Identification of Canonical Tyrosine-dependent and Non-canonical Tyrosine-independent STAT3 Activation Sites in the Intracellular Domain of the Interleukin 23 Receptor*

Background: Activation of STAT3 is the major signaling pathway of IL-23. Results: The study provides detailed characterization of IL-23-dependent signal transduction with the focus on STAT3 phosphorylation. Conclusion: Canonical tyrosine-dependent and non-canonical tyrosine-independent STAT3 activation sites are located within the IL-23R. Significance: This may be the first step in elucidating the signal transduction of IL-23, an important cytokine for TH17 cell development. Signaling of interleukin 23 (IL-23) via the IL-23 receptor (IL-23R) and the shared IL-12 receptor β1 (IL-12Rβ1) controls innate and adaptive immune responses and is involved in the differentiation and expansion of IL-17-producing CD4+ T helper (TH17) cells. Activation of signal transducer and activator of transcription 3 (STAT3) appears to be the major signaling pathway of IL-23, and STAT binding sites were predicted in the IL-23R but not in the IL-12Rβ1 chain. Using site-directed mutagenesis and deletion variants of the murine and human IL-23R, we showed that the predicted STAT binding sites (pYXXQ; including Tyr-504 and Tyr-626 in murine IL-23R and Tyr-484 and Tyr-611 in human IL-23R) mediated STAT3 activation. Furthermore, we identified two uncommon STAT3 binding/activation sites within the murine IL-23R. First, the murine IL-23R carried the Y542PNFQ sequence, which acts as an unusual Src homology 2 (SH2) domain-binding protein activation site of STAT3. Second, we identified a non-canonical, phosphotyrosine-independent STAT3 activation motif within the IL-23R. A third predicted site, Tyr-416 in murine and Tyr-397 in human IL-23R, is involved in the activation of PI3K/Akt and the MAPK pathway leading to STAT3-independent proliferation of Ba/F3 cells upon stimulation with IL-23. In contrast to IL-6-induced short term STAT3 phosphorylation, cellular activation by IL-23 resulted in a slower but long term STAT3 phosphorylation, indicating that the IL-23R might not be a major target of negative feedback inhibition by suppressor of cytokine signaling (SOCS) proteins. In summary, we characterized IL-23-dependent signal transduction with a focus on STAT3 phosphorylation and identified canonical tyrosine-dependent and non-canonical tyrosine-independent STAT3 activation sites in the IL-23R.


Signaling of interleukin 23 (IL-23) via the IL-23 receptor (IL-23R) and the shared IL-12 receptor ␤1 (IL-12R␤1) controls innate and adaptive immune responses and is involved in the differentiation and expansion of IL-17-producing CD4 ؉ T helper (T H 17) cells. Activation of signal transducer and activator of transcription 3 (STAT3) appears to be the major signaling
pathway of IL-23, and STAT binding sites were predicted in the IL-23R but not in the IL-12R␤1 chain. Using site-directed mutagenesis and deletion variants of the murine and human IL-23R, we showed that the predicted STAT binding sites (pYXXQ; including Tyr-504 and Tyr-626 in murine IL-23R and Tyr-484 and Tyr-611 in human IL-23R) mediated STAT3 activation. Furthermore, we identified two uncommon STAT3 binding/activation sites within the murine IL-23R. First, the murine IL-23R carried the Y 542 PNFQ sequence, which acts as an unusual Src homology 2 (SH2) domain-binding protein activation site of STAT3. Second, we identified a non-canonical, phosphotyrosine-independent STAT3 activation motif within the IL-23R. A third predicted site, Tyr-416 in murine and Tyr-397 in human IL-23R, is involved in the activation of PI3K/Akt and the MAPK pathway leading to STAT3-independent proliferation of Ba/F3 cells upon stimulation with IL-23. In contrast to IL-6induced short term STAT3 phosphorylation, cellular activation by IL-23 resulted in a slower but long term STAT3 phosphorylation, indicating that the IL-23R might not be a major target of negative feedback inhibition by suppressor of cytokine signaling (SOCS) proteins. In summary, we characterized IL-23-dependent signal transduction with a focus on STAT3 phosphorylation and identified canonical tyrosine-dependent and non-canonical tyrosine-independent STAT3 activation sites in the IL-23R.
Some of the proinflammatory functions of IL-23 are related to the induction of terminal differentiation and proliferation of IL-17-producing CD4 ϩ T helper (T H 17) cells (7). T H 17 cells are involved in the pathogenesis of inflammatory autoimmune diseases, including rheumatoid arthritis, multiple sclerosis, psoriasis, and inflammatory bowel disease (8). Consequently, IL-23 (p19)-or IL-23 receptor (IL-23R)-deficient mice are resistant to autoimmune and inflammatory disorders, such as experimental autoimmune encephalomyelitis (9), collagen-induced arthritis (10), and inflammatory bowel disease (11,12). Accordingly, targeting IL-23 signal transduction pathways has become a promising therapeutic strategy. However, clinical trials with the anti-p40 antibody ustekinumab, which neutralizes both IL-23 and the related IL-12, have not been reproducibly effective in patients with Crohn disease (13,14). IL-23 also appears to promote tumor incidence and growth (15), but at least one report indicates that IL-23 has also anti-tumorigenic activity against pediatric B-acute lymphoblastic leukemia (B-ALL) cells (16), demonstrating (e.g. for Crohn disease and opposing roles in cancer development) that the biology of IL-23 is still incompletely understood. IL-23 is mainly produced by myeloid dendritic cells after Toll-like receptor activation (17)(18)(19) and by activated proinflammatory type 1 macrophages (20). Little is known about regulation of IL-23 receptor expression, but a number of cell types have been described to express IL-23 receptor chains, including CD4 ϩ T cells of the T H 17 lineage, ␥␦ T cells, macrophages, dendritic cells, and innate lymphoid cells, although these cells are often almost unresponsive to IL-23 due to low expression of the IL-23-specific IL-23R (21,22).
Signaling within the IL-12 family occurs via receptor chains that are structurally homologous to the gp130 family. IL-23 induces dimerization of the IL-12R␤1 receptor, which is also one receptor for IL-12, and the IL-23-specific IL-23R (23). Binding of IL-23 to signal-transducing type I transmembrane ␤-receptors induces the activation of noncovalently ␤-receptor-bound Janus kinases (JAKs) and subsequent signaling pathways, including signal transducers and activators of transcription (STAT) transcription factors (23), phosphoinositide 3-kinase (PI3K) (24), and NF-B (24). So far, there are no data about the involvement of MAPK in IL-23 signal transduction. Cho et al. (24) investigated the effect of a MEK inhibitor on the induction of IL-17 but did not find any influence. STAT and SHP2 phosphorylation by Janus kinases depends on the ability of the Src homology 2 (SH2) 2 domain of the STAT factors and the SHP2 interaction domain to interact with phosphorylated tyrosines embedded in factor-specific binding sites of the cytokine receptor. The IL-23R is regarded as the only signal-transducing component of the IL-23 receptor complex, and, similar to IL-12 signaling, IL-12R␤1 is required for high affinity binding (25). The murine and human IL-23R proteins are 644-and 629-amino acid residue-long type I transmembrane proteins, respectively, with a 66% identity on the amino acid level. The intracellular domains of murine and human IL-23R comprise 247 and 252 amino acid residues, respectively. It has been suggested that IL-23 signaling is mediated through three of seven intracellular tyrosine residues of the IL-23R (Tyr-416, Tyr-504, and Tyr-626 in mice and Tyr-397, Tyr-484, and Tyr-611 in humans). The Y m416/h397 EDI sequence was predicted to be a potential SHP2 binding site (23,26), and Y m626/h611 FPQ was predicted to be a potential STAT1 and STAT3 binding site (23,27,28). The postulated SHP2 binding site within the IL-23R might lead to the activation of the MAPK and the PI3K cascade, as is known for IL-6 signal transduction (29). The GY m504/h484 KPQIS sequence has similarities in the IL-12R␤2 known to bind to STAT4 (GY(L/V)PS (30,31)). The STAT phosphorylation patterns (STAT1, STAT3, STAT4, and STAT5) of IL-12 and IL-23 are similar. However, IL-23-induced STAT4 phosphorylation is much weaker compared with IL-12 (23). For IL-23 signaling, STAT3 appears to be the primary mediator (23). Therefore, we focused on the analysis of STAT3 and MAPK/PI3K activation by murine and human IL-23R. Whereas the predicted MAPK/ PI3K activation site was verified, STAT3 activation is far more complex than expected, with canonical tyrosine-dependent and non-canonical tyrosine-independent activation modes.
Construction of Expression Plasmids-cDNAs coding for murine IL-23R (gene ID 209590) and IL-12R␤1 (gene ID 16161) were amplified from total RNA extracts derived from mouse T cells and cloned into the pcDNA3.1 expression vector (Invitrogen). A C-terminal c-Myc tag was added for detection in cell lysates by Western blot. To create a murine/human chimeric IL-23R with human IL-23R signal transduction, named hIL-23R(mET/hC), the intracellular part of the human receptor (amino acids 377-629) was amplified by polymerase chain reaction (PCR) from a cDNA clone (IMAGE ID 4132295, Source BioScience LifeSciences (Berlin, Germany)) and inserted into pcDNA3.1-mIL-23R, where the appropriate coding sequence (amino acids 396 -644) had been removed. Mutations of tyrosine to phenylalanine in murine IL-23R and the chimera were generated by PCRs using Phusion high fidelity DNA polymerase (FINNZYMES, Thermo Scientific) followed by DpnI digestion of methylated template DNA. Deletion variants of the murine IL-23 receptor (mIL-23R) and the chimera were cloned into pcDNA3.1 using standard PCR. The sequences of the oligonucleotides used in this study will be provided upon request. Further, all cDNAs were subcloned into the eukaryotic expression vector p409 (35) for transient transfection of HeLa cells. For retroviral transduction of Ba/F3-gp130 cells, two retroviral plasmids with different resistance genes have been used. The plasmid pMOWS encodes the puromycin resistance gene (pMOWS-puro (33)), whereas pMOWS-hygromycin mediates hygromycin resistance (pMOWS-hygro (36)). Expression cassettes coding for IL-23R variants were inserted into pMOWSpuro, and those for the murine IL-12R␤1 were inserted into the pMOWS-hygro plasmid. All generated expression plasmids have been verified by sequencing.
Stimulation Assays-For detection of phospho-STAT3 in cotransfected HeLa cells, 24 h after transfection, cells were washed with PBS and cultured for 16 h in serum-free DMEM. Afterward, cells were incubated in the absence or presence of HIL-23 as indicated and lysed in 50 mM Tris-HCl, pH 7.5, 150 mM NaCl, 2 mM EDTA, 1 mM NaF, 1 mM Na 3 VO 4 , 1% Nonidet P-40, and 1% Triton X-100, supplemented with complete protease inhibitor mixture tablets (Roche Applied Science). For detection of phospho-STAT3, phospho-STAT1, phospho-Erk1/2, and phospho-Akt, Ba/F3-gp130 cells, retrovirally transduced with cDNAs for mouse IL-12R␤1 and IL-23R variants, were washed three times with sterile PBS and starved for at least 4 h in serum-free DMEM. Cells were stimulated with HIL-23 as indicated and harvested by centrifugation, and the pellet was directly frozen in liquid nitrogen. For analysis of phospho-Erk1/2 and phospho-Akt, cells were pretreated for 1 h with signaling pathway inhibitors, such as PD98059 (MEK inhibitor) and LY294002 (PI3K inhibitor). Cells were lysed as described above. The protein concentrations of the cell lysates were determined by a BCA protein assay (Pierce) according to the manufacturer's instructions. Activation of STAT3, Erk1/2, and Akt was determined by immunoblotting using antibodies against the phosphorylated proteins.
SDS-PAGE and Western Blot-Equal amounts of proteins from cell lysates were separated by SDS-PAGE under reducing conditions and transferred to a PVDF membrane using a Trans-Blot Turbo transfer system (Bio-Rad). The membrane was blocked in 5% fat-free dried skimmed milk in TBS-T (10 mM Tris-HCl, pH 7.6, 150 mM NaCl, 0.05% Tween 20) and probed with the primary antibody in 5% fat-free dried skimmed milk in TBS-T (STAT3, STAT1, c-Myc, Erk1/2, and Akt) or 5% BSA in TBS-T (phospho-STAT3, phospho-STAT1, phospho-Erk1/2, phospho-Akt, mIL-23R, and mIL-12R␤1) at 4°C overnight. The blots were washed and incubated with the secondary peroxidase-conjugated antibody or streptavidin-HRP for 1 h before applying the ECL Prime Western blotting detection reagent (GE Healthcare). The ChemoCam Imager (INTAS Science Imaging Instruments GmbH, Göttingen, Germany) was used for signal detection according to the manufacturer's instructions. Membranes were stripped in 62.5 mM Tris-HCl (pH 6.8), 2% SDS, and 0.1% ␤-mercaptoethanol for 30 min at 60°C, blocked again, and reprobed with another primary antibody.
Surface and Intracellular STAT3 Staining-To detect surface expression of the IL-23R variants, cells were washed with FACS buffer (PBS containing 1% BSA) and incubated at 5 ϫ 10 5 cells/100 l of FACS buffer supplemented with 2.5 g of monoclonal anti-mouse IL-23R antibody (R&D Systems) for 2 h on ice. After a single wash with FACS buffer, cells were incubated in FACS buffer containing a 1:100 dilution of Alexa Fluor 647conjugated Fab goat anti-rat IgG (Dianova) for 1 h at 4°C. Finally, cells were washed once with FACS buffer, resuspended in 500 l of FACS buffer, and analyzed by flow cytometry (BD FACSCanto II flow cytometer, BD Biosciences). The resulting data were evaluated using the FCS Express software (De Novo Software, Los Angeles, CA). To detect surface expression of the mouse IL-12R␤1, cells were prepared as described above and incubated with 100 l of FACS buffer containing 10 l of PEconjugated anti-mouse IL-12R␤1 (R&D Systems) for 2 h at 4°C. For intracellular staining of phosphorylated STAT3, 1 ϫ 10 6 Ba/F3 cells were washed three times with sterile PBS, starved for at least 4 h in serum-free DMEM, and stimulated with HIL-23 as indicated. Cells were harvested by centrifugation and fixed in 2% (w/v) paraformaldehyde at 37°C for 15 min, followed by permeabilization in 90% (v/v) methanol for 30 min on ice. Cells were washed (PBS containing 0.5% BSA) and stained for phospho-STAT3 and STAT3 using Alexa Fluor 647 mouse anti-Stat3 (Tyr(P)-705) and PE-conjugated mouse anti-Stat3 antibodies (BD Biosciences) according to the manufacturer's instructions. Flow cytometry was performed using a BD FAC-SCanto II flow cytometer (BD Biosciences), and data were analyzed using FACS DIVA software (BD Biosciences).
Proliferation Assay-Ba/F3-gp130 cell lines were washed three times with sterile PBS and resuspended in DMEM containing 10% FCS at 5 ϫ 10 4 cells/ml. The cells were cultured for 2 days in a final volume of 100 l with or without cytokines, as indicated. The CellTiter-Blue cell viability assay (Promega, Karlsruhe, Germany) was used to estimate the number of viable cells following the manufacturer's protocol. Fluorescence (excitation 560 nm, emission 590 nm) was recorded using the Infinite M200 PRO plate reader (Tecan, Crailsheim, Germany) immediately after the addition of the 20 l/well CellTiter-Blue reagent (time point 0) and up to 1 h after incubation under standard cell culture conditions. Fluorescence values were normalized by subtraction of time point 0 values. For comparison of independent Ba/F3-gp130 cell lines, the n-fold proliferation was calculated by setting the negative control (cells growing without cytokines) of each Ba/F3 cell line to a value of 1. All of the values were measured in triplicate per experiment.

IL-23 Induces
Long Term STAT3 Activation-The cytoplasmic region of mIL-23R contains seven tyrosine residues; six are conserved between mIL-23R and hIL-23R. Mouse Tyr-416, Tyr-504, and Tyr-626 and human Tyr-397, Tyr-484, and Tyr-611 were defined as potential SH2 domain-binding sites for binding of SHP2, STAT4, and STAT1/3, respectively (23) (Fig.  1A). Tyr-542 is unique in mIL-23R, whereas Tyr-463 only exists in hIL-23R. Both tyrosines are not embedded in classical SH2 domain-binding sites (YXXQ) with Y 542 PNFQ and Y 463 PQ, respectively (Fig. 1A). Here, we concentrated on the activation/ phosphorylation of STAT3 by the IL-23R. To study the functional role of the tyrosine residues of the murine and human IL-23R in STAT3 activation, IL-23-responsive Ba/F3-gp130 FIGURE 1. The cytoplasmic domains of mouse and human IL-23R induce STAT3 activation. A, the mouse IL-23 receptor complex consists of IL-12R␤1, containing one tyrosine residue, box 1 and box 2 motifs, and the unique IL-23R with seven tyrosine residues. Six tyrosines are conserved in mice and humans, and putative binding sites for SHP2, STAT4, and STAT1/3 have been postulated. Tyrosine residues are indicated and numbered in IL-23R. The non-conserved tyrosines Tyr-542 (mouse) and Tyr-463 (human) are underlined. B, proliferation of stably transduced Ba/F3-gp130 cells with cDNAs coding for murine IL-23R, hIL-23R(mET/hC) containing the cytoplasmic domain of the human receptor, murine IL-12R␤1, and co-transduced Ba/F3 cells. Equal numbers of cells were cultured for 2 days in the presence of 10 ng/ml HIL-6, 0.2% HIL-6 conditioned cell culture supernatant, or 0.2% HIL-23 or without cytokine. Ba/F3-gp130 cells were used as a control. Proliferation was measured using the colorimetric CellTiter-Blueா cell viability assay. C, stably transduced Ba/F3 cells were washed three times, starved, and stimulated with 0.2% HIL-23 for 10 min. Cellular lysates were prepared, and equal amounts of proteins (50 g/lane) were loaded onto SDS gels. Western blots were performed using antibodies specific for phospho-STAT3 and STAT3. For positive control, Ba/F3-gp130 cells were stimulated for 10 min with 0.2% HIL-6 and analyzed. Western blots are shown for one representative experiment. Error bars, S.D. JULY 5, 2013 • VOLUME 288 • NUMBER 27 cell lines have been generated, expressing the murine receptor chains on their cell surface (supplemental Fig. 1, A and B). Further, a murine/human chimeric IL-23R with hIL-23R signal transduction has been constructed, in which the extracellular and transmembrane region (ET domain) of the mIL-23R was fused to the cytoplasmic region (C) of hIL-23R, named hIL-23R(mET/hC) (Fig. 1A). Proliferation of Ba/F3 cells depend on IL-3 and activation of STAT5. After stable transfection with a gp130 cDNA, proliferation of Ba/F3-gp130 cells depends on IL-6 and soluble IL-6R or Hyper-IL-6 and activation of STAT3 (34) (Fig. 1B). After stable transduction of the IL-23 receptor chains mIL-12R␤1 and mIL-23R or hIL-23R(mET/hC), Ba/F3-gp130-mIL-12R␤1-mIL-23R, and Ba/F3-gp130-mIL-12R␤1-hIL-23R(mET/hC), cells were able to proliferate in the presence of HIL-23, whereas Ba/F3-gp130 cells expressing either mIL-23R, the hIL-23R(mET/hC), or the mIL-12R␤1 did not grow (Fig. 1B). The dependence on IL-23 of these Ba/F3 cells was confirmed by analysis of STAT3 phosphorylation. HIL-23 induced STAT3 phosphorylation only in cells expressing mIL-23R or hIL-23R(mET/hC) plus mIL-12R␤1 (Fig. 1C). Next, a time course experiment was carried out to investigate the dynamics of STAT3 phosphorylation in Ba/F3-gp130-mIL-12R␤1-mIL-23R cells stimulated with HIL-23 or HIL-6 for up to 240 min. STAT3 phosphorylation was analyzed by flow cytometry and Western blotting at the indicated time points (Fig. 2, A and B). Interestingly, induction of STAT3 phosphorylation in Ba/F3-gp130-mIL-12R␤1-mIL-23R cells upon HIL-23 stimulation is slow with a peak at 60 min, but it is prolonged in comparison with HIL-6 stimulation, which leads to a maximal induction after 5-10 min and a decline after 60 min (Fig. 2, A and B).

STAT3 Activation in IL-23 Signal Transduction
As a second cellular model to investigate IL-23-dependent STAT3 phosphorylation, we chose HeLa cells. Recently, it was shown that STAT3 was activated in HeLa cells co-transfected with hIL-12␤1 and hIL-23R (37). In co-transfection experi-ments, we first examined expression and signaling of mIL-12R␤1 and mIL-23R in HeLa cells (supplemental Fig. 2A). IL-23-induced STAT3 phosphorylation was only observed in HeLa cells co-transfected with mIL-12R␤1 and mIL-23R but not after transfection of only one receptor chain (supplemental Fig. 2A). Surface expression of mIL-23R and mIL-12R␤1 was demonstrated by FACS analysis with the use of antibodies directed against the extracellular domains of mouse IL-12R␤1 and mouse IL-23R (supplemental Fig. 2B). IL-23-induced STAT3 phosphorylation was also detected in HeLa cells co-transfected with mIL-12R␤1 and hIL-23R(mET/hC) (supplemental Fig. 2C). Taken together, our results indicate that Ba/F3 and HeLa cells are good cellular systems to investigate the biological function of the mIL-23R and the hIL-23R chimera hIL-23R(mET/ hC) with respect to proliferation and STAT3 phosphorylation.

IL-23-induced STAT3 Activation Is Not Restricted to the Predicted Tyr-626 SH2-binding Site Motif in Murine IL-23R and the Predicted Tyr-611 SH2-binding Site Motif in Human IL-23R-
Two potential STAT binding sites (pYXXQ) in the cytoplasmic region of the IL-23R have been postulated (STAT4, mouse Tyr-504 and human Tyr-484; STAT1/3, mouse Tyr-626 and human Tyr-611 (23)). To examine the role of these individual tyrosine motifs in mIL-23R and hIL-23R(mET/hC), receptor mutants with Tyr 3 Phe substitutions were generated (Fig. 3A). The non-conserved tyrosine residues of mouse (Tyr-542) and human (Tyr-463) IL-23R have been included because they are part of motifs similar to the STAT binding motif (mouse Y 542 PNFQ, human Y 463 PQ) and also the potential SHP2 binding sites (mouse Y 416 EDI, human Y 397 EDI).

STAT3 Activation in IL-23 Signal Transduction
explain the differences between HeLa and Ba/F3 cells concerning STAT3 phosphorylation. One possible explanation could be that there is only a minimal phosphorylation of STAT3 in HeLa cells upon IL-23 stimulation due to the mutations of Tyr-542 and Tyr-626, which is not distinguishable from the background STAT3 activation.
As seen for the murine IL-23R, the appropriate double and triple mutants of the hIL-23R(mET/hC) (Y463F/Y611F and Y463F/Y484F/Y611F) clearly led to phosphorylation of STAT3 and proliferation of Ba/F3-gp130-mIL-12R␤1-hIL-23R(mET/ hC) cells upon stimulation with HIL-23 (Fig. 4, A and B). Contrary to the mIL-23R, stimulation of HeLa cells co-transfected with mIL-12R␤1 and the double or triple mutant of hIL-23R(mET/hC) with HIL-23 still led to STAT3 phosphorylation (supplemental Fig. 4B). Because IL-23R that did not carry any SH2-binding motif was still able to induce STAT3 phosphorylation, we were not able to determine which tyrosine within the human IL-23R was responsible for STAT3 phosphorylation in this experiment. We could, however, conclude that an additional noncanonical activation mechanism other than classical and atypical SH2-binding sites within the murine and human IL-23R must induce STAT3 phosphorylation after IL-23 stimulation.

STAT3 Activation in IL-23 Signal Transduction
suppressed by pretreatment with MEK inhibitor PD98059 or PI3K inhibitor LY294002 (Fig. 7). For control of Erk and Akt stimulation assays, appropriate Ba/F3 cell lines were further analyzed according to STAT3 activation in the absence or presence of Jak inhibitor P6. Comparable Western blot data were obtained for mIL-23R-⌬415 containing no tyrosine residue within the IL-23R (data not shown). Our results imply that proliferation of Ba/F3-gp130-mIL-12R␤1-mIL-23R (mutant) cells that fail to induce STAT3 phosphorylation after HIL-23 stimulation, such as mIL-23R-⌬503, is maintained by activation of Erk1/2 and PI3K/Akt.

DISCUSSION
Receptors that activate STAT3 are characterized by one or more YXXQ motifs (39,40). Mutation of the tyrosine or glutamine residue of this consensus sequence from receptors such as gp130 and IL-9R abrogated canonical STAT3 recruitment and phosphorylation, demonstrating that the recognition by the SH2 domain of STAT3 is critical for its activation (40,41). Here, we have identified the tyrosine-dependent canonical and tyrosine-independent non-canonical STAT3 activation sites within the murine and human IL-23R. Our study reveals three major findings. First, the tyrosine Y m416/h397 EDI motif is involved in the activation of MAPK/PI3K pathway. Second, the Y m626 FPQ motif was confirmed as the canonical STAT3 binding site in murine IL-23R, and we identified an additional STAT3 binding site, which slightly deviates from the consensus SH2-binding site YXXQ, because it contains three instead of two amino acids between the tyrosine and the glutamine (YXXXQ). This motif is only found in murine IL-23R (YPNFQ) and not in the rat (YPNFT) and other organisms and accordingly also not in the human IL-23R (HPNFA) (supplemental Fig. 6B). To the best of our knowledge, it has not been described so far that an SH2binding site can contain three amino acids between Tyr and Gln. Third, we identified an additional motif between amino acid residues 554 and 570 in the murine IL-23R that facilitates non-canonical tyrosine-independent STAT3 phosphorylation after IL-23 stimulation. Deletion of this 17-amino acid motif depleted STAT3 phosphorylation, as was demonstrated by immunoblotting and intracellular staining. However, complete abrogation of STAT3 activation was only shown by flow cytometry. Ba/F3 cells containing four Tyr 3 Phe mutations and the deletion of the non-canonical tyrosine-independent motif still proliferated in the presence of IL-23, perhaps indicating that IL-23 signal transduction is a complex network not only involv- ing Jak/STAT, MAPK, and PI3K/Akt. Previous studies also revealed the involvement of NF-B in IL-23 signaling (24). Our experiments and homology alignments indicate that this motif is conserved and accordingly also present in the human IL-23R. Protein motif analysis did not lead to the identification of a general motif between positions 553 and 571 (plus 10 amino acid residues surrounding this core sequence). Further, we failed to detect any apparent amino acid sequence identity between the identified 17-amino acid residue sequence, needed to mediate tyrosine-independent activation of STAT3, and other cytokine receptors. Two amino acid residues N-terminal of position 554 (Leu), a putative CK1 phosphorylation site (S 548 ASS) was identified, which is highly conserved within the IL-23Rs of all species analyzed. However, this site is N-terminal from the deletion ⌬554 and therefore not likely to be involved in STAT3 activation. The related CK2 has recently been described to be mandatory for STAT signaling of oncostatin M (OSM) (42), which is an IL-6 type cytokine. A general involvement of CK2 in signaling of the IL-6/IL-12 type cytokine family cannot not be excluded at the moment. Thus far, CK1 has not been described as being involved in STAT3 signaling. Moreover, the motif for CK1 (SXX(S/T)) is frequently observed in proteins, and putative CK1 sites do not necessarily represent real target sites. Further, it was suggested that PI3K/Akt may be involved in STAT3 phosphorylation (24). However, we showed that the PI3K/Akt-mediated STAT3 activation is not related to the here described non-canonical tyrosine-independent STAT3 phosphorylation, because this also occurs in a mutant (Y416F/ Y504F/Y542F/Y626F) that was not able to activate PI3K/Akt after IL-23 stimulation, and stimulation of Ba/F3-gp130-mIL12R␤1-mIL-23R cells with IL-23 in the presence of the PI3K inhibitor LY294002 did not deplete STAT3 phosphorylation (data not shown).
In the case of the G-CSFR, a mutant lacking all intracellular tyrosines was shown to recruit and activate STAT3 via an unknown mechanism, albeit at higher concentrations of G-CSF compared with tyrosine-dependent activation (43). The authors did not detect a constitutive association of STAT3 with the G-CSFR. They speculate that an intermediate molecule, interacting with the C-terminal region of the receptor, contains a phosphotyrosine binding site for the SH2 domain of STAT3 (43). In the case of IL-23R, we were also not able to co-immunoprecipitate STAT3 and IL-23R (data not shown), indicating that STAT3 and IL-23R are, as in the case of G-CSFR, not con- stitutively associated. Recently, Dumoutier et al. (44) demonstrated that the C-terminally located 84 amino acid residues of IL-22R contain a thus far undefined motif, which allows constitutive association with STAT3, most likely via the coiled-coil domain of STAT3. Importantly, IL-22-induced tyrosine-independent activation of STAT3 is facilitated by this 84-amino acid sequence, and mutation of all cytoplasmic tyrosine residues of the IL-22R only partially affects STAT3 activation. The authors speculated that receptor preassociation with STAT3 assures a faster response or efficient STAT3 activation in cells with lower endogenous STAT3 expression. Association of STAT3 with IL-23R to enable fast STAT3 activation is also not supported by our finding that IL-23-induced activation of STAT3 is slow compared with IL-6-induced activation of STAT3 via gp130. Therefore, we assume that binding of STAT3 to the receptor for both canonical and non-canonical STAT3 activation is cytokine-induced. However, the detailed mechanism has to be elaborated in further studies. SH2-independent recruitment of STAT3 might serve to avoid negative feedback by proteins such as suppressor of cytokine signaling 3 (SOCS3), which can compete with STAT factors for phosphotyrosines (46). IL-6-gp130-STAT3 activation is rapidly switched off by SOCS3-negative feedback regulation (47). We and others showed that IL-23induced STAT3 activation is not switched off (48), suggesting that SOCS3 is not a negative regulator of IL-23 signaling. However, these data were obtained after overexpression of the IL-23R. Chen et al. (49) reported that SOCS3 negatively regulates IL-23 signaling during T H 17 development in primary T cells. Therefore, it remains to be seen whether the non-canonical STAT3 activation plays a role in escape from negative SOCS3 feedback inhibition. STAT3 plays a major role during differentiation and proliferation of T H 17 cells. The frequency of T H 17 cells was reduced in experimental autoimmune encephalomyelitis (EAE)-resistant mice with a conditional deletion of STAT3 (50,51), and deletion of SOCS3 increased the number of T H 17 cells (49). It has been demonstrated that IL-6 is needed for STAT3 activation during the initial but not late phase of T H 17 differentiation and not for T H 17 proliferation (7). It might be speculated that the rapid SOCS3-mediated negative feedback down-regulation of IL-6-mediated STAT3 activation provokes the need for second line cytokines such as IL-23 to ensure prolonged STAT3 activation during late phase of T H 17 differentiation and proliferation that is not completely negatively regulated by SOCS proteins. Sustained STAT3 activation can contribute to tumor development (52), and the IL-6/gp130 signaling pathway is a candidate for constitutive STAT3 activation in tumors (53). Therefore, STAT3 activation has to be strictly regulated, which might not be properly ensured by IL-6 signaling because the responsible receptors are more commonly expressed. IL-23R expression is very limited and not found, for example, on naive T cells. However, during differentiation of T cells to T H 17 cells, these cells start to express IL-23R FIGURE 9. Characterization of the non-canonical STAT3 activation motif in the intracellular domain of IL-23R. A, two IL-23R variants either with a deletion of 17 amino acids (Y416F/Y504F/Y542F/Y626F-⌬554 -570) or with the addition of the respective motif (Y416F-⌬503ϩ543-582) were generated and stably transduced into Ba/F3-gp130-mIL-12R␤1 cells. Resulting Ba/F3 cell lines were washed three times with PBS, starved, and stimulated for 10 min with IL-23. Equal amounts of protein were loaded onto SDS gels. Western blotting was performed using antibodies specific for phospho-STAT3 and STAT3. Data are representative for two experiments. The presented Western blots originate from different membranes and are therefore separated by black lines. B, equal numbers of stably transduced Ba/F3 cells were cultured for 2 days in the presence of 0.2% HIL-6 or 0.2% HIL-23 or without cytokine. Proliferation was measured with the colorimetric CellTiter-Blueா cell viability assay. Values represent the mean value of three repetitions and were normalized. For comparison, n-fold proliferation was calculated by setting the negative control of each Ba/F3 cell line to a value of 1. Data are representative of at least two independent experiments. C, IL-23R variants were washed three times with PBS and starved for 2 h in serum-free DMEM. 1 ϫ 10 6 cells were stimulated for 30 min with 0.2% HIL-6 or HIL-23, harvested by centrifugation, fixed in 2% (w/v) paraformaldehyde, and permeabilized in 90% (v/v) methanol. Cells were stained for phospho-STAT3 and STAT3 overnight and analyzed by flow cytometry. Two Ba/F3-gp130-mIL-12R␤1-mIL-23R/Y416F/Y504F/ Y542F/Y626F-⌬554 -570 cell passages were analyzed, cultivated in the presence of either HIL-6 (I) or HIL-23 (II). Error bars, S.D. and become a target for IL-23 to enable target-oriented and sustained activation of STAT3. However, it remains to be seen whether the non-canonical STAT3 activation motif is needed for efficient T H 17 differentiation.