Eotaxin-1/CCL11 is involved in cell migration in rheumatoid arthritis

Eotaxin-1 (CCL11) induces the migration of different leukocyte types by interacting with CCR3. In rheumatoid arthritis (RA), fibroblast-like synoviocytes (FLS) are pathogenic effectors and a major CCR3-expressing cell. The aim of this study was to investigate the expression and function of CCL11 in RA FLS. The expression of CCL11 and CCR3 was evaluated by ELISA, immunofluorescence and quantitative PCR analysis. The CCL11 levels in serum and synovial fluids (SFs) from RA patients were significantly higher than those in serum from healthy controls and SFs from osteoarthritis patients. CCL11 and CCR3 were expressed in the RA synovial tissue lining layers. The secretion of CCL11 in RA FLS-conditioned medium and the mRNA expression of CCL11 and CCR3 were induced by TNF-α. Furthermore, CCL11 induced the mRNA expression of CCL11 and CCR3. Application of a CCR3 antagonist reduced TNF-α-induced CCL11 secretion from RA FLS. CCL11 induced the migration of RA FLS and monocytes. RA FLS migration was decreased by treatment with CCL11 siRNA. The migration of monocytes to medium conditioned with CCL11 siRNA-transfected and TNF-α-stimulated RA FLS was reduced. These data indicate that the self-amplification of CCL11 via CCR3 may play an important role in cell migration in RA.


CCL11 and CCR3 are expressed in RA ST and RA FLS.
We measured the expression of CCL11 and CCR3 in RA ST using immunohistochemistry. We found that CCL11 and CCR3 were expressed in RA cells lining STs ( Fig. 2A). Next, we assessed CCL11 and CCR3 expression in RA FLS using immunohistochemistry. We demonstrated that CCL11 and CCR3 were expressed in RA FLS and that their expression levels were increased by recombinant TNF-α stimulation (Fig. 2B). We then measured the secretion of CCL11 in RA FLS-conditioned medium using ELISA and the mRNA expression of CCL11 and CCR3 using qPCR. The secretion of CCL11 in RA FLS-conditioned medium (Fig. 3A) and the expression of CCL11 mRNA (Fig. 3B) were increased by TNF-α stimulation in a time-dependent manner at 12, 24 and 48 h (n = 7 patients). The peak expression of CCR3 mRNA was induced by TNF-α stimulation at 12 h (n = 7 patients, Fig. 3C).
CCL11 stimulation induces CCL11 and CCR3 mRNA. Because CCL11 expression was increased by TNF-α stimulation in a time-dependent manner, we hypothesized that CCL11 expression was induced by CCL11 itself and/or other factors. We assessed the mRNA expression of CCL11 and CCR3 in recombinant CCL11stimulated RA FLS using qPCR. The expression of CCL11 mRNA at 48 h was induced by CCL11 self-stimulation compared with control treatment (no stimulation) (Fig. 4A). The expression of CCR3 mRNA at 48 h was also induced by CCL11 stimulation compared with control treatment (Fig. 4B). We then measured the effect of a CCR3 antagonist (SB328437) on the secretion of CCL11 in TNF-α-stimulated RA FLS-conditioned medium. Treatment with the CCR3 antagonist reduced the TNF-α-induced CCL11 secretion from RA FLS (Fig. 4C).

CCL11 induced the migration of RA FLS.
To determine whether CCL11 induced the migration of RA FLS, we performed transwell migration assays. CCL11-stimulated cells migrated significantly more efficiently than unstimulated cells (mean ± SEM, number of cells per field; 35.8 ± 5.2 and 24.5 ± 4.6, respectively, p < 0.05, Fig. 5A). We next explored the possibility that reduced expression of CCL11 could inhibit the migration of RA FLS. We used siRNA directed against CCL11. The knockdown of CCL11 in RA FLS was confirmed by the mRNA The levels of CCL11 in the serum from RA patients (n = 26) were higher than those in the serum from healthy controls (HCs) (n = 28). (B) The levels of CCL11 in SFs from patients with RA (n = 15) were higher than those in SFs from patients with osteoarthritis (OA) (n = 16). The data values are expressed as the mean ± SEM. (C) The CCL11 levels were positively correlated with the levels of TNF-α, CCL2/MCP-1 and CCL3/MIP-1α in SFs of RA patients. *p < 0.05 was significant. www.nature.com/scientificreports/ expression of CCL11, which was lower than that in control cells (Fig. 5B). We found that CCL11 siRNA-treated RA FLS showed decreased migration toward 5% FBS compared with control siRNA-treated cells (mean ± SEM, number of cells per field; 40.0 ± 3.6 and 60.0 ± 5.6, respectively, p < 0.05, Fig. 5C).

CCL11 induced the migration of THP-1 cells.
Next, we performed in vitro THP-1 chemotaxis assays to determine whether CCL11 induced monocyte migration. The number of cells migrating toward CCL11 was significantly larger than the number of cells migrating toward the control (mean ± SEM, number of cells per field; 46.5 ± 4.2 and 22.0 ± 2.9, respectively, p < 0.05, Fig. 6A). To confirm the effect of siRNA treatment against CCL11 in RA FLS, we evaluated THP-1 cell chemotaxis toward medium conditioned by CCL11 siRNA-transfected and TNF-α stimulated RA FLS. The knockdown of CCL11 in RA FLS-conditioned medium was confirmed by ELISA. The secretion of CCL11 in CCL11 siRNA-transfected and TNF-α-stimulated RA FLS-conditioned medium was lower than that in control siRNA-transfected RA FLS-conditioned medium (n = 3, Fig. 6B). In addition, the number of cells that migrated toward CCL11 siRNA-transfected and TNF-α stimulated RA FLSconditioned medium was lower than the number that migrated toward control siRNA-transfected and TNF-α stimulated RA FLS-conditioned medium (mean ± SEM, number of cells per field; 31.9 ± 3.6 and 66.9 ± 16.0, respectively, p < 0.05, Fig. 6C).  www.nature.com/scientificreports/

Discussion
In this study, we demonstrated that the levels of CCL11 in serum from RA patients were higher than those in the serum from HC, and the levels of CCL11 in SFs from patients with RA were higher than those in SFs from patients with OA. Furthermore, the CCL11 levels in SFs were positively correlated with the levels of TNF-α, CCL2, and CCL3. These data suggest that CCL11, like other CC chemokines, is involved in joint inflammation in RA. We measured the concentration of CCL11 in serum and SF from the patients without any treatment at the time of diagnosis. We also collected both SF and serum from 15 patients at same day. The levels of CCL11 in SF were lower than those in serum. However, the levels of CCL11 in SF were positively correlated with the levels of TNF-α and the serum levels of CCL11 were not. Therefore, we think the CCL11 in SF is more useful inflammatory mediator than that in serum. These results were different from previous study reported by Liu et al. 20 . We guess the serum CCL11 level is affected by some factors other than arthritis. We compared the concentration of serum CCL11 from 16 patients at pretreatment and post treatment at 6 months, and we could not find any significant differences. And serum CCL11 levels in about a half of patients were increased. We guess serum levels of CCL11 might be affected by treatment or disease duration. Syversen et al. reported that higher serum levels of CCL11 predicted less radiographic progression in early RA patients 24    www.nature.com/scientificreports/ TNF-α reduced CCL11 release 20 . This result is different from ours. Although there are individual differences in CCL11 secretion from RA FLS by TNF-α stimulation, we demonstrated TNF-α increased CCL11 secretion in a time-dependent manner using 7 patients. And we also demonstrated significant increase of CCL11 by TNF-α stimulation at 24 h using the RA FLS from 3 patients which secrete CCL11 well in the experiment using CCR3 antagonist. RA FLS are one of the cellular sources of CCL11 and are also susceptible to CCL11 stimulation in the inflammatory environment of RA. The time course study showed that CCL11 induced the expression of CCR3 and also its own expression. Inhibition of TNF-α-induced CCL11 secretion from RA FLS by a CCR3 antagonist indicates that TNF-α-induced CCL11 expression depends on activation of CCR3 and ligands of CCR3 including CCL11. It has been reported previously that CCL11 upregulates the expression of CCR3 mRNA, and CCL11-induced CCR3 expression is reduced by a CCR3 antagonist in RA FLS 20 . Therefore, CCR3 plays an important role in the secretion of CCL11 from RA FLS. Recombinant CCL11 induced the migration of RA FLS, and the migration of RA FLS treated with siRNA against CCL11 decreased. These results suggest CCL11 from RA FLS might induce the migration of RA FLS. CCL11 induced the migration of monocytes, and the migration of monocytes to CCL11 siRNA-transfected and TNF-α-stimulated RA FLS-conditioned medium was reduced. Therefore, CCL11 contributes to cell recruitment to RA joint inflammation. RA FLS are key mediators of inflammation and joint destruction in RA. These cells aggressively invade the extracellular matrix, producing proteases and inflammatory cytokines and chemokines. Upon TNF-α stimulation, CCL11 secreted from RA FLS should bind to CCR3, which is also overexpressed, and induce the upregulation of CCL11 and CCR3 expression on RA FLS. This self-amplification mechanism of CCL11 via CCR3 may contribute to aggressive behavior of RA FLS.
In conclusion, the increased levels of CCL11 in the serum and SF, enhanced expression of CCL11 and CCR3 on RA FLS induced by TNF-α, and induction of monocyte and RA FLS migration by CCL11 indicate that CCL11 plays an important role in the pathogenesis of RA. The suppression of CCL11 and CCR3 expression could be one of the therapeutic effects of anti-TNF agents on RA FLS-induced inflammation in RA.

Methods
Patients. RA ST samples were obtained from patients undergoing arthroplasty. The patients with RA were diagnosed using the American College of Rheumatology 1987 revised criteria for RA. Serum was collected from 26 patients with RA before the initial treatment and from 28 healthy controls. The healthy controls were Japanese blood donors matched for gender and age. SFs were collected from 15 patients with RA before the initial treatment and 16 patients with osteoarthritis (OA). All specimens were obtained with informed consent and collected following approval from the Showa University Institutional Review Board. All methods were carried out in accordance with relevant guidelines and regulations.
Cell culture. Fresh ST samples were minced and digested in tissue enzyme digestion solution as described previously 26 . FLS were cultured in RPMI-1640 medium supplemented with 10% fetal bovine serum (FBS) (Omega Scientific), 100 units/mL penicillin, and 100 µg/mL streptomycin at 37 °C in a humidified 5% CO 2 atmosphere. The cells were seeded in 6-well plates (BD Biosciences, Bedford, MA, USA) at a density of 2 × 10 5 cells per well and were maintained in complete medium. For all experiments in this study, FLS were used between passages 4 and 9, and cells were synchronized in 0.1% bovine serum albumin (BSA) (serum starvation medium) for 24 h prior to analysis or functional assays.
THP-1 cells (a human acute monocyte leukemia cell line) were purchased from the American Type Culture Collection (Manassas, VA, USA). THP-1 cells were cultured in complete RPMI 1640 medium as described previously. www.nature.com/scientificreports/ Enzyme-linked immunosorbent assay (ELISA). ELISAs were performed as described previously 27 .
The levels of CCL11 in the cell supernatants of RA FLS-conditioned medium stimulated with 50 ng/mL TNF-α (R&D Systems, Minneapolis, MN, USA) with or without 100 µM of the specific CCR3 antagonist SB328437 (Abcam, Cambridge, MA, USA) were measured using an ELISA kit (R&D Systems) following the manufacturer's protocol. The plates were developed using tetramethylbenzidine as a substrate (TMB, Sigma-Aldrich, ST. Louis, MO, USA), and the absorbance was recorded using a microplate reader. The levels of CCL11, MCP-1/CCL2, MIP-1α/CCL3, RANTES/CCL5 and TNF-α in serum and SFs were also measured using ELISA kits (R&D Systems).
Quantitative real-time reverse transcription polymerase chain reaction (PCR). Following cell synchronization for 24 h, cells were left unstimulated or were stimulated for 12, 24 or 48 h with 50 ng/mL TNF-α. Total RNA was extracted using RNeasy Mini RNA isolation kits (Qiagen, Valencia, Spain) in accordance with the manufacturer's protocol. Complementary DNA (cDNA) was synthesized using a High-Capacity cDNA Reverse transcription Kit (Applied Biosystems, Waltham, MA, USA) per the manufacturer's protocol. CCL11, CCR3, and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) primers were purchased from Qiagen. Quantitative PCR was performed using SYBR Green quantitative PCR master mix (Qiagen) on a C1000 thermal cycler (Bio Rad, Hercules, CA, USA) as described previously 5 . The efficiency of the primer assays was guaranteed by the manufacturer to be > 90%. Each reaction was measured in triplicate, and the data were normalized to the expression levels of the housekeeping gene GAPDH. The ratio of each mRNA relative to the GAPDH mRNA was calculated using the ΔΔ threshold cycle method. Transwell migration assays for RA FLS. The transwell migration assays were performed in Transwell systems. RA FLS (5 × 10 5 ) were resuspended in assay media (RPMI with 0.5% BSA) with 200 ng/mL CCL11 and allowed to migrate through uncoated transwell chambers in response to 5% FBS for 4 h. Equal numbers of live RA FLS (5 × 10 5 ) that were pretreated with CCL11 siRNA or control siRNA for 24 h were resuspended in only assay media and allowed to migrate through uncoated transwell chambers in response to 5% FBS for 4 h. After 4 h, cells were stained with DAPI (Life Technologies, Carlsbad, CA, USA) for 20 min at room temperature. The fluorescence of the migrating cells on each membrane was visualized using an Olympus microscope IX71. Images were acquired from four nonoverlapping fields per membrane, and invading cells in each field were counted visually. Each experiment included four membranes per sample.

Immunohistochemical analysis.
Chemotaxis assays for THP-1. Chemotaxis assays were performed as described previously 28 . For the THP-1 chemotaxis assays, a 48-well Boyden chamber with a 5 μm polycarbonate membrane was used. The lower wells were filled with the stimulus solution. SF [1:50 dilution (in 0.5% BSA/RPMI) was used as the positive control, and 0.5% BSA/RPMI was used as the negative control. THP-1 cells in 0.5% BSA/RPMI at 1.2 × 10 6 /mL were added to the upper wells and incubated at 37 °C for 90 min in an incubator. After incubation, the membrane was stained with Diff-Quick. The migrated cells were counted by a blinded observer. Three high-power (400 ×) fields were counted in each well, and the results were expressed as the number of cells per high-power field.
Statistical analysis. The data were analyzed using Student's t test for normally distributed variables and the Mann-Whitney U test for non-normally distributed variables. The data are reported as the mean ± standard error of the mean (SEM). A p value less than 0.05 was considered statistically significant.

Data availability
The data that support the findings of this study are available from the corresponding author upon request.