CTRP4 binds to the interleukin-6 Receptor and ameliorates autoimmune encephalomyelitis by suppressing Th17 cell differentiation

CTRP4, a secreted protein, plays an important role in protecting against sepsis and energy metabolism; however, its physiological functions in autoimmune disease remain unknown. In this study, we demonstrate that Th17 cell-associated experimental autoimmune encephalomyelitis was greatly exacerbated in Ctrp4 −/− mice because of increased Th17 cell inltration. In vitro, Ctrp4 deciency enhanced the ability of naïve CD4 + T cells to differentiate into Th17 cells. Mechanistically, CTRP4 interfered with the binding of IL-6 to its receptor IL-6R by directly binding to IL-6R and suppressed the activation of STAT3-related pathways in response to IL-6. Furthermore, treatment of induced EAE mice with recombinant CTRP4 protein ameliorated the symptom. In conclusion, our results indicate that CTRP4, as an endogenous regulator of the IL-6 receptor signaling pathway, may be a potential therapeutic intervention for multiple sclerosis.


Introduction
Multiple sclerosis is an in ammatory disorder of the central nervous system (CNS) identi ed by chronic demyelination and axonal damage. Upon activation in the peripheral lymphoid organs, the autoimmune T cells enter the CNS through the blood brain barrier and are reactivated, resulting in the in ltrated leukocytes and disseminated in ammation, demyelination and severe disease symptoms 1, 2, 3 . One of the most frequent animal models of multiple sclerosis is experimental autoimmune encephalomyelitis (EAE) 3 . Studies using EAE model suggest that the initiation of EAE is associated with peripheral priming of myelin-speci c autoimmune Th1 and Th17 cells 2 . Furthermore, dysregulated Th17 cell activity has been shown to induce the pathogenesis of multiple sclerosis and correlate with disease severity and relapse frequency 4 .Th17 cells display considerable plasticity and are categorized as pathogenic or nonpathogenic depending on the cytokine milieu. TGFβ and IL-6 drive classical Th17 cell differentiation 5,6 , while Th17 cells with pathogenic functions are poised to develop strong tissue-destructive properties upon on IL-23 and IL-1β expression, which induce the expression of the effector cytokines GM-CSF and IFNγ. These cell populations impair the blood brain barrier and promote neuronal damage 7,8 .
Interleukin-6 (IL-6) is a key cytokine to initiate the transcriptional program of Th17 cells, which is indispensable for the EAE disease process. IL-6 signaling is activated by the formation of a complex between IL-6 and different forms of IL-6R. Several modes of gp130 activation are initiated and mainly recruit STAT3, which is followed by downstream signaling activation. Classic signaling is triggered by IL-6 binding to the membrane form of IL-6R. Soluble IL-6R is cleaved from membrane-bound IL-6R by the proteases ADAM10 and ADAM17 or is produced upon translation of an alternatively spliced IL6R mRNA 9,10 . IL-6 binds to soluble IL-6R and is required for cells that do not express IL-6R on the surface, a process termed trans-signaling. A third mode termed as trans-presentation has been identi ed: dendritic-cellspeci c IL-6 and membrane IL-6R are presented to gp130-expressing T cells and induce pathogenic Th17 cells 11,12 . Interfering with the IL-6-STAT3 signaling axis appears to have effects on the disease progression of EAE in different ways 13 .
C1q/TNF-related protein 4(CTRP4), featuring two highly conserved complement C1q domains connected by a short linker, is a classical secreted protein of 329 amino acids 14 . CTRP4 played a critical role in a DSS-induced mouse colitis-associated colorectal cancer model by mediating the in ammatory response 15 . Loss of protective functions in a CTRP4-de cient sepsis model through promoting of TLR4 internalization, activation of downstream pathways and in ammatory cytokine release in macrophages 16 . There is also evidence suggesting that CTRP4 generated by neurons plays a role in energy metabolism by modulating food intake and body weight 17,18 . These results suggested that CTRP4 is an anti-in ammatory cytokine, but the precise mechanism underlying its role in the autoimmune disease is unknown.
This study was prompted by our observation that immunized Ctrp4 −/− mice immunized to induce EAE exhibited severe symptoms. We investigated this molecular mechanism starting from the nding that CTRP4 suppressed IL-6-mediated STAT3 activation by interacting with IL-6R, interfering with the binding of IL-6 to its receptor, IL-6R. In addition, we elucidated the treatment e cacy of CTRP4 in EAE. This study has important implications in the understanding of CTRP4 function and the development of MS therapy.

Experimental Mice
The generation of Ctrp4 −/− mice has been previously described 16 . To produce Il6ra-knock out mice, we crossed Il6ra / mice with EIIa Cre , in which IL-6R is depleted in all tissues. B6.SJL mice were purchased from the Jackson Laboratory. All mouse strains had a C57BL/6J background, and six-to twelve-week-old mice were used for most of experiments. Littermates with the appropriate genotypes were used as ageand sex-matched controls. All mice were bred and maintained under SPF conditions in an animals facility.
All animal experiments were conducted according to the guidelines of the Institutional Animal Care.

EAE model establishment:
The MOG 35 − 55 peptide with the amino acid sequence MEVGWYRSPFSRVVHLYRNGK was synthesized by Synpeptide. We immunized 8-to10-week-old female mice subcutaneously with 200 µg of MOG 35 − 55 emulsi ed with incomplete Freund's adjuvant (Cat#F5506, Sigma) containing 5 mg/mL heat-killed Mycobacterium tuberculosis (Cat#231141, BD Bioscience), followed by an tail vein injection of 200 ng of pertussis toxin (Cat#179B, List Biological Laboratories)on day 0. An intraperitoneal injection of 200 ng of pertussis toxin was administered on day 2. The mice were examined daily for clinical signs of disease and evaluated as previously described 19 .
Quantitative RT-PCR Total RNA samples were extracted with Trizol reagent and reverse transcribed to cDNA according to the manufacturer's insructions. Quantitative RT-PCR was performed using SYBR Green (Transgene) with a Roche LightCycler480 system.
All the primers used for real-time PCR are listed in supplemental table 1. The conditions for real-time PCR were 40 cycles at 94°C for 15 s followed by 40 cycles at 60°C for 60 s. The target gene expression levels were normalized to Gapdh mRNA levels. Western blot analysis Cell lysates were prepared in RIPA buffer containing protease inhibitor cocktail and phosphatase inhibitor cocktail, and were centrifuged at 3500 rpm for 10 min at 4°C. Proteins were subjected to PAGE with 10% gels and transferred to nitrocellulose membranes and probed with antibodies against phospho-STAT3 Immunoprecipitation was performed as we previously described 20 . Brie y, various cells were lysed in an lysis buffer (20 mM Tris-HCl, pH 7.5; 150 mM NaCl; 0.5 mM EDTA; 1.5 mM MgCl 2 ; 0.1%NP-40; 10% glycerol and protease inhibitor cocktail) at 4°C for 30 min. The lysates were centrifuged for 10 min at 13,000 × g. Protein concentrations were determined, and then equal amounts of cell lysates were immunoprecipitated with anti-IL6R, or anti-CTRP4 antibodies bound to protein G Sepharose (GE Healthcare), or anti-FLAG M2 A nity Gel (Cat#A2220,Sigma). After extensive washing ve times with lysis buffer, the immunoprecipitates were subjected to SDS-PAGE gel. Radio-ligand binding assay CTRP4 protein was puri ed from CTRP4-CHO cells as described 16  for 24 hr at 4°C with shaking, and 25%PEG was added, followed by counting on a γ counter.

ELASA
Human IL-6R protein (R&D) was precoated onto plates and maintained overnight at 4°C and then incubated with human IL-6 or CTRP4 protein in duplicate. The plates were washed twice with PBS/0.05% Tween 20. Each well in the plates was blocked with 200 µl of 1% BSA in PBS for 2 h. The plates were incubated with anti-IL-6R primary antibody. After washing the plates, detection was performed by streptavidin-HRP and tetramethylbenzidine (TMB). The plate was monitored for 10-20 min for color change and read at OD405. Supernatants from cell cultures were collected after activation under different conditions and secreted cytokines in the supernatants were measured by CBA assay (Biolegend).
Histological assay of spinal cord sections Diseased animals were perfused at the peak of disease with cold PBS followed by 4% paraformaldehyde for xation. The spinal cords were dissected and xed with 4% PFA, dehydrated and embedded in para n. 5µm-thick sections were stained with haematoxylin & eosin or Luxol fast blue. For immuno uorescence staining with anti-CD4 and anti-IL-17, sections were incubated at 55°C for 30 min for antigen retrieval. The sections were pretreated with a 0.3% solution of H 2 O 2 to block endogenous peroxidase activity. The sections were then incubated with 10% goat serum in PBS-T at room temperature for 30 min. The sections were then incubated with primary antibodies diluted in blocking buffer for 24 h at 4°C. The primary antibodies were anti-IL-17A (Invitrogen, 14-7179-80, 1:100 dilution) and anti-CD4 (Invitrogen, 14-9766-82, 1:100 dilution). The sections were labelled with biotin-tagged anti-IgG (1:200 in PBS-T; Vector Laboratories), followed by incubation with avidin-biotin complex (Vector Laboratories) and biotinylated tyramide. Antibody binding was detected with Alexa Fluor 555-labelled streptavidin against IL-17 at room temperature for 1 h. After staining with DAPI to visualize cell nuclei, the slides were mounted and analysed by uorescence microscopy. Lentiviral infection IL-6R and gp130 cDNA was cloned into a lentivirus vector TG006. For lentivirus production, HEK293T cells were transfected with 10 µg of TG006-gp130 transfer vector in 1.5mLof opti-MEM and 5 µg of psPAX2 and 5 µg of VSVG packing vector. Lentivirus particles were harvested 72 h after transfection, ltered and added to Ba/F3 cells in the presence of polybrene (8 µg/ml). After 24 h, the medium containing the viral particles was replaced by viral particle-free DMEM medium. The cells were cultured and maintained in DMEM containing 2 µg/ml puromycin.

Statistical analysis
Experimental results were analysed for signi cance using one-way ANOVA or Student's t-test assuming unequal variance, except for the EAE clinical data, which were analysed by Mann-Whitney U test on each day. Statistical analyses were performed using GraphPad Prism, version 6.0. P-values < 0.05 were considered statistically signi cant.

Results
Increased numbers of Th17 cells in the periphery of Ctrp4 −/− mice In our previous study, we already observed preferential expression of Ctrp4 in T cells, which inspired us to further analyse the function of Ctrp4 in T cell development and differentiation. To this end, we rst examined the distribution of thymocytes within the four major thymic populations in the Ctrp4 −/− mice and WT controls. We found that eight-week-old CTRP4-de cient mice had undetectable changes in the percentage of CD4 + T cells and had normal thymocyte development, suggesting a minimal role of CTRP4 in T cell development( Fig. 1A-B). Considering that the size of the peripheral T cell pool is stably maintained under homeostatic conditions and that the most of peripheral T cells are in a naïve state with CD44 lo CD62L hi , a proportion of memory-like T cells, characterized by CD44 hi CD62L lo markers, increase with age because of homeostatic proliferation induced by self-peptide/MHC ligands. Notably, a marked increase in memory-like CD4 + T cells in Ctrp4 −/− mouse and a reduction in naïve CD4 + T cells were observed, which commonly tend to indicate autoimmune in ammation (Fig. 1C). However, the ow cytometry analysis showed no differences in the proportion of memory-like CD8 + T in the spleen (Fig. 1D).Then, we examined the effector T cell subsets in the periphery of WT and Ctrp4 −/− mice.
Among the CD44 hi memory T cell population in the spleen, the IL-17A-producing Th17 cell subset was profoundly increased in the Ctrp4 −/− mice ( Changes in different CD4 + T cell subsets were further measured by Th cell signature genes in CD4 + T cells. When Ctrp4 −/− T cells were compared with WT cells, we found that the mRNA expression of Th17 cell lineage-speci c genes (Il17a, Il17f, Rorc and Cxcl3) was upregulated signi cantly ( Fig. 1F), whereas Tbx21, Gata3 and Foxp3 gene expression was unaltered (supplemental Fig. 1C). In summary, these data suggest that CTRP4 may be involved in T cells differentiation toward the Th17 cell lineage and suggest the possibility that CTRP4 plays a role in autoimmunity.

CTRP4 de ciency in mice exacerbates EAE symptom
T cell responses to self-antigens contribute to the development of many autoimmune and in ammatory diseases. To gain further insight into the pathophysiological roles of CTRP4 in the Th17 cell-mediated autoimmune disease, we generated the EAE models with CTRP4-su cient and -de cient mice by immunizing them with myelin oligodendrocyte glycoprotein (MOG 35 − 55 ) to mimic human MS. Compared with the wild-type mice, disease progression in the Ctrp4 −/− mice showed advanced onset and developed more severe clinical symptoms and more body weight loss, suggesting a role for CTRP4 in reducing EAE induction ( Fig. 2A). Haematoxylin and eosin or Luxol fast blue staining revealed that more immune cell in ltration and increased demyelination and axon degeneration in the spinal cord (Fig. 2B). To investigate the population of lymphocytes in ltrating the CNS, we isolated mononuclear cells from the CNS. The immunohistological analysis and ow cytometry showed more CD4 + T cells in neuronal tissue in the Ctrp4 −/− mice 18 days after immunization (Fig. 2C, supplemental Fig. 2A). The phenotype of Ctrp4knockout mice was characterized by an increased frequency and absolute number of CNS-in ltrating CD45 + F4/80 + macrophages during the peak phase of EAE (Fig. 2D). Additionally, the number of IL17A + and IL17A + IFNγ + double-producing CD4 + T cells, which re ects the disease severity, was increased in Ctrp4-de cient mice as compared with the number in the spinal cords from CTRP4-expression counterparts ( Fig. 2E-F, supplemental Fig. 2B). Furthermore, the in ltration of Th1 and Treg cells was not different between WT and CTRP4-de cient mice (Fig. 2E). To investigate whether the augment of CNSin ltrating CD4 + cells in the Ctrp4 −/− mice was due to increased peripheral Th17 cells or not, we further analysed the proportions of CD4 + T cells and Th17 cells in the draining lymph nodes and spleen. Increased frequencies of Th17 cells corresponded with an increased number of CNS-in ltrating Th17 cells at the peak of the EAE symptoms, suggesting that CTRP4 may in uence the number of Th17 cells not only in the CNS but also in the peripheral lymphoid tissues (supplemental Fig. 2D-G).
To determine the role of Ctrp4 in regulating the recall response of MOG 35 − 55 -speci c T cells, we stimulated T cells isolated from draining LNs at the early effector phase of EAE induction and cultured them ex vivo. Upon re-stimulation with MOG peptides, we found that MOG 35 − 55 -stimulated T cells exhibited substantially increased proliferation in the Ctrp4-de cient mice (Fig. 2G). Consistently, the production of IL-17 and IFN-γ was signi cantly increased in the Ctrp4 −/− mice, which correlated with the clinical score (Fig. 2H). To further elucidate the subsets of MOG 35 − 55 -peptide-speci c CD4 + T cells, CFSE low CD4 + T cells were gated, and an increased percentage of MOG 35 − 55 -peptide-speci c Th17 cells was observed in sample obtained from CTRP4-de cient mice (Fig. 2I).We thus speculated that the loss of CTRP4 promotes EAE disease progression primarily by regulating Th17 cells. 3. CTRP4 de ciency facilitates IL-6-driven Th17 cell differentiation We further examined whether the increased accumulation of Th17 cells in the CNS in Ctrp4 −/− mice was attributed to the different abilities of Th17 cells to migrate, expand, survive, or in ltrate in the absence of CTRP4. We rst sorted and cultured naïve CD4 + T cells obtained from Ctrp4 −/− and WT mice and cultured them under classical Th17 cell-polarizing condition and analysed their cytokine pro les. The loss of Ctrp4 resulted in a signi cant difference in the percentage of IL17A-producing cells and increased the production of IL-17A in the cultures (Fig. 3A). Consistent with this nding, CD4 + T cells lacking of Ctrp4 exhibited increased Rorc mRNA expression under Th17 cell-differentiating conditions, while Ctrp4 −/− T cells remained unchanged under Th0 cell-differentiating conditions (Fig. 3B). Moreover, IL-23 is critical for the differentiation of pathogenic Th17 cells. Exposure of naïve CD4 + T cells to the cytokine combination of IL-1β + IL-6 + IL-23 results in the acquisition of a pathogenic Th17 cell phenotype. Thus, we next examined whether CTRP4 de ciency affected the differentiation of pathogenic Th17 cells. More IL-17producing CD4 + cells from Ctrp4 −/− mice were observed under pathogenic conditions than were observed among the cells obtained from WT mice, and IL-17A was much more abundant in the Ctrp4 −/− cultures than in the control cultures (Fig. 3A). Th17-associated genes, such as Rorc, Il17a and Il17f, were signi cantly upregulated in the Ctrp4 −/− CD4 + T cells, implying that elevated IL-17 secretion can be partially attributed to the increased expression of their mRNAs. In addition, IL23R expression on T cells is essential for Th17 cell stability and pathogenicity. The mRNA expression of Il23r and Ifng was also increased in CD4 + T cells from the Ctrp4 −/− mice (Fig. 3C).
Next, we examined whether Ctrp4 de ciency affects T cell proliferative capacity. We performed ow cytometry analysis of CFSE dilution in CD4 + T cells stimulated with TCR activation or Th17 cell-polarizing cytokines. Ctrp4 −/− CD4 + T cells showed a similar proliferative capacity (Fig. 3D-E). In addition, we evaluated the apoptotic capacity of CTRP4-de cient CD4 + T cells under the same conditions, and similarly, no signi cant differences was detected (Fig. 3F-G). With regard to Th17 cell migration, Th17 cells highly express the chemokine receptors CCR6 and CCR2, which allow their recruitment to the CNS 21,22 . Our data demonstrate that there was no difference in the proportion of cells that expressed CCR2 and CCR6 after Ctrp4 was depleted, suggesting that Ctrp4 did not alter the ability of Th17 cells to tra c to sites of in ammation (supplemental Fig. 3A). Consistent with this nding, the CNS of EAE-induced Ctrp4de cient expressed levels of various chemokines that mediate immune-cell recruitment that were similar to levels expressed by WT-immunized mice (supplemental Fig. 3B). Further, the absence of Ctrp4 in Th17 cells did not lead to changes in the expression of the key activation markers CD25, CD44, CD69 or CD103 (supplemental Fig. 3C). Collectively, these ndings further support the possibility that Ctrp4 directs CD4 + T cell fate choice toward Th17 cell differentiation without affecting Th17 cell proliferation, apoptosis, migration or activation.

CTRP4 inhibits IL-6 activation by directly binding with IL-6R
Many studies have reported that IL-6 signaling is required for the differentiation of Th17 cells, and we hypothesized that CTRP4 probably interacts with IL-6R to suppress IL-6-induced Th17 cell differentiation. We used computational molecular docking experiments performed using Discovery Studio software to analyse the interaction modes on the basis of X-ray crystal structure of IL-6R obtained from the PDB database. The docking results supported our hypothesis (supplemental Fig. 4A). To test this hypothesis, we performed confocal immuno uorescence microscopy. The results revealed CTRP4 colocalization with IL-6R in both the cytoplasm and membrane of Jurkat cells (Fig. 4A). The direct interaction was further con rmed by CTRP4 and IL-6R co-immunoprecipitating from primary CD4 + T cells, and reverse immunoprecipitation experiments showed the same results (Fig. 4B). To con rm interactional binding, recombinant CTRP4 protein was labelled with an 125 I tracer. Saturation-binding assays demonstrated the direct binding between CTRP4 and IL-6R with a K D of 3.941 nM (Fig. 4C,F). We then evaluated the ability of unlabelled CTRP4, IL-6 and OSM of gp130 family to replace 125 I -labelled CTRP4 in the competitionbinding assays. Indeed, unlabelled CTRP4 and IL-6 strongly competed with radio-iodinated CTRP4 for binding with IL-6R with an IC 50 value of 77.25 nM and 5.233 nM, respectively, while OSM did not exert a competitive effect (Fig. 4D). To determine the region of IL-6R necessary for binding CTRP4, HEK293T cells were transfected with Myc-tagged CTRP4 and different FLAG-tagged domains of IL-6R. Because the D3 domain of IL-6R covers the majority of the IL-6 interface area 23,24 , the D3 domain was found to be critical for the IL-6R association with CTRP4, as expected (Fig. 4E).
While these results clearly suggest that CTRP4 interacts with IL-6R, we sought to investigate whether CTRP4 disturbed the formation of the IL-6-to-IL-6R complex for further activation. Thus, HEK293T cells transfected with pmCherry-tagged IL-6 and EGFP-tagged IL-6R, and the signi cant colocalization was observed in the cytoplasm and membrane. However, the association was decreased in the presence of CTRP4 (Fig. 4H). This results was consistent with the ELASA results. The addition of recombinant CTRP4 protein impaired the interaction of IL-6 and IL-6R in an in vitro binding assay (Fig. 4G). Similarly, coimmunoprecipitation demonstrated that CTRP4 interfered with the binding of IL-6 to IL-6R (Fig. 4I).Collectively, these results demonstrated that CTRP4 may function to suppress IL-6/IL-6R signaling pathway activation during Th17 cell differentiation. 5. CTRP4 prevents IL-6-mediated STAT3 phosphorylation by interacting with IL-6R IL6 induces Th17 cell differentiation through the activation of the STAT3 pathway 25,26 , especially the phosphorylation of STAT3 at tyrosine 705 27,28 . To delineate the molecular mechanism underlying the observed inhibitory effect of CTRP4 on Th17 cell differentiation, we assessed the signaling activity of JAK/STAT pathways. First, we determined that the expression of IL-6R and gp130 was comparable between WT and Ctrp4-de cient CD4 + T cells upon TCR stimulation (supplemental Fig. 4A). Furthermore, IL6 stimulation led to increased levels of pSTAT3 in WT cells, and we observed signi cantly higher levels of pSTAT3 at tyrosine 705 in Ctrp4 −/− CD4 + T cells stimulated for 5 and 15 min with physiological concentrations of IL6 (Fig. 5A). Consistent with the results, enhanced STAT3 phosphorylation in the Ctrp4 −/− T cells was observed by ow cytometry (Fig. 5B). Among the other signaling molecules tested, pJAK2, pERK and pAkt had higher expression than CD4 + T cells from WT mice, suggesting that the IL6-STAT3 signaling pathway is constitutively hyper-activated after the loss of CTRP4 in CD4 + T cells (Fig. 5A). To address the molecular mechanisms by which CTRP4 inhibits the development of EAE in vivo, we assessed the expression of the IL6 receptor in the context of the EAE model. Flow cytometry analysis revealed that the expression of IL6R and gp130 in the CD4 + T cells from CTRP4-de cient mice was similar to that in the CD4 + T cells from WT mice (supplemental Fig. 4B-D). Of note, the de ciency of CTRP4 in CD4 + T cells resulted in a signi cant increase in Y705-phosphorylated STAT3 in the spinal cord tissue of EAE mice compared to the level in the WT mice, while the abundance of phosphorylated STAT1 in the CD4 + T cells was not altered (Fig. 5C).
Previous studies have shown that the IL6 signaling cascade is initiated by the binding of IL6 to membrane-bound IL6R and gp130, which is called a classical signal 11 . First, we used Ba/F3 cells, which are an IL-3-dependent mouse pro-B-cell line that lacks both endogenous IL-6R and gp130 expression 29,30 , to establish cells with gp130 and IL-6R through stable retroviral expression, and then we analysed their proliferation rate in response to IL-6 signaling (supplemental Fig. 5A). IL-6 alone induced the proliferation of Ba/F3-gp130-IL-6R cells, while rhCTRP4 substantially reduced IL-6-mediated proliferation, suggesting that CTRP4 affects classical IL-6 signaling (Fig. 5D). Furthermore, we investigated the effect of rhCTRP4 on IL-6 trans-signaling. After establishing stable expression of gp130 in Ba/F3 cells, the cells proliferated in response to hyper-IL-6 but not in response to IL-6 (supplemental Fig. 5B-C). Exogenous CTRP4 suppressed hyper-IL-6-mediated Ba/F3-gp130 growth (Fig. 5E). This results was validated with MEFs obtained from IL-6R-knockout mice, which do not express membrane-bound IL-6R; therefore, IL-6 stimulation relies on trans-signaling (Fig. 5F). Recombinant CTRP4 protein treatment reduced basal pSTAT3 levels, and reversed IL-6-or hyper-IL-6-induced STAT3 activation (Fig. 5G). Together, these results suggested that CTRP4 in uences proliferation by competing with IL-6 for membrane-bound or soluble IL-6R. 6. rhCTRP4 is effective for the treatment of EAE As CTRP4 is a secreted protein, we rst veri ed whether the addition of rhCTRP4 impairs Th17 cell differentiation. The frequency of the IL17-producing cell population in culture was decreased in the presence of rhCTRP4 to the culture in vitro (Fig. 6A). The CD4 + T cells treated with rhCTRP4 under Th17 cell-differentiating conditions signi cantly reduced the Rorc and Il17a transcript levels (Fig. 6B), as also illustrated by a markedly decreased IL17A generation (Fig. 6C). To determine whether the degree of the suppressive effect correlated with the dose of rhCTRP4, we treated CD4 + T cells with various concentrations of rhCTRP4. The results indicated that higher doses of rhCTRP4 inhibited Th17 cell differentiation to a lesser extent, suggesting that rhCTRP4 impaired the differentiation of naïve CD4 + T cells towards Th17 cells in a dose-dependent manner (Fig. 6D-E). Consistent with the fact that CTRP4 had an impact on the IL-6 downstream pathway, pretreatment with rhCTRP4 abrogated STAT3 activation in naïve CD4 + T cells in response to IL6 (Fig. 6F). Additionally, the IL-6-triggered STAT3 phosphorylation increase was abolished as early as 5 min after exposure to rhCTRP4 and sustained the inhibitory effect through 60 min (Fig. 6G). Similarly, the enhanced phosphorylation of JAK2 and STAT3 by IL-6 can be attenuated in a dose-dependent manner to normal levels by adding rhCTRP4 (Fig. 6H).Taken together, these results further con rmed that the IL6/STAT3 signaling is negatively regulated by CTRP4.
Next, we wondered whether rhCTRP4 is effective in alleviating established EAE disease in Ctrp4 −/− mice. We immunized mice with the MOG peptide, followed by daily intraperitoneal injection of rhCTRP4 on the day of immunization. As illustrated in Fig. 7A, treatment with rhCTRP4 showed signi cant e cacy and resulted in reduced severity of EAE compared with that observed in control mice. The histological analysis of the spinal cord revealed that administration of rhCTRP4 was accompanied by decreased in ammation and demyelination in the affected spinal cord (Fig. 7B). The mice had a marked reduction in the number of CD4 + cells and Th17 cells in ltrating the CNS upon therapeutic rhCTRP4 administration ( Fig. 7C-D). The immunohistological analysis of pSTAT3 in the spinal cord revealed a reduced number of pSTAT3-positive cells in rhCTRP4-treated mice (Fig. 7E). Accordingly, treatment with rhCTRP4 suppressed the phosphorylation of STAT3 and JAK2 in CD4 + T cells at the peak of the disease (Fig. 7F). We also isolated T cells from primed mice and detected their reactivity towards antigen. When challenged with the MOG peptide, the T cells derived from the lymph nodes of the EAE mice treated with rhCTRP4 displayed a markedly dampened proliferative response (Fig. 7G).IL-17A and IFNγ generation was reduced in the MOGprimed T cells from rhCTRP4-injected mice (Fig. 7H). In summary, therapeutic delivery of rhCTRP4 ameliorated the clinical severity of EAE associated with reduced encephalitogenic effector T cells responses.

CTRP4 prevents EAE disease in a STAT3-dependent manner
We have shown that Th17 cells accumulate in Ctrp4 −/− mice in a manner dependent on the activation of ampli ed IL-6. We wondered whether the suppressive effect of CTRP4 on STAT3 activation contributes to the protection of the host resistant to EAE. To target STAT3 signaling, we used S3I-201,which is a smallmolecule inhibitor that interacts with the STAT3-SH2 domain to suppress STAT3 homodimer formation and STAT3 binding to DNA 31 . After immunizing Ctrp4-Knockout mice with MOG peptide, the mice were treated either with DMSO or with S3I-201. The results showed that S3I-201 administration resulted in the mitigation of the autoimmune disease compared with disease level in the control mice (Fig. 8A), as re ected by a signi cant reduction in immune cell in ltration in the spinal cord (Fig. 8B). Similar to the inhibitory effects of CTRP4 in vitro, inhibition of STAT3 has a signi cant impact on ameliorating disease caused by the depletion of CTRP4 in vivo.

MOG-reactive T cells that expand in the presence of rhCTRP4 ameliorated EAE
We next determined whether CTRP4 can in uence the pathogenicity of the autoantigen-speci c CD4 + T cell response in vivo. For this purpose, we immunized B6.SJL mice with MOG 35 − 55 emulsi ed in CFA. Ten days later, we isolated lymphoid cells from the spleen and draining lymph nodes before re-stimulating the cells with MOG peptide, IL-23 and either BSA or rhCTRP4. Compared to the CD4 + T cells stimulated with BSA, the CD4 + T cells treated with rhCTRP4 showed signi cantly reduced the frequency of MOG-speci c Th17 cells. In agreement with the ow cytometry data, IL-17A release was diminished in the culture supernatants of the CD4 + T cells treated with IL-23 and rhCTRP4 (Fig. 8C).
To further determine whether rhCTRP4 incubation affects the pathogenicity of MOG-reactive CD4 + T cells, expanded MOG-reactive CD4 + T cells were adoptively transferred into irradiated C57BL/6J mice. The majority of the recipient mice injected T cells stimulated with rhCTRP4 exhibited signi cantly milder symptoms than the mice injected with BSA-treated T cells (Fig. 8D). We observed that the numbers of host CD4 + T cells and CD4 + CD45.1 + donor T cells were both comparable in the CNS in the two groups (Fig. 8E, F).Consistent with this nding, the frequency and absolute number of IL-17-producing cells among the CD45.1 + donor T cell population stimulated with IL-23 plus rhCTRP4 were similar to those of the donor T cells stimulated with IL-23 plus BSA. Notably, given that the comparable number of CD45.1 + CD4 + IL-17A + donor T cells, we speculate that Th17 cell pathogenic phenotype was impaired.
Accordingly, the IL-17 producer were able to co-express GM-CSF and IFNγ, an ability that was linked to the encephalitogenic potential of the Th17 cells and that was reduced in the recipients of the T cells treated with exogenous rhCTRP4 (Fig. 8E, G). Collectively, these results showed that pretreatment with rhCTRP4 during the ex vivo expansion of MOG-reactive CD4 + T cells rendered the T cells more encephalitogenic and more pro cient in inducing autoimmune CNS in ammation.

Discussion
Previous studies showed that, CTRP4, as a secreted protein, conferred protection against endotoxic shock and diminished the susceptibility to the development of acute colitis and colitis-associated colorectal cancer 17 . Here, we demonstrated the immunomodulatory properties of CTRP4 in the IL-17-related autoimmune disease EAE, on which, to our knowledge, no functional reports exist to date. In the present study, the protective function of CTRP4 is highlighted by the nding that a lack of CTRP4 leads to much more severe clinical signs of EAE symptoms and histopathological severity, as well as increased frequency of IL-17A + and IFNγ + IL-17A + T cells in the CNS. In the absence of CTRP4, similarly increased levels of accumulated Th17 cells were simultaneously observed in the CNS and periphery after EAE induction. This phenotype may be explained by the possibility that Ctrp4 acts directly on naïve CD4 + T cell to drive their differentiation towards Th17 cell lineage but does not in uence the proliferation or apoptosis of Th17 cells. In line with this notion, a lack of Ctrp4 did not alter the expression of CCR6 or CCR2 in CD4 + T cells, which are known to be critical for the migration of Th17 cells into the CNS when local in ammation occurs. MOG-reactive T cells that were expanded in the presence of rhCTRP4 triggered less profound symptoms when the passive EAE induction approach was used. Hence, our ndings con rmed the role of CTRP4 functions as an anti-in ammatory factor 14,15,16,32 .
Providing a molecular explanation for why CTRP4 modulates Th17 cell differentiation, we identi ed a previously unknown interaction between CTRP4 and IL-6R. Activation of STAT3 by IL-6 requires IL-6 binding to IL-6R to form the heterodimeric IL-6-IL-6R complex that interacts with gp130.In this study, we found that de ciency of Ctrp4 did not affect the expression levels of IL-6R or gp130 on the cell surface, suggesting that this is not the mechanism by which CTRP4 affects IL-6 activity. The coimmunoprecipitation and radio-ligand binding assay results indicated that CTRP4 binds with IL-6R. In addition, we found that CTRP4 antagonized the association of IL-6 with IL-6R by interacting with IL-6R and impaired the IL-6-mediated activation of STAT3, which is an essential regulator of the lineage commitment to Th17 cells 33 . This notion is also supported by additional pieces of evidence, such as the ELASA results showing that CTRP4 abrogated the interaction of IL-6 and IL-6R. Although the binding a nity between IL-6R and CTRP4 was relatively low compared with the a nity of IL-6 for IL-6R, a higher CTRP4 concentration in the serum compared with that of IL-6 affords more opportunity for CTRP4 to bind. Although our results support the conclusion that the greatly exacerbated EAE symptoms of Ctrp4 −/− mice were due to alterations in the differentiation of CD4 + T cells, the suppressive effect of CTRP4 cannot extended to other IL-6R-expressing cells, including immune cells and nonimmune cells. Further study is needed to gain a full understanding of the role of CTRP4 in multiple cell types and the universal role of the CTRP4 and IL-6R interaction in in ammatory disease.
The biological effects of IL-6 are highly complex, as they can be mediated via multiple pathways.
Classical IL-6 signaling primarily involves gp130 to activate downstream pathways upon membranebound IL-6R stimulation. Soluble IL-6R binding IL-6 induces signaling and takes place in cells that do not express IL-6R through trans-signaling 34 . Our data reveal that CTRP4 disturbs not only the classical pathway but also trans-signaling. Previous studies have shown that IL-6 bound to mIL-6R expressed on dendritic cells is presented to gp130-expressing T cells through trans-presentation 35 . Because rhCTRP4 can suppress Th17 cell differentiation in the absence of dendritic cells, we ruled out the possibility of IL6 trans-presentation in the process. It is possible that CTRP4 is involved in the plasma-cell differentiation of B cells or T follicular helper cells given the critical role of IL-6, and further studies will be needed to investigated these pathways. STAT3 is speci cally required for the response to IL-6 and plays a crucial function in CD4 + T cells differentiation 36 . At the molecular level, STAT3 stimulates the expression of the Th17 cell lineagespecifying RORγ and RORα, which are required for Th17 development 37 . Our ndings suggest that CTRP4 suppressed the activation of STAT3 in CD4 + T cells in vitro and in vivo. The fact that severe symptoms in Ctrp4 −/− mice can be attenuated in the presence of a STAT3 inhibitors was further con rmed, suggesting that CTRP4 plays a role in EAE in a STAT3-dependent manner. Thus, CTRP4 limited naïve CD4 + T cell he differentiation toward Th17 cell lineage.
Because suppression of IL-6R is an attractive strategy for the treatment of autoimmune diseases involving autoreactive immune cells, we wanted to know whether rhCTRP4 injection is effective for the treatment of EAE. In consistent with the results from Ctrp4-knockout mice, when rhCTRP4 was administered by intraperitoneal injection, a bene cial treatment effect was observed. In summary, CTRP4 is a promising candidate of the medicine for the treatment of multiple sclerosis.  Data are shown as mean ± SEM and are from one of three independent experiments with similar results.

Figures
Statistical signi cance was determined using two-tailed unpaired Student t-test, *P < 0.05, **P < 0.001, ***P < 0.0001. (C-H)Statistical signi cance was determined using two-tailed unpaired Student t-test, *P < 0.05, **P < 0.001, ***P < 0.0001.  were added to plate-bound IL-6R. Dose-dependent binding of CTRP4 to IL-6R was detected by anti-IL-6R mAb but was not observed in addition of BSA to IL-6R-coated plates. (G)Speci city of binding of soluble rhCTRP4 to solid-phase IL-6R was performed by competitive ELASA. Increasing concentrations of rhCTRP4 were incubated with a constant concentration (300ng/mL) of IL-6 and the mixture were added to solid-phase IL-6R plates. The binding was detected by anti-IL-6R mAb. Binding is expressed as the mean OD450±S.D. of duplicate samples and is shown for one example representative of three independent experiments. (H-I) HEK293T cells were co-transfected with EGFP-IL-6R and pmCherry-IL-6 in the presence or absence of CTRP4 overexpression. The interaction of IL-6 to IL-6R were detected by laser confocal microscopy(H) and co-immunoprecipitation(I). (F,G) Data are shown as mean ± SEM and are from one of three independent experiments with similar results. Statistical signi cance was determined using two-way ANOVA with Bonferroni's post-test. *p < 0.05; **p < 0.01.

Figure 7
In vivo administration of rhCTRP4 attenuates the clinical severity (A)Active EAE was induced in C57BL/6 mice by subcutaneous immunization with MOG35-55 peptide in CFA and PTX injection.rhCTRP4 and control(BSA) were administered intraperitoneally on day onset (days 10) and every day until sacri ced.
Clinical scores are depicted. n=10/group, **P < 0.001, Data are shown as mean ± SEM and analysed by  Schematic diagram represents the role of CTRP4 in naïve CD4+ T cells, in which CTRP4 inhibits the binding of IL-6 and IL-6R by blocking downstream STAT3 pathway Th17 differentiation is dependent on IL-6 inducing activation of STAT3 transcriptional factor. STAT3 induces the expression of the "master regulator" gene of Th17 cell differentiation, RORγt, which then induces the expression of Th17 effector cytokine such as IL-17A. Loss of CTRP4 promotes IL-6-mediated IL-6R-gp130 complex formation, recruitment and phosphorylation of JAK2, with subsequent phosphorylation and nuclear localization of STAT3 to activate STAT3 transcription targets, leading to much more severe clinical disease symptom in