Ms. Y.-M. Moon and Dr. J. Lee contributed equally to this work.
Gene Associated With Retinoid-Interferon-Induced Mortality 19 Attenuates Murine Autoimmune Arthritis by Regulation of Th17 and Treg Cells
Article first published online: 25 FEB 2014
Copyright © 2014 by the American College of Rheumatology
Arthritis & Rheumatology
Volume 66, Issue 3, pages 569–578, March 2014
How to Cite
Moon, Y.-M., Lee, J., Lee, S.-Y., Her, Y.-M., Ryu, J.-G., Kim, E.-K., Son, H.-J., Kwok, S.-K., Ju, J. H., Yang, C.-W., Park, S.-H., Kim, H.-Y. and Cho, M.-L. (2014), Gene Associated With Retinoid-Interferon-Induced Mortality 19 Attenuates Murine Autoimmune Arthritis by Regulation of Th17 and Treg Cells. Arthritis & Rheumatology, 66: 569–578. doi: 10.1002/art.38267
- Issue published online: 25 FEB 2014
- Article first published online: 25 FEB 2014
- Accepted manuscript online: 18 NOV 2013 11:18AM EST
- Manuscript Accepted: 31 OCT 2013
- Manuscript Received: 30 APR 2013
- Ministry for Health and Welfare, Republic of Korea. Grant Number: Korean Health Technology R&D Project grant A092258
- Ministry of Education, Science, and Technology, Republic of Korea
- National Research Foundation of Korea. Grant Number: Basic Science Research Program grant 2012-0006135
STAT-3 is a key transcriptional factor in the interleukin-6 (IL-6)–mediated differentiation of Th17 cells. Because Th17 is believed to be a central player in rheumatoid arthritis (RA), we sought to evaluate whether an endogenous inhibitor of the STAT3 gene, GRIM-19 (gene associated with retinoid–interferon–induced mortality 19), could attenuate the progression and severity of murine collagen-induced arthritis (CIA) through suppression of Th17 cells and, reciprocally, could increase expression of Treg cells.
Overexpression of GRIM-19 was produced either by intravenous/intramuscular administration of a GRIM-19 overexpression vector in DBA1/J mice or by development of GRIM-19–transgenic (Tg) mice on a C57BL/6 background. Clinical signs were scored for arthritis severity, and mouse splenocytes, serum, and joint tissue were obtained for immunostaining and histologic analyses.
The numbers of CD4+IL-17+ cells and CD4+pSTAT3+ cells were decreased, while the numbers of CD4+CD25+Foxp3+ cells and CD4+pSTAT5+ cells were increased, in both GRIM-19 vector–transfected and GRIM-19–Tg mice. Administration of the GRIM-19 overexpression vector into mice with CIA markedly suppressed the clinical and histologic signs of arthritis in the affected joints. Similarly, when CIA was induced in GRIM-19–Tg mice, the arthritis phenotype was markedly attenuated and the expression of inflammatory cytokines (IL-1β, IL-6, tumor necrosis factor α, and IL-17) in the arthritic joints was also significantly reduced. Moreover, bone marrow–derived monocyte/macrophages obtained from GRIM-19–Tg mice showed attenuated RANKL–induced osteoclastogenesis in vitro.
GRIM-19 improved the clinical and histologic features of CIA and also inhibited osteoclast formation. These findings suggest that GRIM-19 may be a novel treatment agent for RA.
Rheumatoid arthritis (RA) is a systemic autoimmune disease characterized by hyperplasia of synovial tissue and progressive destruction of articular cartilage and bone ([1-3]). Although the exact pathogenesis of RA is not clearly understood, cell-mediated immune responses, particularly those involving T cells, are suggested to play a key role (). Before the discovery of Th17, Th1 cells were believed to drive the inflammatory process in RA. However, since subsequent evidence has revealed that the cytokine milieu of the arthritic joint is abundant in tumor necrosis factor α (TNFα) and interleukin-1β (IL-1β) and is relatively deficient in interferon-γ (IFNγ), the outdated concept of a Th1 cell–driven disease has been replaced by the paradigm that Th17 is a central player in the pathogenesis of RA ([5, 6]).
Th17 cell differentiation is largely dependent on the IL-6–mediated JAK/STAT-3 pathway (). STAT-3 is a particularly critical transcription factor in terms of its role in the function of Th17 effector T cells. In particular, it is involved in the transcription of retinoic acid receptor–related orphan nuclear receptor γt and IL-17.
In addition to its role associated with Th17, STAT-3 is involved in chemokine production and cell migration, thereby promoting recruitment of T cells to the inflamed tissue (). STAT-3 is also reportedly essential in assisting T cells to gain the capacity to help B cells (), a process that has been implicated in the pathogenesis of RA through the production of autoantibodies (). Moreover, STAT-3 is suggested to enhance osteoclastogenesis via a direct effect on osteoclast precursors or an indirect effect mediating RANKL induction (). All of these findings make STAT-3 a promising target for the treatment of RA, in that it targets 3 important cell types implicated in the pathogenesis of RA: T cells, B cells, and osteoclasts.
An approach involving the inhibition of the STAT3 gene was originally suggested as a therapy for cancer, based on evidence of the constitutive activation of STAT3 in cancer cells. However, the use of chemical inhibitors of STAT3 (e.g., AG490) in a murine model of cancer raised the safety issue of unwanted cytotoxic effects (). In contrast, endogenous STAT3 inhibitors showed comparable efficacy and better safety profiles. Similarly, compared with chemical inhibitors, endogenous STAT3 inhibitors may also be of greater use in treating RA. Among the various endogenous STAT3 inhibitors, SOCS-3 (suppressor of cytokine signaling 3) and PIAS-3 (protein inhibitor of activated STAT3), which are involved in the regulatory feedback mechanism of STAT3 activation, have been largely investigated (). Recently, GRIM-19 (gene associated with retinoid–interferon–induced mortality 19) was reported as another STAT3 inhibitor ().
GRIM-19 was originally identified as a novel tumor suppressor protein (). It is a 16-kd protein, and its transcription is induced by IFNα/IFNβ and retinoic acid or lipopolysaccharides in most cells. Several lines of evidence have demonstrated a loss of GRIM-19 protein expression in cancer cells ([15-17]). A proapoptotic effect of GRIM-19 has been reported, mediated by its direct binding to STAT3 and resulting in de-repression of the antiapoptotic effect of STAT3. Furthermore, in cancer cells, overexpression of GRIM-19 has been shown to suppress cell motility and metastasis by inhibition of adhesion kinases ().
Based on these observations of STAT3 inhibition by GRIM-19 in cancer cells, we hypothesized that GRIM-19 overexpression would suppress the differentiation of Th17 cells through the inhibition of STAT3. In addition, we sought to investigate whether this Th17 cell modulation could ameliorate the progression and severity of arthritis in a murine collagen-induced arthritis (CIA) model. The impact of GRIM-19 on osteoclasts, which is implicated in bone destruction in RA, is also addressed.
MATERIALS AND METHODS
Male DBA/1J and C57BL/6 mice (ages 6–8 weeks) (Orient, Korea) were maintained for the induction of CIA. All experimental procedures were examined and approved by the Animal Research Ethics Committee at the Catholic University of Korea (approval no. CUMC-2012-0151-01). GRIM-19–transgenic (Tg) mice on a C57BL/6 background (Macrogen) were generated by microinjection of a transgene. The presence of the transgene in the founders was confirmed by polymerase chain reaction (PCR), using genomic DNA extracted from the tail.
Induction and assessment of arthritis
To induce CIA, 0.1 ml of an emulsion containing 100 μg of bovine type II collagen (CII) and Freund's complete adjuvant (Condrex) was injected intradermally into the base of the tail as a primary immunization. Two weeks later, a booster injection of 100 μg CII dissolved and emulsifed 1:1 with Freund's incomplete adjuvant (Difco) was administered into the hind leg. Thereafter, mice with CIA were injected intravenously with 100 μg of a GRIM-19 overexpression vector in 2 ml of saline over a 10-second period, 7 days after the CII immunization. After 8 days, the same mice received an intramuscular injection of 100 μg of the GRIM-19 overexpression vector in the leg. Vectors were also administered into the thigh area and transfected via electrical stimulation (electroporation) on days 14 and 18. The severity of arthritis was then recorded using mean scores on an arthritis index, on a scale of 0–4, as reported previously ().
Mouse splenocytes were isolated through a mesh, and the red blood cells were lysed. To purify the CD4+ T cell fraction, cell suspensions were incubated with CD4-coated magnetic beads, and the cells were isolated on magnetic-activated cell-sorting separation columns (Miltenyi Biotec).
RNA preparation and real-time PCR
Total RNA was extracted using TRI Reagent (MRC) according to the manufacturer's instructions. Two micrograms of total RNA was reverse-transcribed using the Transcriptor First Strand cDNA Synthesis kit (Roche Applied Science). The levels of messenger RNA (mRNA) expression were estimated using real-time PCR with StepOnePlus (Applied Biosystems).
Cells were lysed in Halt protein lysis buffer with Halt phosphatase inhibitor (Thermo Scientific Pierce). Proteins were separated using sodium dodecyl sulfate–polyacrylamide gel electrophoresis, and then transferred to a Hybond enhanced chemiluminescence membrane (GE Healthcare) for Western blot analysis using the Snap i.d. Protein Detection System (Millipore).
Measurement of IgG subtypes
Blood was obtained from the orbital sinus of mice, and the serum was stored at −20°C until used. Anti-IgG2a antibodies were measured in the serum using a mouse IgG2a enzyme-linked immunosorbent assay (ELISA) quantitation kit (Bethyl Laboratories).
Antibodies against IL-17 and IL-10 were obtained from R&D Systems. The concentrations of IL-17 and IL-10 in cell culture supernatants were measured using sandwich ELISAs (Bethyl Laboratories), according to the manufacturer's instructions. A standard curve was drawn by plotting the optical density versus the log of the concentration of IL-17 or IL-10.
Histologic assessment and confocal microscopic analysis of immunostained joint tissue
A hind leg of each mouse was fixed with 1% formalin, decalcified in EDTA, and embedded in paraffin wax. The sections were then stained with hematoxylin and eosin (H&E), Safranin O, and tartrate-resistant acid phosphatase (TRAP). Immunostaining of the joint tissue was performed as described previously (). Differentiated T cells were centrifuged onto slides using a Cytospin III (Shandon Scientific). Images were acquired using a Zeiss microscope (LSM 510 Meta).
Intracellular staining for flow cytometry
Mouse lymph node cells or spleen cells were stimulated for 4 hours with phorbol 12-myristate 13-acetate and ionomycin (both from Sigma) in the presence of GolgiStop (BD PharMingen). The cells were then analyzed on a fluorescence-activated cell sorter (FACSCalibur; BD PharMingen). The data were analyzed with FlowJo software (TreeStar).
Evaluation of osteoclast formation
Mouse bone marrow–derived monocyte/macrophages (BMMs) were obtained for preparation of osteoclast precursor cells. Evaluation of osteoclast formation was carried out using previously described methods (). A commercial TRAP kit (Sigma-Aldrich) was used in accordance with the manufacturer's instructions, and cells were counterstained with hematoxylin. TRAP-positive cells containing at least 3 nuclei were scored as osteoclasts.
Results were calculated using GraphPad Prism 4.0 software and are presented as the mean ± SD of at least 3 experiments. P values less than 0.05 were considered statistically significant. Data were compared using either two-factor analysis of variance, with Bonferroni post hoc test, or the Mann-Whitney U test, as appropriate.
Suppression of Th17 cell differentiation by GRIM-19 via STAT3 inhibition
Expression of pSTAT3 was significantly lower in the GRIM-19 vector–transfected mouse cells than in mock vector–transfected control cells. In contrast, expression levels of STAT-3 protein were not affected by GRIM-19 (Figure 1A).
To investigate the effect of GRIM-19 on Th17/Treg cell differentiation under inflammatory conditions, the GRIM-19 overexpression vector was administered intravenously and intramuscularly to mice with CIA. Confocal immunostaining showed that the number of CD4+pSTAT3+ (Tyr705 and Ser727) cells was decreased and the number of CD4+pSTAT5+ cells was increased in the spleens of GRIM-19 vector–treated mice (Figures 1B and C). The expression of IL-17 mRNA was reduced, whereas Foxp3 mRNA expression was increased, in the mouse splenocytes following GRIM-19 treatment (Figure 1D).
Consistently, when the lymph node cells from mice with CIA were cultured in a Th17-polarizing condition and analyzed by flow cytometry, the frequency of CD4+IL-17+ cells was lower and the frequency of CD4+CD25+Foxp3+ cells was higher in GRIM-19–treated mice than in control mice (Figure 1E). Thus, GRIM-19 seemed to prevent Th17 cell differentiation by inhibiting the phosphorylation of STAT3, whereas it enhanced Treg cell differentiation by augmenting the phosphorylation of STAT5.
Significant amelioration of the severity of CIA after administration of GRIM-19.
We next investigated whether this regulatory effect of GRIM-19 on Th17 cells could suppress arthritis in the CIA model. Administration of the GRIM-19 vector to mice ameliorated the severity of arthritis, as demonstrated by a reduction in the mean arthritis scores and attenuation of the histologic features of arthritis in the affected joints (Figures 2A and C). Serum levels of total IgG2a and CII-specific IgG2a were also markedly reduced with GRIM-19 administration (Figure 2B).
The expression levels of inflammatory cytokines as well as of RANK and vascular endothelial growth factor (VEGF) in the arthritic joints were also markedly suppressed by GRIM-19 administration (Figure 2D). In addition, the number of CD25+Foxp3+ cells was increased and the number of IL-17+ cells was reduced in GRIM-19–treated mice (Figure 2E). Therefore, suppression of Th17 cells by inhibition of STAT3 seems to ameliorate the progression and severity of autoimmune arthritis in this murine CIA model.
Characterization of Tg mice expressing GRIM-19.
To investigate the in vivo effects of overexpressing GRIM-19, Tg mice were produced. GRIM-19 overexpression in the splenocytes of Tg mice was confirmed by measuring the mRNA and protein levels of GRIM-19 and evaluating the expression of GRIM-19 in mouse spleen tissue by confocal immunostaining (Figures 3A–C). The IL-6–mediated phosphorylation of STAT3 (Figure 3D), the expression levels of IL-17 in naive (CD4+CD62L+CD25−) T cells from Tg mice with the Th17-polarizing condition (Figure 3E), and the extent of cell proliferation following T cell receptor stimulation of splenocytes (Figure 3F) were all significantly lower in Tg mice compared to wild-type (WT) C57BL/6 mice. The frequency of CD4+IL-17+ T cells and the levels of IL-17 mRNA expression were reduced under the Th17-polarizing condition in the Tg mice as compared to that in WT mice (Figures 3G–I). Under the same condition of Th17 polarization, levels of Foxp3 mRNA and production of IL-10 were both increased (Figures 3I and J). Thus, overexpression of GRIM-19 reduced the phosphorylation of STAT3, and consequently, the production of IL-17 was reduced while the expression levels of Foxp3 and IL-10 were increased, suggesting that GRIM-19 has an immune-regulatory function.
Attenuation of CIA in GRIM-19–Tg mice
Next, CIA was induced in GRIM-19–Tg mice to demonstrate the inhibitory effect of GRIM-19 on autoimmune arthritis in vivo. After the induction of CIA in GRIM-19–Tg mice, only mild arthritis developed, without further progression (Figure 4A). Histologic sections of the knee joints of GRIM-19–Tg mice, which were stained with H&E, toluidine blue, and Safranin O, also showed less severe cartilage loss compared to the knee joints of WT mice (Figure 4B). Furthermore, the expression levels of inflammatory cytokines, including IL-1β, IL-6, TNFα, and IL-17, were reduced in Tg mice (Figure 4C). In addition, the expression of RANK and VEGF was decreased in the joints of Tg mice (Figure 4C). The serum level of CII-specific IgG2a was also significantly lower in Tg mice than in WT mice (Figure 4D).
Confocal immunostaining of the spleens of mice showed that the number of CD4+IL-17+ Th17 cells was decreased and the number of CD4+CD25+Foxp3+ Treg cells was increased in the spleens of Tg mice (Figure 4E). CD4+pSTAT3+ T cells (from the Tyr705 and Ser727sites) were less frequently observed in GRIM-19–Tg mice, whereas the number of CD4+pSTAT5+ T cells was increased in GRIM-19–Tg mice (Figure 4F). These findings substantiate the notion that GRIM-19 prevents inflammation and arthritis through regulation of the Th17/Treg cell balance in vivo.
Inhibition of osteoclastogenesis in vitro by GRIM-19.
Because bone resorption by osteoclasts is associated with the development of bone erosion in RA, we sought to investigate the effect of GRIM-19 on osteoclastogenesis. To determine the suppressive effect of GRIM-19, the mRNA levels of several osteoclast-associated molecules, including cathepsin K, NF-ATc1, TRAP, and matrix metalloproteinase 9 (MMP-9), in mixed joint cells of GRIM-19–treated mice were measured with real-time PCR. The mRNA levels of these molecules were significantly lower in GRIM-19–treated mice than in mock vector–treated controls (Figure 5A). To verify the inhibitory effect on osteoclastogenesis, BMM cells from WT or GRIM-19–Tg mice were differentiated into osteoclasts in the presence of macrophage colony-stimulating factor and/or RANKL. BMM cells from GRIM-19–Tg mice formed fewer TRAP+ osteoclasts compared to those from WT mice (Figures 5B and C). In addition, mRNA levels of osteoclast-associated molecules, including TRAP, carbonic anhydrase 2, MMP-9, and NF-ATc1, were significantly lower in Tg mice (Figure 5D).
In the present study, we demonstrated that the endogenous STAT3 inhibitor GRIM-19 suppressed Th17 cells by inhibiting the phosphorylation of STAT3. GRIM-19 also increased the phosphorylation of STAT5, and thereby increased the number of Treg cells. In addition, GRIM-19 directly inhibited osteoclastogenesis from BMM cells in vitro and reduced the expression of osteoclast-associated molecules in the murine arthritic joints. Notably, CIA was attenuated in GRIM-19 vector–treated or GRIM-19–Tg mice. Overall, GRIM-19 exerted an antiarthritic effect in mice with CIA, through its role in the above-described contributory mechanisms.
Although the pathogenesis of RA is not fully understood, it has been verified that IL-6–mediated Th17 cell differentiation is an essential factor (). Thus, it is clear that STAT-3, which is a critical transcription factor of Th17, is a promising target. Ju and colleagues recently reported that IL-17 and STAT3 were highly expressed in the synovium of patients with RA, which supports the rationale for targeting IL-17 and STAT3 in RA (). Intriguingly, knockdown of STAT3 was shown to increase the levels of STAT5 and Treg cells. Consistent with these findings, a condition of GRIM-19 overexpression in mice resulted in a reduction in pSTAT3 and the Th17 cell population and a simultaneous reciprocal increase in pSTAT5 and the Treg cell population. It is not fully understood why the phosphorylation of both molecules showed a reciprocal change. Whether there exists a regulatory mechanism by which STAT3 directly inhibits the activation of STAT5 needs further investigation.
In contrast to the well-delineated mode of action of these 2 molecules, the inhibitory mechanism of GRIM-19 has not been fully elucidated. Nevertheless, previous findings suggest that GRIM-19 inhibits STAT3 specifically and directly. In a study by Zhang et al (), the results suggested that GRIM-19 does not affect phosphorylation of tyrosine or serine residues and is bound directly to the transcription activation domain. The authors argued that GRIM-19 does not interfere with nuclear translocation of STAT3. In contrast, Lufei and colleagues () demonstrated that GRIM-19 suppressed the nuclear translocation of STAT3 by binding to the N-terminus of STAT3. Regardless, it is clear that GRIM-19 suppresses the function of STAT3.
In contrast to the results in the study by Zhang et al, our results showed that GRIM-19 inhibited the phosphorylation of STAT3. Phosphorylation of both Ser727 and Tyr705 was reduced in the GRIM-19 overexpression condition. Li and colleagues recently added evidence to support our findings, demonstrating that down-regulation of GRIM-19 was associated with an increased level of pSTAT3 in hepatocellular carcinoma (). Although the exact molecular mechanism has not been identified, one plausible explanation may be that the product of the STAT3 target gene (i.e., IL-17) itself results in the signal cascade that phosphorylates STAT3 (). Therefore, inhibition of STAT3 may decrease the transcription of the STAT3 target gene and subsequently reduce STAT3 phosphorylation.
Treatment strategies for RA have 2 main goals: to suppress the inflammation of the joint, and to prevent bone destruction. Our data demonstrated that inhibition of STAT3 not only suppresses inflammation related to Th17, but also alleviates the bone destruction process by inhibiting osteoclasts. Inflammation and osteoclastogenesis are actually mechanistically linked, because IL-1β, TNFα, and IL-6 are powerful inducers of osteoclast formation (). However, progressive bone destruction takes place despite the suppression of inflammation in patients with RA ([25, 26]). Thus, direct inhibition of osteoclasts by GRIM-19 endows a further advantage as a new treatment agent for RA.
We reported previously that IL-23–mediated osteoclastogenesis was dependent on NF-κB and STAT3 (). Nevertheless, the exact molecular mechanism was not addressed in that study. The STAT3 inhibitor PIAS-3 was recently reported to inhibit osteoclasts directly via the suppression of microphthalmia-associated transcription factor (MITF), leading to suppressed levels of NF-ATc1 and c-Fos independent of RANKL expression (). Consistent with this, the expression of NF-ATc1 mRNA was reduced in the mixed joint cells of GRIM-19–Tg mice with CIA in our study. However, it seems that MITF-associated suppression is a unique characteristic of PIAS-3 (). Elucidation of the molecular mechanism involved in the suppression by GRIM-19 requires further study.
STAT-3 is a latent cytoplasmic protein when dephosphorylated. However, recent studies have shown that it is also found in the mitochondrial membrane ([30, 31]). Mitochondrial STAT-3 is known to increase the activity of complexes I and II of the electron transport chain in mitochondria, facilitating ATP production. In this regard, GRIM-19 plays a critical role in importing STAT-3 into mitochondria as a chaperone protein and is colocalized with STAT-3 in the mitochondria. Phosphorylation of the Ser727 residue, which is required for full activation of STAT-3, is also essential for GRIM-19 to bind STAT-3 and import it into mitochondria (). Thus, the nuclear action of pSTAT-3 may be further hampered by this mitochondrial import of GRIM-19. The relationship between the mitochondrial electron transport chain and autoimmune arthritis, in terms of the contributory roles of GRIM-19 and STAT-3, will be our next topic of investigation.
Because GRIM-19 is involved in apoptosis and cellular growth, one can expect that Tg mice may manifest an aberrant phenotype related to the cell cycle. However, GRIM-19–Tg mice exhibited no apparent morphologic differences compared with WT mice during the experimental period. This may have been because various pathways other than those involving GRIM-19 are associated with STAT3 regulation. Because adequate expression of STAT3 is crucial to maintain life, GRIM-19 appears to suppress STAT3 activation to a level at which only pathologic inflammation can be regulated. This suggests that GRIM-19 has marked advantages as a new therapeutic agent for RA.
In conclusion, GRIM-19, an endogenous STAT3 inhibitor, improved the clinical and histologic features of CIA. GRIM-19 suppressed Th17 cell differentiation, whereas it augmented Treg cell differentiation, by inhibiting STAT3 phosphorylation and enhancing STAT5 phosphorylation. GRIM-19 also inhibited osteoclast formation in arthritic murine joints and osteoclast differentiation in vitro. These findings suggest that GRIM-19 may be a promising therapeutic agent for treatment of RA.
All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Cho had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study conception and design. Moon, J. Lee, S.-Y. Lee, Kwok, Ju, Yang, Park, H.-Y. Kim, Cho.
Acquisition of data. Moon, J. Lee, S.-Y. Lee, Her, Ryu, E.-K. Kim, Son, Ju.
Analysis and interpretation of data. Moon, J. Lee, S.-Y. Lee, Kwok, Ju, Yang, Park, Cho.