Nuclear Factor (NF)-κB-dependent Thyroid Hormone Receptor β1 Expression Controls Dendritic Cell Function via Akt Signaling*

Despite considerable progress in our understanding of the interplay between immune and endocrine systems, the role of thyroid hormones and their receptors in the control of adaptive immunity is still uncertain. Here, we investigated the role of thyroid hormone receptor (TR) β1 signaling in modulating dendritic cell (DC) physiology and the intracellular mechanisms underlying these immunoregulatory effects. Exposure of DCs to triiodothyronine (T3) resulted in a rapid and sustained increase in Akt phosphorylation independently of phosphatidylinositol 3-kinase activation, which was essential for supporting T3-induced DC maturation and interleukin (IL)-12 production. This effect was dependent on intact TRβ1 signaling as small interfering RNA-mediated silencing of TRβ1 expression prevented T3-induced DC maturation and IL-12 secretion as well as Akt activation and IκB-ϵ degradation. In turn, T3 up-regulated TRβ1 expression through mechanisms involving NF-κB, suggesting an autocrine regulatory loop to control hormone-dependent TRβ1 signaling. These findings were confirmed by chromatin immunoprecipitation analysis, which disclosed a new functional NF-κB consensus site in the promoter region of the TRB1 gene. Thus, a T3-induced NF-κB-dependent mechanism controls TRβ1 expression, which in turn signals DCs to promote maturation and function via an Akt-dependent but PI3K-independent pathway. These results underscore a novel unrecognized target that regulates DC maturation and function with critical implications in immunopathology at the cross-roads of the immune-endocrine circuits.

Despite considerable progress in our understanding of the interplay between immune and endocrine systems, the role of thyroid hormones and their receptors in the control of adaptive immunity is still uncertain. Here, we investigated the role of thyroid hormone receptor (TR) ␤ 1 signaling in modulating dendritic cell (DC) physiology and the intracellular mechanisms underlying these immunoregulatory effects. Exposure of DCs to triiodothyronine (T 3 ) resulted in a rapid and sustained increase in Akt phosphorylation independently of phosphatidylinositol 3-kinase activation, which was essential for supporting T 3 -induced DC maturation and interleukin (IL)-12 production. This effect was dependent on intact TR␤ 1 signaling as small interfering RNA-mediated silencing of TR␤ 1 expression prevented T 3 -induced DC maturation and IL-12 secretion as well as Akt activation and IB-⑀ degradation. In turn, T 3 up-regulated TR␤ 1 expression through mechanisms involving NF-B, suggesting an autocrine regulatory loop to control hormone-dependent TR␤ 1 signaling. These findings were confirmed by chromatin immunoprecipitation analysis, which disclosed a new functional NF-B consensus site in the promoter region of the TRB1 gene. Thus, a T 3 -induced NF-B-dependent mechanism controls TR␤ 1 expression, which in turn signals DCs to promote maturation and function via an Akt-dependent but PI3K-independent pathway. These results underscore a novel unrecognized target that regulates DC maturation and function with critical implications in immunopathology at the crossroads of the immune-endocrine circuits.
The endocrine and immune systems are interconnected via a bidirectional network in which hormones affect immune function, and, in turn, immune responses are reflected in neuroendocrine changes. This bidirectional communication is possible as both systems share common ligands (hormones and cytokines) and their specific receptors (1). Thyroid hormones (TH) 5 play critical roles in differentiation, growth, and metabolism. The classic genomic actions of TH are mediated by nuclear TH receptors (TR) that act mainly as hormone-inducible transcription factors. Several TR␣ and TR␤ isoforms are encoded by the TRA and TRB genes, respectively. The TR␣ 1 , TR␣ 2 , TR␤ 1 , and TR␤ 3 isoforms are widely expressed, whereas TR␤ 2 is predominantly restricted to the hypothalamus-pituitary axis (2). Recent emerging evidence has also characterized the interactions of TR with co-repressor proteins, namely the nuclear co-repressor and the silencing mediator of retinoid and TH receptors. These effects involve histone deacetylase activity that mediates TR silencing in the absence of triiodothyronine (T 3 ) and several co-activator proteins that exhibit histone acetylase activity in the presence of this hormone (2). However, the notion of classical or genomic mechanisms as unique actions mediated by TRs has been challenged in the past decade by descriptions of TH actions that involve extranuclear (nongenomic) effects in a variety of cell types. These TH-dependent pathways are associated to extranuclear TR localized within the cytoplasm and the plasma membrane (3)(4)(5) and to TH-dependent effects mediated by the cell surface ␣ v ␤ 3 integrin (6). Several cytoplasmic T 3 actions mediated by TR are linked to activation of the PI3K pathway in alveolar cells (7) and human fibroblasts (8). Moreover, activation of Akt, a critical component of cell growth and survival (9), has been detected in pancreatic islet ␤ cells upon engagement of TR␤ 1 and activation of PI3K-p85 (10).
Despite considerable progress in understanding the interplay between distinct hormones and the immune cell network, the role of TH in the control of immune cell physiology has received scarce attention with studies almost exclusively focused on effector B and T lymphocytes (11,12). However, the role of TR signaling in the initiation of adaptive immunity remains elusive.
Dendritic cells (DCs) are highly specialized antigen-presenting cells that recognize, process, and present antigens to naive T cells for the induction of antigen-specific immune responses (13). Given the remarkable plasticity of these cells, manipulation of their function to favor the induction of DCs with immunogenic or tolerogenic properties could be exploited to stimulate or attenuate immune responses (14). After in vitro or in vivo exposure to lipopolysaccharides (LPS) or other microbial products, DCs undergo activation and maturation through different signaling pathways, including MAPKK1/ERK, which favors DC survival, and the Akt and NF-B pathways, which allow for DC maturation (15,16). Signaling through NF-B also determines the increased expression of major histocompatibility complex (MHC) II and co-stimulatory molecules, release of proinflammatory cytokines and chemokines, and DC migration and recruitment. This coordinated process leads to sustained T cell stimulatory capacity and IL-12 production, which result in the induction of protective Th1 immunity (15). Recently, we have provided the first evidence indicating a role for TH in the control of DC maturation (17). Our initial results demonstrated the expression of TR, mainly the ␤ 1 isoform, on bone marrow-derived murine DCs and the role of T 3 in driving DC maturation (17).
In this study, we investigated the signaling pathways leading to T 3 effects within the DC compartment, their impact in the immunogenicity of these cells, and the relevance of TR␤ 1 signaling in this process. Our results demonstrate a novel link among NF-B-dependent TR␤ 1 expression, T 3 -induced Akt activation, and the regulation of DC physiology with critical implications in immunopathology.

EXPERIMENTAL PROCEDURES
Mice-Female C57BL/6 mice (B6; H-2b) were obtained from Ezeiza Atomic Center (Buenos Aires, Argentina). Mice were maintained under specific pathogen-free conditions and used at 6 -10 weeks old. Animal protocols were in compliance with the Guide for the Care and Use of Laboratory Animals published by the National Institutes of Health and the local institutional animal care committee.
DC Preparation and Culture-DCs were obtained as described by Inaba et al. (18). Briefly, bone marrow progenitors were collected from the femurs of C57BL/6 mice, cultured in RPMI 1640 medium, 10% fetal calf serum depleted of TH by treatment with resin AG-1-X8 (Bio-Rad) in the presence of granulocyte-macrophage colony-stimulating factor from supernatants of the J558 cell line, and fed every 2 days. At day 10 of cell culture, Ͼ85% of the harvested cells expressed MHC II, CD40, CD80, and CD11c but not Gr-1. Immature DCs (iDCs) were cultured with T 3 (5 nM, DC T3 ) or LPS (100 ng/ml; Escherichia coli strain 0111:B4; Sigma; DC LPS ). Parallel cultures were maintained without stimuli and used as controls (DC Control ). T 3 was purchased from Sigma and prepared according to the manufacturer's recommended protocol. The concentration of T 3 used in the experiments led to 60 -80% of thyroid receptor occupancy, a situation in which most of the biological actions of TH take place under physiological conditions (2). To rule out endotoxin contamination of the T 3 preparation, we checked endotoxin content after reconstitution of the hormone, which raised levels lower than 0.03 IU/ml (limit of detection) by using the Limulus amebocyte lysate assay (Sigma).
Immunofluorescence Microscopy-Bone marrow-derived DCs were generated as described above and cultured on coverslips for 3 days. After treatments, cells were fixed in 4% (w/v) paraformaldehyde, permeabilized in 0.25% (v/v) Triton X-100 in PBS, blocked for 1 h in PBS, pH 7.4, plus 3% (w/v) bovine serum albumin fraction V (Fisher), incubated with primary antibodies (mouse anti-TR␤ 1 sc-738, rabbit anti-Akt 1/2/3 sc-8312, and rabbit anti-PI3K p85␣ sc-423, Santa Cruz Biotechnology) at a dilution of 1:100 for 1 h, washed, and further incubated with Alexa-conjugated goat anti-mouse and goat antirabbit secondary antibodies (Molecular Probes) for 1 h at a dilution of 1:1000. Nuclei were stained with 4,6-diamidino-2phenylindole for 5 min, and samples were washed in PBS and mounted on glass slides using Fluoromount-G (Southern Biotechnology Associates, Birmingham, AL) for examination using an inverted fluorescence microscope equipped for confocal microscopy (Olympus Flowview FV500, Hamburg, Germany). Images were captured using Openlab 3.1 software (Improvision, Lexington, MA) at a magnification of ϫ1000. Negative controls, including omission of primary antibodies, were also performed.
Flow Cytometry-Cells were washed twice with PBS supplemented with 2% (v/v) fetal calf serum and resuspended in 10% (v/v) fetal calf serum in PBS. Cells were then incubated with the following fluorochrome-conjugated monoclonal antibodies for 30 min at 4°C: fluorescein isothiocyanate-anti-CD11c, PE-anti-IA/IE (MHC II), PE-anti-CD40, PE-anti-CD80, and PE-anti-CD86 (Pharmingen). To determine TR␤ 1 expression, DCs were fixed with 1% (w/v) paraformaldehyde, treated with fluorescence-activated cell sorter permeabilizing solution (BD Biosciences), stained with a mouse anti-TR␤ 1 antibody (sc-738, Santa Cruz Biotechnology) at a dilution of 1:100, and further incubated with an Alexa-conjugated goat anti-mouse secondary antibody (Molecular Probes) for 1 h at a dilution of 1:1000. Intracellular cytokine detection was assessed by flow cytometry as described (20) using PE-conjugated anti-IL-12 monoclonal antibody (from BD Biosciences). Briefly, DCs incubated with T 3 or LPS in the absence or presence of 1 M BAY 11-7082 or 2 nM sulfasalazine (NF-B inhibitors), 10 M Akt 1/2 kinase inhibitor (with no inhibitory activity on PI3K) added only once at the beginning of the individual experiments were exposed to brefeldin A (10 g/ml; Sigma) for the last 6 h of cell culture. Cells were then fixed with 1% (w/v) paraformaldehyde, treated with fluorescence-activated cell sorter permeabilizing solution, and stained with an optimal concentration of anti-cytokine monoclonal antibody or an appropriate isotype control (all from BD Biosciences). Cells (at least 10,000 viable cells) were then analyzed in a FACSAria flow cytometer (BD Biosciences) using Flowjo software (Tree Star, Ashland, OR).
RNA Interference-Cells were plated onto six multiwell plates and grown in complete medium. After 24 h, cells were transfected with siRNA SmartPool for thyroid hormone receptor ␤, Akt, Rel A, or PI3K (p85␣) (Dharmacon, Lafayette, CO). Dharmacon also provided the negative control siRNA (scramble siRNA). Transfection was conducted as described previously (21). Briefly, 200 pmol of annealed siRNA or scramble siRNAs were incubated with 7 l of GenePorter (Gene Therapy Systems) in a volume of 100 l of RPMI 1640 medium (serumfree) for 30 min and added to 900 l of DC culture (2 ϫ 10 6 cells) as described above. After 4 h of incubation, an equal volume of RPMI 1640 medium supplemented with 20% (v/v) fetal calf serum was added. After 24 -48 h, transfected DCs were washed and used for subsequent experiments. Silencing of target genes was checked by immunoblot analysis.
Preparation of Nuclear and Cytoplasmic Extracts-Nuclear and cytoplasmic DC extracts were obtained by subcellular fractionation as described previously (22). The supernatant containing cytoplasm was collected and frozen at Ϫ80°C or used immediately. The nuclear pellet was resuspended in 50 l of ice-cold buffer (20 mM HEPES, pH 7.9, 0.4 M NaCl, 1 mM EDTA, 1 mM EGTA, 1 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride), and the tube was vigorously rocked at 4°C for 15 min on a shaking platform. The nuclear extract was centrifuged for 5 min in a microcentrifuge at 4°C, and the supernatant was frozen in aliquots at Ϫ80°C or used immediately.
ChIP and PCR Analysis of Precipitated DNA-Formaldehyde cross-linking, whole cell extract preparation, immunoprecipitation, DNA purification, and PCR analysis were performed as described previously (23) with some modifications (24,25). Antibodies used for immunoprecipitation include anti-NF-B-p65 (10 g, sc-8008; Santa Cruz Biotechnology) and an unrelated isotype control (2 l whole serum). Sonication treatment resulted in average DNA fragment sizes of 0.5-1 kb. PCR (25-35 cycles) was carried out in a 20-l volume with 1:100, 1:10, and 1:1 of the immunoprecipitated material and 1:10,000 of the input material. PCR products were resolved in 2% (w/v) agarose gel and visualized with 0.004 g/ml ethidium bromide. Oligonucleotide PCR primers used for ChIP analysis amplified a DNA fragment (206 kb) homologous to the promoter of the TRB1 gene encompassing the putative NF-B consensus site (sense, TGATCAGAGTGAGGCTGTGC, and antisense, GCTGGTGGAGTCCTGAGTTC).
Real Time PCR Quantitation of Co-immunoprecipitated Promoter Fragments-The relative proportions of co-immunoprecipitated promoter fragments were determined by real time PCR using the primers listed above. PCR was performed using the iCycler iQ real time PCR detection system (7500 real time PCR system. Applied Biosystems) and a SYBR Green-based kit for quantitative PCR (Bio-Rad). The relative proportions of coimmunoprecipitated promoter fragments were determined based on the threshold cycle (C T ) value for each PCR. Real time PCR data analysis followed the methods described in ChIPquantitative PCR primer assays, quantitative real time PCR assays for chromatin immunoprecipitation analyses (user manual, SuperArray Bioscience Corp.). For promoter studies, a ⌬C T value was calculated for each sample by subtracting the C T value for the input (to account for differences in amplification efficiencies and DNA quantities before immunoprecipitation) from the C T value obtained for the immunoprecipitated sample. A ⌬⌬C T value was then calculated by subtracting the ⌬C T value for the T 3 -treated sample from the ⌬C T value of the corresponding control sample. Fold changes in occupancy (NF-B-p65 ChIP relative to control ChIP) were then determined by raising 2 to the Ϫ⌬⌬C T power. For promoter fragment analysis, each sample was quantified in triplicate on at least two separate occasions and from at least three independent immunoprecipitation experiments. Mean Ϯ S.D. values were determined for each fold difference, and these values were subsequently used in two-tailed paired t tests to determine statistical significance, as reported in the figures. A melt curve analysis was performed for each sample after PCR amplification to ensure that a single product of expected melt curve characteristics was obtained.
Statistical Analysis-Statistical analysis was performed using the Student's paired t test. Wilcoxon nonparametric test for paired data was used to determine the significance of the timeresponse curves. p values less than 0.05 were considered statistically significant. To adjust the significance level for multiple comparisons, Bonferroni correction was applied using a corrected significance level of 0.017. All experiments were performed at least in triplicate.

RESULTS
T 3 Triggers PI3K-independent Akt Activation in Bone Marrow-derived Mouse DCs-Given the immunostimulatory effects of T 3 and the preferential cytoplasmic localization of TR␤ 1 FIGURE 1. T 3 induces activation of Akt in a PI3K-independent manner. Bone marrow-derived iDCs differentiated with granulocyte-macrophage colonystimulating factor for 7 days were exposed to T 3 (5 nM) or LPS (100 ng/ml). After different time periods, cells were harvested, and lysates were prepared and analyzed by Western blot for phosphorylated (pAkt) or unphosphorylated Akt (Akt 1), phosphorylated ERK 1/2, p38, or JNK. In parallel DCs were preincubated with LY249002 or wortmannin (PI3K inhibitors), CHX, or H89 (PKA inhibitor) for 30 min before exposure to T 3 or LPS. A, T 3 induces rapid phosphorylation of Akt (after 5 min of exposure), which persists for at least 18 h in a PI3K-and CHX-insensitive manner. Lower panel shows the same blot probed for total Akt to check equal loading of samples. B, phosphorylation of Akt occurs within 15 min following exposure to LPS and is inhibited by preincubation with specific PI3K inhibitors. C, in contrast to LPS, T 3 is unable to induce phosphorylation of ERK 1/2, p38, or JNK on DCs. on DCs (17), in this study we investigated the intracellular pathways triggered by TR␤ 1 signaling on these cells and their relevance in DC physiology. Because the Akt signaling pathway fine-tunes the immunostimulatory function of DCs (26), we first examined the activation of this pathway in T 3 -stimulated (DC T3 )and control (DC Control ) DCs. Bone marrow-derived iDCs were cultured in the presence or in the absence of T 3 (5 nM) for 5, 15, 30, and 60 min and 18 h. A significant increase in

Control of DC Function by TR␤ 1 Signaling
Ser-473 phosphorylation of Akt was detected as early as 5 min following exposure of DCs to T 3 with a peak detected at 30 min of incubation (versus control). The increase in Akt phosphorylation persisted even at 18 h of addition of the thyroid hormone (Fig. 1A). Remarkably, T 3 -induced Akt activation seemed to be independent of PI3K, as pretreatment of iDCs with PI3K inhibitors (wortmannin and LY294002) was not capable of preventing Akt phosphorylation (Fig. 1A). In turn, T 3 effects did not require de novo synthesis of proteins because the addition of CHX to iDC cultures did not alter Ser-473 phosphorylation of Akt (Fig. 1A). Of note, Akt phosphorylation was barely observed on DCs exposed to LPS (Fig. 1B).
Given the involvement of other major signaling pathways during DC maturation (27), we further examined the phosphorylation of ERK 1/2, p38, and JNK following treatment with T 3 as well as the effect of the inhibition of cAMP-activated protein kinase (protein kinase A (PKA)) signaling on T 3 -induced Akt phosphorylation. In contrast to LPS, T 3 could not trigger ERK 1/2, p38, or JNK phosphorylation when added for different time periods to DC cultures (Fig. 1C). Moreover, H89, a potent inhibitor of PKA, could not prevent T 3 -induced Akt phosphorylation (Fig. 1D). In addition, as PP2A (protein phosphatase 2A) negatively regulates Akt activity in various systems (28), we further explored whether a reduction of PP2A C levels could underlie T 3 -mediated up-regulation of Akt. Of note, the levels of PP2A C were not reduced following treatment of DCs with T 3 . On the contrary, PP2A C was slightly increased compared with DCs cultured with medium alone (supplemental Fig. 1). Moreover, exposure to T 3 did not alter the normal interaction between Akt and PP2A C (supplemental Fig. 2), disregarding, at least in part, the participation of this serine/threonine phosphatase in T 3 -mediated induction of Akt. Collectively, these results support the concept of an Akt-dependent PI3K-independent mechanism underlying T 3 modulation of DC physiology. In . TR␤ 1 does not co-localize with PI3K-p85 in bone marrow-derived iDC. Bone marrow-derived iDCs were cultured for 18 h with or without T 3 (5 nM) (DC T3 and DC Control , respectively). A, immunofluorescence staining. DCs were immunostained for TR␤ 1 (green fluorescence) and PI3K-p85 (red fluorescence). All images were acquired and analyzed by confocal microscopy as described under "Experimental Procedures." B, co-immunoprecipitation analysis. Immunoprecipitation (IP) was performed in whole cell lysates using an anti-TR␤ 1 antibody or an irrelevant isotype control antibody. Western blot (WB) analysis for PI3K (p85␣) was performed as described under "Experimental Procedures." The identities of the bands were corroborated with the input control. Blots are representative of three independent experiments. addition, this mechanism does not appear to involve other major signaling pathways, including ERK 1/2, p38, JNK, PKA, or PP2A.

TR␤ 1 Co-localizes and Interacts with Akt in a Ligand-independent Manner-Given the pronounced changes induced by T 3 in Akt phosphorylation, we further examined the possible interactions between TR␤ 1 and Akt in DC T3 and DC Control .
Confocal microscopy and co-immunoprecipitation analyses revealed a striking co-localization of TR␤ 1 and Akt both in DC Control as well as in DC T3 (Fig. 2, A-D). This effect was similar in the absence or presence of T 3 and was specific as shown by the lack of signal when we used an irrelevant isotype control antibody for immunoprecipitation (Fig. 2, B-D). This finding suggests that TR␤ 1 and Akt may interact physically in the cytoplasm of DCs independently of the presence of T 3 , supporting the notion of nongenomics actions of T 3 involving Akt activation. Consistent with the lack of effect of PI3K inhibitors on T 3 -induced Akt phosphorylation, we could observe no co-localization (Fig. 3A) or physical interaction (Fig. 3B) between TR␤ 1 and PI3K, neither in the absence nor in the presence of T 3 , as revealed by confocal microscopy and co-immunoprecipitation.
In several biological systems, Akt is recruited to the plasma membrane where it is activated by phosphorylation. Once activated, pAkt is translocated to different cell compartments, including the nucleus (10). Immunoblot analysis of pAkt in the nuclear fraction of DCs revealed the ability of T 3 to trigger rapid recruitment of pAkt (Ser-473) to the nucleus following exposure to T 3 (Fig. 4). Moreover, a significant proportion of TR␤ 1 shuttled to the nuclear fraction ( Fig. 2A and Fig. 4), in agreement with TR␤ 1 cytoplasmic-nuclear translocation described in other cell types (29). Collectively, these observations reinforce the concept of a PI3K-independent Akt activation mediated by T 3 in bone marrow-derived iDCs.

TR␤ 1 Transduces PI3K-independent Akt-and NF-B-dependent Signals That Mediate DC Maturation and IL-12
Production-The contribution of Akt and NF-B pathways to T 3 -mediated effects was substantiated by functional experiments. These involved evaluation of cell surface phenotypic markers and IL-12 production on iDCs exposed to T 3 or LPS in the presence or absence of Akt or NF-B inhibitors or following siRNA-mediated silencing of Akt, NF-B, or PI3K. Similar to DCs matured in the presence of LPS, exposure of DCs to T 3 resulted in enhanced percentage of cells expressing the co-stimulatory molecules CD40, CD80 (B7.1), and CD86 (B7.2), as well as enhanced expression (mean fluorescence intensity) of these cell surface molecules compared with DCs cultured in the absence of maturation stimuli (Fig. 5A). Remarkably, this effect was prevented when DCs were matured with either T 3 or LPS in the presence of a specific NF-B inhibitor (BAY 11-7082). However, pharmacological disruption of Akt specifically prevented T 3 but not LPS effects (Fig. 5A), suggesting that T 3 triggers DC maturation via Akt and NF-Bmediated mechanisms. The involvement of these pathways was also observed at the level of cytokine production. Exposure to T 3 resulted in a substantial increase in the frequency of IL-12producing CD11c ϩ DCs as well as in the amounts of secreted IL-12p70; these effects were significantly prevented by specific inhibitors of Akt (Akt 1/2 kinase inhibitor; p Ͻ 0.01), NF-B (BAY 11-7082; p Ͻ 0.01) (Fig. 5, B and C), or sulfasalazine (data not shown). Remarkably, similar findings were observed when Akt and NF-B expression was interrupted by specific siRNA (Fig. 6, B and C). Small interfering RNA-mediated silencing of Akt or NF-B successfully prevented T 3 -induced DC maturation, as shown by cell surface phenotypic markers (Fig. 6B) and IL-12 production (Fig. 6C). In contrast, T 3 effects could not be abrogated when DCs were exposed to this hormone in the presence of ERK 1/2, p38, or JNK inhibitors (data not shown). Moreover, siRNA-mediated silencing of PI3K could not abrogate T 3 effects (Fig. 6, B and C). The specificity and efficacy of all siRNAs used are shown in Fig. 6A. Thus, T 3 delivers maturation signals to iDCs via selective modulation of Akt (independently of PI3K) and NF-B.
Critical Role of TR␤ 1 Signaling in T 3 -induced DC Maturation-To investigate whether TR␤ 1 is critical in T 3 -induced DC maturation, we interrupted TR␤ 1 signaling on these cells by siRNAmediated silencing strategies. Transfection of TR␤ 1 -siRNA but not scramble siRNA almost completely abrogated TR␤ 1 expression, as shown by flow cytometric analysis of permeabilized DCs (Fig. 7A) and immunoblot analysis of total cell lysates (Fig.  7B). More importantly, siRNA-mediated TR␤ 1 silencing substantially prevented T 3 -induced DC maturation, as shown by a lower percentage of cells expressing MHC II, CD40, CD80, and CD86, and lower expression levels of these cell surface molecules (Fig. 7C). Furthermore, blockade of TR␤ 1 signaling completely prevented T 3 -induced IL-12 secretion (Fig. 7D), suggesting a critical function of this hormone receptor in transducing intracellular signals that endow DCs with the ability to polarize Th1 responses. This finding was confirmed with four separate siRNA duplexes (supplemental Fig. 3).
To further examine the involvement of TR␤ 1 in regulating T 3 -induced Akt and NF-B activation, we stimulated control or Western blot analysis of phosphorylated Akt (pAkt), Akt, and TR␤ 1 in protein extracts obtained from nuclear and cytoplasmic fractions of DCs exposed to T 3 for 30 min or 18 h. The expression of HDAC1 (nuclear) and ␣-tubulin (cytosolic) was analyzed as controls to check equal gel loading and exclude contamination of the nuclear fraction with cytoplasmic components. However, minimal cytoplasmic contamination of the nuclear fraction, as shown by ␣-tubulin expression, did not alter the conclusions raised. Blots are representative of three independent experiments. TR␤ 1 -siRNA-transfected DCs with T 3 for 30 min or 18 h and analyzed Akt phosphorylation and IB-⑀ degradation. Remarkably, TR␤ 1 silencing completely abrogated T 3 -induced Akt phosphorylation and IB-⑀ degradation, although it did not alter basal Akt or IB-⑀ expression. (Fig. 8). Likewise, TR␤ 1 or Akt silencing prevented T 3 -induced IB-␣ degradation (supplemental Fig. 4). Taken together, these results imply a critical role of intact TR␤ 1 expression in coupling T 3 binding, Akt-and NF-B-dependent signaling, and DC maturation.
T 3 Regulates TR␤ 1 Expression on DCs via NF-B-It is well established that TR␤ 1 expression is differentially regulated by T 3 in a tissue-specific manner (30) and that TH responses are dependent not only on TH but also on TR levels (2). Hence, we studied the effects of T 3 on TR␤ 1 expression on iDCs. Notably, treatment of iDCs with T 3 (5 nM) for 30 min considerably increased TR␤ 1 protein expression, which remained elevated up to 18 h and returned to control levels after 24 h of exposure to T 3 (Fig. 9A) (data not shown). Remarkably, pharmacological inhibition of the NF-B pathway by BAY 11-7082 as well as blockade of NF-B expression by siRNA (Rel A) silencing substantially prevented T 3 -induced up-regulation of TR␤ 1 (Fig. 9A, left and right panels), suggesting involvement of this transcription factor as a regulator of T 3 -induced TR␤ 1 expression in bone marrow-derived DCs.
The gene encoding mouse TR␤ is located in chromosome 14 (31) from 18,493,474 to 18,870,600 bp. Analysis of the promoter region corresponding to the start site of the TRB1 coding sequence (ATG, 18,814,391 bp) up to Ϫ1300 bp revealed a putative NF-B consensus site from Ϫ644 to Ϫ652 bp (GGATGTCCC). ChIP analysis using PCR probes that amplify a 206-bp fragment of the TRB1 promoter harboring the putative NF-B consensus site and DNA obtained from anti-NF-B (p65) immunoprecipitated chromatin revealed a positive signal (Fig. 9B). Remarkably, this signal was considerably amplified following treatment of iDCs with T 3 for 30 min and 18 h (Fig. 9, B and C). Our results identify a functional consensus region for NF-B in the promoter region of the TRB1 gene, which could be involved in T 3 -induced NF-Bdependent regulation of TR␤ 1 expression in bone marrow-derived iDCs. These results were confirmed by functional assays using a luciferase reporter fused to various fragments of the TRB1 promoter (containing or not the NF-B consensus site) co-transfected together with an NF-B expression vector in COS-7 cells (supplemental Fig. 5). These findings provide the first evidence of TRB1 as an NF-B target gene with critical implications in the control of inflammatory responses.

DISCUSSION
DCs are critical "decision-making" cells that must integrate signals from several pathways and receptors, including those arising from engagement of uptake and patternrecognition receptors, pro-inflammatory and anti-inflammatory cytokines, chemokines, and hormones to determine the type and magnitude of adaptive immune responses (32). Recently, we demonstrated an essential role for T 3 in promoting DC maturation and Th1-type cytokine secretion (17). As the T 3 -TR␤ 1 interaction may represent an attractive target for rational manipulation of the immunogenicity of DCs (32), here we investigated the molecular mechanisms and signaling pathways underlying these immunostimulatory effects. Collectively, the results presented in this study underscore a key role of TR␤ 1 signaling in coupling the thyroid system with the initiation of adaptive immunity, through intracellular pathways involving selective activation of Akt and NF-B. In addition, we identified TRB1 as a novel NF-B target gene with critical implications in the development of inflammatory responses.

Control of DC Function by TR␤ 1 Signaling
The ability of T 3 to access the cytoplasmic and nuclear compartments and transactivate TH-regulated genes is well established. This action, classically referred to as genomic effect, occurs within hours to days, which is consistent with the typical hormone functions, including regulation of cell growth, development, and metabolism (2). However, a different mechanism of action, termed nongenomic or extranuclear effect, has also been reported, which promotes rapid responses in cells and occurs within minutes or even seconds of exposure to TH. Such effects have actually been known for many years, although the underlying mechanisms are still poorly understood (6). Several intracellular pathways have been described to be responsible for TH-mediated cytoplasmic actions, including the MAPK, protein kinase C, and PI3K-protein kinase B/Akt pathways. The co-localization of TR␤ 1 with Akt, but not with PI3K, as well as the lack of effect of PI3K inhibitors in T 3 -induced Akt phos- phorylation and IL-12 production, supports the concept of a PI3K-independent mechanism of Akt activation. Moreover, the lack of ERK 1/2, p38, or JNK phosphorylation together with the sustained T 3 -induced Akt phosphorylation even in the presence of a PKA inhibitor disregard the participation of other major signaling pathways in T 3 effects. In turn, the co-localization of TR␤ 1 and Akt was independent of the addition of T 3 , whereas Akt phosphorylation required T 3 binding to its specific receptor. Hence, other molecular and/or biochemical events may be elicited following T 3 binding to TR␤ 1 . Several reports suggested PI3K-independent mechanisms of Akt activation induced by various agents, including forskolin, chlorophenylthio-cAMP, prostaglandin-E 1 , and 8-bromo-cAMP, which were shown to activate Akt through PKA. In this regard, Akt can be activated by a Ca 2ϩ -calmodulin-dependent kinase activity in vitro or by cellular stress through association with Hsp27. In addition, isoproterenol, a ␤-adrenergic agent, can activate Akt in a wortmannin-resistant manner (33). In turn, many proteins have been reported to interact with Akt to favor relevant biological functions, including Hsp27, in which interaction with Akt leads to inhibition of apoptosis in neutrophils (34). It is unlikely that T 3 -bound TR␤ 1 could phosphorylate Akt, as it lacks kinase activity. Nevertheless, the formation of a multiple protein complex associated with T 3 -bound TR␤ 1 -Akt that includes an enzyme with kinase properties may be responsible for the rapid Akt phosphorylation triggered by T 3 . However, the precise intracellular events triggered by T 3 -TR␤ 1 interactions and/or the molecular identity of this putative Akt kinase remain elusive. Yet, the lack of effect of CHX on T 3 -induced Akt phosphorylation does not support the involvement of a newly synthesized protein taking part in this process. Recently, other authors also reported T 3 -mediated cytoplasmic actions in ␤ pancreatic cells through mechanisms involving TR␤ 1 and Akt activation; yet this effect involved co-localization of PI3K with TR␤ 1 and was sensitive to PI3K inhibitors (10). On the other hand, although T 3 treatment did not induce changes in PPA2-Akt interaction or reduction in PPA2 expression, inhibition of phosphatase activity cannot be completely excluded as an alternative mechanism responsible for T 3 -induced Akt phosphorylation.
Remarkably, T 3 -dependent Akt activation was rapidly induced on DCs and lasted for several hours. This finding is of physiological and therapeutic relevance as Akt activation has been shown to be of critical importance for promoting DC survival (35). Hence, manipulation of Akt activation by T 3 -TR␤ 1 signaling opens new avenues for therapeutic harnessing of the inherent immunogenicity of DCs to develop more efficient DCbased tumor vaccines (35).
The increased production of IL-12 in T 3 -treated DCs was completely abolished by pharmacological or siRNA-mediated disruption of either Akt or NF-B pathways. These results contrast with those obtained using LPS as a stimulus, where only NF-B inhibitors significantly prevented the increase in the frequency of IL-12-producing CD11c ϩ DCs (36,37). In this regard, activation of Akt entails a complex series of events involving additional proteins. When phosphorylated, Akt can rapidly translocate to specific intracellular compartments and amplifies specific signaling processes. This sequential series of events is necessary to generate fully activated Akt and has been demonstrated also in experimental models (10).
In our study, T 3 treatment triggered translocation of active Akt to the nucleus, thus allowing this kinase the possibility to further regulate the expression and/or activity of nuclear proteins involved in DC physiology and survival. This finding suggests a dual role of T 3 -induced Akt activation, acting both at the cytoplasmic and nuclear compartments. Although in the absence of T 3 , TR␤ 1 expression was substantially lower in the nucleus compared with the cytoplasm of DCs (17), a significant pool of TR␤ 1 shuttled to the nucleus following exposure to this thyroid hormone, in agreement with other cell types (29). Hence, the participation of the classical genomic TR␤ 1 pathway cannot be excluded.
Given a number of reports emphasizing the independence of the TH receptor in T 3 -mediated cytoplasmic functions (6), a critical issue we wished to address was whether TR␤ 1 was essential or not for T 3 -mediated immunostimulatory effects. By means of siRNA-mediated silencing strategies, we found an essential role for TR␤ 1 in the control of DC maturation, signaling, and function, as shown by reduced expression of cell surface co-stimulatory molecules, diminished IL-12 production, impaired Akt phosphorylation, and unaltered IB-⑀ and IB-␣ levels following addition of T 3 to TR␤ 1 knockdown iDCs. These findings confirm the relevance of cytoplasmic TR␤ 1 expression in mediating T 3 effects within the DC compartment and predict important deficits in the initiation of adaptive immunity in patients suffering the TH resistance syndrome who carry nonfunctional TR␤ 1 (38). In this regard, Baumann et al. (29) demonstrated that TR␤ 1 rapidly shuttles between the nucleus and the cytoplasm. The possible cross-talk between nongenomic and genomic TH signaling is complex and still remains to be fully ascertained; yet these effects appear to act synergistically to modulate cellular processes. The postulated nongenomic signaling pathway may be complemented by nuclear actions of the thyroid hormone, which may amplify the system by generating second messengers and activating multiple signaling cascades. Illustrating this concept, the genomic and nongenomic effects of TH also appear to be strikingly synergistic within the mitochondrial compartment (39). These findings reveal the complexity of the TR signaling and suggest that a cytoplasmic TR with high homology to its nuclear counterpart might be implicated in T 3 -induced DC maturation. Small interfering RNAs, designed to specifically act on nuclear TR␤ 1 mRNA, could silence both cytoplasmic and nuclear TR␤ 1 without discrimination (data not shown). Furthermore, the commercial monoclonal antibody raised against a nuclear TR (see "Experimental Procedures") clearly recognized a cytoplasmic protein responsible for T 3 effects, supporting our hypothesis of a cytoplasmic TR with striking similarities to nuclear TR.
The formation of ligand-bound TR complexes is the critical step in the regulation of TH functions and the regulation of TR levels by its specific ligand T 3 is isoform-and cell type-dependent (40). In our study, T 3 treatment induced a significant increase in TR␤ 1 expression on DCs, and surprisingly, NF-B inhibitors prevented T 3 -induced TR␤ 1 up-regulation. From the early cloning of the TRB1 gene and the study of its promoter region, several consensus sites have been reported for many transcription factors (31,41). Here, we identified a functional consensus site for NF-B located Ϫ644 to Ϫ652 bp up to the starting ATG transcription site. Hence, the augmented expression of TR␤ 1 recorded following exposure of iDCs to T 3 may be achieved, at least in part, through stimulation of NF-B signaling, thus facilitating a positive regulatory loop in which T 3 regulates TR␤ 1 expression, and in turn TR␤ 1 mediates T 3 signaling. On the other hand, it has been demonstrated that classical NF-B-regulated promoters become derepressed by recruiting chromatin-associated IB kinase ␣, which is responsible for stimulating nuclear export and degradation of the silencing mediator of retinoid and TH receptors and recruitment of both histone acetyltransferase and ATP-dependent remodeling complexes (42). This mechanism may also contribute to the increased TR␤ 1 expression on DCs after T 3 -induced NF-B activation. Interestingly, other nuclear receptors also appear to be regulated by NF-B, as this transcription factor specifically binds the Ϫ574/Ϫ565 promoter region of the androgen receptor, which mediates repression of its transcription (43).
In conclusion, this study highlights a novel mechanism by which T 3 -TR␤ 1 signaling modulates DC maturation through activation of an Akt-and NF-B-dependent but PI3K-independent pathway. Finally, our data underscore a possible regulatory loop by which T 3 signaling promotes further expression of TR␤ 1 through mechanisms involving functional interactions of the NF-B transcription factor to the promoter region of the TRB1 gene.