Epigenetic regulation of the expression of Il12 and Il23 and autoimmune inflammation by the deubiquitinase Trabid.

The proinflammatory cytokines interleukin 12 (IL-12) and IL-23 connect innate responses and adaptive immune responses and are also involved in autoimmune and inflammatory diseases. Here we describe an epigenetic mechanism for regulation of the genes encoding IL-12 (Il12a and Il12b; collectively called 'Il12' here) and IL-23 (Il23a and Il12b; collectively called 'Il23' here) involving the deubiquitinase Trabid. Deletion of Zranb1 (which encodes Trabid) in dendritic cells inhibited induction of the expression of Il12 and Il23 by Toll-like receptors (TLRs), which impaired the differentiation of inflammatory T cells and protected mice from autoimmune inflammation. Trabid facilitated TLR-induced histone modifications at the promoters of Il12 and Il23, which involved deubiqutination and stabilization of the histone demethylase Jmjd2d. Our findings highlight an epigenetic mechanism for the regulation of Il12 and Il23 and establish Trabid as an innate immunological regulator of inflammatory T cell responses.

Innate immune cells, including dendritic cells (DCs) and macrophages, play an important role in regulating the nature and magnitude of adaptive immune responses[1]. They recognize microbial components via pattern-recognition receptors, including various toll-like receptors (TLRs), which trigger intracellular signaling events that induce the maturation and function of these cells. DCs function as the primary antigen-presenting cells required for activation of naïve T cells[2]. In addition, DCs and macrophages produce a plethora of cytokines that regulate CD4+ T cell differentiation and promote inflammatory responses[3]. However, deregulated production of proinflammatory cytokines by innate immune cells also contributes to autoimmune and inflammatory diseases. The interleukin 12 (IL-12) family cytokines, particularly IL-12 and IL-23, are proinflammatory cytokines produced by dendritic cells, macrophages, as well as fibroblasts in response to microbial infections[4, 5, 6]. IL-12 is composed of the IL-12α and IL-12β subunits and acts as a key mediator of CD4+ T cell differentiation towards T helper (TH) 1 lineage characterized by production of a signature cytokine, interferon-γ (IFN-γ)[7]. IL-23, composed of IL-12β and a specific subunit, IL-23α, functions to amplify and maintain TH17 subset of CD4+ T cells[6, 7]. IL-12 and IL-23 are induced in response to stimulation by multiple TLR ligands and have been linked to autoimmune and inflammatory diseases[6]. However, the mechanism regulating the induction of Il12 and Il23 gene induction is incompletely understood.

The NF-κB transcriptional regulator c-Rel mediates TLR-stimulated expression of Il12 and Il23 genes[8, 9]. In addition to activating NF-κB and other transcription factors, TLR signals induce chromatin remodeling at the Il12b promoter, which may be important for the accessibility of the promoter region by specific transcription factors[10, 11]. As seen with the Il12b gene, chromatin remodeling allows the accessibility of the Il12a promoter by specific transcription factors, such as c-Rel and C/EBP[12]. Similarly, TLR stimulation results in histone modifications of the nucleosome located at the Il23a promoter[13, 14]. Each nucleosome contains the core histones H2A, H2B, H3 and H4, which are characteristically regulated by post-translational modifications including methylation and demethylation[15]. Recent work has indicated Jmjd2d as a demethylase that mediates histone 3 demethylation involved in Il12b induction in DCs[15, 16]. However, how Jmjd2d is regulated remains unclear.

Here we identified the deubiquitinase (DUB) Trabid (TRAF-binding protein domain, also known as Zranb1), as a crucial regulator of TLR-stimulated expression of IL-12 and IL-23. Trabid belongs to the OTU family of DUBs and preferentially hydrolyzes lysine 29 (K29)- and K33-linked ubiquitin chains [17, 18, 19]. In vitro studies using cancer cell lines suggest a role for Trabid in the regulation of Wnt signaling, but this function remains controversial[20, 21]. By employing a gene targeting approach, we show that Trabid deficiency in DCs and macrophages impaired the induction of Il12 and Il23 genes without affecting the induction of many other cytokine genes. Consistently, Trabid deficiency impaired the production of TH1 and TH17 subsets of inflammatory T cells, rendering mice refractory to the induction of experimental autoimmune encephalomyelitis (EAE), an autoimmune neuroinflammatory disease that is dependent on TH1 and TH17 cells. Our data suggest the involvement of an epigenetic mechanism, in which Trabid regulates histone modifications at the Il12 promoter by controlling the fate of a histone demethylase, Jmjd2d.

RESULTS

Trabid is required for induction of EAE

To study the in vivo function of Trabid, we generated germline Zranb1 knockout (called KO here throughout) mice and wild-type control mice by crossing Zranb1fl/fl mice with CMV-Cre mice (). In addition, we crossed the Zranb1fl/fl mice with CD11c-Cre and Cd4-Cre mice to generate DC-conditional Zranb1 KO (Zranb1f/fCd11c-Cre; called DC-cKO here throughout) and T-cell conditional Zranb1 KO (Zranb1f/fCd4-Cre; called T-cKO here throughout) mice, respectively (). The RT-PCR analyses revealed loss of Zranb1 mRNA expression in T cells, B cells, DCs, and macrophages of the germline KO mice and in DCs and T cells of the DC-cKO and T-cKO mice, respectively (). The germline KO mice were born with expected Mendelian ratio, had normal growth and survival (data not shown) and did not show obvious abnormalities in thymocyte development, although they had a moderate reduction in the frequency of naïve T cells in the spleen (). The percentage of regulatory T (Treg) cells among CD4+ single-positive thymocytes and CD4+ splenic T cells was comparable between wild-type and KO mice (). Additionally, deletion of Trabid had little or no effect on the frequency of conventional DCs or plasmacytoid DCs in the bone marrow and spleen ().

To investigate the function of Trabid in regulating immune responses, we employed a T cell-dependent autoimmunity model, EAE, which involves peripheral generation of central nervous system (CNS)-specific TH1 and TH17 subsets of inflammatory T cells and their subsequent migration to the CNS to induce inflammation and demyelination[22,23]. Wild-type mice immunized with the myelin oligodendrocyte glycoprotein (MOG) peptide MOG35–55, along with pertussis toxin developed severe clinical symptoms (), associated with profound immune cell infiltration and demyelination in the CNS (). Compared to wild-type mice, KO mice displayed significantly delayed onset and reduced severity of EAE disease, as well as substantially less immune cell infiltration and demyelination in the CNS (). Flow cytometry analyses revealed fewer CD4+ and CD8+ T cells and CD11b+CD45hi monocytes, both as frequency and absolute number in the CNS of KO mice compared to wild-type mice (), along with increased frequency of CD11b+CD45lo microglia () during the effector phase of EAE. Consistent with reduced inflammatory cell infiltration, we detected reduced expression of the proinflammatory cytokine genes Il6, Tnf, Il12a, Il12b, and Il23a in the CNS of MOG35–55-immunized KO mice compared to MOG35–55-immunized wild-type mice (), further suggesting attenuated induction of inflammation. In addition, the percentage of IL-17+ TH17 cells and IFN-γ+ TH1 cells within the CD4+ T cells infiltrating the CNS was significantly reduced in the KO mice compared to wild-type (). Compared to EAE-induced wild-type mice, EAE-induced KO mice also had a significantly lower frequency of TH1 and TH17 cells in the draining lymph nodes (), and spleen (), as determined by intracellular cytokine staining () and in vitro recall responses stimulated by the MOG peptide (). These results suggest impaired generation of inflammatory T cells upon autoantigen immunization, highlighting a crucial role for Trabid in mediating in vivo TH1 and TH17 cell differentiation.

Trabid functions in DCs to regulate T cell differentiation

To determine if Trabid functions in T cells in a cell-intrinsic manner or regulates the CD4+ T cell responses through an effect on innate immune cells, particularly DCs, we induced EAE in T-cKO and DC-cKO mice. We found comparable EAE disease scores and CNS immune cell infiltration between wild-type and T-cKO mice immunized with MOG peptide ( and ). In addition, EAE-induced T-cKO mice had no defects in the in vivo production of TH1 and TH17 cells or in vitro T cell recall responses to MOG peptide compared to wild-type mice (). Consistently, the in vitro differentiation efficiency of naïve T-cKO and wild-type CD4+ T cells into TH1, TH17, and Treg cells was comparable (). Wild-type and Trabid-deficient naïve T cells also had similar capacity of proliferation and IL-2 and IFN-γ expression in response to in vitro TCR/CD28 stimulation (). These observations indicate a non-cell autonomous effect of Trabid in regulating T cell differentiation and EAE pathogenesis.

DCs play an important role in regulating the activation and differentiation of naïve CD4+ T cells[24]. Similar to KO mice, DC-cKO mice were refractory to EAE induction, which was associated with a profound reduction in the frequency and absolute number of CNS-infiltrating CD4+ T cells and CD11b+CD45hi myeloid cells () and a higher relative frequency of CNS-resident CD11b+CD45lo microglia (), compared to wild-type mice. Intracellular cytokine staining revealed lower frequencies and absolute numbers of TH1 and TH17 cells, and lower numbers of Treg cells, in the CNS of DC-cKO mice compared to wild-type mice following induction of EAE (), corresponding to the overall reduction in immune cell infiltration into the CNS in these mice (). We detected reduced frequency of TH1 and TH17 cells in the draining lymph nodes of DC-cKO mice () and reduced TH1 and TH17 cytokine recall responses in splenic DC-cKO T cells () compared to splenic wild-type T cells, suggesting impaired production of these inflammatory T cells. These data suggest that expression of Trabid in DCs regulated the generation of TH1 and TH17 cells and EAE pathogenesis.

MOG-immunized DC-cKO and wild-type mice had a similar frequency of CD11c+CD11b+ DCs in the draining lymph nodes, which expressed comparable levels of the costimulatory molecules CD80 and CD86 (), suggesting Trabid deletion did not affect DC development, migration or maturation. However, MOG-immunized DC-cKO mice had reduced expression of the inflammatory cytokine genes Il6, Tnf, Il12a, Il12b, and Il23a in the CNS () and reduced expression of Il12a, Il12b and Il23a, although not Il6 and Tnf, in the draining lymph nodes () compared to MOG35–55-immunized wild-type mice. Together, these observations suggested that Trabid might regulate TH1 and TH17 cell responses through IL-12 and IL-23 production in DCs.

Trabid regulates IL-12 and IL-23 expression downstream of TLRs

We next analyzed the effect of Trabid deficiency on TLR-stimulated DC activation and cytokine expression in vitro. Wild-type and DC-cKO bone marrow-derived DCs (BMDCs) expressed a similar level of the costimulatory molecules CD80 and CD86 following in vitro stimulation with the TLR4 ligand lipopolysaccharide (LPS; ) and displayed similar capacity to mediate antigen-stimulated T cell proliferation, as determined based on CFSE dilution in co-cultured OT-II T cells (). However, culture medium from LPS-stimulated DC-cKO BMDCs induced 2-3 fold lower differentiation of naïve CD4+ T cells into TH1 and TH17 cells compared to conditioned medium from LPS-stimulated wild-type BMDCs (). On the other hand, DC-cKO BMDCs induced Treg cell differentiation similarly to wild-type BMDCs, either in the absence or presence of exogenous TGF-β ().

Real-time quantitative RT-PCR (qRT-PCR) detected significantly reduced expression of Il12a, Il12b, and Il23a mRNA, although similar expression of Ifnb, Il1b, IL6, Tnf, Il10, and Nos2 mRNA, in DC-cKO BMDCs stimulated with LPS () or the TLR3 ligand poly(I:C) () compared to LPS- and poly(I:C)-induced wild-type BMDCs. Similar results were obtained by ELISA to assess the amount of secreted cytokines () and by qRT-PCR using FLT3 ligand-differentiated BMDCs stimulated with the TLR9 ligand CpG (). Furthermore, exogenous IL-12 and IL-23 could rescue the defect of DC-cKO DCs in inducing TH1 and TH17 cell differentiation compared to wild-type DCs (). We also performed studies using bone marrow-derived macrophages (BMDMs) prepared from M-cKO mice and mouse embryonic fibroblasts (MEFs) prepared from Zranb1 germline KO mice. Compared to wild-type control cells, the M-cKO BMDMs and KO MEFs had reduced mRNA expression of Il12 genes when stimulated with LPS () and lipofectamine-transfected poly(I:C) (), respectively. Collectively, these results suggest that Trabid mediates induction of IL-12 and IL-23, thereby promoting TH1 and TH17 differentiation.

Trabid is required for recruitment of c-Rel to the Il12 promoter

To elucidate the molecular mechanism by which Trabid mediates induction of IL-12 family of cytokines, we examined the role of Trabid in regulating TLR-mediated activation of MAP kinases (MAPKs), IκB kinase (IKK) and downstream transcription factors known to regulate Il12 and Il23 gene induction[25, 26, 27]. Compared to wild-type BMDCs, DC-cKO BMDCs did not show appreciable defect in LPS-stimulated phosphorylation of IKK and its substrates IκBα and p105 () or LPS-stimulated activation of the MAPKs p38, extracellular signal-regulated kinase (ERK), and c-Jun N-terminal kinase (JNK) (). Furthermore, we observed similar levels of nuclear translocation of the NF-κB transcription factors c-Rel, p50, and p65, as well as CEBPβ and STAT3 in LPS-stimulated DC-cKO and wild-type BMDCs ().

Because c-Rel has a very important role in the transcription of IL-12 cytokines [8, 9, 28], we used chromatin immunoprecipitation (ChIP) to examine whether Trabid regulated the TLR-stimulated recruitment of c-Rel to Il12 gene promoters in DCs. Stimulation of wild-type BMDCs with LPS induced the binding of c-Rel in the promoter of Il12b gene, particularly between −220 and −16, a region known to contain an NF-κB-binding site (), while c-Rel recruitment to this region was barely observed in DC-cKO BMDCs (). Similar observations were made in BMDCs stimulated with CpG ().

We also examined the recruitment of c-Rel to several other cytokine gene promoters that contain κB sites. LPS strongly induced the binding of c-Rel to the promoter of Il12a and Il12b, but not to that of Ifnb, in wild-type BMDCs, and the LPS-stimulated c-Rel recruitment to the Il12a and Il12b promoters was significantly reduced in DC-cKO BMDCs ( and ). We found a low level of LPS-induced c-Rel binding to the Il1b promoter; however this occurred in both wild-type and DC-cKO BMDCs (). Another NF-κB member, p50, was recruited to both the Il12 promoters and Ifnb promoter in LPS-stimulated wild-type BMDCs, and the recruitment of p50 to the Il12 promoters, but not to the Ifnb promoter, was attenuated in DC-cKO BMDCs (). Thus, although Trabid was dispensable for the nuclear translocation of NF-κB factors, it was required for recruiting c-Rel and p50 to Il12 the promoters.

Trabid regulates histone modifications at the Il12b promoter

When transiently expressed in HEK293 cells, Trabid did not physically interact with c-Rel or inhibit c-Rel ubiquitination (data not shown), and it did not alter the transcriptional activity of c-Rel in a luciferase reporter assay (). We next investigated a possible epigenetic function for Trabid. Gene transcription and silencing are associated with active histone modifications, such as histone 3 lysine 4 trimethylation (H3K4me3), and repressive histone modifications, such as H3K9 dimethylation (H3K9me2) and trimethylation (H3K9me3), respectively[29,30, 31]. We found that LPS induced H3K4me3, downregulated H3K9me2 and H3K9me3, and did not appreciably alter H3K27me3, in the Il12b promoter of wild-type BMDCs (). Compared to wild-type BMDCs, the DC-cKO BMDCs had attenuated induction of H3K4me3 and significantly inhibited removal of the repressive histone marks, H3K9me2 and H3K9me3, in both the Il12b promoter () and the Il12a promoter (). Of note, Rel KO BMDCs did not show defect in the LPS-induced removal of H3K9me2 and H3K9me3 at the Il12b promoter compared to wild-type BMDCs (), suggesting that these Trabid-dependent epigenetic changes were independent of c-Rel recruitment.

To assess the functional role of Trabid-mediated histone modifications at the Il12 promoters, we determined the binding level of RNA polymerase II (Pol II) preinitiation complex components, including Pol II, serine-5 phosphorylated Pol II (Pol II pS5) and the general transcription factor subunits TFIIB and TFIID, in the Il12/Il23 promoters. Compared to the wild-type BMDCs, the DC-cKO BMDCs had attenuated recruitment of these Pol II complex components to the TATA boxes of Il12a, Il12b, and Il23a promoters in response to LPS stimulation (). In contrast, the wild-type and DC-cKO BMDCs did not show differences in LPS-stimulated recruitment of the Pol II complex components to the Il1b and Il6 promoters (Supplementary Fig. 6). Collectively, these results suggest a role for Trabid in regulating TLR-induced histone modifications and assembly of RNA Pol II complex at the promoter of Il12 genes.

The DUB activity of Trabid is required for its function

To determine whether the DUB catalytic activity of Trabid was required for Il12 promoter regulation, we reconstituted Zranb1 KO MEFs with a retroviral vector encoding wild-type Trabid or a catalytically inactive Trabid mutant (C443A). C443A Trabid showed similar, or even more abundant, nuclear translocation as wild-type Trabid (). However, wild-type Trabid, but not C443A Trabid, rescued the poly(I:C)+lipofectamine-induced removal of H3K9me2 and H3K9me3 () as well as the recruitment of c-Rel () at the Il12b promoter in KO MEFs. Consistently, reconstitution of KO MEFs with wild-type Trabid, but not C443A Trabid, greatly promoted poly(I:C)-induced Il12b, Il12a and Il23a mRNA expression in these cells () but had no effect on the induction of the control gene Il6 (). In similar experiments using BMDCs, reconstitution of DC-cKO BMDCs with wild-type Trabid, but not C443A Trabid, rescued the induction of Il12a, Il12b, and Il23a genes (). Thus, the DUB catalytic activity of Trabid is essential for its function in the regulation of histone modifications and Il12 gene induction.

Trabid regulates the demethylase Jmjd2d

We surmised that Trabid might regulate the ubiquitination of a factor involved in regulation of histone modification at the Il12 promoter. In this regard, H2B monoubiquitination at K120 is known to promote H3 methylation[26]. However, although LPS upregulated H2B K120 monoubiquitination in wild-type BMDCs, a similar level of H2B K120 monoubiquitination was detected in DC-cKO cells (). Because Jmjd2 demethylases have been shown to control removal of H3K9me2 and H3K9me3 histone marks[32], we screened all four Jmjd2 proteins using ChIP assays and found that LPS stimulation of wild-type BMDCs induced the binding of Jmjd2d, but not the other Jmjd2 members, to the Il12b promoter (). Furthermore, LPS-stimulated recruitment of Jmjd2d to the Il12a and Il12b promoters was largely lost in DC-cKO BMDCs compared to wild-type BMDCs (). Jmjd2d was only weakly associated with the Il1b and Il6 promoters in LPS-stimulated wild-type BMDCs, and this association was not inhibited in DC-cKO BMDCs (). As seen with the inducible removal of H3K9me2/me3, the LPS-stimulated Jmjd2d recruitment to Il12b promoter was not affected by c-Rel deficiency (), consistent with the idea that chromotin modifications precede c-Rel recruitment.

We next examined the functional contribution of Jmjd2d to the induction of Il12 genes. Incubation of wild-type BMDCs with a selective Jmjd2 inhibitor, 5-carboxy-8HQ, potently inhibited the LPS-stimulated expression of Il12 and Il23 genes, but not that of Il1b and Il6 genes (). Furthermore, 5-carboxy-8HQ had little or no effect on the induction Il12 and Il23 in DC-cKO BMDCs, consistent with the impaired promoter-recruitment of Jmjd2d in these mutant cells. Jmjd2d knockdown with two different shRNAs also led to significant inhibition of Il12 and Il23 gene induction in wild-type BMDCs ().

We next investigated the role of Trabid in regulating the ubiquitination and stability of Jmjd2d. The steady-state expression of Jmjd2d in wild-type BMDCs was low, but it was elevated upon LPS stimulation (). However, the amount of Jmjd2d protein was only slightly increased following LPS stimulation in DC-cKO BMDCs (). This phenotype was not due to altered expression of Kdm4d gene (encoding Jmjd2d), because LPS induced comparable amounts of Kdm4d mRNA in wild-type and DC-cKO BMDCs (), indicating a posttranslational mechanism of Jmjd2d protein regulation. Indeed, treatment of LPS-stimulated DC-cKO BMDCs with the proteasome inhibitor MG132 restored the amount of Jmjd2d protein to a level similar to that in LPS-stimulated wild-type BMDCs (), suggesting that Trabid might stabilize Jmjd2d via deubiquitination. In support of this idea, wild-type Trabid, but not the catalytically inactive C443A Trabid mutant, stabilized Jmjd2d in poly(I:C)+lipofectamine-stimulated Zranb1 KO MEFs (). Furthermore, both wild-type and C443A Trabid physically associated with endogenous Jmjd2d in reconstituted KO MEFs ( and ).

To directly examine the role of Trabid in regulating Jmjd2d ubiquitination, we analyzed the accumulation of ubiquitinated Jmjd2d in wild-type and DC-cKO BMDCs stimulated with LPS in the presence of the proteasome inhibitor MG132. DC-cKO BMDCs had markedly more abundant Jmjd2d ubiquitination than the wild-type BMDCs (), suggesting that Trabid negatively regulated Jmjd2d ubiquitination. When transfected into HEK293 cells, wild-type Trabid, but not C443A Trabid, inhibited Jmjd2d ubiquitination (), and the transfected Jmjd2d was abundantly conjugated with both K11- and K29-linked polyubiquitin chains (). Consistent with prior in vitro findings that Trabid preferentially cleaves K29-linked ubiquitin chains[18, 19], transfected Trabid efficiently reduced the K29 ubiquitination of Jmjd2d (), although Trabid also had an inhibitory effect on K11 ubiquitination of Jmjd2d, albeit more weakly than on K29 ubiquitination (). Retroviral transduction of Jmjd2d in DC-cKO BMDCs increased the amount of Jmjd2d protein level in these cells () and significantly increased the expression of Il12a, Il12b, and Il23a genes (). Collectively, these data suggest that in TLR-stimulated DCs Trabid deubiquitinates and stabilizes Jmjd2d, thereby regulating histone modifications and expression of Il12a, Il12b, and Il23a genes ().

Discussion

The IL-12 family of cytokines plays a crucial role in the polarization of CD4+ T cells to TH1 and TH17 cells and the pathogenesis of EAE[33]. In the present study, we identified the DUB Trabid as an essential regulator of Il12 gene expression in both innate immune cells and fibroblasts. Trabid deficiency impaired TLR-stimulated expression of IL-12 and IL-23 in vitro and suppressed T cell polarization and EAE pathogenesis in vivo. Mechanistically, Trabid modulated the TLR-induced histone modifications at the promoter region of Il12 genes.

Epigenetic factors play a crucial role in the regulation of Il12 genes[34, 35, 36]. In particular, the histone demethylases Aof1 and Jmjd2d have been implicated in the removal of H3K9me2 and H3K9me3 from the Il12b promoter, thereby promoting stimulus-induced recruitment of transcription factors like c-Rel[16, 37]. Our data suggest a ubiquitin-dependent mechanism of Jmjd2d regulation. Along with its inducible synthesis in TLR-stimulated DCs, Jmjd2d underwent ubiquitination and proteolysis, which was counter-regulated by the deubiquitinase Trabid. Trabid deficiency promoted Jmjd2d degradation and inhibited the removal of the transcriptionally repressive histone marks, H3K9me2 and H3K9me3 at the promoter region of Il12 genes. Trabid deletion also inhibited the induction of the transcriptionally active histone mark H3K4me3 at the promoter region of Il12 genes. Previous studies have shown that H3K9 and H3K4 methylations are mutually regulated, with H3K9 methylation inhibiting H3K4 methylation and vice versa[38]. It is thus possible that Trabid may regulate both types of histone methylation events via the same mechanism, although the involvement of two independent mechanisms cannot be excluded.

Our data suggest that Trabid is dispensable for TLR-stimulated c-Rel activation but is required for c-Rel recruitment to the promoter of Il12 genes. It is likely that the promoter recruitment of c-Rel may rely on histone modifications, because these epigenetic events are also dependent on Trabid. In further support of this possibility, c-Rel was not required for TLR-stimulated Jmjd2d recruitment to the Il12 promoter or the histone modification events. We noticed that Trabid regulated H3K9 di- and tri-methylations in a broad region of the Il12b locus, whereas c-Rel was specifically recruited to the promoter-proximal region of Il12b known to contain a κB enhancer. Thus, although histone modifications might promote c-Rel recruitment to the Il12 promoter, the specificity of this latter event is likely guided by the κB enhancer. The molecular mechanism underlying the inducible recruitment of Jmjd2d to specific promoters is currently unclear.

A prior in vitro study using cancer cell lines suggests a role for Trabid in regulating TCF-mediated transcription in Wnt signaling pathway[20], although this function of Trabid has remained controversial [21]. Trabid has also been shown to interact with and inhibits K63 ubiquitination of Drosophila TAK1, reducing the expression of downstream genes encoding antimicrobial peptides in flies[42]. TAK1 is a master kinase that mediates activation of IKK and its downstream transcription factor NF-κB, as well as the activation of MAPKs[43]. We found that Trabid was dispensable for TLR-stimulated activation of IKK-NF-κB and MAPKs in mouse DCs. Consistently, Trabid deficiency did not affect the induction of various NF-κB-target genes, such as Tnfa, Il1b and Il6 but led to specific defect in the induction of Il12 genes. Such distinct activity by Trabid may be due to differences between insect and mammalian systems.

The connection between Trabid and Jmjd2d raises the question of whether Trabid regulates additional genes. While this question needs to be addressed in future studies, our current work revealed that Trabid was dispensable for the induction of several other genes, including Ifnb, Tnf, Il1b, Il6, Il10 and Nos2. Our data further suggested that Trabid-mediated IL-12 and IL-23 induction plays an important role in mediating inflammatory T cell responses and the pathogenesis of EAE, an animal model of the autoimmune neuroinflammatory disease multiple sclerosis. These findings implicate Trabid and Jmjd2d as potential therapeutic targets for the treatment of inflammatory diseases, such as multiple sclerosis.

ONLINE METHODS

Mice

Zranb1-targeted mice(Zranb1tm1a(EUCOMM)Hmgu) were generated at Knockout Mouse Project (KOMP) by targeting exon 3 of Zranb1 gene using a FRT-LoxP vector (Supplementary Fig. 1a). Zranb1-floxed mice (in C57BL/6 × 129/Sv mixed background) were generated by crossing the Zranb1-targeted mice with FLP deleter mice (Rosa26-FLPe; Jackson Laboratory). The Zranb1-floxed mice were further crossed with CMV-Cre, Cd4-Cre, Cd11c-Cre, and Lyz2-Cre mice (all from Jackson Laboratory, C57BL/6 background) to generate Zranb1 germline KO, T-cKO (Zranb1f/fCd4-Cre), DC-cKO (Zranb1f/fCd11c-Cre), and M-cKO (Trabidf/fLyz2-Cre) mice, respectively. Heterozygous mice were bred to generate littermate controls and KO (or conditional KO) mice for experiments. Outcomes of animal experiments were collected blindly and recorded based on ear-tag numbers of the experimental mice. Genotyping was performed as indicated in Supplementary Fig. 1. Mice were maintained in specific pathogen-free facility, and all animal experiments were conducted in accordance with protocols approved by the Institutional Animal Care and Use Committee of the University of Texas MD Anderson Cancer Center.

Antibodies, plasmids, and reagents

Antibodies used in immunoblotting, immunoprecipitation, and ChIP assays are listed in Supplementary Table 3. Fluorescence-labeled antibodies are described in the section of flow cytometry and cell sorting. Human Trabid cDNA was amplified by PCR from the pHM6-HA-Trabid plasmid (provided by Dr. Mariann Bienz, Medical Research Council, UK) and inserted into the pCLXSN(GFP) retroviral vector[44]. The catalytically inactive Trabid mutant, HA-Trabid C443A, was created by site-directed mutagenesis. Mouse Jmjd2d expression vector (pcDNA-FLAG-mKdm4d) was obtained from Addgene, and the cDNA insert was transferred to the pCLXSN(GFP) vector to generate pCLXSN(GFP)-HA-Jmjd2d. pGIPZ lentiviral vectors encoding a non-silencing shRNA control and two different Jmjd2d shRNAs were obtained from Thermo Scientific.

LPS (derived from Escherichia coli strain 0127:B8) and CpG (2216) were from Sigma-Aldrich. Poly (I:C) was from Amersham, and recombinant murine GM-CSF was from Peprotech. The Jmjd2 inhibitor 5-carboxy-8HQ was purchased from Tocris Bioscience.

Induction and assessment of EAE

For active EAE induction, age- and sex-matched mice were immunized s.c. with MOG35-55 peptide (300μg) mixed in CFA (Sigma-Aldrich) containing 5 mg/ml heat-killed Mycobacterium tuberculosis H37Ra (Difco). Pertussis toxin (200 ng, List Biological Laboratories) in PBS was administered i.v. on days 0 and 2. Mice were examined daily and scored for disease severity using the standard scale: 0, no clinical signs; 1, limp tail; 2, paraparesis (weakness, incomplete paralysis of one or two hind limbs); 3, paraplegia (complete paralysis of two hind limbs); 4, paraplegia with forelimb weakness or paralysis; 5, moribund or death. After the onset of EAE, food and water were provided on the cage floor. For visualizing CNS immune cell infiltration and demyelination, spinal cords of the EAE-induced mice were collected on day 30 and subjected to hematoxylin and eosin (H&E) and luxol fast blue (LFB) staining, respectively. Mononuclear cells were prepared from the CNS (brain and spinal cord) of EAE-induced mice as described[45] and analyzed by flow cytometry.

Flow cytometry, cell sorting, and intracellular cytokine staining

Single-cell suspensions of Spleen, bone marrow or brains and spinal cords from MOG35–55-immunized mice were subjected to flow cytometry using FACSAria (BD Bioscience) and the following fluorescence-labeled antibodies from eBioscience: PB-conjugated anti-IFNγ, anti-CD4, and anti-CD11c; PE-conjugated anti-B220 and anti-IL-17A; APC-conjugated anti-CD11b and anti-CD62L; APC-CY7-conjugated anti-CD8; and FITC-conjugated anti-CD44 and anti-Foxp3.

For intracellular cytokine staining, the CNS-infiltrating T cells were stimulated for 4 h with PMA plus ionomycin in the presence of monensin and then subjected to intracellular IFN-γ, IL-17A staining and flow cytometry analysis.

In vitro CD4+ T cell differentiation

Purified naïve CD4+ T cells (CD44loCD62Lhi) were activated with plate-bound anti-CD3 and anti-CD28 under Th0 (5 μg/ml anti-IL-4 and 5 μg/ml anti–IFN-γ), TH1 (5 μg/ml anti–IL-4 and 10 ng/ml IL-12), TH17 (5 μg/ml anti-IL-4, 5 μg/ml anti-IFN-γ, 20 ng/ml IL-6, and 2.5 ng/ml TGF-β), or Treg cells (5 μg/ml anti-IL-4, 5 μg/ml anti-IFN-γ, and 1 ng/ml TGF-β) conditions. The concentration of anti-CD3 antibody was 5 μg/ml for Th0 and TH1 conditions and 1 μg/ml for TH17 and Treg conditions, whereas the concentration of anti-CD28 was 1 μg/ml for all conditions. After 4 d of activation, Treg cells were quantified by flow cytometry based on staining of Foxp3. For detection of TH1 and TH17 cells, the cells were restimulated for 4 h with PMA and ionomycin in the presence of a protein transport inhibitor, monensin, followed by intracellular staining of IFN-γ and IL-17, respectively.

Generation of BMDCs and BMDMs

Bone marrow cells isolated from the wild-type or DC-cKO mice were cultured in RPMI 1640 medium contained 10% FBS supplemented with GM-CSF (10 ng/ml) for 7 days. The differentiated BMDCs were stained with Pacific blue-conjugated anti-CD11c and isolated by FACS sorter. In some experiments, BMDCs were also generated using medium supplemented with FLT3 ligand (20 ng/ml). Other than stated, GM-CSF-differentiated BMDCs were used in the experiments. BMDMs were generated using MCSF-supplemented medium.

For retroviral infection of Trabid-deficient BMDCs, Retrovirus particles were produced by transfecting HEK293 cells (using calcium-phosphate method) with pCLXSN-(GFP)-HA-jmjd2d or pCLXSN-(GFP) control vector, along with the packaging vectors pCL-ampho and VSV-G. BMDCs, differentiated using GM-CSF-supplemented medium for 3 days, were infected with the recombinant retroviruses for 8 h and then continued cultured under GM-CSF-supplemented medium. After 5 days, the GFP positive cells were sorted by flow cytometry and used for experiments.

Functional assays of DCs in vitro

The function of DCs was measured based on their ability to mediate antigen-stimulated activation of the OVA-specific OTII cells. DCs isolated from wild-type or DC-cKO mice were incubated for 6 h with 5 mg/ml OVA after stimulated with LPS (100ng/ml) for 12h, washed for 3 times with fresh medium, and then mixed with naïve OTII CD4+ T cells (CD44loCD62Lhi) labeled with CFSE. After three days, flow cytometry was performed to measure the proliferation of OTII T cells based on CFSE dilution.

For functional examination of DC-secreted cytokines in regulating T cell differentiation, the wild-type and DC-cKO DCs were stimulated with LPS for 48 h, and the supernatants were collected for inducing T cells differentiation. Naïve CD4+ T cells (CD44loCD62Lhi) isolated from wild-type mice were stimulated for 4 days with plate-bound anti-CD3 (5μg/mL) and anti-CD28 (1μg/mL) plus the DC culture supernatants described above, followed by intracellular cytokine staining and flow cytometry to measure the frequency of IFN-γ-producing TH1 cells and IL-17-producing TH17 cells or the Foxp3+ Treg cells. In some experiments, exogenous IL-12 and IL-23 were provided. Cytokine induction in DCs was also measured by ELISA and qRT-PCR.

Jmjd2d knockdown

Lentiviral particles were produced by transfecting HEK293 cells (using calcium-phosphate method) with a pGIPZ lentiviral vector, encoding either a non-silencing shRNA or Jmjd2d-specific shRNAs, along with the packaging vectors psPAX2 and pMD2. BMDCs, differentiated using GM-CSF-supplemented medium for 5 days, were infected with the lentiviruses for 8 h. After 72 h, the infected cells were enriched by flow cytometric cell sorting (based on GFP expression) and subsequently used for experiments.

ELISA and qRT-PCR

Supernatants of in vitro cell cultures were analyzed by ELISA using a commercial assay system (eBioScience). For qRT-PCR, total RNA was isolated using TRI reagent (Molecular Research Center, Inc.) and subjected to cDNA synthesis using RNase H-reverse transcriptase (Invitrogen) and oligo (dT) primers. qRT-PCR was performed in triplicates, using iCycler Sequence Detection System (Bio-Rad) and iQTM SYBR Green Supermix (Bio-Rad). The expression of individual genes was calculated by a standard curve method and normalized to the expression of Actb. The gene-specific PCR primers (all for mouse genes) are shown in Supplementary Table 1.

Immunoblotting, immunoprecipitation, and Ubiquitination assays

Whole-cell lysates or subcellular extracts were prepared and subjected to immunoblotting and immunoprecipitation assays as described[44]. For ubiquitination assays, cells were pretreated with MG132 for 2 h and then lysed with a Nonidet P-40 lysis buffer (50 mM Tris-HCl, pH7.5, 120 mM NaCl, 1% Nonidet P-40, 1 mM EDTA, 1 mM DTT) containing 6 M urea and protease inhibitors. Jmjd2d was isolated by IP with antibodies for Jmjd2d (93694, Abcam) and the ubiquitinated Jmjd2d was detected by immunoblotting using an anti-ubiquitin (P4D1). For transfection models, FLAG-Jmjd2d was transfected into HEK293 cells along with HA-ubiquitin or HA-ubiquitin mutants along with other indicated expression vectors, and the transfected cells were pretreated with MG132 for 2 h and then lysed for Jmjd2d IP followed by detecting ubiquitinated Jmjd2d by IB using anti-ubiquitin.

ChIP assays

ChIP assays were performed with DCs (6 × 107) stimulated for 6 h with LPS or MEFs (2×107) stimulated for 9 h with poly(I:C) plus lipofectamine. The cells were fixed with 1% formaldehyde and sonicated as previously described[46]. Lysates (from 2×107 cells in 3 ml) were subjected to IP with the indicated antibodies, and the precipitated DNA was then purified by Qiaquick columns (Qiagen) and quantified by quantitative PCR (QPCR) using a pair of primers that amplify the target region of the indicated promoter (Supplementary Table 2). The precipitated DNA is presented as percentage of the total input DNA. For histone modification analyses, the DNA bound by modified histone 3 is presented as percentage of total histone 3-bound DNA. For all of the genes analyzed, the promoter region covering a well-defined κB site was selected for QPCR assays to analyze c-Rel recruitment and histone modifications. For examining recruitment of general transcription factor components, the TATA box region of the promoters was selected for ChIP assay QPCRs.

Luciferase reporter gene assays

HEK293 cells (2 × 105) were transfected, by calcium phosphate precipitation, with a firefly luciferase reporter driven by the Il12b promoter, along with the indicated cDNA expression vectors as well as the control pRL-TK promoter-driven Renilla luciferase reporter. At 36 h post-transfection, cells were collected for dual luciferase assays (Promega). Firefly luciferase activities were normalized based on the Renilla luciferase activities.

Statistical analysis

Statistical analysis was performed using Prism software. Kruskal-Wallis nonparametric test followed by Dunn’s post analysis were performed for analysis of EAE scores and two-tailed unpaired t-tests were performed for all other data analyses. P values less than 0.05 were considered significant, and the level of significance was indicated as *P<0.05, **P<0.01. In the animal studies, 4 mice are required for each group based on the calculation to achieve a 2.3 fold change (effect size) in two-tailed t-test with 90% power and a significance level of 5%. All statistical tests are justified as appropriate, and data meet the assumptions of the tests. The variance is similar between the groups being statistically compared.