Calcineurin Is Expressed and Plays a Critical Role in Inflammatory Arthritis¹

Rheumatoid arthritis (RA)³ is characterized by a tumor-like expansion of the synovium, which is composed of proliferating synoviocytes, infiltrating leukocytes, and new blood vessels (1). In the synovium, synoviocytes actively participate in the chronic inflammatory responses of RA as a major cell population of the invasive pannus (2). They strongly express a variety of activation markers, including surface molecules (e.g., MHC class II, VCAM-1), and thereby can efficiently present Ags to T cells. The synovial fibroblasts isolated from RA patients have the potential to produce matrix degrading enzymes and several cytokines such as IL-1β, IL-6, and IL-8 (3). Moreover, synovial fibroblasts proliferate abnormally and invade the local environment and exhibit the characteristics of tumor cells such as somatic mutations in H-ras and p53 (4, 5).

Calcineurin is a heterodimeric phosphatase formed by the association of a 60-kDa catalytic subunit (calcineurin A) and a 19-kDa calcium-binding subunit (calcineurin B) (6). There are two major isoforms of calcineurin A, designated as calcineurin Aα and Aβ (7), which have unique function and characteristics. Calcineurin Aα is widely distributed among the tissues, whereas calcineurin Aβ is predominantly expressed in the cells of the lymphoid lineage and mediates the immune response (8). An increase in protein tyrosine phosphorylation and the cytoplasmic-free Ca²⁺ responses triggers the calcineurin phosphatase activity. These processes have been suggested to be associated with lymphocyte abnormalities in autoimmune diseases such as systemic lupus erythematosus (9, 10). In immune cells, calcineurin controls the activity of a wide range of transcription factors, including NF-AT, NF-κB, c-fos, and Elk-1 (6). As a result, calcineurin plays a crucial role in T cell activation, cell growth, apoptosis, neuron depotentiation, and angiogenesis. In addition, its overexpression has been implicated in the pathogenesis of cardiomyopathy and stroke (11). However, the expression and function of calcineurin in a pathologic lesion of chronic inflammatory diseases, such as RA, remain to be defined.

In an attempt to determine the role of calcineurin in inflammatory arthritis, we investigated the expression and function of calcineurin in the rheumatoid synovium and synoviocytes, the actual sites of chronic inflammation. Our results suggest that calcineurin plays an important role in synoviocyte activation and arthritis progression in vivo, and such a function is tightly linked to dysregulated intracellular Ca²⁺ store and Ca²⁺ response triggered by proinflammatory cytokine. Moreover, the selective inhibition of calcineurin by the overexpression of calcineurin-binding protein (Cabin) 1, a natural calcineurin antagonist, hampered synoviocyte activation, thereby providing a novel strategy for the treatment of calcineurin-dependent inflammatory diseases, such as RA.

The FLS were prepared from the synovial tissues of patients with RA or osteoarthritis (OA) who had undergone total joint replacement surgery. The isolation of the FLS from the synovial tissues was performed as described previously (12). The FLS, from passages 3 through 7, were seeded in 24-well plates (Nunc) at 3 × 10⁴ cells/well or in a 100-mm culture dish at 5 × 10⁵ cells in DMEM supplemented with 10% FCS (Invitrogen Life Technologies) and cultivated at 37°C for 24 h. To eliminate the influence of FCS on calcineurin activity, the cells were washed with DMEM and then incubated again in serum-free DMEM supplemented with insulin-transferrin-selenium A (ITSA; Invitrogen Life Technologies) for 48 h, until the medium was replaced with fresh DMEM/ITSA. The cells were stimulated with the medium alone, TNF-α (R&D Systems) and IL-1β (Endogen) in attempt to induce the synthesis of IL-6 and matrix metalloproteinase (MMP) from the synoviocytes at the onset of culture. In some experiments, FLS were incubated in the presence of various concentrations of cyclosporin A (CsA; Calbiochem), ranging from 4 to 800 nM. After 24 or 72 h of incubation, the cell-free supernatants were collected and stored at −20°C until assayed.

Synovial fluid from RA patients was collected by arthrocentesis into sterile tubes, diluted 1/5 with PBS immediately after collection, and passed through sterile gauze. SFMC were isolated by density gradient centrifugation on Ficoll-Hypaque (SG 1077). The cell viability was >95% as determined by trypan blue exclusion. SFMC were resuspended in complete medium, which consisted of RPMI 1640 supplemented with 10% FCS, 100 U/ml penicillin, 100 mg/ml streptomycin, and 2 mM l-glutamine. Each culture was performed in triplicate at a density of 5 × 10⁵/well in 96 microtiter wells (Nunc) at 37°C in a 5% CO₂ atmosphere. Variable concentrations (4–800 nM) of CsA were added to the wells at the onset of culture. After 24 h of incubation, the concentrations of IL-17 and IL-10 in the cell-free supernatants were measured by ELISA.

Immunohistochemical staining of the synovium was performed on 5-μm sections of formalin-fixed paraffin-embedded blocks. The sections were mounted on superfrost glass slides, deparaffinized in xylene, and rehydrated in a graded series of ethanol, followed by microwave Ag retrieval. The endogenous peroxidase activity was blocked by 3% hydrogen peroxide. After blocking the nonspecific binding by treating the slides with 10% normal goat serum at 37°C for 60 min, the slides were incubated with mouse anti-human calcineurin Ab (Sigma-Aldrich) at a 1/100 dilution overnight at 4°C. The sections were washed and incubated with the secondary Ab, biotinylated goat anti-mouse IgG (DakoCytomation). After washing and incubating the sections with peroxidase-conjugated streptavidin (synovial fluid mononuclear cell) at room temperature for 30 min, 3,3′-diaminobenzidine was added to reveal the Ag. The sections were counterstained with Mayer hematoxylin, dehydrated, cleared, and mounted. The negative control tissue was prepared in the same manner described above, except that primary Ab was omitted or replaced by isotype control Ab (IgG1; R&D Systems).

The cells were lysed in lysis buffer (0.5% Triton X-100, 300 mM NaCl, 50 mM Tris-HCl (pH 7.6), containing 1 mM PMSF, 10 μg/ml aprotinin, and 1 μg/ml leupeptin), and the resulting lysates were cleared by centrifugation. The quantity of cellular proteins in the clarified supernatant was determined by Bradford protein assays (Bio-Rad). Eighty micrograms of proteins was resolved on 10% SDS-PAGE and then transferred to polyvinyl difluoride membranes (Amersham Biosciences). The membranes were incubated with the rat anti-calcineurin mAb (Sigma-Aldrich), followed by HRP-conjugated anti-mouse or anti-goat IgG (Santa Cruz Biotechnology). The protein was visualized using a chemiluminescence reaction (ECL; Amersham Biosciences).

The calcineurin enzymatic activity was measured using a phosphatase assay kit (BIOMOL), according to the manufacturer’s protocol. The data are presented as a concentration (nM) of phosphate released per 1 mg of cytosolic protein.

The total RNA was extracted with RNAzolB (Biotex Laboratories) from cultures of synoviocytes (1 × 10⁶ cells), according to the manufacturer’s instructions. The cDNA synthesis was performed using RevertAid M-murine leukemia virus reverse transcriptase (MBI Fermentas), random hexamer (TaKaRa), and 2 μg of the total RNA. The PCR were then conducted using 2 μl of the cDNA product in a total volume of 25 μl for 35 cycles. The primers for amplifying the calcineurin Aα, Aβ, and GAPDH mRNA are as follows: calcineurin Aα mRNA, 5′-CGACAGGAAAAAAATTTGCTGGAT-3′ and 5′-TTGTTTGGCTTTTCCTGTACATG-3′ (673 bp); calcineurin Aβ mRNA, 5′-AACCATGATAGAAGTAGAAGCTC-3′ and 5′-CACACACTGCTGGATAGTTA TAA-3′ (568 bp); and GAPDH mRNA, 5′-CGATGCTGGGCGTGAGTC-3′ and 5′-CGTTCAGCTCAGGGATGACC-3′ (498 bp). The PCR products were electrophoresed in 1.5% agarose gel in a 1× TAE buffer (0.04 M Tris-acetate and 1 mM (pH 8.0).

The changes in the [Ca²⁺]_i resulting from the exposure of PMA (Sigma-Aldrich) and the cytokines were measured as described previously (13). Briefly, for the intracellular calcium image, Fluo-3 AM (Molecular Probes) was dissolved (1 mg/ml) in DMSO and stored at −20°C until used. The FLS of RA and OA patients were resuspended in Ca²⁺-free HEPES-buffered physiological saline solution (HPSS; 116 mM NaCl, 5.4 mM KCl, 0.8 mM MgSO₄, 20 mM HEPES, and 10 mM glucose) and were loaded with 2 μl of Fluo-3 AM for 40 min at room temperature. After washing with HPSS buffer, the cells were further incubated for 15 min in the absence of Fluo-3 AM to completely de-esterify the dye. The data was normalized by measuring the residual fluorescence (F₀) at the end of the experiment and calculated as the ratio of the fluorescence of the experiment conditions (F)/F₀. The excitation of Fluo-3 AM was provided by the 488-nm line of an argon laser, and the emission range was 515 nm. The intracellular calcium image was detected by inverted confocal microscopy (Zeiss LSM 510 Meta; Zeiss) with an objective (×20). In some experiment, intracellular Ca²⁺ pool stored in endoplasmic reticulum (ER) pool, and non-ER pool was also assessed, as described previously (14). For this, the FLS were cultured in Ca²⁺-free HPSS, and the remaining Ca²⁺ was removed by the addition of EGTA (0.2 mM) immediately before stimulation.

The FLS undergoing apoptosis were determined by cellular DNA fragmentation ELISA kit (Roche Applied Science), which is based on the quantitative sandwich ELISA principle using two mouse mAbs directed against the DNA and BrdU. The enzymatic activity of caspase-3 was also determined using the apotarget caspase-3/cpp32/colorimetric protease assay kit (BioSource International), according to the manufacturer’s instructions.

The amounts of IL-6, IL-10, and IL-17 released into the culture supernatants were measured by ELISA, as described previously (12). The concentrations of MMP-1 (Amersham Biosciences), MMP-2 (R&D Systems), MMP-3 (Amersham Biosciences), and MMP-9 (R&D Systems) were also measured by ELISA, according to the manufacturers’ instructions. The MMP assay kits detected both the proform and active form and had no significant cross-reactivity with the other MMP.

The human Cabin 1 gene was amplified from the rheumatoid synoviocyte cDNA by a PCR with two pairs of primers encompassing the calcineurin-binding domain sequence (5641–6614). The forward primer has an EcoRI site, and the reverse primer has a SalI site for cloning. The resulting PCR product was cloned into the enhanced GFP expression vector (BD Biosciences), pGFP-C1, to generate the pGFP-Cabin 1 construct, which was verified by sequencing (ABI PRISM 310; Applied Biosystems). The plasmid DNA transfection of the Cabin 1 gene into the MH7A cells, the immortalized synoviocytes with SV40 T Ag (15), was also conducted in 6-well plates using LipofectAMINE 2000 reagent. The day before transfection, 3 × 10⁴ of the MH7A cells were plated in 2 ml of a medium/well. For each well, the LipofectAMINE reagent (1.6 μl) was mixed with the pEGFP-C1 vector or pEGFP-Cabin 1 (0.8 μg) in serum-free Opti-MEM to allow the DNA-LipofectAMINE reagent complexes to form. The complexes were added to the respective wells and mixed by gently rocking the plate back and forth. After 4 h, the cells were incubated in a CO₂ incubator at 37°C for 24 h in RPMI 1640 supplemented with 10% FCS.

Male DBA/1 mice were immunized with native bovine type II collagen (Chondrex) at 8–12 wk of age, as described previously (16). Three weeks after the primary immunization, CsA (4, 10, and 20 mg/kg) dissolved in the vehicle (ethanol-oil 1:7) was injected i.m. into the CIA animals every other day for 3 wk. The control mice received the vehicle alone. The incidence and severity of arthritis in the four groups of mice were determined by visual inspection. The mice were observed two to three times per week for the onset, duration, and severity of the joint inflammation over a period of 12 wk after the primary immunization. Each limb, except the hind foot that received the booster immunization, was assessed on a 0- to 4-point scale, as described earlier (16).

At day 42, all three paws and ankles, except the hind foot that received the booster immunization, were harvested from each mouse and fixed overnight in 10% formalin. The paws and ankles were then decalcified in 30% citrate buffered formic acid at 4°C, dehydrated in a graded series of ethanol and xylene, and embedded in paraffin. Thin sections (4-μm thick) were stained with H&E. Immunohistochemical staining of calcineurin in the arthritis joints was performed using the anti-calcineurin Ab (BD Biosciences) by a similar method to that used for calcineurin staining in the human synovium.

To investigate the correlation of calcineurin activity with the severity of arthritis, synovial cells and draining lymph node cells were taken from the mice at the onset, peak, and diminution phases of CIA. The tissues were pooled from four to five mice and minced into single-cell suspensions in RPMI 1640. The synovial cells were separated from the tissues by a similar method used to isolate the human FLS. The cells were incubated again in serum-free DMEM supplemented with ITSA for 24 h. The calcineurin activities in the lymph node cells (1 × 10⁷ cells) and primary-culture synovial cells (2 × 10⁶ cells) were measured using a phosphatase assay kit (BIOMOL).

Data are expressed as the mean ± SD. Comparisons of the numerical data between groups were performed by the paired or unpaired Mann-Whitney U test. Values of p < 0.05 were considered statistically significant.

To investigate the distribution and localization of calcineurin in joint tissues, we performed the immunohistochemical staining of the synovium of three RA patients using an Ab to calcineurin A, which is a catalytic subunit of the calcineurin complex (17). Calcineurin A expression was observed in all three RA synovial tissue sections. The positive staining was mainly seen in the lining layer of the hyperplastic synovium (Fig. 1, A–D), infiltrating leukocytes (Fig. 1, B and C), and endothelial cells in the sublining (Fig. 1,D), highlighting the ubiquitous expression of calcineurin. Calcineurin A was also expressed in the lining layer of OA synovium (n = 3) (Fig. 1, E–H), but its expression level was relatively weak compared with that of RA synovium. As a negative control, isotype Ab did not show any immunoreactivity with the calcineurin Ag in the RA synovium (data not shown).

The lining layer of the synovium is composed of macrophage-like synoviocytes and FLS. We next compared the basal expression of calcineurin A in the cultured FLS of RA patients and OA patients. Western blot analysis revealed the presence of calcineurin A protein in six of eight RA FLS and in four of eight OA FLS with similar passage numbers (Fig. 2,A). Interestingly, the two RA patients, who show no calcineurin activity in FLS by Western blot analysis, had normal levels of erythrocyte sedimentation rate and serum C-reactive protein determined immediately before the operation, in contrast to the rest of patients with polyarthritis and active synovitis with high erythrocyte sedimentation rate and C-reactive protein. Overall, the RA FLS showed a 9-fold higher amount of calcineurin A than OA FLS (OD ratio (calcineurin/β-actin): 0.36 ± 0.09 vs 0.04 ± 0.02, p = 0.03), judging from the expression level in relation to β-actin as the internal control. The enzymatic activity of calcineurin A, which was measured by the phosphatase assay, was also stronger in the RA FLS (n = 8) than in the OA FLS (n = 8; p = 0.02) (Fig. 2,B). RT-PCR analysis for the calcineurin A isoforms revealed that all FLS from RA (n = 5) and OA patients (n = 5) expressed both calcineurin Aα and Aβ mRNA (Fig. 2 C). Given the predominant expression of calcineurin Aβ in lymphoid lineage cells (7, 8), these data support the notion that synoviocytes should be treated as immune cells, actively participating in inflammatory response within the joints (1, 2, 3, 4, 5).

In inflamed joints, rheumatoid synoviocytes are exposed to various proinflammatory cytokines, which are interdependently linked to promote and maintain the inflammatory environments. An experiment was conducted to determine the effect of the proinflammatory cytokines, IL-1β and TNF-α, on the calcineurin phosphatase activity in RA FLS. The calcineurin activity in the FLS was increased rapidly as early as 5 min after the treatment with IL-1β (10 ng/ml), reached a peak level after 2 h, and remained elevated up to 24 h (Fig. 3,A). Time-dependent increase in calcineurin activity was also observed after TNF-α (10 ng/ml) stimulation (Fig. 3,B). In contrast, IL-10 (10 ng/ml), an anti-inflammatory cytokine, failed to increase the calcineurin activity even after 6 h or more of incubation (Fig. 3,B). It is well established that elevated intracellular Ca²⁺ concentration stimulates the calcineurin activity and enables it to dephosphorylate the cytosolic transcription factor (18). Accordingly, the influence of pro- and anti-inflammatory cytokines on Ca²⁺ release in FLS needs to be determined. The result showed that both PMA (100 nM) and TNF-α (10 ng/ml) caused a rapid and transient increase in the release of Ca²⁺ in FLS (Fig. 3, C, D, and E), which was completely blocked by pretreating the cells with BAPTA (20 μM), an intracellular Ca²⁺ chealtor (Fig. 3,F and data now shown). Moreover, RA and OA FLS pretreated with TNF-α (10 ng/ml) showed a prolonged and greater Ca²⁺ response to PMA compared with the cells treated with PMA alone (Fig. 3,G). These data suggest that TNF-α not only increases the calcineurin activity via the release of intracellular Ca²⁺ but prior exposure to it sensitizes the FLS to external stimuli. However, an anti-inflammatory cytokine IL-10 (10 ng/ml) did not induce the release of Ca²⁺ (Fig. 3,H), which was similar to its lack of effect on the calcineurin activity. Of note, RA FLS (n = 4) showed a higher degree of [Ca²⁺]_i release than OA FLS (n = 4) when stimulated with either TNF-α (10 ng/ml) or PMA (100 nM) (Fig. 3, D and E). Given the higher IL-1β and TNF-α expression levels in RA joints than in OA joints, the increased calcineurin activity in the RA FLS compared with OA FLS may have been caused by the difference in the cytokine levels in the joints.

There are two major Ca²⁺ pools in T cells: ER pool and non-ER pool. The ER Ca²⁺ pool can be measured by treating the cells with thapsigargin, a specific inhibitor of the ER Ca²⁺ ATPase pump responsible for Ca²⁺ reuptake into the ER (19). In a Ca²⁺-free condition, a saturating dose of ionomycin releases Ca²⁺ from all intracellular organelles. However, ionomycin added immediately after depleting ER Ca²⁺ pool releases Ca²⁺ from the remaining non-ER Ca²⁺ stores (14). As shown in Fig. 4,A, there was no significant difference in [Ca²⁺]_i rise between RA and OA FLS when the cells were stimulated with thapsigargin (500 nM) immediately after the addition of EGTA (Fig. 4,A). However, RA FLS showed a significantly higher Ca²⁺ response to ionomycin (5.6 μM) than OA FLS (Fig. 4,B), demonstrating the enlarged total (ER plus non-ER) Ca²⁺ store in RA FLS. After the thapsigargin-induced [Ca²⁺] signal had returned to baseline, subsequent addition of ionomycin to the FLS resulted in a higher Ca²⁺ rise in RA FLS compared with OA FLS (Fig. 4 C), indicating that the amount of Ca²⁺ remaining after the depletion of ER Ca²⁺ is greater in RA FLS. Taken together, these data suggest that Ca²⁺ store in non-ER compartment, including mitochondria and the nuclear envelope, but not in thapsigargin-sensitive ER pool, is increased in RA FLS, which may underlie the altered Ca²⁺ signaling, thus making these cells hyperresponsive to external stimuli such as TNF-α.

To determine the role of calcineurin in the activation of synoviocytes, we investigated the effect of calcineurin inhibitors on the production of IL-6 and MMP, which are the major proinflammatory cytokine and enzyme produced by the FLS. When added to the FLS culture, CsA (40–800 nM), a calcineurin inhibitor, partially inhibited the IL-1β- or TNF-α-stimulated production of IL-6 in a dose-dependent manner (Fig. 5,A). Another calcineurin inhibitor, FK506, showed a similar pattern in blocking IL-6 production to CsA (Fig. 5,B). The inhibitory effect of CsA was not caused by cell death because the degree of apoptosis, which was determined by DNA fragmentation ELISA, was not altered by CsA (40–4000 nM) (Fig. 5,C). The viability of the FLS, determined by the MTT assay or caspase-3 activity, was not influenced by any of CsA concentrations tested (40–800 nM) either (data not shown). The IL-1β or TNF-α-stimulated productions of MMP-1 and -2 were almost completely repressed to their spontaneous levels by the addition of CsA (Fig. 5,D), suggesting that synthesis of MMP-1 and -2 is more sensitive to CsA than that of IL-6. In contrast, MMP-3 and -9 productions from the same cells were virtually unaffected by CsA, decreasing by <15% of the change increase by IL-1β or TNF-α (Fig. 5 D and data not shown), suggesting that calcineurin differentially regulates the productions of MMP-1, -2, -3, and -9 from RA FLS.

In the rheumatoid synovium, FLS can be activated by the Ag-stimulated lymphocytes or macrophages via cell-cell contact, as well as by soluble mediators, such as IL-17, released by these immune cells (1, 2, 3, 11, 20, 21). It has been reported that CsA blocks the PMA plus ionomycin-triggered IL-17 secretion by mononuclear cells (22) but stimulates IL-10 production by dendritic cells (23). As shown in Fig. 5,E, the addition of CsA increased the spontaneous production of IL-10 by SFMC but decreased the production of IL-17 in a dose-dependent manner. At a 800 nM concentration of CsA, the mean levels of the cytokines were 306 ± 94 pg/ml for IL-10 (5.0-fold over the constitutive levels) and 32 ± 5 pg/ml for IL-17 (24.6% of the initial response). Moreover, CsA also inhibited the IL-17 production by SFMC stimulated with LPS or PMA plus calcium ionomycin (Fig. 5,F), suggesting that it directly inactivates rheumatoid T cells to inhibit the IL-17 production, a T cell-derived cytokine. In contrast, CsA synergistically increased the LPS-induced IL-10 production by SFMC while decreasing the PMA + ionomycin-induced IL-10 production (Fig. 5 F). This suggests that IL-10 response to CsA may be different depending on the kinds of mitogen. Considering the importance of cell-cell communication in the activation of FLS (21), CsA may inactivate the FLS indirectly by regulating the activity of the infiltrating mononuclear cells, possibly leading to the enfeeblement of the intercellular cross-talk.

Pharmacologic calcineurin inhibitors might have multiple biological effects, independent of calcineurin inhibition. For this reason, we sought to specifically abolish the calcineurin activity by overexpressing the calcineurin inhibitory domain of Cabin 1, a natural noncompetitive inhibitor of calcineurin (24, 25). In accordance with a previous report (24), transfecting the Cabin 1 gene (5641–6614), which contains the critical region that interacts with calcineurin, into Jurkat T cells resulted in the suppression of cell proliferation upon a PMA and phytohemagglutinin treatment, as determined by [³H]thymidine incorporation assay (data not shown). We postulated that Cabin 1 peptide overexpression may attenuate the activation of RA FLS because the expression levels of Cabin 1 mRNA and protein in the FLS were not different between RA and OA patients (Fig. 6,A), despite the higher calcineurin activity in the RA FLS (Fig. 2, A and B). After transfection of the GFP-Cabin 1 gene (5641–6614) into immortalized synoviocytes (MH7A cells) (15), the GFP-Cabin 1 fusion peptide (66 kDa), as well as the original Cabin 1 protein (220 kDa), could be detected by Western blotting analysis using the anti-Cabin 1 Ab (Fig. 6, B and C). The calcineurin phosphatase activity was almost completely extinguished in the cells transfected with the GFP-Cabin 1 peptide gene compared with untransfected cells or the cells transfected with GFP gene only (Fig. 6,D). The IL-6 production, which was stimulated with the medium alone or with IL-1β (0.1 ng/ml), was diminished partially by the fusion protein (transfected vs untransfected cells; 2.6 ± 0.2 vs 1.0 ± 0.1 ng/ml for medium alone, 5.1 ± 0.4 vs 3.1 ± 0.2 for IL-1β-stimulation) (Fig. 6,E). Moreover, the Cabin 1 peptide potently inhibited the spontaneous production of MMP-2 from the synoviocytes (transfected vs untransfected cells; 31.1 ± 0.3 vs 5.2 ± 0.1 ng/ml). On the other hand, Cabin 1 peptide failed to abolish the productions of MMP-3 and -9 (Fig. 6,F), which is in parallel with the data on the regulation of MMP-1, -2, -3, and -9 by CsA (Fig. 5 D). These results, showing that Cabin 1 peptide blocked the productions of IL-6 and MMP-2, suggest that the targeted inhibition of calcineurin may be an effective strategy for mitigating rheumatoid inflammation and eliminate the concern that CsA might inactivate the FLS in a manner independent of calcineurin.

The next goal was to further analyze the pathologic role of calcineurin in the in vivo arthritic condition. To this end, we generated the CIA in mice, which developed as early as 3 wk after primary immunization, peaked at 5–7 wk and thereafter spontaneously resolved by 12 wk (data not shown) (16). CsA was injected into these mice for 3 wk every other day, 3 wk after primary immunization. The optimal dose of CsA was determined in a preliminary study in which CsA, ranging from 4 to 20 mg/kg, was given to CIA mice (Fig. 7,A). The maximal suppression of arthritis was achieved with a concentration of 10 mg/kg CsA. A histological examination of the joints on day 42 after immunization showed that the paws and ankles of the CsA-treated mice showed a lower degree of inflammation, synovial hyperplasia, bone destruction, and pannus formation compared with the vehicle-treated mice (Fig. 7,B). The calcineurin distribution in the affected joint tissues was examined using immunohistochemical methods. The joints from arthritic mice exhibited intensive staining with anti-calcineurin Ab, notably in the invading pannus, leukocyte infiltrates, and fibroblast-like stromal cells (Fig. 7,C). The CsA treatment reduced the number of calcineurin-positive cells in the joints (Fig. 7,C), indicating that calcineurin participated directly in cartilage and bone destruction. To further explore the temporal relationship between calcineurin activity and the progression of arthritis, the synovial cells and draining lymph node cells were taken from the mice at three distinct stages of the disease, namely the onset, peak, and diminution phase of CIA. The calcineurin phosphatase activity in both the synovial cells and lymph node cells began to increase at day 21 compared with the cells from the normal mice. The activity peaked at day 42 and ultimately decreased to the basal level at day 84 (Fig. 7,D), which coincided with the course of the severity of arthritis, as assessed by the clinical and histological parameters. In addition, the CsA (4 and 10 mg/kg) treatment for 3 wk partially relented the peak increase in calcineurin activity in the synovial and lymph node cells, as determined on day 42 (Fig. 7 E). Overall, these results suggest that calcineurin is critically involved in the initiation and progression of inflammatory arthritis in vivo.

In this study, we showed that Ca²⁺ and calcineurin play a critical role in synoviocyte activation, as well as in the progression of chronic inflammatory arthritis. Previous studies reported that Ca²⁺ influx and intracellular Ca²⁺ are involved in regulating the MMP production by normal synovial cells (26, 27). It was also shown that the activation of the protein kinase C and Ca²⁺ influx are essential processes in the transduction of IL-1β by rabbit synoviocytes to cause membrane depolarization (28). However, the tripartite relationship among Ca²⁺ release, calcineurin signaling, and synoviocyte activation in RA patients has been documented poorly. Our study demonstrated that calcineurin expression in the FLS was higher in RA patients than in OA patients. The increase in calcineurin activity was triggered by proinflammatory cytokines such as IL-1β and TNF-α and tightly linked to the abnormal Ca²⁺ release upon external stimuli. Calcineurin was also shown to play a key role in the productions of IL-6 and MMP-1 and -2 by FLS. Moreover, in an animal model of RA, calcineurin was strongly expressed in the synoviocytes of the invading pannus and correlated well with the severity of arthritis. Collectively, these data point to a view that Ca²⁺ and calcineurin are critical mediators of synovial inflammation.

Prior exposure of T cells to TNF-α is known to alter Ca²⁺ signaling (29, 30). Our data revealed that pretreatment of TNF-α sensitized the FLS to PMA, a protein kinase C activator, resulting in shifting the Ca²⁺ response in FLS from OA type to RA type (Fig. 3 G). It is unclear what pathway is involved in PMA- or TNF-α-induced Ca²⁺ signals in FLS. In our experiment using RA FLS, the PMA-induced [Ca²⁺]_i increase was similarly noted, although the remaining Ca²⁺ was removed by the addition of EGTA (data not shown). This suggests that the PMA, which in general facilitates Ca²⁺ entry, may affect intracellular Ca²⁺ storage in RA FLS, as it did in some cell types (31, 32). Given the enlarged Ca²⁺ store in RA FLS, prolonged exposure to TNF-α could modify some proteins critical for Ca²⁺ storage, which may render these cells more sensitive to PMA. RA T cells or neutrophils in the joints also have altered Ca²⁺ signaling (14, 33), supporting our hypothesis that proinflammatory conditions may contribute to the dysregulated Ca²⁺ store and Ca²⁺ signaling. Alternatively, the observed alteration of intracellular Ca²⁺ store may be the primary event that underlies the increased Ca²⁺ and calcineurin response to external stimuli, such as PMA and TNF-α. Further study is necessary to understand the molecular and subcellular actions of PMA and TNF-α in FLS.

Our work also underscores the importance of Ca²⁺ and calcineurin signaling as the driving force for perpetuating the chronic inflammation. Rheumatoid synoviocytes exhibit a unique transformed and aggressive phenotype, which is possibly elicited by chronic exposure to a genotoxic environment with reactive oxygen species, growth factors, and proinflammatory cytokines (1, 2, 3, 4, 5). As evidenced in this study, the Ca²⁺ hyperresponsiveness of the synoviocytes to exogenous stimuli may provoke abnormal inflammatory responses in the joints. The resultant proinflammatory cytokines, in concert with the abnormal Ca²⁺ response to these cytokines, can synergistically instigate calcineurin activity in the FLS. In this case, it seems likely that the cytokine-induced calcineurin activity is not absolutely dependent on [Ca²⁺]_i rise because TNF-α (10 ng/ml) or IL-1β (10 ng/ml) additively increased the calcineurin activity in RA FLS pretreated with a saturating dose (50 μM) of A23187, which can instigate the [Ca²⁺]_i maximally (data not shown). It is also possible that the calcineurin activity is escalated via cell-cell interactions because FLS are in intimate contact with the infiltrating leukocytes in the synovium. Then, the elevated calcineurin is capable of promoting the secretion of the MMP and IL-6, which contribute to joint destruction and the further recruitment of inflammatory cells. Calcineurin may dictate the complex inflammatory cascades in the center of the self-perpetuating cycle.

CsA has been used in the treatment of many chronic inflammatory diseases, including RA (34). The principal action mechanism of CsA involves the suppression of cytokine production originating from T cells (35). However, T cells, which are the main target cells of CsA, are relatively scanty in rheumatoid synovium, and CsA has been used to control inflammation in RA without knowing the exact mechanisms. The evolving concept on the pathogenesis of RA has focused on the T cell-independent cytokine networks and the aggressive behavior of the rheumatoid synoviocytes (2, 3, 4, 5). This work demonstrates that calcineurin in synovial fibroblasts can be targeted for the treatment of RA. The concentration of CsA used in our culture system is physiologically relevant in that the trough serum concentration of CsA is ∼80–400 nM (36, 37, 38). Therefore, it seems apparent that inactivation of FLS by blocking calcineurin activity is one of the major mechanisms explaining how CsA works in patients with RA.

The action of CsA is not restricted to T cells and is dependent on cell types and stimuli. For example, the inhibition of TNF-α production by CsA is observed in B cells (39) but not in LPS-treated monocytes (40). CsA markedly enhances LPS-induced IL-10 release but completely prevents anti-CD3-induced IL-10 release in mice (41). This study shows that CsA differentially regulates the productions of MMP-1, -2, -3, and -9 by RA FLS, as well as the spontaneous IL-17 and IL-10 productions by SFMC. Moreover, CsA addition to the LPS-stimulated SFMC resulted in the superinduction of IL-10 secretion, although it completely blocked the PMA + ionomycin-induced IL-10 production. Similar gene induction has been reported for TGF-β, IL-5, and IL-13 in T cells treated with CsA (42, 43, 44). It is unclear how CsA treatment is related to cytokine gene induction. One possible explanation would be the inhibition of the production of nuclear factors binding to negative regulatory sequences of corresponding gene or lack of inhibition of regulatory proteins that promote gene expression. As a result, although the inhibition of calcineurin by CsA blocks certain signals important for cytokine gene expression such as IL-17, other aspects of signal for regulatory genes such as IL-10, normally suppressed by calcineurin, may be simultaneously up-regulated.

Previous reports have shown that the targeted inhibition of calcineurin by Cain, the mouse homolog of human Cabin 1, regulates the synaptic endocytosis of neurotransmitter vesicles (45). Moreover, adenoviral gene transfer of Cain prevents the agonist-induced cardiomyocyte hypertrophy (46), suggesting that overexpressed Cain mitigates cellular hyperactivation in various cell types other than T cells. In this study, we observed first that the overexpression of Cabin 1 can suppress the productions of IL-6 and MMP-2 in rheumatoid synoviocytes, as well as T cell proliferation. Major immunosuppressants, CsA and FK506, specifically inhibit the calcineurin activity via the binding to cyclophilin and FKBP12, respectively (47), which may function as an antioxidant. For example, cyclophilin increases the peroxiredoxins activity and protects cardiac myoblasts from CsA-induced toxicity (48). Transgenic mice overexpressing cyclophilin are resistant to CsA-induced nephrotoxicity via its peptidyl-prolyl cis-trans isomerase activity (49). In this respect, a strategy to block the calcineurin activity directly without affecting cyclophilin is needed to minimize the side effects of CsA. Our finding provides a rationale for evaluating the Cabin 1 as a novel strategy for controlling chronic arthritis with minimal side effects.

In conclusion, our data suggest that the abnormal increase in the Ca²⁺ level and calcineurin signaling may contribute to synoviocyte activation and joint destruction by inducing the productions of the proinflammatory cytokine and MMP. These findings provide a new insight into the pathogenic mechanisms of RA and emphasize the importance of Ca²⁺ and calcineurin as potential candidates for the therapeutic modulation of the chronic inflammatory conditions, including RA, atherosclerosis, and transplantation rejection.

We thank Drs. Dae-Myung Jue and Sue-Yun Hwang for critically reviewing the manuscript.

The authors have no financial conflict of interest.