Role of Endoplasmic Reticulum Stress in Rheumatoid Arthritis Pathogenesis

Rheumatoid arthritis (RA) is characterized by a tumor-like expansion of the synovium, which is composed of proliferating synoviocytes and infiltrating leukocytes, including T cells and B cells; these are likely activated by autoantigens (1). In RA joints, various inflammatory cells, including innate immune cells (e.g., mast cells, macrophages, dendritic cells [DCs], and natural killer cells), adaptive immune cells (T- and B cells), endothelial cells, and fibroblast-like synoviocytes (FLS), are activated (1-5). In particular, interleukin (IL)-17 producing T cells (the so-called T_H17 cells) have emerged as one type of immune cell that is associated with the initiation and perpetuation of RA (6), and the modulation of IL-17 has been demonstrated to be effective for suppressing arthritis (6). These innate and adaptive immune cells interact via an array of cytokines and/or cell-to-cell contacts, which can also activate each other, leading to secretion of diverse cytochemokines, growth factors, and reactive oxygen species, which ultimately constructs persistent pro-inflammatory cascades (1-6).

The endoplasmic reticulum (ER) is the site of biosynthesis for all secreted and membrane proteins (7). The lumen of the ER is a unique environment, critical for proper folding of proteins destined for secretion or display on the cell surface (7). Homeostasis in the ER is maintained by a coordinated adaptive program, unfolded protein response (UPR) and ER-associated degradation (ERAD) (8-11). However, a variety of disturbances, including mutations that predispose proteins to misfolding in both substrate and pathway chaperones, altered cellular metabolism, and infection, can increase protein misfolding (7, 8). The accumulation of unfolded proteins in the ER leads to a condition known as ER stress (8-11). During ER stress, the glucose-regulated protein of 78 kDa (GRP78), a molecular chaperone also known as binding immunoglobulin protein (BiP), initiates a signaling cascade of UPR (9).

After initiation by GRP78, the main UPR signaling is propagated by three ER-localized protein sensors: inositol-requiring transmembrane kinase-endoribonuclease-1α (IRE1α), double-stranded RNA-dependent protein kinase-like ER kinase (PERK), and activating transcription factor 6 (ATF6) (12, 13). In the resting state, GRP78 binds the N-termini of IRE1α, PERK, and ATF6, preventing their activation. Upon activation, GRP78 binds to unfolded or misfolded proteins, and it releases IRE1α, PERK and ATF6, triggering UPR signaling. The intrinsic ribonuclease activity of IRE1α results in the production of X-box binding protein-1 (XBP-1), a transcription factor that induces the expression of genes involved in restoring protein folding or in degrading unfolded proteins (9, 14). On the other hand, PERK activates initiation factor 2α phosphorylation, preventing general protein synthesis through translation repression. However, under sustained ER stress, the cell may fail to resolve the protein-folding defect and to restore homeostasis in the ER; the resultant UPR then initiates programmed cell death, such as apoptosis and autophagy (12, 13).

The ER stress response has also been recognized in a wide range of diseases, including cancer, hypoxia, ischemia/reperfusion injury, heart disease, neurodegenerative disorders, inflammatory bowel disease, obstructive airway disease, diabetes, and infection (7, 10, 11, 15-17). In particular, defects in the ER stress response have been implicated in chronic autoimmune inflammatory diseases (8, 10). Microarray analysis using the muscle tissue of patients with myositis has revealed that the expression of GRP78 is increased (18), suggesting that the ER response is involved in skeletal muscle damage in autoimmune myositis. Additionally, GRP78 is a target of auto-reactive B and T cell responses in a murine model of anti-Ro (SS-A) autoimmunity (19). Moreover, misfolded human leukocyte antigen-B27 (HLA-B27) has been suggested to promote spondyloarthropathy through abnormal ER stress responses (20-25). During assembly with β2m and the peptide in ER, HLA-B27 heavy chain has a tendency to misfold and to form aberrant disulfide-linked dimers (24). Enhanced accumulation of misfolded heavy chains can activate the abnormal UPR, and induce pro-inflammatory responses in spondyloarthropathies (25).

Collectively, earlier studies (8-10, 14, 19-21, 23, 25) have suggested that there is cross-talk between the ER stress response and chronic autoimmune inflammation, and that ER stress can induce or modify the phenotype of inflammatory diseases. However, the role of the ER stress response in the pathogenesis of RA remains to be defined. In fact, diverse stressful conditions, including hypoxia, low glucose, and the pro-inflammatory cytokine milieu, are frequently observed in the RA joints (26), and these might act as ER stressors. In this review, we integrate the current knowledge on the possible link between the ER stress response and RA, a representative chronic inflammatory disease, focusing on the role of GRP78, and propose potential pathophysiological effects and therapeutic implications of GRP78 in RA pathogenesis.

In various tumor models, GRP78, a central regulator of the ER stress response, plays an important role in resisting stressful host microenvironments and facilitating cell survival (27-33). For example, hypoxia-induced ER stress causes enhanced GRP78 expression, which increases cellular survival and adaptation to the microenvironment in colorectal cancer (28). Surface GRP78 on prostate cancer cells can bind activated α2-M and promote cellular proliferation and survival by activation of extracellular signal-regulated kinases (ERK) 1/2, p38 mitogen-activated protein kinases, phosphatidylinositide 3-kinases/protein kinase B, and nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB) signaling cascades, as well as by elevation of the GRP78 expression itself (29-31). ER transmembrane GRP78 forms a complex with caspase-7 or caspase-12, and inhibits caspase-7 activation-induced cell death (32). GRP78 also co-localizes with Raf-1 on the outer membrane of mitochondria to maintain mitochondrial permeability and thus protect cells from ER stress-induced apoptosis (33).

The pathologic hallmark of RA is the "invasive pannus" that results from synoviocyte hyperplasia (1-4). Rheumatoid synoviocytes, which consist of FLS and synovial macrophages, exhibit invasive characteristics reminiscent of cancer cells, destroying cartilage and bone. They are responsible for many aspects of RA pathology, such as synovial proliferation, perpetuation of chronic inflammation, and joint destruction (34-36). In the inflamed RA synovium, it was demonstrated that the FLS of RA patients (RA-FLS) exhibit considerable morphologic alterations (34-37). They have abundant cytoplasm, a dense rough ER, and large pale nuclei. It has been postulated that chronic exposure of FLS to a combination of inflammatory cytokines, growth factors, and chronic hypoxia results in continued activation of FLS, exhibiting some features of tumor cells, such as anchorage-independent growth, alterations in their response to apoptotic stimuli, and migration/invasion toward articular cartilage and bone. Moreover, as seen in Fig. 1, RA-FLS were more resistant to apoptosis induced by ER stressors (e.g., tunicamycin and thapsigargin) than FLS of osteoarthritis (OA) patients. These features of RA-FLS are referred to as "tumor-like' transformation or dedifferentiation (36, 37). However, it remains unclear how RA-FLS exhibit a transformed phenotype.

Fig. 1
Rheumatoid arthritis synoviocytes are less sensitive to ER stress-induced apoptosis than osteoarthritis synoviocytes. (A and B). FLS (4×10³ cells) of RA (n = 5) and of OA patients (n = 5) were treated with tunicamycin (Tm) or thapsigargin (Tg). Cell viability was assessed by MTT (tetrazolium) assay. Data are the mean±SD of five independent experiments performed in triplicate, and are presented as percentages versus untreated cells. ^*P < 0.05 compared with OA-FLS. (A) Cell viability determined 3 hr after treatment with thapsigargin (1 or 10 µM). (B) Time-dependent response to 10 µM of thapsigargin. (C and D). ER stress-induced synoviocyte apoptosis. (C) The apoptosis of the FLS (3×10⁴ cells) of RA (n = 3) and of OA patients (n = 3) was induced by treating cells with thapsigargin (10 µM) or tunicamycin (20 µg/mL) for 1 hr. Degrees of apoptosis were assessed by cellular DNA fragmentation ELISA. Results are the means±SD of three independent experiments, and are expressed as fold increases versus basal levels. ^*P < 0.05 compared with OA-FLS. (D) APOPercentage Apoptosis Assay. FLS were treated with thapsigargin (10 µM) or tunicamycin (20 µg/mL) for 2 hr. Apoptotic cells were bright pink (left panel). Representative digital images of three independent experiments are shown. Scale bars: 100 µm. Levels of apoptosis determined using APOPercentage apoptosis assay kits are also presented as pixel numbers using Adobe Photoshop (right panel). ^*P < 0.05 compared with OA-FLS. ER, endoplasmic reticulum; FLS, fibroblast-like synoviocytes; RA, rheumatoid arthritis; OA, osteoarthritis; SD, standard deviation; ELISA, enzyme-linked immunosorbent assay.

• Click for larger image

Recently, we have provided a glimpse of evidence on how RA-FLS exhibit abnormal proliferation in inflamed joints. We were the first to demonstrate that ER stress-associated gene signatures, induced by chronic hypoxia and pro-inflammatory cytokines, are responsible for the abnormal proliferation of RA-FLS and angiogenesis, namely "pannus formation" (38). Down-regulation of GRP78, a master regulator of the ER stress response, increases apoptosis of RA-FLS. Conversely, overexpression of GRP78 prevents RA-FLS from apoptotic death induced by an ER stressor (38). Moreover, GRP78 controls synoviocyte proliferation and angiogenesis in vivo (38). Mice with Grp78 haploinsufficiency exhibit the suppression of experimentally induced arthritis, and develop a limited degree of synovial proliferation and angiogenesis (38). We suggest that the ER chaperone GRP78 is critical for synoviocyte proliferation and angiogenesis, the pathologic hallmark of RA, and may be responsible for FLS transformation. In conclusion, our findings provide new insights into the role of GRP78 in the pathogenesis of RA, and explain how normal synoviocytes develop an aggressive phenotype in RA joints.

Accumulating evidence suggests that ER stress is involved in the proinflammatory process (9). For example, the pro-inflammatory cytokines, including IL-1β and tumor necrosis factor alpha (TNF-α), have been reported to induce the ER stress response in hepatocytes, leading to the activation of CREBH, a transcription factor that stimulates the expression of proteins involved in the acute inflammatory response, such as serum amyloid P-component and C-reactive protein (69). In murine fibrosarcoma cells, TNF-α was found to trigger UPR, increasing the expressions of XBP1 and GRP78 (70). Selective abrogation of GRP78 by subtilase cytotoxin blunts activation of the pro-inflammatory NF-κB signal pathway, and protects mice from endotoxic lethality and CIA (71). We have also demonstrated that pro-inflammatory cytokines can induce GRP78 expression in RA-FLS (38). Together, the previous findings (9, 38, 69-71) indicate a link between ER stress and inflammation, suggesting that ER stress is one of the major mediators of chronic inflammation.

GRP78 expression is not limited to the ER, but is significantly identified on cell surface (72). As seen in Fig. 2, we identified that FLS expressed GRP78 on the cell surface in addition to ER, and that surface GRP78 levels were higher in RA-FLS than in OA-FLS. Interestingly, a recent report has shown that citrullinated GRP78 on the surface of monocyte/macrophage acts as a receptor for ACPA to enhance activation of the inflammatory NF-κB pathway and production of inflammatory cytokine TNF-α (65). Therefore, it is possible that ACPA may bind to GRP78 on RA-FLS or RA synovial macrophages, and then trigger cyto-chemokine production by inducing NF-κB. The resultant increase in pro-inflammatory cytokines may further induce GRP78 expression in RA-FLS and FLS proliferation (38), constructing a feed-forward cycle of rheumatoid inflammation. If this is the case, therapeutic agents targeting surface GRP78 can be effective for the selective incapacitation of invasive RA-FLS, as they were for some types of cancer (27).

Fig. 2
GRP78 is expressed in the ER and membrane of synoviocytes. (A and B). Immuno-fluorescence staining of GRP78 in FLS. RA synoviocytes were permeabilized, and stained with anti-GRP78 antibody and CellLight ER-RFP, an ER marker. Images were obtained by confocal microscopy. (a) phase contrast image, colocalization of GRP78 (b, green) with ER marker (c, red) is shown in orange (d, merge). Scale bars: 100 µm. (C and D). FACS analysis of synoviocytes obtained from OA (C) or RA patients (D). Cells were stained with DyLight 488-conjugated anti-GRP78 antibody, and were analyzed by flow cytometry. Red histograms correspond to specific labeling for surface GRP78 and gray histograms indicate isotypic control antibody. GRP78, glucose-regulated protein of 78 kDa; FLS, fibroblast-like synoviocytes; RA, rheumatoid arthritis; ER, endoplasmic reticulum; FACS, fluorescence-activated cell sorting; OA, osteoarthritis.

• Click for larger image

An endogenous intracellular chaperone molecule, released at times of acute or chronic physiological stress as a form of exosome, necrotic or apoptotic debris, can contribute to immune-modulating signals within the immune network through a variety of mechanisms (73). As reported previously (61), we found that cell-free GRP78 was frequently found in the synovial fluid of RA patients (Fig. 3A). Interestingly, when GRP78 was added to synovial mononuclear cells of patients with RA, the production of IL-17 and TNF-α was increased (Fig. 3B). Consistent with Corrigall et al.'s study (61), such an increase was not noted with OA or normal PBMCs (Fig. 3B). In addition, exogenous GRP78 increased IL-23 production by lipopolysaccharides-stimulated DCs, while simultaneously decreasing IL-10 production by these cells (Fig. 3C), indicating that the GRP78-induced increase in IL-17 production was mediated by the modulation of IL-10 and IL-23 production from mature DCs. Moreover, GRP78 treatment to immature DCs unregulated the expression of co-stimulatory molecules, such as CD40 and CD80. The CD40 and CD86 expressions in DCs stimulated with TNF-α were also additively increased by the treatment with GRP78 (Fig. 3D). These data provide additional evidence for the GRP78-induced increase in chronic inflammatory responses in RA.

Fig. 3
Recombinant GRP78 induces pro-inflammatory response in rheumatoid mononuclear cells. (A) Expression of GRP78 in the synovial fluid of RA patients (n = 8), which was determined by Western blot analysis. (B) GRP78-induced production of IL-17 and TNF-α by synovial fluid mononuclear cells of RA patients (n = 3) versus peripheral blood mononuclear cells of OA (n = 3). Mononuclear cells (1×10⁶) were stimulated with recombinant GRP78 for the indicated time. Cytokine concentrations in the culture supernatants were determined by ELISA. Data are the mean±SD, and are presented as the fold increase as compared with media only. (C) Increase in IL-10 and IL-23 production by recombinant GRP78. RA mononuclear cells (1×10⁶) were stimulated with recombinant GRP78 in the presence of LPS (1 µg/mL) for 24 hr. The IL-10 and IL-23 levels in the culture supernatants were determined by ELISA. (D) GRP78-induced upregulation of co-stimulatory molecules on dendritic cells (DCs). RA mononuclear cells (1×10⁶) were stimulated with recombinant GRP78 (rGRP78) in the presence or absence of TNF-α (10 ng/mL) for 24 hr, and the expressions of CD40, CD80, and CD86 on immature DCs were analyzed by flow cytometry. RA, rheumatoid arthritis; rGRP78, recombinant glucose-regulated protein of 78; IL, interleukin; OA, osteoarthritis; TNF-α, tumor necrosis-factor alpha; ELISA, enzyme-linked immunosorbent assay; SD, standard deviation; LPS, lipopolysaccharide.

• Click for larger image

Taken together, pro-inflammatory cytokines up-regulate GRP78 expression in RA-FLS. The increased GRP78 expression, in turn, could further activate RA-FLS by interacting with ACPA as a surface form and as a soluble form by triggering IL-17 production and co-stimulatory molecule expression in RA synovial mononuclear cells.

GRP78 is induced in hypoxic endothelial cells (74), and is up-regulated by vascular endothelial growth factor (VEGF) treatment (75). GRP78 knockdown significantly suppresses VEGF-induced activation of ERK1/2, phosphoinositide phospholipase C, and VEGF receptor-2 (VEGFR-2) as well as VEGF-induced endothelial cell proliferation (75). Several lines of evidence have shown that GRP78 promotes tumor angiogenesis. Kringle 5 (K5) of human plasminogen can function as a binding partner of GRP78 on the cell surface of proliferating endothelial cells (74). Conditional knockout mice of GRP78 in the endothelial cells can cause dramatic reduction of tumor angiogenesis (76). Knockdown of GRP78 expression in human endothelial cells reduces angiogenesis by suppressing cell proliferation, survival, and migration (76). It has also been demonstrated that cell-surface GRP78-targeting peptide has an anti-angiogenic effect (77, 78). The bacterial AB5 subtilase cytotoxin can specifically cleave GRP78 at a single amino acid, abolishing GRP78 function rapidly and specifically (77). Conjugation of GRP78 with the plasminogen K5 or extracellular Par-4 promotes endothelial apoptosis, which suggests that cell-surface GRP78-targeting peptide can be utilized as a potential anti-angiogenesis therapy.

Angiogenesis is highly active in RA, particularly in the early onset of the disease (3, 4). The newly formed vessels can maintain the chronic inflammatory state by transporting the inflammatory cells to the site of synovitis, as well as supplying nutrients and oxygen to the synovium (3, 4). Of many angiogenic factors, VEGF plays a central role in "pannus formation" (3, 4). In RA, VEGF appears in increased amounts in the sera, synovial fluids, and inflamed synovium of patients (79), and thus constitutes a potential candidate for therapeutic modulation. Treatment with anti-VEGF antibody has been shown to attenuate CIA in mice (80). Again, specific inhibition of VEGF by soluble VEGF receptors reduced the disease severity in murine CIA (80). Our group has shown that GRP78 deficiency inhibits VEGF165 stimulated endothelial cell proliferation (38). In addition, VEGF₁₆₅-induced tube formation, migration, and chemotaxis of endothelial cells are also markedly reduced by knockdown of GRP78. These results, together with previous reports (38, 79, 80), indicate that GRP78 directly mediates VEGF₁₆₅-induced migration, chemotaxis, and endothelial cell proliferation. Thus, anti-GRP78 inhibitors could be effective for suppressing the excessive angiogenesis frequently noted in RA joints.

In addition to UPR, ERAD is also required to avoid ER stress in the cells (8-11). The UPR relieves ER stress by inducing ER chaperones to increase the protein-folding capacity of the ER, as well as by inhibiting general protein translation. In contrast, the ERAD eliminates misfolded or unassembled proteins that accumulate in the ER through the ubiquitin-proteasome system (8-11). Unless two compensatory mechanisms of UPR and ERAD work properly, ER stress causes cell damage, and eventually cell death (8-11). Synoviolin is one of the ER-resident E3 ubiquitin ligases involved in ERAD, and is implicated in RA pathogenesis (81, 82). Several studies have shown the relationship between GRP78 and synoviolin (72, 83). In stressed cells, increased GRP78 expression is associated with activation of P58IPK and other co-chaperones, which enhances ERAD in the ER lumen (72). In zebrafish embryonic cell line ZF4, endogenous IGF1 is induced as XBP-1 splicing during ER stress, and XBP-1 not only increases GRP78 but also induces synoviolin (83). Such findings (72, 81-83) suggest that the ERAD system is closely related to UPR.

As mentioned above, RA-FLS are the major cell population in tumor-like expansion and invasive pannus. In the inflamed joints, RA synovial cells have to keep producing large amounts of proteins for the progression of inflammation. In this context, ERAD may be a necessary processing system for ER homeostasis (84). Indeed, ERAD is aberrantly unregulated in RA (85). A recent study has demonstrated that overexpression of synoviolin causes arthropathy with synovial hyperplasia, whereas knockdown of synoviolin results in increased apoptosis of synovial cells and less sensitivity to CIA in mice (85). Enhanced ERAD may efficiently remove unfolded protein in ER, which results in the indirect suppression of UPR activation (85). This notion is supported by previous findings that mouse embryonic fibroblasts that lack synoviolin show increased susceptibility to ER stress-induced apoptosis (81, 86). The previous reports on ERAD (85, 86) are consistent with our data in that dysregulated ER responses critically contribute to synovial hyperplasia and the development of chronic arthritis. Thus, it would be interesting to investigate whether two biological processes, UPR and ERAD, affect each other to induce RA.

The possible role of ER stress in RA pathogenesis is summarized in Fig. 4. Micro-environmental stresses such as hypoxia, glucose deprivation, reactive oxygen species, and pro-inflammatory cytokines, may increase ER stress in both innate immune cells (e.g., DCs and FLS) and adaptive immune cells (e.g., T and B cells) in inflamed joints. In particular, during ER stress, GRP78 expression is increased in RA-FLS. The increased GRP78 expression promotes FLS survival and proliferation, resulting in synovial proliferation. The induction of GRP78 by ER stress may lead to an increase in GRP78 in the ER lumen as well as promotion of GRP78 re-localization from the ER to the cell surface; in this case, cell surface GRP78 can be a target for ACPA and may act as an auto-antigen for T and B cells. Moreover, extracellular GRP78, detected at high levels in RA joints, may contribute to the development of auto-reactive T cells and increase the production of IL-17 and TNF-α in RA synovial mononuclear cells. In addition, citrullinated GRP78 on monocytes/macrophages binds to ACPA, and stimulates the production of pro-inflammatory cytokines, such as TNF-α, which further increases GRP78 expression in RA-FLS. Increased GRP78 expression in RA-FLS, in turn, could amplify the inflammatory cascade by escalating pannus formation. Finally, GRP78 directly stimulates VEGF-induced migration/chemotaxis and endothelial cell proliferation, which facilitate synovial angiogenesis.

Fig. 4
Hypothetical model for the role of GRP78 in the pathogenesis of rheumatoid arthritis. GRP78, glucose-regulated protein of 78 kDa; ROS, reactive oxygen species; TNF-α, tumor necrosis factor-alpha; IL-1β, interleukin-1beta; ER, endoplasmic reticulum; BiP, binding immunoglobulin protein; FLS, fibroblast-like synoviocytes; APC, antigen presenting cell; Ag, antigen; Ab, antibody; ACPA, anti-cyclic citrullinated peptide antibodies; GF, growth factor.

• Click for larger image

GRP78 is traditionally regarded as a major ER chaperone (7, 9, 10, 15). However, increasing evidence indicates that GRP78 exists outside the ER, in the cytoplasm and cell membrane, and plays a critical role in cell survival, tumor angiogenesis, metastasis, and resistance to cancer therapy (7, 8, 10, 27, 72). In this regard, our finding that GRP78 is present on the surface of FLS may open the door to novel therapeutic approaches that specifically target synoviocyte proliferation and endothelial cells, the pathologic hallmark of RA. For example, conjugation of toxin- or apoptosis-inducing agents with synthetic peptides that can bind to GRP78, such as WIFPWIQL (73), may inhibit synovial proliferation, angiogenesis, and the pannus formation. In addition, it can be expected that extracellular GRP78 could suppress RA activity by inducing T cell tolerance and also by competing with membrane GRP78 for binding of the anti-GRP78 antibody. We are currently investigating such possibilities.