GLK-IKKβ signaling induces dimerization and translocation of the AhR-RORγt complex in IL-17A induction and autoimmune disease

The kinase GLK induces dimerization of AhR-RORγt, leading to induction of the cytokine IL-17A and autoimmune responses.


INTRODUCTION
Autoimmune diseases, which are chronic, debilitating, and lifethreatening, arise from immune system overactivation. T helper 17 cells [T H 17, interleukin-17A (IL-17A)-producing CD4 + T cells] and IL-17A play critical roles in the pathogenesis of autoimmune diseases (1,2). Induction of IL-17A occurs in the synovial fluids from rheumatoid arthritis (RA) patients (3) and in the renal biopsies from systemic lupus erythematosus (SLE) patients (1). IL-17A facilitates differentiation of osteoclasts, which lead to arthritis (3). IL-17A also promotes B cell proliferation and class switch recombination, contributing to autoantibody production and autoimmune responses (4). Furthermore, IL-17A overexpression causes tissue damage by inducing infiltration of neutrophils and macrophages through multiple chemokines (1,2,5). IL-17A knockout (KO) or IL-17A blockade abolishes disease development in animal autoimmune models such as experimental autoimmune encephalomyelitis (EAE) and collagen-induced arthritis (CIA) (2). Thus, understanding the mechanism(s) controlling IL-17A transcription may lead to the discovery of novel therapeutic targets for autoimmune diseases.
MAP4K3 (also named GLK) is a mammalian Ste20-like serine/ threonine kinase that was first identified as an upstream activator for the c-Jun N-terminal kinase pathway (22,23). GLK plays an important role in T cell receptor (TCR) signaling by directly phosphorylating and activating PKC, leading to activation of IKK [inhibitor of nuclear factor B (NF-B) kinase ] and NF-B in T cells (24). Moreover, clinical samples from patients with autoimmune diseases, such as SLE, RA, or adult-onset Still's disease, show markedly increased GLK expression in T cells; the frequencies of GLK-expressing T cells are positively correlated with disease severity (24)(25)(26). Consistent with the correlation of high levels of GLK expression with autoimmune disease, GLK-deficient mice are resistant to T H 17-mediated EAE induction or CIA induction and display lower T H 17 responses (24,26). Besides T H 17, differentiation of T H 1 and T H 2 is also impaired in GLK-deficient T cells because of 1 IL-17A overproduction and subsequent autoimmune phenotypes in mice.

GLK induces IL-17A transcription by activating AhR and RORt
Next, we studied the mechanism of GLK-induced IL-17A in T cells. The levels of IL-23 receptor and phosphorylated STAT3 were not increased in T cells of Lck-GLK Tg mice ( fig. S4, A and B), suggesting that IL-17A overexpression is not due to enhancement of IL-23 signaling or IL-6/STAT3 signaling. Consistent with the IL-17A protein levels, mRNA levels of IL-17A were significantly increased in the purified T cells of Lck-GLK Tg mice compared to those of wild-type mice ( Fig. 2A). We studied whether IL-17A overexpression is due to transcriptional activation of the IL-17A promoter. IL-17A promoter activities in Jurkat T cells were enhanced by GLK overexpression but not by GLK kinase-dead (K45E) mutant (Fig. 2B). Next, we studied the bindings of individual IL-17A transcription factors to the IL-17A promoter (Fig. 2, C and D). ChIP analyses showed that bindings of AhR and RORt (−877) to the IL-17A promoter were induced in T cells of Lck-GLK Tg mice (Fig. 2D), whereas bindings of STAT3, IRF4, KLF4, and BATF to the IL-17A promoter were not enhanced (Fig. 2D). The binding of RORt to the −120 region of the IL-17A promoter was not significantly induced (Fig. 2D); others reported similar findings (27,28). Consistent with ChIP data, mutation of the AhR-binding element or the RORt-binding site (−877) abolished the GLK-enhanced IL-17A reporter activity, whereas mutation of the STAT3-binding site (Fig. 2E) or the RORt-binding site (−120) did not affect the GLK-induced IL-17A reporter activity ( fig. S4C). Notably, GLK overexpression induced AhR response element-driven reporter (XRE-Luc) activity (Fig. 2F), whereas RORt (−877) or STAT3 response element-driven reporter (RORt-Luc or SIE-Luc) activity was unaffected (Fig. 2F). These results suggest that GLK signaling induces IL-17A transcription by activating AhR and maybe RORt.

GLK controls IL-17A production and autoimmune responses through AhR
To further demonstrate the role of AhR in promoting IL-17A production in Lck-GLK Tg mice, we bred Lck-GLK Tg mice with T cell-specific AhR conditional KO (AhR cKO: AhR f/f ;CD4-Cre) mice. Serum IL-17A levels were drastically reduced in Lck-GLK/ AhR cKO (Lck-GLK;AhR f/f ;CD4-Cre) mice, whereas serum TNF- and IFN- levels were unaffected by AhR deficiency (Fig. 3A). Levels of ANA, anti-dsDNA antibody, and RF were also decreased in Lck-GLK/AhR cKO mice compared to those in Lck-GLK Tg mice ( fig. S5A). Histology staining showed that AhR KO abolished induction of nephritis and spleen abnormality in Lck-GLK Tg mice ( fig. S5B). AhR KO also suppressed infiltration of inflammatory immune cells in the liver of Lck-GLK Tg mice ( fig. S5B). The data indicate that AhR plays a critical role in the IL-17A overproduction and autoimmune responses in Lck-GLK Tg mice.

PKC phosphorylates AhR at Ser 36 and induces AhR nuclear translocation
Next, we studied the mechanism of GLK-induced AhR binding to the IL-17A promoter. The confocal images (using two different anti-AhR antibodies; Fig. 3B and fig. S5C) and subcellular fractionation experiments (Fig. 3C) showed that AhR nuclear translocation The serum levels of autoantibodies in 20-week-old Lck-GLK and Lck-GLK/IL-17A KO mice were determined by ELISAs. The levels are presented relative to the value from one of the Lck-GLK mice. n = 6 per group. (E) IL-17A expression was attenuated by GLK shRNA. Murine primary splenic T cells were transfected with green fluorescent protein (GFP)-human GLK shRNA and a control GFP vector. The transfected T cells were stimulated with anti-mouse CD3 antibodies for 3 hours and then determined by flow cytometry at day 3 after transfection. Data show the events of IL-17A-producing T cells (GFP-gated). WT, wild-type littermate controls; Lck-GLK, T cell-specific GLK Tg mice; Lck-GLK/IL-17A KO, Lck-GLK;IL-17A-deficient mice; ANA, antinuclear antibody; -double-stranded DNA (dsDNA), anti-dsDNA antibody; RF, rheumatoid factor; APC, allophycocyanin. Data shown are representative of three independent experiments. *P < 0.05, **P < 0.01 (two-tailed Student's t test).  was enhanced in T cells of Lck-GLK Tg mice. In addition, we examined whether GLK signaling induces AhR nuclear translocation by enhancing phosphorylation of AhR. There is only one commercial anti-phospho-AhR antibody that detects phospho-Ser 36 AhR; however, the role of Ser 36 phosphorylation in AhR nuclear translocation has not been demonstrated (29). Immunoblotting analyses using the anti-phospho-AhR antibody for AhR phosphorylation showed that AhR Ser 36 phosphorylation was enhanced in T cells of Lck-GLK Tg mice, as well as in anti-CD3-stimulated T cells ( Fig. 3D and fig. S5D). These data suggest that GLK overexpression (and TCR signaling) may induce AhR activity by enhancing AhR Ser 36 phosphorylation and nuclear translocation in T cells.

of 15
Next, we investigated which kinase is responsible for phosphorylation and nuclear translocation of AhR in T cells of GLK Tg mice. SGK1 (serum/glucocorticoid-regulated kinase 1) can stabilize T H 17 population (30). The basal or TCR-induced SGK1 activation was unchanged in Lck-GLK T cells ( fig. S5D), suggesting that SGK1 is not involved in GLK-induced AhR phosphorylation. GLK signaling in T cells induces kinase activities of PKC, IKK, and IKK (24). To determine which kinase phosphorylates AhR, we immunoprecipitated GLK, PKC, IKK, IKK, and AhR each from individually transfected human embryonic kidney (HEK) 293T cells and subjected them to in vitro kinase assays. The data showed that AhR Ser 36 phosphorylation was drastically induced by PKC in vitro (Fig. 3E). We confirmed the specificity of the antibody for AhR Ser 36 phosphorylation using a wild-type AhR or an S36A mutant transfectant by immunoblotting ( fig. S5E). Immunofluorescence and confocal imaging analyses showed that PKC overexpression enhanced AhR nuclear translocation in Jurkat T cells The interaction between PKC and AhR was induced in purified T cells of Lck-GLK Tg mice (Fig. 3F). Moreover, the proteinprotein interaction/ALPHA (amplified luminescent proximity homogeneous assay) technology assays showed an interaction (<200 nm) between PKC and AhR, but not between GLK and AhR ( fig. S6A). Furthermore, the fluorescence resonance energy transfer (FRET) analysis showed a direct interaction (1 to 10 nm) between PKC and AhR in PKC/AhR-cotransfected Jurkat T cells ( fig. S6B).
To study the subcellular localization and protein interaction (<40 nm) in vivo, we performed in situ PLA using probes corresponding to PKC and AhR. The PLA data showed strong signals in the cytoplasm of T cells from Lck-GLK Tg mice ( Fig. 3G and  fig. S6C). In situ PLA using probes corresponding to Myc and Flag tags showed similar results in Myc-PKC-and Flag-AhRoverexpressing HEK293T cells ( fig. S6D). In vitro binding assays with purified PKC and AhR proteins further confirmed this direct interaction ( fig. S6E). Furthermore, purified PKC phosphorylated AhR at Ser 36 , whereas purified kinase-dead (K409W) mutant of PKC did not (Fig. 3H). These data demonstrate that PKC directly interacts with and phosphorylates AhR at Ser 36 , leading to AhR nuclear translocation.
To verify the role of PKC in AhR nuclear translocation and its in vivo function, we generated PKC KO mice using transcription activator-like effector nuclease (TALEN) technology ( fig. S7, A to C). We then bred Lck-GLK Tg mice with PKC KO mice to generate Lck-GLK;PKC −/− mice. As expected, PKC KO abolished GLK-induced AhR Ser 36 phosphorylation in T cells (Fig. 3I). Immunofluorescence and confocal imaging analyses showed that AhR was detected abundantly in the nucleus of T cells from Lck-GLK Tg mice (Fig. 3J). In contrast, AhR expression was detected in the cytoplasm, but not in the nucleus, of T cells from wild-type and Lck-GLK Tg/PKC KO (Lck-GLK;PKC −/− ) mice (Fig. 3J). The serum levels of IL-17A and autoantibodies were significantly decreased in Lck-GLK;PKC −/− mice compared to those in Lck-GLK Tg mice (fig. S7, D and E). Moreover, the inflammatory phenotypes were abolished in Lck-GLK Tg/PKC KO mice ( fig. S7F). Together, these results indicate that GLK induces IL-17A production by activating PKC-AhR signaling in T cells.

AhR interacts with RORt and transports RORt into the nucleus
Paradoxically, both AhR-and RORt-binding elements are required for the GLK-induced IL-17A reporter activity (Fig. 2E); however, GLK induced the activity of AhR, but not RORt, response element ( Fig. 2F). We suspected that AhR may facilitate induction of RORt activity. We first studied whether GLK induces RORt binding to the IL-17A promoter through AhR. ChIP data showed that the GLK-induced RORt binding to the IL-17A promoter was abolished in AhR KO T cells (Fig. 4A). The data suggest that AhR facilitates the binding of RORt to the IL-17A promoter. Next, we studied whether AhR interacts with RORt. The interaction between endogenous AhR and RORt was drastically enhanced in T cells of Lck-GLK Tg mice (Fig. 4B). Confocal imaging analysis showed colocalization of AhR and RORt in the nucleus of T cells from Lck-GLK Tg mice (Fig. 4C). ChIP analysis using an anti-RORt antibody also showed the binding of RORt to the AhRbinding element of the IL-17A promoter in T cells of Lck-GLK mice (Fig. 4D), suggesting that RORt binds to the AhR-binding element through the AhR-RORt complex formation. Moreover, in situ PLA with probes corresponding to AhR and RORt showed strong interaction signals in the nucleus of T cells from Lck-GLK Tg mice compared to those from wild-type mice (Fig. 4E). These data indicate direct interaction between AhR and RORt in the nucleus of T cells from Lck-GLK Tg mice. In addition, similar to AhR nuclear translocation, GLK-induced RORt nuclear translocation was abolished by PKC KO (Fig. 4C). The data support the idea that PKC-phosphorylated AhR recruits RORt into the nucleus. AhR-mediated RORt nuclear translocation is further confirmed by AhR KO; RORt localized exclusively in the cytoplasm in T cells of Lck-GLK;AhR f/f ;CD4-Cre mice even in the presence of functional PKC (Fig. 4C, bottom). Thus, this result rules out the possibility that PKC directly regulates RORt nuclear translocation. Collectively, these results indicate that AhR directly interacts with and transports RORt into the nucleus in GLK-overexpressing T cells.

IKK-mediated RORt Ser 489 phosphorylation induces AhR-RORt interaction
Surprisingly, in situ PLA showed an interaction between AhR and RORt even in the cytoplasm of Lck-GLK;PKC −/− T cells (Fig. 4E). This result suggests that the interaction between AhR and RORt is not regulated by AhR phosphorylation. Interaction between AhR and RORt is still detectable in Lck-GLK;PKC −/− T cells (Fig. 4E); this result may be due to a compensatory signaling event in PKC KO mice. To identify the kinase that stimulates the interaction between AhR and RORt, we initially tested the potential role of GLK, IKK, or IKK in the induction of AhR-RORt interaction. We cotransfected the kinase GLK, PKC, IKK, or IKK with AhR plus RORt into HEK293T cells, followed by coimmunoprecipitation assays. The data showed that IKK overexpression enhanced the interaction between AhR and RORt, whereas overexpression of GLK, PKC, and IKK did not (Fig. 4F). Next, we tested whether IKK stimulates RORt phosphorylation, which then induces the interaction of RORt with AhR. Flag-tagged RORt was purified from HEK293T cells cotransfected with RORt plus either IKK or vector. Glutathione S-transferase (GST) pulldown assays showed that purified GST-tagged AhR recombinant proteins strongly interacted with the purified RORt proteins from RORt plus IKK-cotransfected cells (Fig. 4G). The data suggest that IKK stimulates a direct interaction between AhR and RORt by inducing RORt phosphorylation. Conversely, PLA data showed that IKK KO abolished the GLK-induced interaction between AhR and RORt in T cells of Lck-GLK Tg mice (Fig. 5A). In addition, confocal imaging analysis showed that IKK KO specifically abolished nuclear translocation of RORt, but not AhR, in GLK Tg T cells (Fig. 5B). Consistently, serum IL-17A levels were also decreased in Lck-GLK;IKK f/f ;CD4-Cre mice compared to those in Lck-GLK Tg mice (Fig. 5C). Next, we studied whether IKK interacts directly with RORt. GST and His pulldown assays using purified recombinant GST-tagged IKK and His-tagged RORt proteins showed a direct interaction between IKK and RORt (Fig. 5D).
To probe whether IKK phosphorylates RORt, we individually immunoprecipitated Flag-tagged RORt, IKK, and IKK kinasedead (K44M) mutant from HEK293T transfectants and then subjected them to in vitro kinase assays. The data showed that IKK, but not IKK-K44M mutant, induced RORt phosphorylation in vitro (Fig. 5E). To identify the IKK-targeted RORt phosphorylation site, we isolated in vitro phosphorylated Flag-tagged RORt, followed by mass spectrometry (MS) analyses. Ser 489 was identified as the RORt phosphorylation site by IKK (Fig. 5F). To demonstrate phosphorylation of the RORt Ser 489 site, we generated an anti-phospho-RORt (Ser 489 ) antibody, which specifically recognized RORt wild type but not S489A mutant when cotransfected with IKK (Fig. 5G). Immunoblotting using this phosphoantibody showed that RORt Ser 489 phosphorylation was induced in T cells of Lck-GLK Tg mice, and the phosphorylation was abolished by IKK KO (Fig. 5H). Consistently, RORt-S489A mutant failed to interact with AhR under IKK overexpression (Fig. 5I). Together, in GLK-overexpressing T cells, IKK phosphorylates RORt at Ser 489 and induces its interaction with AhR, which then transports RORt into the nucleus and cooperates with RORt to stimulate IL-17A transcription.

Phosphorylated RORt interacts with AhR in TCR-induced or murine autoimmune T cells
Because both GLK and IKK (24) activation and IL-17A production (31,32) are inducible by TCR signaling, we asked whether phosphorylated RORt-mediated AhR-RORt interaction is also induced by TCR signaling. We found that, following IKK activation, anti-CD3 stimulation induced RORt Ser 489 phosphorylation in murine T cells (Fig. 6A). Moreover, coimmunoprecipitation assays and PLA showed that TCR signaling induced the interaction between endogenous RORt and AhR (Fig. 6B and fig. S8A). TCR signaling also stimulated the interaction between AhR and Ser 489 -phosphorylated RORt (fig. S8B). Conversely, IKK cKO abolished the TCR-induced RORt Ser 489 phosphorylation (Fig. 6C) and the AhR-RORt interaction (Fig. 6D) in T cells. These data suggest that IKK-mediated RORt phosphorylation and subsequent AhR-RORt interaction are induced during T cell activation. To investigate whether IKK-mediated RORt phosphorylation regulates IL-17A production after TCR stimulation, we determined secreted cytokines from anti-CD3-stimulated T cells by ELISA. Consistent with previous reports (31,32), TCR signaling induced IL-17A production in T cells (Fig. 6, E and F). IKK cKO abolished TCR-induced IL-17A production in T cells (Fig. 6E), supporting the notion that TCR-activated IKK induces IL-17A production in normal T cells. As expected, TCR-induced levels of several IKK/NF-B-mediated cytokines (IL-2, IFN-, IL-4, IL-6, and TNF-) (table S1) were also reduced in T cells of IKK cKO mice (Fig. 6E). To further verify the IKK-RORt-IL-17A pathway, we used primary splenic T cells from T cell-specific RORt cKO mice. Unlike IKK cKO, RORt cKO only abolished IL-17A production upon TCR stimulation (Fig. 6F). We verified the abolishment of RORt expression in T cells of RORt cKO mice by immunoblotting analysis (Fig. 6G). The data suggest that IKK-RORt activation mainly induces IL-17A production during TCR signaling, while IKK-NF-B activation regulates the production of multiple cytokines, including IL-2, IFN-, IL-4, IL-6, and TNF- ( Fig. 6H and fig. S9).

DISCUSSION
IL-17A is a major proinflammatory cytokine enhanced in the sera of patients with autoimmune diseases. Identification of upstream kinases and transcriptional mechanisms for IL-17A production should facilitate development of novel therapeutic approaches for IL-17A-mediated diseases. Here, we report that GLK, PKC, and IKK are key kinases controlling IL-17A transcription through a bifurcate signaling cascade in pathogenic T H 17 cells.
A key finding of this report indicates that RORt is a novel target of IKK; IKK directly phosphorylates RORt at Ser 489 , leading to IL-17A induction. To our knowledge, this IKK-phosphorylated Ser 489 is the first identified pathogenic phosphorylation site of RORt. IL-17A induction is concomitant with increased RORt mRNA levels under T H 17 polarization conditions in vitro (28,33,34); however, RORt mRNA levels are unchanged during TCR-induced IL-17A induction (32). Because of the strong induction of RORt expression during in vitro T H 17 differentiation (28,33,34), one might think that RORt proteins are not expressed before T H 17 differentiation. However, RORt mRNA levels in murine T H 0 cells are detectable, although very low (35,36). Moreover, RORt proteins in human T H 0 cells (36) or murine naïve CD4 + T cells (37)(38)(39) are detectable by immunoblotting. The RORt protein levels in naïve CD4 + T cells are one-fifth of those in T H 17 cells (38). Furthermore, data using RORt-GFP reporter mice and flow cytometry also show that some CD3 + T cells in the spleen, lymph nodes, bone marrow, lung, and skin express RORt proteins (40), and less than 50% of these infiltrating RORt + T cells express IL-17A (40). These data indicate that RORt proteins are expressed in naïve T cells. Collectively, either overexpression or Ser 489 phosphorylation of RORt can lead to IL-17A transcriptional activation. Ser 489 -phosphorylated RORt interacts with AhR and is transported into the nucleus by AhR. Notably, both TCR-stimulated or autoimmune T cells displayed induction of RORt Ser 489 phosphorylation and AhR-RORt   (table S1). NF-B is required for TCR-induced production of multiple cytokines; however, the GLK-IKK-NF-B cascade alone is not sufficient for the induction of multiple cytokines. Collectively, GLK overexpression or TCR signaling induces IL-17A transcription through AhR and RORt in T cells.
interaction. These findings suggest that IKK-mediated RORt Ser 489 phosphorylation is a common transcriptional activation mechanism of IL-17A induction in T cell activation or autoimmune diseases. Ubiquitination of RORt at Lys 446 negatively regulates IL-17A, but not IL-17F, production by blocking its interaction with the coactivator SRC1 (41). In contrast, RORt K63-linked ubiquitination at Lys 69 is required for IL-17A transcription (42). In addition, CD5 signaling may contribute to RORt nuclear translocation in T H 17 differentiation in vitro (43). IL-6 plus IL-23 stimulation also induces RORt nuclear translocation in murine neutrophils in vitro (44). Thus, it is likely that RORt Ser 489 phosphorylation also cross-talks with or is involved in other signaling events.
One of the exciting findings in this report is that AhR is responsible for nuclear translocation of RORt. PKC directly activates AhR by phosphorylating AhR at Ser 36 during TCR signaling and GLK signaling. Ser 36 phosphorylation of AhR mediates the nuclear translocation of the AhR-RORt complex, and both transcription factors then induce IL-17A transcription. Paradoxically, GLK overexpression only enhanced AhR (but not RORt) response element-driven reporter activity. Nevertheless, the mutation of either AhR-binding site or RORt-binding site within the IL-17A promoter blocked GLK signaling-induced IL-17A transcription. It is likely that, besides nuclear translocation of RORt by AhR, the binding of AhR to the AhR response element or indirectly binding to other regions within the IL-17A promoter may also facilitate the binding of RORt to the IL-17A promoter. The binding of AhR with its ligand triggers the nuclear translocation and DNA binding activity of AhR; distinct ligands stimulate different physiological functions of AhR in T cells (45,46). AhR activation by a natural ligand results in enhanced T H 17 population and severe EAE symptoms in mice (45,47). Besides TCR signaling, it is possible that AhR Ser 36 phosphorylation may also regulate IL-17A production in ligand-induced or other unliganded signaling pathways.
GLK is required for TCR signaling and production of several cytokines such as IFN-, IL-4, and IL-2 (24); however, Lck-GLK Tg mice showed selective overproduction of IL-17A, but not other NF-B-mediated cytokines. This dichotomy may be attributed to the fact that, besides NF-B, other transcription factors (such as NFAT1 or AP-1) are also required for the transcriptional activation of IL-2, IL-4, IFN-, IL-6, or TNF- during TCR signaling (table S1) (48,49). Thus, activation of NF-B alone without activation of other critical transcription factors in T cell is not sufficient to induce the production of IL-2, IL-4, IFN-, IL-6, or TNF- (Fig. 6H). T cell-specific IKK Tg mice do not display any significant induction of IL-2 or IFN- in resting T cells, while basal levels of the T cell activation markers CD25 and CD69 are enhanced (50). Consistently, activated IKK in GLK Tg T cells is not sufficient to induce NF-B-mediated cytokines. In addition, STAT3 activation is required for the transcription of IL-17F, IL-21, IL-22, and IL-23R in T H 17 cells (39,(51)(52)(53); conversely, IL-6 or IL-21 induces STAT3 activation in T H 17 cells (39,51,53). It is likely that the reason for the lack of IL-17F/IL-21/IL-22 induction and IL-23R overexpression in GLK Tg T cells may be the lack of STAT3 activation. Besides IL-17A, IL-17F and IL-22 are also regulated by RORt (53). However, the Lys 466 -ubiquitinated RORt regulates the production of IL-17A, but not IL-17F (41). Our results in this study showed the binding of the AhR-RORt complex to the −877, but not the −120, region of the IL-17A promoter. These findings suggest that the AhR-RORt complex does not bind to the RORt-binding elements that regulate the transcription of IL-17F and IL-22. Collectively, the GLK-IKK-RORt axis is not sufficient to induce multiple cytokines because of the lack of activation of other critical transcription factors or the lack of AhR-RORt complex-binding elements. Last, it is also plausible that GLK overexpression may suppress the production of multiple cytokines through an unknown inhibitory mechanism.
Our study reveals a critical pathogenic mechanism of GLKinduced IL-17A transcription in autoimmune diseases. In summary, GLK signaling induces Ser 36 phosphorylation and nuclear translocation of AhR through PKC. AhR interacts with RORt and transports RORt into the nucleus of T cells. The GLK-PKC-AhR axis is responsible for the nuclear translocation of RORt, but not for the interaction of RORt with AhR. On the other hand, the interaction between RORt and AhR is controlled by IKK, which phosphorylates RORt at Ser 489 . Similarly, TCR signaling also induces phosphorylation of AhR and RORt, as well as the interaction and nuclear translocation of the AhR-RORt complex. Thus, TCR or GLK signaling induces IL-17A transcription through the novel RORt Ser 489 phosphorylation and subsequent AhR-RORt interaction/ nuclear translocation. Collectively, the GLK-PKC/IKK-AhR/RORt pathway plays a critical role in the pathogenesis of IL-17A-mediated autoimmune diseases (Fig. 6H and fig. S9). This pathway may also be involved in other IL-17A-mediated inflammatory diseases that are induced by other IL-17A + cell types (for example, group 3 innate lymphoid cells). Thus, inhibitors of GLK or the AhR-RORt complex could be used as IL-17A-blocking agents for the treatment of IL-17A-mediated diseases. Furthermore, since GLK overexpression is also correlated with cancer recurrence (54,55), studying the potential involvement of GLK-IL-17A signaling in cancer recurrence/ metastasis may help future development of cancer therapy.

Mice
All animal experiments were performed in the Association for Assessment and Accreditation of Laboratory Animal Care International (AAALAC)-accredited animal housing facilities at the National Health Research Institutes (NHRI). All mice were used according to the protocols and guidelines approved by the Institutional Animal Care and Use Committee of NHRI. Floxed AhR mice (JAX 006203), IL-17Adeficient mice (JAX 016879), floxed RORt mice (JAX 008771), and floxed IKK mice (EMMA 001921) were purchased from the Jackson Laboratory (JAX) or the European Mouse Mutant Archive (EMMA). The three aforementioned mouse lines were backcrossed for 10 generations onto the C57BL/6 background. The data presented in this study were performed on sex-matched, 4-to 26-week-old littermates. For T cell development analyses, 5-week-old, sex-matched mice were used. All mice used in this study were maintained in temperaturecontrolled and pathogen-free cages.

Generation of Lck-GLK Tg mice and PKC KO mice
A full-length human GLK coding sequence was placed downstream of the proximal Lck promoter (fig. S1A). Lck-GLK Tg mice in C57BL/6 background were generated using pronuclear microinjection by NHRI Transgenic Mouse Core. Two independent Lck-GLK Tg mouse lines were used. Lck-GLK Tg mouse line #1 was used in all the Lck-GLK Tg experiments, except for the studies in figs. S1B and S2E, in which Lck-GLK Tg mouse line #2 was used. PKC KO mice were generated by TALEN-mediated gene targeting ( fig. S7, A to C); the nucleotides 5′-GGTGGAACACTAAAAATAATATGTCTTAGAGCCCCAT-ACATACAGTGTTTGTCTTTTGTCATTTTTCTAGGGAA-CAACCATGTCACCGTTTC-3′ of the PKC intron 1 and exon 2 were deleted in the mutated allele. For TALEN-mediated gene targeting in mice, embryo microinjection of TALEN mRNA was performed by NHRI Transgenic Mouse Core.

Reagents and antibodies
GLK antibody (-GLK-N) was generated by immunization of rabbits with peptides (murine GLK epitope: 4 GFDLSRRNPQEDFELI 19 ; identical to human GLK protein sequences 4 to 19) and was used for Figs. 2E, 3 (D and E), and 4F. Anti-GLK monoclonal antibody (clone C3) was generated by immunization of mice with peptides (murine GLK epitope: 514 EQRGTNLSRKEKKDVPKPI 533 ) and was used for
The mutant constructs were verified by DNA sequencing. The plasmids pGL4.43-Luc2P-XRE (AhR-response XRE-Luc), pGL4.47-Luc2P-SIE (STAT3-responsive SIE-Luc), pGL4.32-Luc2P-NF-B-RE-Hygro, and pGL4 luciferase reporter vector were purchased from Promega. The plasmid for the RORt (−877) response element-driven reporter was constructed by cloning four copies of the RORt (−877) response element into the pGL4.43 luciferase reporter vector. For in vitro binding assays, purified AhR proteins were isolated from HA-AhRtransfected HEK293T cells, followed by HA-peptide elution. Purified recombinant GST-PKC and GST-IKK proteins were purchased from SignalChem. Purified recombinant 6×His-ROR proteins were purchased from MyBioSource. Recombinant proteins of GST-tagged PKC K409W proteins were isolated from Escherichia coli (BL21) and then purified by GST pulldown assays. Purified Flag-tagged RORt protein was immunoprecipitated and then eluted with Flag peptides from lysates of HEK293T cells that were cotransfected with Flag-RORt plus either CFP-IKK or vector. Purified recombinant protein of GST-tagged AhR was purchased from Abnova.

Luciferase reporter assays
The 2-kb IL-17A promoter-driven firefly luciferase reporter plasmid and a Renilla luciferase control plasmid (pRL-TK) were cotransfected into Jurkat T cells. After 24 hours, 1 × 10 6 cells were harvested, washed with phosphate-buffered saline, and resuspended in 60 l of RPMI 1640 plus 60 l of lysis buffer. Data represent the mean of the ratios of the firefly luciferase activity to the Renilla luciferase activity.

ALPHA technology
ALPHA technology/protein-protein interaction assays were performed according to the manufacturer's protocol from PerkinElmer Life Sciences, as described in a previous publication (57). When the proteindonor pair was within 200 nm, a luminescent signal was detected by an EnVision 2104 Multilabel Plate Reader (PerkinElmer Life Sciences).

FRET assays FRET signal in live cells was detected by an EnVision 2104 Multilabel
Plate Reader (PerkinElmer Life Sciences). The reaction was excited by light passing through a 430-nm filter (with 8 nm bandwidth), and the intensity of emitted fluorescence passing through a 530-nm filter (with 8 nm bandwidth) was recorded. If the protein-protein pair was in close proximity (1 to 10 nm), a 530-nm signal would be detected. The FRET efficiency was calculated by a formula, efficiency = (1 − FDA/FD) × 100% (where FDA is the relative fluorescence intensities of the donor in the presence of the acceptor and FD is the relative fluorescence intensities of the donor in the absence of the acceptor).
In situ PLA PLAs were performed using the Duolink In Situ Red Starter kit (Sigma-Aldrich) according to the manufacturer's instructions. Briefly, cells were incubated with rabbit or mouse primary antibodies for each molecule pair (AhR plus PKC, AhR plus RORt, or Flag plus Myc), followed by species-specific secondary antibodies conjugated with oligonucleotides (PLA probes). After ligation and amplification reactions, the PLA signals from each pair of PLA probes in close proximity (<40 nm) were visualized as individual red dots by a fluorescence microscope (Leica DM2500) or a confocal microscope (Leica TCS SP5 II). Each red dot represents a direct interaction.

Immunoprecipitation, GST/His pulldown, and immunoblotting analyses
Immunoprecipitation was performed by preincubation of 0.5 to 1 mg of protein lysates with 1 g of antibody for 1 hour at 4°C, fol-lowed by the addition of 20 l of protein A/G Sepharose beads for 3 hours. For GST and His pulldown assays, GST-and His-tagged proteins plus their interacting proteins were incubated for 3 hours with glutathione Sepharose beads (GE Healthcare) and Ni Sepharose beads (GE Healthcare), respectively. The immunocomplexes or GST/His pulldown complexes were washed with lysis buffer (1.5 mM MgCl 2 , 0.2% NP-40, 125 mM NaCl, 5% glycerol, 25 mM NaF, 50 mM tris-HCl, and 1 mM Na 3 VO 4 ) three times at 4°C, followed by boiling in 5× loading buffer at 95°C for 3 min. The immunoblotting analyses were performed as described previously (57).

Cell transfections, T cell stimulation, in vitro kinase assays for PKC or IKK, and flow cytometric analyses
These experiments were performed as described previously (24).

Immunofluorescence and confocal imaging
Cells were fixed in cold methanol for 2 min. After permeation with 0.25% Triton X-100 for 1 hour, the cells were blocked with 5% bovine serum albumin for 2 hours. The cells were incubated with primary antibodies (1:200 dilution) for 24 hours and then with secondary antibodies (1:500 dilution) for 2 hours. The secondary antibodies donkey anti-rabbit immunoglobulin G (IgG)-Alexa Fluor 568 and goat anti-mouse IgG-Alexa Fluor 488 were purchased from Abcam and Life Technologies, respectively. The cover slides were mounted in Fluoroshield with DAPI (GeneTex) and analyzed using a Leica TCS SP5 confocal microscope.

Liquid chromatography-MS
After in vitro kinase assays, protein bands of Flag-tagged RORt were collected from InstantBlue (GeneMark)-stained SDS-polyacrylamide gel electrophoresis gels. Proteins were digested with trypsin and subjected to liquid chromatography-MS/MS analyses by an LTQ Orbitrap Elite hybrid mass spectrometer as described previously (57).

Statistical analyses
All experiments were repeated at least three times. Data are means ± SEM. The statistical significance between two unpaired groups was analyzed using two-tailed Student's t test. P values of less than 0.05 were considered statistically significant.

SUPPLEMENTARY MATERIALS
Supplementary material for this article is available at http://advances.sciencemag.org/cgi/ content/full/4/9/eaat5401/DC1 Fig. S1. Normal T cell and B cell development in Lck-GLK Tg mice. Fig. S2. Inflammatory phenotypes and enhanced TH17 differentiation in Lck-GLK Tg mice. Fig. S3. Autoimmune responses in Lck-GLK Tg mice are abolished by IL-17A deficiency. Fig. S4. GLK transgene does not regulate IL-23 receptor expression, STAT3 phosphorylation, and RORt-binding element at the −120 region of the IL-17A promoter. Fig. S5. PKC controls Ser 36 phosphorylation-mediated AhR nuclear translocation and AhR-mediated autoimmune responses. Fig. S6. PKC directly interacts with AhR in the cytoplasm of Lck-GLK T cells. Fig. S7. Autoimmune responses in Lck-GLK Tg mice are reduced by PKC KO. Fig. S8. TCR signaling induces in vivo interaction between AhR and RORt. Fig. S9. Schematic model of AhR/RORt-mediated IL-17A transcription in T cells of Lck-GLK Tg mice with different gene-KO backgrounds. Table S1. Transcription factors of NF-B-mediated cytokines. References (58,59)