4‐Phenylbutyric acid, a potent endoplasmic reticulum stress inhibitor, attenuates the severity of collagen‐induced arthritis in mice via inhibition of proliferation and inflammatory responses of synovial fibroblasts

Abstract

4‐Phenylbutyric acid (4‐PBA) exerts potent pharmacological effects, including anti‐inflammatory properties, via inhibition of endoplasmic reticulum (ER) stress. However, it is not known whether 4‐PBA attenuates the severity of rheumatoid arthritis. The present study aimed to determine whether the inhibition of ER stress by 4‐PBA ameliorated experimentally induced arthritis. The proliferation of synovial fibroblasts (SFs) and expression of matrix metalloproteinases (MMPs) were evaluated in the presence of interleukin (IL)‐1β with or without 4‐PBA. The effect of 4‐PBA on the phosphorylation of Mitogen‐activated protein kinase (MAPK) and the activation of Nuclear factor‐κB (NF‐κB) in IL‐1β‐stimulated SFs was assessed. In an in vivo study, the effects of 4‐PBA were investigated using DBA/1 mice with collagen‐induced arthritis (CIA). Clinical, histological, and serological assessments of CIA treated with 4‐PBA were performed to determine the therapeutic effect of 4‐PBA. In vitro, 4‐PBA inhibited the proliferation and expression of IL‐1β‐stimulated SFs and MMP‐1 and MMP‐3 through the suppression of both the phosphorylation of MAPKs and NF‐κB in IL‐1β‐stimulated SFs. The 4‐PBA treatment markedly attenuated the severity of arthritis in CIA mice. The 4‐PBA treatment ameliorated joint swelling and the degree of bone erosion and destruction and decreased the level of inflammatory cytokines and MMP‐3 and Cox‐2. Furthermore, remarkable improvements in histopathological findings occurred in 4‐PBA‐treated mice. These findings suggested that 4‐PBA could attenuate the severity of arthritis in CIA mice by partially blocking the phosphorylation of MAPKs and the activation of NF‐κB in SFs. Thus, through the inhibition of ER stress, 4‐PBA may be a potent agent for the treatment of RA.

INTRODUCTION

The endoplasmic reticulum (ER), the largest cellular organelle, serves many pivotal roles in cellular activity and survival, including protein synthesis and folding, transport and storage of calcium, and lipid synthesis. Thus, maintaining homeostasis of the ER is closely related to cell survival. Impairment of ER function by intra- or extracellular pathological conditions can result in an accumulation of unfolded or misfolded proteins in the ER—a condition known as ER stress. ER stress elicits activation of unfolded proteins and dissociation of the 78-kDa glucose-regulated protein (GRP78; a chaperone protein controlling activation of the ER-transmembrane signaling molecule) from the three transmembrane sensors in the ER, which are inositol-requiring enzyme 1 (IRE1), activating transcription factor 6 (ATF6), and PKR-like ER protein kinase (PERK). Activation of these ER-transmembrane signaling molecules leads to ER-associated degradation activation, which results in proteasome-dependent degradation of misfolded proteins if the unfolded protein response (UPR) signaling pathway is successful. However, if the UPR signaling pathway is unsuccessful, cell execution is initiated, and previous studies suggested that CHOP, a member of the CCAAT/enhancer-binding protein (C/EBP) family, is associated with ER stress-induced apoptosis.^1,2

Several recent studies have implicated ER stress in the pathogenesis of inflammatory diseases including lipopolysaccharide (LPS)-induced lung inflammation model, high-fat diet-induced insulin resistance model, or chronic graft-versus-host disease.^1–4 Activation of ER stress was documented in these inflammatory models, and reduction of ER stress with 4-phenylbutyric acid (4-PBA) ameliorated the severity of systemic inflammation in these various inflammatory conditions.

Rheumatoid arthritis (RA) is a chronic autoimmune disease associated with debilitating joint inflammation and progressive joint disability. The key pathological hallmark of RA is synovial inflammation and proliferation.⁵ At the primary inflammatory site, various cell lineages infiltrate into the synovium; in particular, the synovial lining fibroblasts display unusually aggressive features. Rheumatoid arthritis synovial fibroblasts (RASFSs) exhibit tumor-like properties, such as increased proliferation and survival, and release proinflammatory mediators and matrix metalloproteinase (MMP), which perpetuate joint inflammation and cartilage destruction.^5,6 Furthermore, RASFSs facilitate osteoclastogenesis by releasing receptor activator of nuclear factor-κB ligand (RANKL).⁷ In the synovial tissue of RA patients, the markers of ER stress were more highly expressed than those of osteoarthritis.⁸ Furthermore, overexpression of GRP78, which is the central regulator of ER homeostasis, suppressed apoptosis of RASFSs, while its downregulation had the opposite effect.⁹ The apparent presence of ER stress in RASFSs and anti-inflammatory properties of 4-PBA prompted us to speculate that inhibition of ER stress with 4-PBA may have a therapeutic effect on RA.

Thus, we investigated the effect of ER stress inhibition by 4-PBA on interleukin (IL)-1β-induced proliferation and inflammatory response of RASFSs to determine whether 4-PBA can ameliorate experimentally induced arthritis. Furthermore, we assessed the potential mechanisms underlying the therapeutic efficacy of 4-PBA in RA.

MATERIALS AND METHODS Reagents and antibodies

The 4-PBA (Sigma-Aldrich, St. Louis, MO, USA) was dissolved in water, a 1-M stock solution. Primary antibodies against the following proteins were used: ERK 1/2 (1:1000; Cell Signaling Technology; cat. no. #9102), p-ERK 1/2 (1:1000; Cell Signaling Technology; cat. no. #9101), JNK (1:1000; Cell Signaling Technology; cat. no. #9252), p-JNK (1:1000; Cell Signaling Technology; cat. no. #9251), p38 (1:1000; Cell Signaling Technology; cat. no. #9212), p-p38 (1:1000; Cell Signaling Technology; cat. no. #9211), NF-κB (p65) (1:1000; Cell Signaling Technology; cat. no. #94764), p-NF-κB (p65) (1:1000; Cell Signaling Technology; cat. no. #3033), IkBα (1:1000; Cell Signaling Technology; cat. no. #9242), AKT (1:1000; Cell Signaling Technology; cat. no. #9272), COX-2 (1:1000; Cell Signaling Technology; cat. no. #4842), β-actin (1:1000; Cell Signaling Technology; cat. no. #4967), MMP-1 (1:2000; Enzo Life sciences; cat. no. ADI-905-472), TIMP-1 (1:2000; Enzo Life sciences; cat. no. ADI-905-584), MMP-2 (1:1000; Bioworld; cat. no. BS1236), MMP-3 (1:1000; Bioworld; cat. no. BS1238), MMP-13 (1:1000; Bioworld; cat. no. BS1231), TIMP-2 (1:1000; Bioworld; cat. no. BS1366), GRP78 (1:1000; Bioworld; cat. no. BS1154), CHOP (1:1000; Bioworld; cat. no. BS1136), and ATF6α (1:1000; Santa Cruz Biotechnology; cat. no. sc-22799).

Isolation and culture of RASFSs

RASFSs were obtained at the time of total knee arthroplasty from patients who fulfilled the American College of Rheumatology criteria for RA, as previously described.¹⁰ SFs (fourth to sixth passage) were used for the experiments, and morphologically, the homogenous shape was confirmed under inverted microscope. We ensured the purity of the SF culture using flow cytometry with positive expression of phycoerythrin (PE)-conjugated anti-CD90 antibody and negative expression of fluorescein isothiocyanate PE-conjugated anti-CD14, anti-CD3, and anti-CD19 antibodies. All patients provided informed consent for the use of their tissue for research purposes, and the study was approved by the Jeonbuk National University Hospital Ethical Committee (CBNU 2015-04-025).

Cell proliferation and viability assay

RASFSs (1 × 10⁵ cells/well) were seeded in a 96-well plate. After 3 h of attachment, the cells were incubated for 24 or 48 h with various concentrations of 4-PBA and/or IL-1β (1 ng/ml). Cell proliferation was assessed using Cell Counting Kit-8 (CCK-8; Enzo Life Sciences, NY, USA).

To assess the cytotoxicity of 4-PBA, various concentrations of 4-PBA (1–80 mM) were added to the RASFS culture (5 × 10³ cells/well) containing 100 μl of Dulbecco's modified Eagle medium with 10% fetal bovine serum at 37°C in a 5% CO₂ incubator after the 3-h attachment period. After 24 h of incubation, cell viability was measured using CCK-8.

Induction of collagen-induced arthritis (CIA)

Male DBA1/J mice (7–9 weeks old) were purchased from Japan SLC Inc. The mice were randomly divided into three groups: control (DBA/1 mice without CIA induction), CIA mice treated with vehicle (phosphate buffered saline [PBS]), and CIA mice treated with 4-PBA (15 mice/group). Mice were maintained in an environmentally controlled room (temperature, 22 ± 2°C; humidity, 55% ± 5%) with 12-h light/dark cycles. Water and food were freely available to the mice. Male DBA1/J mice were immunized with 100 μg of Bovine Type II Collagen (Chondrex, Redmond, WA, USA) emulsified in 2 mg/ml of Complete Freund's adjuvant (Chondrex) on day 0 and day 21, by intradermal tail injection. Intraperitoneal injection of 4-PBA (240 mg kg⁻¹ day⁻¹) or PBS was administered daily for 3 weeks, after the second immunization.

The incidence and severity of arthritis was assessed twice per week by monitoring the clinical appearance of erythema and swelling of the front and hind paws. A semiquantitative scoring system using an electrical caliper was adopted, as described previously.¹¹ The mice were euthanized 21 days after the second immunization. All mice were euthanized following the guidelines from the American Veterinary Medical Association on Euthanasia using sodium pentobarbital after the experiment.

The study was approved by the Institutional Animal Care and Use Committee of Jeonbuk National University (Approval No. CBNU 2018-020), and the animal care and experimental procedure were performed according to the Guide for the Care and Use of Laboratory Animals published by the National Institutes of Health.

Micro–computed tomography (CT) imaging

Foot joints were scanned using a SkyScan 1076 micro-CT instrument (SkyScan, Kontich, Belgium) to assess structural bony changes. The images of CIA mice were reconstructed into three-dimensional volume with an effective voxel size of 18 μm and analyzed for structural parameters, such as bone volume fraction (BV/TV), using NRecon software and CT-Analyzer (version 1.10.0.1; SkyScan).

Histological assessment

The murine paws were fixed in 10% formalin, dehydrated in ethanol, and embedded in paraffin. The paraffin blocks were cut into 5-μm sections and stained with hematoxylin and eosin, Safranin O, or tartrate-resistant acid phosphatase (TRAP). The severity of synovial inflammation, cartilage damage, and erosion was scored as previously described.^12,13

Immunoblotting

Western blot analysis was performed as previously described,¹⁴ with anti-β-actin antibody to assess equal loading. The protein levels of ERK 1/2, p-ERK 1/2, JNK, p-JNK, p38, p-p38, NF-κB (p65), p-NF-κB (p65), IκBα, AKT, COX-2, β-actin, MMP-1, TIMP-1, MMP-2, MMP-3, MMP-13, TIMP-2, GRP78, CHOP, and ATF6α were detected by western blot analysis with the corresponding antibodies.

Cells and tissues were lysed in lysis buffer (20 mM HEPES, pH 7.2, 1% Triton X-100, 150 mM NaCl, 0.1 mM phenylmethylsulfonyl fluoride, 1 mM EDTA, and 1 μg/ml aprotinin). The protein concentration was measured using the Bradford assay. Proteins separation was performed by 9%–15% sodium dodecyl sulphate–polyacrylamide gel electrophoresis (SDS-PAGE); proteins were subsequently transferred to a polyvynylidene difluoride membrane (GE Healthcare Life Sciences, Pittsburgh, PA, USA). After blocking with 5% skimmed milk, the membranes were probed with primary antibodies overnight at 4°C and incubated with horseradish peroxidase-conjugated secondary antibodies (goat anti-rabbit; 1:3000; Enzo Life Sciences; cat. no. ADI-SAB-300). Reactive proteins were visualized using enhanced chemiluminescence (GE Healthcare Life Sciences, Pittsburgh, PA, USA) using Fusion Fx7 (Vilber Lourmat, France).

Preparation of extract of the ankle joints

Ankle joint samples were collected after euthanasia, soft tissue was removed, and the remaining joint samples were immediately frozen in liquid nitrogen. Using a blender at full speed at 4°C for 30 s, the tissue samples (30 μg) were homogenized in 600 μl of tissue lysis buffer (1:100; Thermo Fisher Scientific, Rockford, IL, USA) containing a protease inhibitor cocktail (Thermo Fisher Scientific). After storage at 4°C for 30 min, the homogenate was centrifuged at 13,000 rpm and 4°C for 15 to produce protein extract. The supernatant for enzyme-linked immunosorbent assay (ELISA) was stored at −20°C.

Cytokine analysis

Mouse blood samples for cytokine analyses were collected at the time of euthanasia, 21 days after the second immunization. The concentration of IL-6 (Enzo Life Science; cat.no. ADI-900-045) and tumor necrosis factor α (TNF-α) (Enzo Life Science; cat.no. ADI-900-047) in the murine serum and ankle joint extracts was measured using an ELISA kit (Enzo Life Sciences, NY, USA) in accordance with the manufacturers' instructions.

Statistical analysis

Student's t test or analysis of variance with a Tukey's post hoc test was used to compare means. P-values <0.05 were considered statistically significant. The results were expressed as mean ± SEM.

RESULTS 4-PBA suppressed the proliferation of IL-1β-stimulated RASFSs

We first investigated the effect of increasing 4-PBA concentration on proliferation and viability of IL-1β-stimulated RASFSs (n = 3–5). Cell proliferation was significantly inhibited at 4-PBA concentrations higher than 20 mM for 24 h (Figure 1(A)) and 4 mM for 48 h (Figure 1(B)) of incubation without IL-1β, respectively. To assess whether the suppressive effect resulted from the toxicity of 4-PBA, we assessed RASFS viability after 24-h treatment with increasing concentrations of 4-PBA without IL-1β. Doses up to 40 mM did not cause significant cell death (Figure 1(C)). Thus, in the following experiments, we treated RASFSs with 20 mM of 4-PBA (the concentration showing maximum inhibitory effect on RASFS proliferation without cytotoxicity) (Figure 1(D)).

4-PBA caused a decrease in the expression of ER stress markers in RASFSs

A previous study suggested that the synovium of RA patients has higher levels of ER stress compared with that of Osteoarthritis (OA) patients.⁹ To determine whether the RASFSs stimulated with IL-1β, which are in similar condition with the inflamed synovium, express the elevated level of ER stress and the effect of 4-PBA on the expression of ER stress markers of the RASFSs, RASFSs (n = 3–5) were treated with 4-PBA in the presence or absence of IL-1β.

As shown in Figure 2(A), IL-1β-stimulated RASFSs showed significantly higher expression of GRP78 and downstream signaling markers including ATF6, AKT, and CHOP compared with unstimulated cells. As expected, 4-PBA attenuated the enhanced expression of GRP78 and CHOP (Figure 2(A) and (B)), whereas the expression of ATF6 and AKT did not differ significantly following 4-PBA treatment. These results verified increased ER stress in stimulated RASFSs and the inhibitory effect of 4-PBA in attenuation ER stress, at least in part.

4-PBA downregulated MMP-1, MMP-3, and MAPK expression in IL-1β-stimulated RASFSs

MMPs, which are released by RASFSs in response to the increased level of IL-1β, play a major role in cartilage degradation of synovial joints.¹⁵ Thus, we assessed the expression level of proteins including COX-2, MMPs, and TIMPs in RASFSs (n = 3–5) with or without IL-1β stimulation. We found that IL-1β increased the expression of MMP-1, MMP-2, and MMP-3 and that 4-PBA decreased the IL-1β-induced enhancement of MMP-1 and MMP-3 expressions (Figure 3(A)). Expression of COX-2, which is an enzyme that induces prostaglandin production and is closely associated with the inflammatory cascade, was not significantly different between the 4-PBA and vehicle-treated cells. Activation of Mitogen-activated protein kinase (MAPK) and NF-κB signaling pathways in RASFSs was also assessed. 4-PBA inhibited the phosphorylation of ERK, p38, and JNK in the MAPK pathway, as well as inhibiting the phosphorylation of NF-κB p65 following IL-1β stimulation (Figure 3(B)). Taken together, these data suggest that 4-PBA treatment decreases the production of MMP-1 and MMP-3 by RASFSs stimulated with IL-1β and inhibits the phosphorylation of NF-κB, which induces the expression of various proinflammatory genes.

4-PBA attenuated the severity of arthritis in CIA murine model

First, we assessed the expression of ER stress markers from ankle joints extracts to find out whether 4-PBA decreases the level of ER stress in vivo as well as in vitro. The mice were divided into three groups: control (DBA/1 mice without CIA induction), CIA mice treated with vehicle (PBS), and CIA mice treated with 4-PBA (15 mice/group). Western blot analysis revealed that elevated levels of GRP78, ATF6, AKT, and CHOP from CIA mice, but 4-PBA-treated CIA mice showed significantly decreased expressions of GRP78 and ATF6 (Figure 4(A) and(B)). Although these results were not statistically significant, the expression of AKT and CHOP from 4-PBA-treated mice showed a decreased trend compared with that of CIA mice. These results demonstrate that the CIA induction leads to the elevation of ER stress levels and 4-PBA effectively reduces the enhanced ER stress in CIA mice.

To extend our findings in vivo, we assessed whether 4-PBA could dampen the inflammatory responses in a CIA murine model. Mice treated with 4-PBA had significantly lower paw thickness score and clinical score compared with vehicle-treated mice (Figure 5(A)). We found that 4-PBA treatment was able to delay the onset of arthritis compared with vehicle treatment (Figure 5(B)). RASFSs release RANKL, which promotes osteoclastogenesis and bony erosion in RA. Micro-CT analysis of metatarsophalangeal (MTP) joints, metatarsal bone, and tarsal bone was performed to analyze the effect of 4-PBA on bone erosion and joint destruction. CIA mice showed the narrowing of the MTP and tarsometatarsal joint space and erosion of the metatarsal and tarsal bone (Figure 5(B)). The erosive, destructive joint damage was markedly attenuated in the 4-PBA-treated mice, as shown by the analysis of BV/TV (Figure 5(B)).

4-PBA reduced the production of IL-6 and TNF-α and lowered the expression of MMP-3 and COX-2 in the CIA murine model

According to the aforementioned data, 4-PBA could decrease the level of MMPs in RASFSs and ameliorates the degree of arthritis. Thus, we sought to investigate the effects of 4-PBA on the inflammatory cytokine production in the murine serum and joint extract of the mice. We measured the concentration of IL-6 and TNF-α, which are the major inflammatory cytokines, using ELISA and found a reduced production of IL-6 and TNF-α in the joint extract and serum of 4-PBA-treated CIA mice compared with those of vehicle-treated mice, respectively (15 mice/group) (Figure 6(A)). Moreover, western blot of the joint extract showed that the expression of MMP-3 and COX-2 was significantly lower in the 4-PBA-treated group compared with that in vehicle-treated group (Figure 6(B)). The expression of MMP-1, MMP-2, and MMP-13 showed no statistical differences between the two groups. Collectively, these results suggest that 4-PBA treatment has a potential for alleviation of arthritis in CIA mice in part via the suppression of inflammatory responses.

4-PBA reduced bone destruction and inflammatory infiltration in CIA mice

To further validate these in vivo results, we assessed histopathological evaluation of tibiotalar joint sections of 4-PBA-treated mice compared with that of the vehicle-treated group (15 mice/group). It showed the inflammatory cells infiltration and bone and cartilage destruction, with an increased number of osteoclasts (arrows, Figure 7(A)). The joint inflammation in CIA mice was remarkably alleviated by 4-PBA treatment (Figure 7(B)), with reduced cartilage destruction and erosion, as well as a decreased number of osteoclasts, compared with that of the vehicle-treated group. Based on the aforementioned experiments, these results indicate that ER stress is involved in the inflammatory cascade of RASFSs at least in part and ER stress inhibition using 4-PBA attenuates synovial inflammation and arthritis in CIA mice.

DISCUSSION

The 4-PBA, a low-molecular-weight chemical chaperon, attenuates ER stress via increasing protein folding capacity and suppressing the UPR activation in the ER.^16,17 Many studies implicated ER stress in the pathogenesis of various diseases and demonstrated the therapeutic potential of 4-PBA by using it as a reducer of ER stress.^3,18–20 In the present study, we showed that the expression of ER stress markers was significantly enhanced in IL-1β-stimulated RASFSs, which is consistent with previous reports^9,21 and indicates that ER stress is implicated in the pathogenesis of RA. RASFSs treated with 4-PBA had significantly lower levels of ER stress indicators, which means that 4-PBA reduces ER stress in these cells.

We observed that the expression level of GRP78, a key indicator of ER stress, was markedly decreased by 4-PBA treatment both in vitro (Figure 2) and in vivo (Figure 4), confirming that 4-PBA serves as an ER stress modulator in SFs and CIA mice, which was in good agreement with our expectations. It is worth mentioning that 4-PBA treatment resulted in reduced ER stress in this inflammatory arthritis model through its actions as a chemical chaperone, thus providing evidence that amelioration of arthritis in CIA mice resulted from the inhibition of ER stress.

However, the level of CHOP expression was decreased by 4-PBA treatment in SFs, but not in CIA joint extracts. Possible explanations for the differences in CHOP expression are outlined in the following: (1) different sampling times of SFs and mouse joint extracts as a consequence of the experimental circumstances, that is, in vivo and in vitro sampling; and (2) differences in the nature of the experimental specimens, that is, collection of cells or tissues.

Multiple studies have reported that ER stress can activate inflammatory responses in local or systemic disease. IL-6, a key inflammatory cytokine in RA pathophysiology,²² is abundant in the serum and synovial fluid of RA patients, and its levels are positively correlated with disease activity. Furthermore, IL-6 has been implicated in joint erosion, and its expression correlates with that of proMMP-3, a proteinase closely associated with articular cartilage destruction. Blocking IL-6 with a humanized anti-IL-6R monoclonal antibody, tocilizumab, is currently used as a treatment modality of RA.²³ The crosstalk between ER stress and IL-6 release has been reported previously.²⁴ Krupkova et al. showed that IL-6 release was increased following ER stress activation and that the crosstalk between ER stress and IL-6 induction was enhanced by phosphorylation of p38 MAPK in vivo.²⁴ In agreement with previous findings, CIA mice treated with an ER stress blocker showed lower levels of IL-6 and decreased arthritis severity compared with the vehicle-treated mice. Moreover, decreased levels of MMP-3, which is positively correlated with that of IL-6, were observed in the CIA mice treated with 4-PBA. These findings suggest that ER stress inhibition plays a role in the modulation of inflammatory responses, including IL-6 production and contributes to the alleviation of arthritis in CIA mice.

Lower levels of TNF-α and COX-2 in the serum and joint extract of 4-PBA-treated CIA mice, respectively, also reflect the potential therapeutic efficacy of 4-PBA. Enhanced COX-2 expression following ER stress activation has been reported in various studies. Zhang et al. reported that the transcriptional and translational expressions of COX-2 in vascular smooth muscle cells were suppressed with 4-PBA, which demonstrates the therapeutic potential of ER stress blockade against vascular complications.²⁵ In a lupus nephritis model, 4-PBA treatment decreased the expression of ER stress markers in cultured podocytes, thus alleviating lupus nephritis-induced podocyte damage.²⁶ COX-2 is a critical inducer of prostaglandin E2 release under pathological conditions. Thus, nonsteroidal anti-inflammatory drugs, which are commonly used to treat the symptoms of inflammatory conditions, such as arthritis, target COX enzymes to inhibit prostanoid synthesis.²⁷ ER stress activation and COX-2 expression are closely associated, and the attempts to block ER stress were shown to successfully inhibit COX-2 expression. The inhibitory effect of 4-PBA on COX-2 expressions was confirmed in our murine arthritis model.

In the present study, 4-PBA treatment decreased the number of TRAP-positive osteoclasts and suppressed bone erosion and joint destruction in CIA mice compared with that in vehicle-treated animals. There is abundant evidence that a high RANKL/OPG ratio and an increased number of TRAP-positive osteoclasts in the synovium and synovial fluid of RA patients are implicated in the destructive nature of the disease.^28,29 Recently, the association between ER stress and RANKL-mediated osteoclastogenesis and bone erosion was revealed.^30–33 In an earlier report, we found that bone marrow macrophages treated with thapsigargin (an inducer of ER stress) showed considerably increased osteoclast differentiation and resorption pit area, while those treated with 4-PBA displayed significantly suppressed osteoclast formation and a decreased resorption pit area.³⁴ Furthermore, osteoclastogenesis was significantly reduced in osteoclast precursor cells when ER stress-activating genes, such as GRP78, IRE1, and PERK, were knocked down.³⁴ Thus, ER stress and joint destruction, including bone erosion and resorption, are closely associated. The 4-PBA demonstrated a protective role in LPS-induced inflammatory bone loss via reducing osteoclast autophagy and inhibiting NF-κB activation.³⁵ Although we demonstrated the beneficial role of 4-PBA on phenotypic features of osteoclasts only, the aforementioned evidence makes it reasonable to postulate that 4-PBA could have therapeutic potential for ameliorating bone resorption.

There are certain limitations to this study. First, we only used IL-1 for the stimulation of SFs and confirmed that 4-PBA could suppress the proliferation elicited by that. However, other proinflammatory cytokines, such as TNF-α or IFNγ, also exist in the joint cavity of RA patients, and they participate in the stimulation of SF proliferation. It has been revealed that IL-1 is detected at high levels long after the onset of RA in affected areas, whereas TNF-α is predominantly detected during the early stages of the disease. Furthermore, IL-1β is a potent proinflammatory cytokine that stimulates SFs to release MMPs, which in turn play a role in the destruction of cartilage and joint tissue. Thus, we thought IL-1 was suitable for our experiments because (1) it would allow us to assess the suppressive capacity of 4-PBA on SFs stimulated by a potent proinflammatory cytokine, and (2) we used SFs from patients with established RA that are taking medication, but not from naïve patients. However, the suppressive effect of 4-PBA on the proliferation of TNF-α- or IFNγ-stimulated SFs could potentially be different. Thus, further studies using TNF-α- or IFNγ-stimulated SFs are necessary for proving the universal suppression capacity of 4-PBA. Second, we could not exclude the unidentified systemic effects of 4-PBA in ameliorating arthritis. However, modulation of ER stress by 4-PBA in SFs and CIA mice was confirmed in the present study, and the crosstalk between ER stress and inflammation cascades, including NF-kB and MAPK signaling pathways, was revealed by several research studies.^1,5,24 Further studies are warranted for the evaluation of both local and systemic effects of 4-PBA in inflammatory diseases. Third, there are disparity between the in vitro and in vivo study of COX-2 expression. We also found these results surprising and therefore repeated the experiments; however, the results remained the same.

According to several previous studies, COX-2 activity increases in in vivo tests on mouse arthritis models, whereas in in vitro tests using SFs, COX-2 activity does not increase at the basal level but increases when treated with IL-1β.^36,37 This also confirmed that expression of COX-2 increased in vivo and in vitro when subjected to IL-1β; however, the expression differed between in vivo and in vitro experiments when treated with 4-PBA. There may be different reasons explaining why COX-2 expression was not reduced by the 4-PBA treatment in vitro. First, the 4-PBA concentration in vitro may have been insufficient for suppressing COX-2 expression, or the reduction of COX-2 was not obvious because the time was too short to induce suppression of COX-2 synthesis. In most previous studies on inflammatory arthritis, there was an increase in the activity of COX-2 in vivo, and its expression was reduced by a mechanism that inhibits COX-2. In the respective experiment of our study, we showed decreased expression of various inflammatory markers and reduced clinical symptoms. Considering the relationship between prostaglandin and IL-6 production, the decrease in IL-6 suggests decreased COX-2 expression in our experiments. However, additional experiments are needed to determine why COX-2 expression was not reduced in the 4-PBA treatment in vitro. We added this point to the section on study limitations.

In conclusion, we showed that inhibition of ER stress using 4-PBA alleviated inflammatory arthritis in CIA mice. This effect was mediated by the suppression of synovial proliferation; release of MMPs and proinflammatory cytokines; and inflammatory bone loss. Although the underlying mechanisms need to be further elucidated, our data clearly show the therapeutic potential of 4-PBA for decreasing the disease severity of RA.

ACKNOWLEDGMENTS

This research was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT, and Future Planning (Grant/Award Number: 2015006120) and by the fund of the Biomedical Research Institute, Jeonbuk National University Hospital.

CONFLICT OF INTEREST

The authors declare no conflicts of interest.