The Nuclear IκB Family Protein IκBNS Influences the Susceptibility to Experimental Autoimmune Encephalomyelitis in a Murine Model

The nuclear IκB family protein IκBNS is expressed in T cells and plays an important role in Interferon (IFN)-γ and Interleukin (IL)-2 production. IκB-ζ, the most similar homolog of IκBNS, plays an important role in the generation of T helper (Th)17 cells in cooperation with RORγt, a master regulator of Th17 cells. Thus, IκB-ζ deficient mice are resistant to Th17-dependent experimental autoimmune encephalomyelitis (EAE). However, IκB-ζ deficient mice develop the autoimmune-like Sjögren syndrome with aging. Here we found that IκBNS-deficient (Nfkbid−/−) mice show resistance against developing Th17-dependent EAE. We found that Nfkbid−/− T cells have decreased expression of IL-17-related genes and RORγt in response to Transforming Growth Factor (TGF)-β1 and IL-6 stimulation. Thus, IκBNS plays a pivotal role in the generation of Th17 cells and in the control of Th17-dependent EAE.


Introduction
The transcriptional regulator IkB NS is a member of the nuclear IkB family, which also includes IkB-f and Bcl-3. IkB NS expression occurs in T cells and is rapidly induced upon T cell receptor (TCR) stimulation [1]. IkB NS has 7 ankyrin repeat domains that bind the p50 subunit of the DNA-binding protein nuclear factorkappa B (NF-kB), but has no DNA-binding domain [2]. IkB NS interacts with NF-kB to control Interleukin (IL)-6 gene expression in macrophages [3,4]. In T cells, IkB NS positively regulates IL-2 expression, a target of NF-kB [5]. Previously, Schmitz's group showed that IkB NS intrinsically induces Forkhead box P3 (Foxp3) positive regulatory T cells (Tregs) in vivo and in vitro [6]. Foxp3 is a master regulator of Tregs and can be induced by NF-kB activation [7].
TCR and Transforming Growth Factor (TGF)-b signaling are necessary for the generation of both Tregs and T helper (Th)17 cells [8]. IL-17 is a proinflammatory cytokine that plays an important role in autoimmune diseases and against bacterial infections [9]. The nuclear IkB family protein IkB-f can be induced in T cells in response to TGF-b1 and IL-6 stimulation and positively regulates Th17 generation in cooperation with RORct [10]. Therefore, IkB-f-deficient mice are resistant to Th17dependent experimental autoimmune encephalomyelitis (EAE), a model of multiple sclerosis. However, these mice have more effecter memory T cells and produce more Interferon (IFN)-c in the periphery, leading to a Sjögren-like syndrome with age [11].
Here, we demonstrate that IkB NS -deficient (Nfkbid 2/2 ) mice are resistant to Th17-dependent EAE. Further analysis revealed that the percentage of Th17 cells in the draining lymph nodes (LNs) of myelin oligodendrocyte glycoprotein (MOG)-immunized Nfkbid 2/2 mice was decreased relative to that of control mice. In addition, IkB NS -deficient T cells were less capable of generating Th17 cells in response to TGF-b1 and IL-6. Mechanistically, we found that IkB NS -deficient T cells show decreased RORct induction in response to TGF-b1 and IL-6.

Mice
Nfkbid 2/2 mice (having a mixed C57/BL6 6 BALB/c genetic background) were established as described previously [3]. All mice were maintained in specific pathogen-free conditions in the animal facilities of Tohoku University. All animal protocols were approved by the Institutional Committee for the Use and Care of Laboratory Animals of Tohoku University (2013MA029, 2013MA031 and 2013MA032).

EAE induction and analysis of cytokine production
Nfkbid +/+ and Nfkbid 2/2 mice were injected subcutaneously (on the lower back) on day 0 with emulsions containing complete Freund's adjuvant (CFA; BD Difco TM ; Detroit, MI), 100 mg MOG peptide (MEVGWYRSPFSRVVHLYRNGK; MBL International Corporation, Nagoya, Japan), and 0.5 mg Mycobacterium tuberculosis H37RA (BD Difco TM ). In addition, these mice received 500 ng pertussis toxin (Sigma) by i.p. injection to boost immunological responses on day 0 and 2. These mice were observed until day 21 after immunization and clinical signs of EAE were scored according to a previously described protocol [12]. To study cytokine production, draining lymph node cells were derived and cultured for 72 h in the presence of 10 ng/ml MOG peptide.

Plasmids, antibodies, and cytokines
Expression vectors encoding FLAG-tagged mouse RORct and IkB-f were constructed as described previously [10,14]. Mouse IkB NS was inserted into a pcDNA3-FLAG vector at the EcoRI and BamHI sites. The mouse IL-17A promoter (26647 to +1) was inserted into a pGL4.12 vector at the NheI and HindIII sites. pGL4 and pcDNA3 were obtained from Life Technologies (Rockville, MD), and phRL-TK was obtained from Promega Corp. (Madison, WI).

Flow cytometric analysis
Cell surface antigens were stained in the dark at 4uC with antibodies diluted in PBS containing 0.5% BSA (FACS buffer). To study intracellular cytokine production, cells were stimulated with 250 nM ionomycin (BD Bioscience, San Jose, CA) and 50 nM phorbol 12-myristate 13-acetate (Sigma-Aldrich) in the presence of GolgiStop (BD Bioscience) for 4 h at 37uC in 5% CO 2 . Cells were fixed with 4% paraformaldehyde-PBS, permeabilized with 0.5% saponin in FACS buffer, and then stained in the dark at 4uC with the indicated antibodies [11]. Stained cells were analyzed with a Gallios TM flow cytometer (Beckman Coulter, Inc.; Brea, CA) and the data obtained were analyzed with FlowJo software (Tree Star, Inc.; Ashland, OR).

Enzyme-linked immunosorbent assays (ELISAs)
ELISA kits for IL-17A and IFN-c (eBioscience) were used to quantify the respective cytokines in culture supernatants, according to the manufacturer's protocols.

Chromatin immunoprecipitation (ChIP)
Cultured CD4 + T cells were fixed in 1% formaldehyde, exposed to 0.2 M glycine to halt the fixation process, and washed in icecold PBS containing 0.5% BSA. Subsequently, cells were lysed by sonication in SDS lysis buffer containing 1% (wt/vol) SDS, 10 mM EDTA, and 50 mM Tris (pH 8.0). Cellular debris was removed by centrifugation. A ChIP assay was performed using antibodies against acetyl-histone H3 (Lys27) and normal rabbit IgG (Cell Signaling Technology; Danvers, MA) and Dynabeads Protein G (Life Technologies). Immunoprecipitated and input DNA was then analyzed by quantitative PCR using SYBR Premix EX Taq (Takara Bio). The sequences of primers used for quantitative PCR are as follows: 59-GCTGCTGTTTCCTTGA-GAGG-39 and 59-GCTGGATAAGGCAGGAACAG-39 for con-

Histology
Tissues were fixed by immersion in 10% formalin in phosphatebuffered saline and embedded in paraffin blocks. Three-micrometer-thick sections were stained with hematoxylin and eosin (HE staining) or luxol fast blue solution and cresyl violet solution (Klüver-Barrera staining), and then examined by light microscopy. Immunohistochemistry of galectin-3 was described previously [15]. Briefly, the paraffinized sections were antigen revealed by using a 0.01 M citrate buffer (pH 6.0) by the PascalR heatinduced target retrieval system (DAKO). Anti-galectin-3 antibody using at a dilution of 1:100 in 2% BSA/PBS were added on the slides and incubated overnight at 4uC. Anti-galectin-3 antibody was detected with a biotinylated anti-Rat IgG (1:200) for 30 min, followed by incubation with avidin-coupled peroxidase (Vectastain ABC kit, Vector Laboratories) for 30 min. The peroxidase binding sites were detected by staining with 3,39-diaminobenzidine (DAB) in 50 mM Tris-EDTA buffer, pH 7.6.

Luciferase Assays
HEK 293 cells (1610 5 cells) were transfected using the calcium phosphate-DNA coprecipitation method with IL-17A reporter and expression vectors (pcDNA3-RORct, IkB NS , and IkB-f) with pRL-TK-Luc. Twenty-four hours after transfection, the medium was changed and the cells were incubated for a further 24 h. Luciferase activities were measured using the Dual-Luciferase Reporter Assay System, according to the manufacturer's instructions (Promega Corp., Madison, WI). Data shown are the mean 6 SE of duplicate samples from a representative of at least 3 independent experiments.

Statistical Analysis
Paired data were evaluated using the Student's t test.

Nfkbid 2/2 mice maintain immune homeostasis
A previous study showed that IkB NS plays an important role in the generation of Tregs [6]. However, Nfkbid 2/2 mice and bumble mutant mice (harboring a stop codon after exon 4 of the Nfkbid gene) appear healthy and do not exhibit an inflammatory phenotype in the periphery [16]. IkB NS shares greatest homology with IkB-f (43% identity at the gene level). We confirmed that Nfkbid 2/2 mice appear healthy and can live for up to 6 months without disease manifestation. At 8-12 weeks old, the percentage of effector/memory and naïve CD4 + cells in the peripheral LNs and spleen were comparable in Nfkbid +/+ and Nfkbid 2/2 mice ( Fig. 1A and 1B). Although the overall percentages of IFN-c-and IL-17A-producing CD4 + cells in the spleen and peripheral LNs were low (Fig. 1C and 1D), we were still able to demonstrate that the percent of IFN-c-producing CD4 + cells in the peripheral LNs was lower in Nfkbid 2/2 mice compared to Nfkbid +/+ mice (Fig. 1D).
Nfkbid 2/2 mice resist EAE development Next, we generated classical Th17-dependent EAE models by immunizing mice with the MOG peptide [17,18]. Ten to twelve days after MOG immunization, Nfkbid +/+ mice developed EAE, starting with paralysis of the tail, followed by paralysis in the front and hind limbs between days 18 and 21 ( Fig. 2A). However, Nfkbid 2/2 mice showed significantly fewer clinical signs of diseases ( Fig. 2A). Further, Nfkbid 2/2 mice have fewer IL-17Aproducing CD4 + cells and reduced MOG antigen-specific IL-17A production in their draining LNs (Fig. 2B-D). We also confirmed that lymphocyte infiltration and demyelination occurred in the lumber section of spinal cords of Nfkbid +/+ EAE model mice, but not in those of Nfkbid 2/2 EAE models (Fig. 2E). Galectin-3, an activation maker of monocyte/macrophages/microglia [19], was observed as dark-brown staining in injured white matter of spinal cord in EAE model. We found that many galectin-3 positive cells in the lumber section of spinal cords of Nfkbid +/+ EAE model mice, but not Nfkbid 2/2 EAE model mice (Fig. 2E). In addition, adoptive transfer of CD4 + T cells from EAE models of Nfkbid 2/2 mice to naïve Nfkbid +/+ mice caused tail and hind limb paralysis, although these symptoms were less severe than those observed with adoptive transfer of CD4 + T cells from EAE models of Nfkbid +/+ mice (Fig. S1). Thus, T cells serve an intrinsic role in the resistance of EAE in Nfkbid 2/2 mice.

Reduced Th17 cell differentiation in Nfkbid 2/2 T cells
Next, we examined whether Nfkbid 2/2 T cells are capable of differentiating into Th17 cells. We found that Nfkbid 2/2 T cells show decreased expression of IL-17A ( Fig. 3A and 3B) and Th17related genes (Ccr6 and Il-17f; Fig. 3C and 3D) in response to TGF-b1 and IL-6 stimulation. In addition, detection of Acetylated histone H3 on Lysine 27 (H3K27Ac), a histone modification associated with open chromatin configurations, was reduced in the conserved non-coding sequence (CNS) 1, CNS 2, CNS 3, and Il-17f promoter regions in Nfkbid 2/2 T cells under Th17 conditions compared to that observed in Nfkbid +/+ T cells (Fig. 3E-F). These regions potentially regulate Il-17 gene expression [20]. Interestingly, the acetylation status of the CNS 1 region in Nfkbid 2/2 T cells did not change under Th17 conditions compared with Th0 conditions (Fig. 3F). A previous study indicated that the CNS 1 region plays an important role for both IL-17A and IL-17F expression [21]. Thus, Nfkbid 2/2 T cells were impaired in generating Th17 cells in response to TGF-b1 and IL-6 because of reduced histone H3 acetylation in the CNS 1 regions.

IkB NS does not control il17a gene expression
IkB-f, a homolog of IkB NS , can be upregulated in T cells in response to TGF-b1 and IL-6 stimulation, directly binds to CNS1, and positively regulates IL-17A expression in cooperation with RORct [22,23]. Here, we found that IkB NS expression was comparable under Th0 and Th17 conditions (Fig. 4A). In addition, IkB NS expression had no effect on IL-17A reporter activity, even in the presence of RORct (Fig. 4B). Thus, while IL-17A expression may be regulated by IkB-f, the homolog IkB NS does not transcriptionally control IL-17A.

Reduced RORct expression in Nfkbid 2/2 T cells
IkB NS can be induced upon TCR stimulation and can control NF-kB transcriptional activity [6,24]. In T cells, IkB NS deficiency leads to decreased production of IL-2 (a target of NF-kB) in response to TCR stimulation [5]. Thus, Nfkbid 2/2 T cells have a reduced ability to activate NF-kB in response to TCR stimulation. Rel (NF-kB subunit)-deficient T cells fail to generate Th17 cells since RORct induction is diminished in response to TGF-b1 and IL-6 stimulation [25].
To understand the molecular mechanism underlying the control of IL-17A gene expression by IkB NS , we examined the expression of RORct (a master regulator of Th17) and found that it was decreased in Nfkbid 2/2 T cells (Fig. 5). Thus, IkB NS controls NF-kB activation, which plays a pivotal role in RORct expression and Th17 cell differentiation.

Discussion
IkB NS , a member of the nuclear IkB family of proteins, is induced by TCR stimulation in thymocytes [1,26]. Interestingly, * * * * -+ -+ Nfkbid 2/2 mice show a high sensitivity to lipopolysaccharide (LPS)-induced endotoxin shock and 4,4-dimethyl-4-silapentane-1sulfonic acid-induced colitis because Nfkbid 2/2 DCs and macrophages produce large amounts of IL-6 in response to LPS stimulation [3]. It is well known that IL-6 positively regulates Th17 cell generation [27]. In addition, IkB NS may play a pivotal role in IL-10 production from regulatory DCs [28]. Thus, the intrinsic role of IkB NS in T cells may contribute to exacerbating Th17dependent EAE. Although IkB NS is important for Foxp3 + Tregs generation [6], Nfkbid 2/2 mice and bumble mutant mice appear healthy and do not exhibit an inflammatory phenotype in the periphery (Fig. 1) [16]. IkB NS may play a minor role in the immune suppression ability of Tregs.   IkB-f, a homolog of IkB NS , has a transcriptional activation domain and is important for Th17 cell differentiation [10]. IkB NS does not have a similar homologous transcriptional activation domain. In addition, IkB-f expression in T cells is dependent on TGF-b1 and IL-6 stimulation [10]. Thus, the regulation of IL-17A gene expression by IkB-f and IkB NS proceeds by different mechanisms.
Nfkbid 2/2 T cells show reduced proliferation and NF-kB activation in response to TCR [5]. However, a previous study showed that IL-17A promoter activity is dispensable for NF-kB activation [25]. Our ChIP data revealed that acetylation of histone H3 in the CNS 1 region does not change in Nfkbid 2/2 T cells in response to TGF-b1 + IL-6 (Fig. 3F). The CNS 1 region normally controls both IL-17A and IL-17F gene expression [21]. RORct, a master regulator of Th17, has the ability to bind both CNS 1 and the Il-17a promoter region, and it positively regulates IL-17A gene expression [29]. Therefore, our results indicate that Nfkbid 2/ 2 T cells showed impaired Th17 cells differentiation because of a reduction in RORct expression and histone H3 acetylation in the CNS 1 region. In conclusion, we show that IkB NS deficiency causes resistance to Th17-dependent autoimmune disease.

Supporting Information
Figure S1 Passive EAE model using adoptive T cell transfer. Collected draining LNs from the Nfkbiz +/+ and Nfkbiz 2/2 mice at day 12 after MOG immunizations. LN cells were re-stimulated by MOG (10 ng/ml) after 3 days in culture, and CD4 + T cells were isolated using the CD4 + CD25 + Regulatory T cell Isolation Kit (Miltenyi Biotec). Nfkbiz +/+ mice (n = 3-4/ group) were intravenously injected (5 6 10 5 CD4 + T cells/mouse) and EAE symptoms were scored for up to 12 days. In addition, these mice received 500 ng pertussis toxin (Sigma) by i.p. injection to boost their immunological responses on Days 0 and 2. Data shown represent mean + S.E. Paired data were evaluated using the Student's t test. *p ,0.05. (EPS)