RESULTS AND DISCUSSION
Genetic Selection for IB␤ Interacting cDNA Clones in Yeast-The validity of the yeast two-hybrid system (8) in iso-lating IB␤ interactors was tested by examining interaction properties of IB␤ with various components of the NFB complex in yeast. As expected from the published results (1-4), IB␤⌬1, a N-terminal deletion mutant of IB␤ (depicted in Fig.  1), interacted specifically with p65 but not with p50. Similarly, p50 interacted with p65, Bcl-3, and p50, whereas p65 did not interact with Bcl-3 or p65 (Table I). In addition, Bcl-3 readily interacted with IB␤ and IB␤⌬1. 2 Full-length IB␤ was a transcriptional activator in yeast when fused to a heterologous DNA binding domain, whereas IB␤⌬1 was transcriptionally inert. 2 Thus, we used IB␤⌬1 as a bait to isolate IB␤ interactors in the yeast two-hybrid system. One of the strongest IB␤ interactors isolated from a mouse liver cDNA library (13) encoded the RXR sequences from the third cysteine residue of the first zinc finger motif within the DNA-binding domain to the C termini. This result is consistent with our previous report in which IB␤ (initially referred to as Trip9) was first identified as one of a series of TR or RXR interacting proteins (10). These results immediately raised a possibility of a novel, IB␤-mediated crosstalk between NFB and nuclear receptor signaling pathways.
Ligand-stimulated Interactions of IB␤ and RXR/TR-The interaction of IB␤ with receptors was further characterized using the yeast two-hybrid system and GST pull down assays. As shown in Table II, the full-length IB␤ and three N-terminal deletion mutants (depicted in Fig. 1) corresponding to independent isolates of Trip9 (10) all interacted strongly with TR or RXR in a hormone-stimulated manner. In contrast, the related Bcl-3 protein interacted relatively weakly only with RXR in a 9-cis-RA-stimulated manner. Both IB␤ and Bcl-3 failed to interact with glucocorticoid receptor. As shown in Fig.  1, IB␤ contains six ankyrin repeats that constitute the interaction interface with the Rel homology domain of NFB, whereas Bcl-3 contains seven ankyrin repeats. Among these, ankyrin repeats 1, 5, and 6 of IB␤ and ankyrin repeat 3 of Bcl-3 contain a single amino acid motif LXXLL, recently shown to be an interaction interface for the nuclear hormone receptors (16,17). This motif apparently mediates the ligand-dependent interaction of the AF-2 transactivation domain of the receptors with transcription cofactors such as RIP-140, SRC-1, and CBP (16,17). Consistent with this, all of the N-terminal deletion mutants of IB␤ capable of interacting with the receptors retain the last two LXXLL motifs (Fig. 1). Thus, we tested whether IB␤ and the N-terminal deletion mutants interact 2 S. Y. Na and J. W. Lee, unpublished observations.

TABLE I Interactions between various components
of the NFB complex in yeast The indicated B42 and LexA plasmids were transformed into a yeast strain containing an appropriate ␤-galactosidase reporter gene. At least six separate transformants from each transformation were transferred to indicator plates containing 5-bromo-4-chloro-3-indolyl ␤-D-galactopyranoside, and reproducible results were obtained using colonies from a separate transformation. W, white colonies (no interaction); B, blue colonies after 1 day (strong interaction); LB, blue colonies after 3 days (weak interaction). LexA  The indicated B42 and LexA plasmids were transformed into a yeast strain containing an appropriate ␤-galactosidase reporter gene. The LexA fusions to GR, RXR, TR, and TR459 include sequences from the C terminus of the DNA-binding domain of the various receptors to their C termini as described (10,11,13). The respective ligands for GR, RXR, and TR were added to the plates as described (10). At least six separate transformants from each transformation were transferred to indicator plates containing 5-bromo-4-chloro-3-indolyl ␤-D-galactopyranoside, and reproducible results were obtained using colonies from a separate transformation. W, white colonies (no interaction); B, blue colonies after 1 day (strong interaction); LB, blue colonies after 3 days (weak interaction).
with LexA/TR459, a mutant TR-LBD that lacks AF-2 function but retains wild type affinity for thyroid hormone (T3) (10,11). This mutation blocked interaction with all of the IB␤ proteins, suggesting that the interaction interface may involve the AF-2 domain of nuclear receptors and the LXXLL motifs in the ankyrin repeats of IB␤.
To further characterize these interactions in vitro, GST fusions to IB␤ and IB␤⌬1 were expressed, purified, and tested for interaction with various in vitro translated receptor constructs. These include the ligand binding domains of RXR and TR (RXR-LBD and TR-LBD), the ABC domains of TR (TR-ABC), and full-length TRs deleted or point-mutated for the AF2 domain (TR⌬ and TR459, respectively). As shown in Fig. 2, IB␤ and IB␤⌬1 interacted weakly with TR-LBD and RXR-LBD in the absence of ligand. In agreement with the yeast results, however, interactions of IB␤ with TR-LBD or RXR-LBD were significantly enhanced in the presence of each cog-nate ligand. IB␤⌬1 behaved similarly. In contrast, the AF2 mutants TR⌬ and TR459 as well as TR-ABC showed relatively weak and hormone-independent interaction with IB␤ and IB␤⌬1. Thus, these results confirm the importance of the AF-2 for the ligand-dependent interactions and also suggest an additional ligand-independent interaction interface at the N-terminal ABC domains.
Cotransfections of IB␤ Repress the 9-cis-RA-induced Transcriptional Activities of RXR in an LPS-dependent Manner-To assess the functional consequences of these interactions, IB␤ was cotransfected into CV1 cells along with an RXR expression vector and a reporter construct controlled by TREpal, which is transactivated by RXR-RXR homodimers as well as various receptor heterodimers (18). Increasing amounts of cotransfected IB␤ had no significant effect on transcriptional activities of the TREpal-driven reporter, either in the presence or the absence of 9-cis-RA (Fig. 3). Similarly, RXR did not affect the ability of either IB␤ or IB␣ to inhibit transactivation by p65. 3 However, addition of 2 g/ml LPS specifically inhibited 9-cis-RA-induced transcription in an IB␤ dose-dependent manner, with cotransfection of 200 ng of IB␤ decreasing transcriptional activities to near background levels (Fig. 3). In contrast, cotransfection of IB␤ did not affect the transcriptional activity of Gal4-VP16, either in the presence or the absence of LPS, as assessed using the Gal4-TKLuc reporter construct (14) (data not shown). Similarly, IB␤ did not significantly affect ␤-galactosidase expression of the transfection indicator construct pRSV-␤-gal in the presence or the absence of LPS or 9-cis-RA (data not shown). These results are consistent with the proposal that IB␤ translocation into the nucleus is dependent on stimulation by chronic inducers such as LPS (5-7) and suggests that only this nuclear IB␤ is capable of interacting with nuclear receptors. These results along with the yeast and in vitro interaction results suggest that this nuclear IB␤ may mask the AF-2 domain of nuclear receptors from interacting with receptor coactivators. Alternatively, IB␤ bound to the AF-2 domain may have more direct inhibitory interactions with the transcriptional machinery. Consistent with this, LexA/IB␤ was a transcriptional activator in yeast, whereas LexA/IB␤⌬1 was not, suggesting the existence of an autonomous transactivation domain at the N terminus of IB␤ (amino acids 1-173). 2 In addition, full-length IB␤ also showed specific binding to a novel transcription cofactor we recently isolated. In conclusion, we have identified RXR as an IB␤ interactor and shown that this interaction decreases RXR-driven transcriptional activities in an LPS-dependent manner. This antagonism is in marked contrast to the glucocorticoid-dependent inhibition of NFB activities, in which glucocorticoids increase the synthesis of IB␣, which should then sequester NFB in an inactive cytoplasmic form (19). However, it was recently suggested that glucocorticoid-induced IB␣ synthesis and inhibition of NFB activity are two separable biochemical processes (20). Accordingly, glucocorticoid-mediated inhibition of NFB activity may involve other mechanisms such as the IB␤-RXR interactions described here. The antagonistic interaction is consistent with the fact that LPS is one of the best known pro-inflammatory agents (21), whereas retinoids are anti-inflammatory (22)(23)(24). Thus, exploration of these interactions may lead to new insights into mechanisms of inflammatory signal transduction pathways and possibly the development of new anti-inflammatory agents.