A Critical Role for Allograft Inflammatory Factor-1 in the Pathogenesis of Rheumatoid Arthritis

Rheumatoid arthritis (RA)² is a systemic inflammatory polyarthropathy and typically exhibits a chronic progressive course with remissions and relapses of disease activity (1). The chronic persistent arthritis results in joint destruction with an associated loss of function due to joint deformity. Although the etiology of RA is still unclear, it is pathologically characterized by synovial hyperplasia, inflammatory cell infiltration and angiogenesis (2). The mononuclear cell infiltrate within the synovium consists predominantly of CD4 T cells (3). Macrophage-like synovial lining cells and interdigitating synovial fibroblasts proliferate and also express large amounts of class II MHC Ags. These pathological processes lead to the formation of so-called pannus with accompanying bone erosion. Recently, anti-cytokine therapies targeting the cytokines TNF-α, IL-6, and IL-1 have been used to treat patients with RA and it has been demonstrated that such treatments may suppress the accompanying bone destruction as well as the synovitis (4, 5). These cytokines are expressed in the inflamed synovium and stimulate synovial cells to proliferate and produce various metalloproteinases that act to degrade cartilage. It can therefore be appreciated that the suppression of mediators such as cytokines involved in the progression of disease may represent an effective therapy for inducing remission of diseases in patients with RA.

Allograft inflammatory factor-1 (AIF-1) is a cytokine and was originally identified and cloned from rat cardiac allografts undergoing chronic rejection (6). AIF-1 is a 17-kDa, IFN-γ-inducible, Ca²⁺-binding EF-hand protein that is encoded within the HLA class III genomic region in the direct vicinity of TNF-α (6, 7, 8). AIF-1 is a Ca²⁺-binding protein, and because Ca²⁺ is an important second messenger involved in cellular activation of many cells including monocytes, it is likely that AIF-1 may be involved in the regulation of cell cycle progression and cellular activation status. AIF-1 expression has been documented in various human tissues and cells such as macrophage cell lines (9), peritoneal macrophages, spleen, peripheral blood leukocytes, thymus, liver, lung and placenta (8, 10), although its detailed physiological functions remain unclear. AIF-1 is expressed by macrophages and microglial cells in experimental autoimmune encephalomyelitis and neuritis (11, 12). Microglial response factor-1 and ionized calcium-binding adapter molecule-1 are homologs of AIF-1 and have been localized to activated infiltrating macrophages in degenerating and inflamed tissues (13, 14, 15). AIF-1 is constitutively expressed in lymphatic tissue and is up-regulated in monocytes/macrophages under various pathological conditions, such as uveitis and in the carotid artery injury model in the rat (8, 11). Moreover, in human disease, AIF-1 is expressed by microglial cells following a cerebral infarction (16) and by both the glandular and stromal cells of endometrial tissues in endometriosis (17). A recent study also indicates that AIF-1 enhances the activation of lymphocytes (18). Therefore, AIF-1 is thought to play a fundamental role in several cell types involved in chronic immunological inflammatory processes but especially macrophages.

In the present study, we examined the potential role of AIF-1 in RA. We determined and compared AIF-1 expression in synovial tissues derived from patients with either RA or osteoarthritis (OA). We also examined the effect of AIF-1 on the proliferation of RA synoviocytes, and cytokines production by human synoviocytes and peripheral blood monocytes; both critical factors in the pathogenesis and progression of RA. Furthermore, we determined the concentration of AIF-1 and IL-6 in synovial fluid (SF).

This study protocol was approved by the Kyoto Prefectural University of Medicine institutional review board, and informed consent was obtained from each patient. Synovial tissues were obtained during joint replacement surgery from 10 patients with RA (3 males/7 females; mean age ± SE, 57.5 ± 3.9 years) and 10 patients with OA (4 males/6 females; mean age ± SE, 58.6 ± 4.0 years), according to the criteria of the American College of Rheumatology.

Synovial tissues were minced and stirred with 1 mg/ml collagenase (Sigma-Aldrich) in serum-free RPMI 1640 medium (Nissui Pharmaceutical) for 3 h, filtered through a nylon mesh and washed extensively. Cells were cultured in RPMI 1640 medium supplemented with 10% FCS and 1% antibiotics in a humidified incubator at 37°C in the presence of 5% CO₂. The synovial cells used in experiments were from passage 3–6 at which stage they exhibited a fibroblastoid morphology.

Two synthetic peptides, which corresponded to residues 53–71 and 113–129 of human AIF-1 (AIF-1_53–71 and AIF-1_113–129, respectively) as deduced from the nucleotide sequence of the human AIF-1 gene, were obtained with an additional cysteine residue at the N terminus (Biologica). Following purification by reversed phase HPLC, the synthetic peptide (purity >90%) was coupled to keyhole limpet hemocyanin with N-(ε-maleimidocaproyloxy)succinimide (Sigma-Aldrich). The carrier-conjugated peptide was then emulsified with Freund’s complete adjuvant (Difco Laboratories) and injected s.c. (0.5 mg/injection) into rabbits. Rabbits were immunized six times at 10-day intervals. Blood samples were collected 10 days after the last injection, and the specific Ab in the sera was purified using an AIF-1 peptide-coupled cyanogen bromide-activated Sepharose affinity column. The Abs reacted with protein from abdominal adipose and PBMC that is identical to the molecular size of purified recombinant human AIF-1.

Human AIF-1 cDNA was amplified from human peripheral lymphocyte cDNA (BD Clontech) using PCR. The forward and reverse primers were 5′-GTG GAT CCA TGA GCC AAA CCA GGG ATT T-3′ (containing BamHI site) and 5′-CAC TCG AGT CAG ATA GGG CTT TCT TGG CT-3′ (containing XhoI site), respectively. To express AIF-1 as a GST fusion protein, the DNA fragment obtained was inserted in the BamHI/XhoI sites of pGEX-4 (Amersham Biosciences) in frame. The fusion protein was purified by a GST purification system (Amersham Biosciences) and affinity chromatography with anti-AIF-1_113–129 Ab.

To investigate the effect of AIF-1 on the cell proliferation and cytokines induction, rAIF-1 was treated with detoxi-gel endotoxin removing gel (Pierce). Endotoxin detection was performed using Limulus amoebocyte lysate analysis (Wako Pure Chemical), and treated AIF protein was confirmed to contain <0.1 ng/μg of endotoxin.

Anti-rAIF-1 antiserum was raised in a similar manner as anti-synthetic peptide Ab by injecting 50 mg of human rAIF-1 into a rabbit. The human rAIF-1 Ab IgG fraction was prepared by chromatography on a human rAIF-1-coupled cyanogen bromide-activated Sepharose affinity column and biotinylated with 5-(N-succinimidyloxycarbonyl)pentyl d-biotinamide (Dojindo Chemical).

Immunohistochemical staining was performed using the avidin-biotin peroxidase complex system. Synovial tissue specimens were preserved in 10% formalin and embedded in paraffin. Specimens were serially sectioned onto microscope slides at a thickness of 4 μm and then deparaffinized. Tissue sections were immersed for 45 min in 0.3% peroxide in methanol to block endogenous peroxidase activity. Nonspecific binding sites were saturated by exposure to 0.2% BSA and normal goat serum diluted 1/66.7 in PBS (pH 7.4) for 20 min. Primary Abs against human AIF-1_113–129 (1/800 dilution in PBS) and control normal rabbit serum were applied to tissue sections and incubated overnight at 4°C. The slides were then washed with PBS for 10 min. Biotinylated goat anti-rabbit IgG (Vector Laboratories) in 10 ml of PBS was applied to tissue sections, and the slides were incubated at room temperature for 30 min. They were then washed with PBS for 10 min, followed by incubation with prepared avidin DH-biotinylated peroxidase complex (Vector Laboratories) for 45 min. After washing with PBS for 10 min, color was developed by immersing the sections in peroxidase substrate solution containing 0.05% (w/v) 3,3′-diaminobenzidine tetrahydorochloride (Vector Laboratories) and 0.01% H₂O₂ in 0.05 mol/L Tris-buffered saline (pH 7.2) for 5 min. Positive staining was indicated by brownish black deposits, and counterstaining was performed with hematoxylin.

For each of the tissue specimens from RA and OA, the extent and intensity of staining with anti-AIF-1_113–129 Ab in synovial lining cells, macrophages, and fibroblasts were graded on a scale of 0 to 3+ by two blinded observers on two separate occasions using coded slides as previously described (19). A 3+ grade implies maximally intense staining, whereas 0 implies no staining.

Western blot analysis was performed as previously described (20) with a slight modification. Synovial specimens were homogenized, solubilized in lysis buffer (PBS, pH 7.4, containing 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS, and 0.6 mM PMSF). Thirty micrograms of each total protein extract from synovial tissues or SF cells were electrophoresed in 4–20% gradient gel SDS-PAGE. Following electrotransfer to a polyvinylidine difluoride membrane, blots were incubated with anti-rAIF-1 Ab (1/500 dilution), followed by incubation with a HRP-linked anti-rabbit IgG (1/1000 dilution; EY Laboratories). After washing, immunoreactivity was detected by ECL using ECL Plus Western blotting detection reagents (Amersham Biosciences). Images were obtained and quantified using a Bio-Rad VersaDoc imaging system (model 5000) with Bio-Rad Quantity One software. Results are expressed as a ratio (protein of interest/β-actin) to correct for sample loading.

Total RNA was extracted from synovial tissues and SF cells using TRIzol (Invitrogen Life Technologies). The first cDNA synthesis from 1 μg of total RNA was catalyzed by the SuperScript First-Strand Synthesis System for RT-PCR using oligodeoxythymidine_12–18 primer (Invitrogen Life Technologies). The resulting first strand cDNA was used for PCR amplification with forward primer 5′-AGA CGA TCC CAA ATA TAG CAG-3′ and reverse primer 5′-TAG CTC TAG GTG AGT CTT GG-3′ for AIF-1 (10) and with the human G3PDH amplimer set for G3PDH (BD Clontech). The PCR mixture comprised 1 μl of first strand cDNA, 0.8 μM of each of the primers indicated above and 1.25 U of Taq polymerase (Takara Premix Ex Taq; Takara Biochemicals) in a total volume of 50 μl of PCR buffer provided by the manufacturer. After an initial denaturation at 94°C for 5 min, PCR was conducted at 94°C for 45 s, at 58°C (for AIF-1) or 50°C (for G3PDH) for 45 s, and at 72°C for 1 min, followed by a final extension step at 72°C for 6 min. The PCR products were electrophoresed in 2% agarose gel and stained with ethidium bromide. The intensity of each band was measured with an Image Master version 2.01 (Amersham Biosciences). We calculated the intensity of the AIF-1 transcript at 26 cycles relative to that of G3PDH at 22 cycles to compare initial mRNA levels.

To evaluate the effect of AIF-1 on cell growth, RA synoviocytes were seeded on a 96-well cell culture plate (Nunc) at a concentration of 5 × 10³ cells/well in a volume of 100 μl. Twenty-four hours later, each well was treated with 1, 10, and 100 ng/ml recombinant AIF-1 and 10 ng/ml LPS (Escherichia coli LPS; Sigma-Aldrich) in RPMI 1640 medium containing 1% FCS and incubated at 37°C in 5% CO₂. Where indicated, cells were preincubated for 1 h with 5 mg/ml anti-AIF-1_113–129 before incubation with rAIF-1 (100 ng/ml). Cell viability was measured at 96 h colorimetrically using the cell counting kit (Dojindo) and a microplate reader (model 550; Bio-Rad) at the test wavelength of 450 nm and the reference wavelength at 650 nm (21). This assay is based on the cleavage of the 2-(4-iodophenyl)-3-(4-nitrophenyl)-5-(2,4-disulfophenyl)-2H-tetrazolium monosodium salt(WST-1) by mitochondrial dehydrogenase in viable cells. Each measurement was determined in six separate experiments, and the results are presented as a percentage of the value for the control cultures.

For the assay of cytokines production, RA synoviocytes were cultured in RPMI 1640 containing 10% FCS at 37°C in 5% CO₂ in 6-well cell culture plates (Nunc). After reaching confluence, the cells were also incubated with or without 1–100 ng/ml rAIF-1 in a total volume of 500 μl of serum-free RPMI 1640 medium for 48 h. Where indicated, cells were preincubated for 1 h with 5 mg/ml anti-AIF-1_113–129 before incubation with rAIF-1 (100 ng/ml). Each culture was performed in triplicate, and supernatants were separated by centrifugation 2000 rpm for 10 min) and concentrations of IL-6, IL-1β, and TNF-α were measured with the ELISA kits according to the manufacturer’s instructions. ELISA kits used were purchased from Biosource International for IL-6, and R&D Systems for IL-1β and TNF-α.

Human PBMCs were isolated from heparinized whole blood from normal healthy volunteers (n = 5) by Ficoll-Paque (Amersham Biosciences) density gradient centrifugation. The cells were washed three times with sterile PBS (pH 7.4) and resuspended in RPMI 1640 medium supplemented with 10% FCS and 1% antibiotics. The cells were seeded in 12-well flat-bottom culture plates (Nunc) at a density of 5 × 10⁵ cells/well in a total volume of 500 μl of serum-free RPMI 1640 medium (Nissui Pharmaceutical). Cells were incubated with or without 1, 10, or 100 ng/ml rAIF-1 or 10 ng/ml LPS (Sigma-Aldrich). Where indicated, cells were preincubated for 1 h with 5 mg/ml anti-AIF-1_113–129 before incubation with rAIF-1 (100 ng/ml). After incubation at 37°C in a humidified atmosphere of 5% CO₂/95% air for 24 h, the culture supernatants were recovered and stored at −80°C until assay. IL-6 concentrations were measured using a commercial ELISA kit (Biosource International) according to the manufacturer’s instructions.

To investigate the effect of IL-6 on AIF-1 production, RA synoviocytes were cultured in RPMI 1640 containing 10% FCS at 37°C in 5% CO₂ in 6-well cell culture plates (Nunc). After reaching confluence, the cells were also incubated with or without 0.1–100.0 ng/ml rIL-6 (R&D Systems) in a total volume of 500 μl of serum-free RPMI 1640 medium for 48 h. Each culture was performed in triplicate, and supernatants were separated by centrifugation (2000 rpm for 10 min) and kept at −80°C until assay.

SF samples were obtained from 12 patients with RA (5 males/7 females; mean age ± SE, 57.1 ± 4.1 years) and 8 patients with OA (3 males/5 females; mean age ± SE, 58.2 ± 4.8 years), according to the criteria of the American College of Rheumatology. Serum C-reactive protein (CRP) in RA and OA patients was 1.3 ± 0.37 mg/dl (mean ± SE) and 0.53 ± 0.25 mg/dl (mean ± SE), respectively. Samples were centrifuged at 3000 rpm for 10 min and then treated with 1000 U/ml hyaluronidase (Sigma-Aldrich) for 60 min at 37°C.

The AIF-1 concentrations in SF were measured by ELISA as previously described (17) with a slight modification. In brief, microtiter plates were coated with affinity-purified anti-AIF-1_53–71 IgG (2.0 mg/ml) diluted with 10 mM carbonate buffer (pH 9.3) for 2 h at room temperature. After washing, nonspecific binding sites in each well were blocked with 10 mM carbonate buffer containing 0.5% BSA. Standard solution (0–2500 pg/ml human rAIF-1) and SF samples treated with hyaluronidase diluted (1/2) with sample buffer (50 mM Tris-HCl buffer, pH 7.0, containing 200 mM NaCl, 10 mM CaCl₂, 0.1% Triton X-100, and 1% BSA) were added to the wells, and the plate was incubated for 1 h at room temperature. After washing with BSA-free sample buffer, biotinylated anti-rAIF-1 IgG was added to each well. The plate was incubated for 1 h at room temperature, washed, and then incubated for an additional hour at room temperature with streptavidin-HRP diluted 1/10,000. After final washing, the plate was treated for 20 min with a substrate solution of 3,3′,5,5′-tetramethylbenzidine and H₂O₂ was added to each well and allowed to react for 20 min at room temperature. The reaction was stopped by the addition of 1 M phosphoric acid, after which OD values at 450 nm were read with an ELISA plate reader. The detection limit of the assay was 20 pg/ml and intraassay and interassay coefficients of variation were 6.2–9.2% at 115 pg/ml and 4.6–7.2% at 600 pg/ml. The IL-6 concentrations were measured using a commercial ELISA kit (Biosource International) according to the manufacturer’s instructions.

Differences in mRNA and protein levels of AIF-1, the graded scores of the extent and intensity of immunostaining with anti-AIF-1 and SF concentrations of AIF-1 and IL-6 between the two patient groups were analyzed by the Mann-Whitney U test. Difference in induction of IL-6 by AIF-1 was analyzed by the Kruskal-Wallis test followed by Dunn’s multiple comparison test. The results of cell proliferation studies were analyzed by two-way ANOVA followed by Scheffe’s multiple comparison test. Pearson’s correlation coefficient was used to examine the correlation between the concentrations of AIF and IL-6 in SF. p < 0.05 was considered significantly different.

Immunohistochemistry was performed to examine the expression and localization of AIF-1 in synovial tissues derived from patients with either RA or OA. Immunohistochemical staining showed strong expression of AIF-1 by infiltrating mononuclear cells and synovial fibroblasts in RA synovial tissue (Fig. 1, A–C) compared with tissue derived from joints affected by OA (Fig. 1,D). Immunostaining of tissue sections with normal rabbit serum was completely negative in all synovial tissue derived from patients with RA (Fig. 1 E).

We found markedly enhanced expression of AIF-1 in infiltrating mononuclear cells (score; mean ± SE, 2.43 ± 0.12) and fibroblasts (2.33 ± 0.15) in RA synovial tissue (Fig. 2). In contrast, in OA tissues the immunohistochemical score of AIF-1 in infiltrating mononuclear cells (0.58 ± 0.12) and fibroblasts (0.55 ± 0.12) was significantly lower compared with patients with RA (both p < 0.001, Mann-Whitney U test) (Fig. 2).

As shown in Fig. 3 A, AIF-1 protein was detected at 17 kDa as a single band in protein extracts from SF cells and synovial tissues of both patients with RA and OA. The intensity of the AIF-1 protein band was significantly elevated in both the SF cells (1.8-fold, p < 0.01) and the synovial tissue from patients with RA (1.7-fold, p < 0.05) compared with the levels in patients with OA.

RT-PCR analysis indicated AIF-1 mRNA expression in both SF cells and synovial tissues from all 13 patients with RA or OA (Fig. 4 A). The levels of AIF-1 mRNA expression were significantly elevated in the SF cells (1.8-fold, p < 0.01) and the synovial tissues (1.9-fold, p < 0.01) from patients with RA compared with the expression levels in patients with OA.

We examined the effect of AIF-1 upon the cell proliferation of synoviocytes isolated from RA patients. Fig. 5 illustrates the effect of several doses AIF-1 upon cell proliferation of synoviocytes. AIF-1 significantly increased cell growth in a dose-dependent manner. Incubation of synoviocytes with 100 ng/ml AIF-1 for 96 h resulted in a significant increase in cell growth compared with controls (0 ng/ml AIF-1; p < 0.0001) or 1 ng/ml AIF (p < 0.001). The AIF-induced cell proliferation was significantly inhibited by anti-AIF-1_113–129 Ab (5 mg/ml).

As shown in Fig. 6,A, the IL-6 concentration in the culture supernatant significantly increased after stimulation for 48 h by human rAIF-1 in a dose-dependent manner (p < 0.05–p < 0.0001). The AIF-induced IL-6 secretion was completely inhibited by anti-AIF-1_113–129 Ab (5 mg/ml). In contrast, human rAIF-1 had no effect on the secretion of both IL-1β (Fig. 6,B) and TNF-α (Fig. 6 C).

As shown in Fig. 7, the IL-6 concentration in the culture supernatant markedly increased after stimulation for 24 h by human rAIF-1 in a dose-dependent manner (p < 0.001–p < 0.0001). The AIF-induced IL-6 secretion was completely inhibited by anti-AIF-1_113–129 Ab (5 mg/ml).

Next, we examined the effect of IL-6 on AIF-1 production from synoviocytes isolated from RA patients. As shown Fig. 8, AIF-1 secretion was not stimulated by the addition of human rIL-6, although slightly increase of AIF-1 secretion was seen by the addition of high-dose IL-6 (100 ng/ml).

The AIF concentration in SF was significantly elevated in patients with RA (mean ± SE; 237.8 ± 17.5 pg/ml) compared with patients with OA (162.6 ± 17.6 pg/ml, p < 0.05) (Fig. 9,A). Similarly, the IL-6 concentration in SF was significantly elevated in patients with RA (mean ± SEM; 14.05 ± 1.04 ng/ml) compared with patients with OA (7.80 ± 0.62 ng/ml, p < 0.001) (Fig. 9,B). There was a positive correlation between the SF levels of AIF-1 and IL-6 (r = 0.618, p < 0.01) (Fig. 9 C).

In this study, we have demonstrated for the first time that AIF-1 was more strongly expressed in superficial synovial cells, fibroblasts and infiltrating inflammatory mononuclear cells in the synovial tissue of RA patients compared with the synovial tissue of OA patients. The immunohistochemical findings were confirmed by semiquantitative Western blot and RT-PCR studies. Additional noteworthy findings of this study are the demonstration that AIF-1 increased the proliferation of synovial cells and enhanced IL-6 secretion from human RA synoviocytes and PBMCs in a dose-dependent manner.

Synovial cells proliferate like tumors and form a pannus within joints affected by RA. The pannus directly invades the cartilage and bone and activates osteoclasts which produce various proteases leading to severe cartilage and bone destruction. The proliferation of synovial cells is one of the most critical pathological changes in RA. Therefore, our findings suggest that the increased AIF-1 expression may play an important role in the development and progression of RA.

The exact mechanism(s) whereby increased AIF-1 expression influences synovial cell proliferation is unknown. A potential mechanism underlying the cell proliferation induced by AIF (22, 23, 24, 25, 26, 27) may involve the chemical structure containing an EF-hand calcium-binding sequence. Calcium, the primary receptor protein calmodulin and calcium-dependent protein kinases are essential for entry into the cell cycle in response to mitogenic signals (28, 29, 30, 31). The overexpression of calmodulin in mouse C127 cells enhances cell proliferation, predominantly through a reduction in the length of the G₁ phase of cell cycle (32, 33). The overexpression of the calcium-binding protein sphingosine kinase in 3T3 fibroblasts enhances cell proliferation by promoting the G₁ to S phase transition (34). Moreover, the transfection of rat vascular smooth muscle cells with AIF-1 that is able to bind calcium, increased cell proliferation with constitutive AIF-1 expression resulting in a shortening of the cell cycle and aberrant expression of cell cycle proteins (35). These findings suggest that AIF-1 may function as a cell cycle regulator affecting proliferation in several cell types. Thus, it is possible that AIF-1 promotes synovial cell proliferation by shortening the cell cycle of synovial cells in patients with RA.

Various cytokines are involved in the pathogenesis of RA, and TNF-α and IL-1 play central roles (36). TNF-α and IL-1 are key cytokines in inflammatory cytokine cascades that are directly involved in vascularization, outgrowth of cells of the synovial surface layer, inflammatory cell infiltration and destruction of bone and cartilage through activation of MAP kinase and NF-κB. TNF-α is believed to be upstream of IL-1 in the cytokine cascade of RA and is considered to be the most important cytokine underlying disease progression (36). IL-6 is believed to act downstream of TNF-α and IL-1 and is induced by these cytokines. IL-6 has various physiologic activities but also plays a critical role in the pathogenesis of many inflammatory diseases. IL-6 has been reported to activate B cells, T cells and osteoclasts in RA (37), suggesting its involvement in the pathology of RA. Recently, the effectiveness of biological drugs that target various cytokines has been emphasized. For example, infliximab is a mAb to TNF-α and has inhibitory effects on joint destruction (4, 38, 39, 40, 41). Also, the anti-IL-6 receptor mAb tocilizumab has been reported to exert anti-proliferative effects on the synovial membrane, reduce inflammatory mononuclear cell infiltration, and inhibit the differentiation of osteoclasts in RA (5, 42, 43, 44). In vitro studies have demonstrated increased AIF-1 expression by macrophage cell lines in response to IFN-γ, whereas AIF-1 gene transfected mouse macrophage cell lines exhibit increased production of IL-6, IL-10, and IL-12 p40 following LPS stimulation (45). In this study, we have demonstrated increased IL-6 production by human PBMCs cells following AIF-1 stimulation. In contrast, there was no effect upon IL-1β and TNF-α production (data not shown). Thus, AIF-1 may play a critical role in the chronic inflammation of RA synovial tissue by increasing IL-6 production via IL-1β and TNF-α-independent pathways.

In this study, we also found for the first time that the level of AIF-1 in SF derived from RA patients is significantly higher than that in the SF of OA patients. Interestingly, SF AIF-1 levels significantly correlated with SF IL-6 levels. OA is a mild inflammatory disease with cartilage degradation. As noted above in the last paragraph, anti-IL-6 receptor mAb is very effective for RA and regulates various pathological mechanisms of RA. Furthermore, AIF more strongly expressed in synovial tissues in RA compared with OA. These things also suggest that synovial AIF-1 secreted from synovial cells and infiltrating inflammatory mononuclear cells in affected joints result in increased local IL-6 production in an autocrine manner and this may enhance the local immune inflammatory processes. IL-6 expression is activated by the binding of two transcription factors, NF-κB and NF-IL-6, to the regulatory region of the IL-6 gene (46). AIF-1 may potentially activate these transcription factors directly or through other intracellular signals. These findings indicate that AIF-1 is a novel member of the cytokine network involved in chronic synovial inflammation.

In conclusion, we have demonstrated that AIF-1 is expressed in synovial cells and mononuclear cells in RA synovial tissue and increases proliferation of cultured synoviocytes and production of IL-6 by these synoviocytes and PBMCs. AIF-1 plays an important role in the pathogenesis of RA by affecting key processes such as the activation of synovial cell proliferation and the inflammatory cytokine cascade including IL-6 in joints. Recently, RA patients have achieved remission by anti-cytokine therapy, and the prognosis has improved dramatically. Although further work is needed to clarify the molecular mechanism of action of AIF-1 in RA, we suggest that AIF-1 may represent a new molecular target in RA therapy.

The authors have no financial conflict of interest.