HtrA2 suppresses autoimmune arthritis and regulates activation of STAT3

Rheumatoid arthritis (RA) is an autoimmune disease that is related to the induction of T helper (Th)17 cells, which secrete interleukin-17, and activation of the signal transducer and activator of transcription (STAT) 3. The expression of high-temperature requirement protein A (HtrA) 2, a serine protease involved in apoptosis, was decreased in RA patients nonresponsive to drug treatment of RA. The aim of this study was to determine whether overexpression of HtrA2 has a therapeutic effect on RA. Th17 differentiation, osteoclastogenesis, and lymphocyte activation are increased in motor neuron degeneration (mnd)2 mice, which lack HtrA2 activity because of a missense mutation (Ser276Cys) in the protease domain of HtrA2. The inhibitor of HtrA2 also increased Th17 differentiation. On the other hand, HtrA2 induced cleavage of STAT3 and overexpression of HtrA2 attenuated CIA in a mouse model. HtrA2 overexpression inhibited plaque development as well as the differentiation of Th17 in ApoE−/− mice after immunization with proteoglycans to induce a hyperlipidemia-based RA animal model. The therapeutic function of HtrA2 in inflammatory diseases is linked with Th17 development and the STAT3 pathway in splenocytes. These results suggest that HtrA2 participates in immunomodulatory activity where the upregulation of HtrA2 may shed light on therapeutic approaches to RA and hyperlipidemia.

RA drug treatment. It is well documented that gene expression of HtrA2 in peripheral blood mononuclear cells of nonresponders to methotrexate (MTX) is significantly decreased compared with that in responders; therefore, HtrA2 is an attractive candidate to test in nonresponders to MTX 17 .
Motor neuron degeneration (mnd)2 mice lack HtrA2 activity because of the missense mutation Ser276Cys in the protease domain of HtrA2 18 . It has been documented that loss of HtrA2 function results in dysfunction including cell death 19 . Indeed, mnd2 mice show tendrils in the spleen and thymus, and death occurs 40 days after birth 20 . It is also widely believed that this specific gene mutation causes disproportionate inflammation and untimely death 21 .
We hypothesized that HtrA2 is associated with RA pathogenesis mediated by STAT3 and Th17. The current study was performed to identify whether HtrA2 could be a therapeutic target in RA. First, we investigated the role of HtrA2 on STAT3 expression and Th17 differentiation. Second, we analyzed the therapeutic effect of HtrA2 expression in a model of experimental autoimmune arthritis. Finally, we examined whether HtrA2 inhibits the development of plaque and Th17 differentiation in APOE knockout mice immunized with a proteoglycan to induce a hyperlipidemia-based animal model of RA.

Results
Excessive activation of T cells in mnd2 mutation. HtrA2 mnd2 is single point mutation of HtrA2 causing an untimely death 20 and dysregulated inflammation 21 . We hypothesized that HtrA2 mnd2 may be involved in an excessive immune response. HtrA2 mnd2 caused an abnormal small body size and a smaller spleen with fewer cells (Fig. 1A). On the other hand, CD4 + cells in the spleen of HtrA2 mnd2 increased significantly compared to that of wild type (WT) (Fig. 1B). We also investigated whether HtrA2 could downregulate the activation of T cells. In the mixed lymphocyte reaction (MLR), HtrA2 overexpression in LBRM cells significantly decreased T cell proliferation compared with that induced by mock vector in LBRM cells (Fig. 1C). We also found that the numbers of naïve T cells (both CD4 + and CD8 + ) were reduced but those of memory T cells were increased in spleens from The WT and HtrA2 mnd2 mutant mice CD4 + T cell populations were analyzed by flow cytometry (n = 6). (C) A T cell proliferation response assay (MLR) was conducted to measure T cell activity of LBRM transfected with HtrA2 or mock vector (n = 3). (D) CD4 + CD44 + CD62L + and CD4 + CD44 + CD62L + T cell populations in WT and HtrA2 mnd2 mutant mice were analyzed by confocal scanning microscopy (n = 3). (E) CD4 + T cell populations in spleen, lung, and liver from WT and HtrA2 mnd2 mutant mice were examined by confocal scanning microscopy (n = 3). (F) The CD4 + T cell populations in colon from WT and HtrA2 mnd2 mutant mice were measured by immunocytochemistry (n = 3). Data are mean ± standard deviation of three independent experiments (***P < 0.01).

Regulation of STAT3 production by HtrA2 expression.
To identify whether HtrA2 reduces STAT3 expression associated with inflammation 22,23 , we performed an in vitro cleavage assay using GST-fusion proteins of STAT3 and HtrA2. We found that STAT3 was degraded by HtrA2 ( Fig. 2A). However, the number of pSTAT3 705 + and pSTAT3 727 + cells significantly increased in the spleen of HtrA2 mnd2 mice compared to WT (Fig. 2B). CD4 + IL-17 + cells were also promoted in splenocytes of HtrA2 mnd2 mice compared to that in WT (Fig. 2C). In addition, the level of pSTAT3 was increased by IL-6 stimulation, but IL-12 failed to increase pSTAT1 expression (Fig. 2D).
Excess IL-17 production induced by HtrA2 deficiency. As STAT3 directly regulates IL-17 expression 6 , we analyzed IL-17 levels mediated by HtrA2. Gene expression of IL-17 in LBRM decreased significantly by overexpressing HtrA2 (Fig. 3A). IL-17A promoter activity also decreased significantly by HtrA2 overexpression (Fig. 3B). In contrast, siRNA HtrA2 increased IL-17 mRNA significantly (Fig. 3C). We also found that the HtrA2 inhibitor, UCF101, significantly decreased the level of HtrA2 mRNA but enhanced that of IL-17 ( Fig. 3D and E). In addition, CD4 + IL-17 + and CD4 + IFN-γ + cells were increased in spleen, liver, and lung from HtrA2 mnd2 compared with WT. In contrast, CD4 + IL-13 + cells were reduced in these tissues from HtrA2 mnd2 mice compared with WT ( Fig. 3F-H).  LBRM transfected with the HtrA2 or mock vector and the Il17a promoter was cultured for 72 hours under Th17 cell-polarizing conditions. Luciferase activity was examined using a dual-luciferase reporter assay system. (C) Relative IL-mRNA 17 expression was analyzed by real-time PCR using HrA2 siRNA-transfected LBRM cultured for 72 hours under Th17 cell-polarizing conditions. (D) Relative HtrA2 mRNA expression was analyzed by real-time PCR using UCF101 treated LBRM cultured for 72 hours under Th17 cell-polarizing conditions. (E) Relative IL-17 mRNA expression was analyzed by real-time PCR using UCF101 treated LBRM cultured for 72 hours under Th0 cell conditions (stimulation only with anti-CD3 and anti-CD28 without added cytokines) or Th17 cell-polarizing conditions. (F-H) CD4 + IL-17 + , CD4 + IFN-γ + , and CD4 + IL-13 + T cell populations in spleen, liver, and lung from WT and HtrA2 mnd2 mutant mice were measured by confocal scanning microscopy (n = 3). The number of cells was counted in four independent quadrants. Data are mean ± standard deviation of three independent experiments (*P < 0.05, **P < 0.03, ***P < 0.01, n = 3).
HtrA2 overexpression attenuates CIA severity reducing Th17 differentiation. To determine whether HtrA2 has therapeutic activity in RA, we measured the development of CIA in mice injected with either the HtrA2 overexpressing or mock vector once weekly from day 7 after the first immunization. HtrA2 overexpression significantly reduced CIA severity compared to that of mock (Fig. 5A). A histological analysis indicated that HtrA2 overexpression attenuated inflammation and bone and cartilage damage (Fig. 5B). The overexpression of HtrA2 decreased Th17 differentiation in the spleens of mice with CIA (Fig. 5C). Regulatory T (Treg) cell differentiation was not changed in CIA-induced mice injected with the HtrA2 overexpressing vector (Fig. 5D).
HtrA2 overexpression decreased atherosclerotic plaque burden and inflammation in hyperlipidemia-based mouse RA. Metabolic disorders are correlated with non-response to RA treatment 25 . As ApoE −/− mice generate atherosclerotic lesions repeatedly at susceptible locations, particular at branch points of the aorta 26,27 , we examined whether overexpressing HtrA2 is therapeutic and reduces aortic plaque progression in ApoE −/− mice immunized with proteoglycan to induce a hyperlipidemia-based RA animal model. Results showed that HtrA2 overexpression inhibited aortic plaque formation compared to that of mock and enbrel (Fig. 6A). HtrA2 overexpression also decreased the infiltration of immune cells and bone destruction (Fig. 6B). There was a reduction in Th17 differentiation, however, Treg differentiation remained unchanged with HtrA2 overexpression (Fig. 6D).

Discussion
HtrA2 has been investigated primarily as a serine protease with neurodegeneration and apoptosis functions 14,16 . HtrA2 has been suggested to be a candidate associated with non-response to drug treatments for RA 17 . However, minimal information is available regarding the therapeutic effect of HtrA2 in RA. Here, we investigated the therapeutic function of HtrA2 and identified its remedial process in RA.
The most remarkable finding of this study was that HtrA2 ameliorated RA by inhibiting STAT3. As STAT3 enhances IL-17 production and Th17 differentiation 5-7 , inhibiting STAT3 is a promising strategy for RA therapy. In this study, we found that HtrA2 mnd2 significantly increased pSTAT3 expression, whereas HtrA2 induced STAT3 degradation in vitro. As well, HtrA2 revealed curative activity in a CIA and hyperlipidemia-based RA. To our knowledge, this is the first study suggesting the use of HtrA2 in RA therapy by inhibiting STAT3. Furthermore, results found that immoderate T cell activation and proliferation of CD4 + T cells, which can induce uncontrolled inflammation such as an inflammatory storm, and may be involved in early death 20 .
Th17 release of IL-17 has been shown to be positively correlated with the progression of RA. As IL-17 upregulates the inflammatory response 28-30 , Th17 and IL-17 exacerbates inflammation in RA 2,31,32 . Th17 results in osteoclastogenesis and bone annihilation as RA progresses 24,33 . The release of IL-17 from Th17 cells is controlled by IL-13, which can inhibit the expression of proinflammatory cytokines 34,35 . Our data indicates that HtrA2 overexpression decreased the activation of T cells and IL-17 gene expression and promoter activity, however, the HtrA2 inhibitor and HtrA2 mnd2 increased IL-17 mRNA levels as well as the proliferation of Th17 cells. Supporting this, CD4 + IL-17 + and CD4 + IFN-γ + cells were also increased but CD4 + IL-13 + cells were reduced in several tissues from HtrA2 mnd2 mice. Additionally, osteoclastogenesis of HtrA2 mnd2 increased significantly compared to that of WT. As Th17 and IL-17 cause excessive inflammation during RA development, our results suggest a novel therapeutic function of HtrA2 as a treatment for RA.
CD4 + and CD8 + T cells are involved in maintaining homeostasis of the immune system 36 . Because naïve T cells (CD62L hi CD44 lo ) differentiate into memory T cells (CD62L lo CD44 hi ) after exposure to antigen, memory T cells have previously encountered and responded to antigen 37 . Our observation that naïve T cells were decreased whereas memory T cells were increased in HtrA2 mnd2 mice suggests that HtrA2 may play an important role in the immune response.
Metabolic disorders are correlated with being non-responsive to RA treatment 25 and may indicate similar effects to other RA drugs 13 . HtrA2 has been pointed out as a significant target for non-response to MTX 17 . Moreover, HtrA2 mRNA levels in peripheral blood mononuclear cells and CD4 + T cells of healthy individuals and patients with RA from the National Center for Biotechnology Information Gene Expression Omnibus database (GSE15573 and GSE4588) were analyzed. Here, HtrA2 gene expression increased in patients with RA compared to that in healthy controls (Supplementary Figure 1A). However, HtrA2 mRNA expression in patients with familial hypercholesterolemia, a genetic disorder of uncontrolled high cholesterol level related to primary hyperlipidemia 38 , decreased compared to that of controls (Supplementary Figure 1B). Additionally, HtrA2 gene expression levels in patients with RA within the GSE58795 database inclusive to clinical information was analyzed and found that HtrA2 expression of non-responders and moderate responders was significantly downregulated compared to that of responders (Supplementary Figure 1B). These results suggest that HtrA2 expression may be increased to protect against RA pathogenesis in a way similar to that of IL-10 39 . HtrA2 may have a remedial function of reducing plaque formation in the aorta and differentiation of Th17 in hyperlipidemia-based RA. Thus, an HtrA2 deficiency may have no response to RA associated with a metabolic disorder.  A and B) IL-17A expression in CD4 + T cells from WT and HtrA2 mnd2 mutant mice or CD4 + T cells treated with UCF101 or vehicle and incubated for 3 days under Th17 cell-polarizing conditions (stimulation only with anti-CD3 and anti-CD28 with TGF-β and IL-6) was measured by flow cytometry (n = 6). (C) Bone marrow cells from WT and HtrA2 mnd2 mutant mice were cultured with M-CSF (10 ng/ml) and RANKL (50 ng/ml). Cells were fixed and stained for TRAP, and the number of TRAP + cells was counted using a light microscope (original magnification, × 100 n = 6). (D) Relative mRNA expression of osteoclastogenic markers was measured by real-time PCR (n = 6). Data are mean ± standard deviation of three independent experiments (**P < 0.03, ***P < 0.01). Granulocyte-macrophage colony-stimulating factor (GM-CSF) is an immune modulatory cytokine that plays a key role in immune tolerance. There is evidence that GM-CSF ameliorates autoimmunity 40,41 . GM-CSF also increases differentiation of Treg cells, thereby reducing the severity of autoimmune disease 42 . However, GM-CSF can also mediate excessive inflammation and exacerbate autoimmune diseases including RA 34 . Particularly relevant to this study is the demonstration that GM-CSF deficiency protected against CIA progression 41 .
Eye-mediated immune tolerance may also be a good strategy for CIA treatment. It has been suggested that eye-induced immune tolerance to type II collagen has therapeutic benefit in an arthritis model 44 . CD8 + Tregs also perform a therapeutic function in CII-induced disease by inducing eye-induced tolerance 45 . Therefore, the therapeutic activity of HtrA2 in CIA may be related to induction of immune tolerance. However, further study is required to identify whether GM-CSF expression and eye-mediated immune tolerance are affected in CIA mice treated with HtrA2.
Diminished HtrA2 expression may be involved in the non-responsiveness to drugs for treating RA, however, there are few data available regarding the anti-arthritic effects of HtrA2. The new identified HtrA2 functions demonstrate its large role in ameliorating RA progression, which may shed new light on treating non-responders to medication. We have shown that HtrA2 suppressed Th17 differentiation and IL-17 expression by inhibiting STAT3. These preliminary data suggests that HtrA2 could be a crucial candidate for treating patients with RA including non-responders.

Materials and Methods
Animals. Male DBA1/J (Orient, Seoul, Korea) and apolipoprotein E deficient (ApoE −/− ) mice (Jackson Laboratories, Bar Harbor, ME, USA) at 6-8-weeks-old were maintained in groups of five in polycarbonate cages in a specific pathogen-free environment. They were fed standard mouse chow (Ralston Purina, Gray Summit, MO, USA) and water ad libitum. Mice harboring the mutant motor neuron degeneration (mnd) 2 on a C57BL/6 mice background were obtained from Professor Hyangshuk Rhim (Department of Medical Life Sciences, College of Medicine, The Catholic University of Korea, Seoul, Republic of Korea). Heterozygous mnd2 (HT, mnd2/+ ) males were crossed with HT females 3-6 months of age to generate littermates of all three genotypes: wild-type (WT, + /+ ), HT, and homozygous mnd2 (mnd2/mnd2). The genotypes of the mice were analyzed by the polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) method described previously 46 . Ethics statement. The Animal Care Committee of The Catholic University of Korea approved the experimental protocol. All experimental procedures were evaluated and carried out in accordance with the protocols approved by the Animal Research Ethics Committee at the Catholic University of Korea. All procedures performed followed the ethical guidelines for animal studies. Type II collagen immunization to induce arthritis. Collagen induced arthritis (CIA) was induced in DBA1/J mice (n = 10/group) by injecting 100 μ g bovine type II collagen (Chondrex Inc., Redmond, WA, USA) dissolved overnight in 0.1 N acetic acid (4 mg/ml) in complete or incomplete Freund's adjuvant (Chondrex) via the base of the tail. The immunization was performed intradermally. Clinical scoring of arthritis. Peripheral joints were examined for arthritis visually twice weekly. Arthritis was graded based on the method of Williams et al. 47 with the following five grades: 1) Grade 0: no evidence of erythema or swelling; 2) grade 1: erythema and mild swelling confined to the mid-foot (tarsals) or ankle joint; 3) grade 2: erythema and mild swelling extending from the ankle to the mid-foot; 4) grade 3: erythema and moderate swelling extending from the ankle to the metatarsal joints; and 5) grade 4: erythema and severe swelling encompassing the ankle, foot, and digits. The final value represented the average index from all four legs recorded by two independent observers.

Induction of hyperlipidemia-based RA and preparation of atherosclerotic lesions. ApoE −/−
mice were maintained in a temperature controlled (23 ± 2 °C) room. ApoE −/− strain mice were injected with proteoglycan. The ApoE −/− mice were fed a western type diet (Teklad Adjusted Calories Western-type diet), containing 21% fat by weight (0.15% by weight cholesterol and 19.5% by weight casein without sodium cholate) for 16 weeks. The aorta was uncovered longitudinally from the aortic root to the iliac bifurcation. The stained aorta tissues were kept in fixation solution until images were captured.

Injection of agents.
Eight days after immunization, mice with autoimmune arthritis were injected intravenously with 100 μ g HtrA2 or mock vector in 2 ml saline over a 10-sec period. The same mice received an intramuscular injection of 50 μ g HtrA2 or mock vector in the leg using electrical stimulation (electroporation). Enbrel (Pfizer, Berlin, Germany) was injected into ApoE −/− mice. The injection was conducted subcutaneously.
HtrA2 siRNA transfection. Small interfering RNA (siRNA) constructs for HtrA2 siRNA and non-targeting siRNA (Dharmacon, Lafayette, CO, USA) were obtained using siGENOME SMARTpool reagents (Dharmacon). siRNA was transfected using the Amaxa 4D-nucleofector X unit and the DN-100 program according to the manufacturer's recommendations (Lonza, Cologne, Germany). Real-time quantitative polymerase chain reaction (PCR). Total RNA was extracted using TRIzol reagent (Molecular Research Center. Cincinnati, OH, USA). The concentration of RNA in each sample was measured using a NanoDrop ND-1000 (Thermo Fisher Scientific, MA, USA). Total RNA (2μ g) was reverse transcribed into cDNA using the Transcriptor First Strand cDNA Synthesis Kit (Roche Applied Science). Gene expression were estimated using a LightCycler 2.0 instrument (Roche Diagnostic, Mannheim, Germany) and ver. 4.0 software. All reactions were performed with LightCycler FastStart DNA Master SYBR Green I (Takara, Shiga, Japan) following the manufacturer's instructions. Relative mRNA levels were normalized to that of β -actin. The primer sequences used to amplify the mouse genes are listed in Supplementary Table 1.

Intracellular
Protein purification. The pGST-STAT3 and pGST-HtrA2 plasmids were grown overnight in 20 ml LB containing 100 μ g/ml ampicillin (LBAmp). The cultures were transferred to a fresh LBAmp medium and grown for an additional 90 min at 37 °C, followed by inducing the GST-fusion proteins with 0.5 mM isopropyl-β -D-1 thiogalactopyranoside for 2 additional hours. Cells were harvested and resuspended in 1 ml EBC buffer [50 mM Tris-HCl (pH 8.0), 120 mM NaCl, and 0.5% NP40] containing 100 μ g/ml lysozyme. The cell suspension was lysed by sonication for 30 sec on ice and the lysates were collected by centrifugation at 12,000 rpm for 1 min. The GST-fusion proteins were purified from the cell lysates under non-denaturing conditions by selective binding to glutathione-Sepharose beads for 20 min at room temperature. Protein quantity and purity were analyzed by comparing with bovine serum albumin of known concentrations by sodium dodecyl sulfate-polyacrylamide gel electrophoreses (SDS-PAGE), followed by Coomassie Brilliant Blue staining (0.1% Brilliant Blue R, 45% methanol, and 10% acetic acid).
Scientific RepoRts | 6:39393 | DOI: 10.1038/srep39393 In vitro cleavage assay. GST-HtrA2 (0.2 μ M) was incubated with GST-STAT3 (1 μ M) in 100 μ l of cleavage buffer [50 mM Tris-HCl (pH 7.5), 1 mM DTT] for 16 hr at 37 °C for the in vitro cleavage reaction. The proteins bound to the beads and the supernatant were separated by centrifugation at 12,000 rpm for 5 min and resolved by 15% SDS-PAGE and analyzed by IB assays. Statistical analysis. All data are expressed as mean ± standard deviation. The statistical analysis was performed using SPSS 10.0 for Windows (IBM Corp., Armonk, NY, USA). Numerical data of the groups were compared using one-way analysis of variance (ANOVA) and the nonparametric Mann-Whitney test. Differences in the mean values of various groups were analyzed by ANOVA with a post-hoc test. P-values < 0.05 were statistically significant.