IL-17^−/− congenic C57BL/6 mice (hereafter referred to as IL-17^−/− mice) originally provided by Dr. Iwakura from Tokyo University were bred and maintained under specific pathogen-free conditions in accredited animal facilities at Hanyang University. C57BL/6 mice were purchased from Orient Bio. K/BxN mice were obtained by crossing KRN TCR transgenic mice on a C57BL/6 background (K/B) with NOD mice (15). The study protocol was approved by the Institutional Animal Care and Use Committee of Hanyang University. All animal experiments were carried out in accordance with the committee guidelines and regulations

Serum was collected from 8~12-week-old arthritic K/BxN mice. Wild-type (WT) and IL-17^−/− mice at 7 weeks of age were injected intraperitoneally with K/BxN serum (150 µl/mouse). Arthritic signs were evaluated in a blinded manner, and disease severity was assessed using a previously described scoring system (15). Ankle thickness of both hind paws was measured axially across the malleoli using a caliper. Hind paws were removed on day 12 after serum transfer, fixed, and decalcified in 5.5% EDTA in phosphate-buffered formalin. The specimens were embedded in paraffin, sectioned, and stained with H&E.

Hind paw tissues were fixed, embedded in paraffin, and sectioned at 7 mm in thickness. Standard immunohistochemical methods were then applied. Goat rabbit anti-mouse IL-17 Ab (eBioscience, San Diego, CA, USA) was used at the appropriate dilution. The sections were incubated with secondary Ab and exposed to avidin-biotin-peroxidase complexes and 3.3′-diminobenzidine (all from Vector Laboratories, Burlingame, CA, USA), followed by counterstaining with 1% methyl green solution.

To prepare single cell suspensions from synovial cells of the mice, the synovial tissues around ankle joints were collected, sliced into small pieces, digested with 50 µg/ml Liberase (Roche, Switzerland) at 37℃ for 40 min, and then filtered through cell strainers with a 70-µm pore size. The synovial cells were cultured in RPMI 1640 (Welgene, Gyeongsangbuk-do, South Korea) medium containing 10% FBS (Gibco, Detroit, MI, USA) in the presence or absence of 10 ng/ml IL-23 (BD Biosciences, San Jose, CA, USA) or immune complexes for 48 h. Immune complexes were prepared as described previously (16). The cells were stimulated with a mixture of 40 ng/ml phorbol myristate acetate (PMA) (Sigma-Aldrich, St. Louis, MO, USA), 1 µg/ml ionomycin (Sigma-Aldrich), and Golgi-Stop (BD Biosciences) for the last 5 h of incubation to detect IL-17 expression by FACS.

IL-17 expression by synovial cells was detected by FACS as previously described (17). mAbs to mouse IL-17, Sca1, Thy1, CD11b, CD11c, Gr-1, NK1.1, and c-Kit were purchased from eBioscience, and mAbs to mouse CD4 and γδ TCR were obtained from BD Biosciences.

RNA was purified from synovial cells using TRIzol (Invitrogen, Carlsbad, CA, USA) and assayed by RT-PCR as described previously (18). The following primer sequences were used: IL-17, 5′-TCC AGA AGG CCC TCA GAC TA-3′ and 5′-AGC ATC TTC TCG ACC CTG AA-3′; IL-23, 5′-ATC CAG TGT GAA GAT GGT TG-3′ and 5′-GGA GTT GGC TGA GTC CTA GT-3′; β2 microglobulin (β2M), 5′-TGA CCAGCT TGT ATG CTA TC-3′ and 5′-CAG TGT GAG CCA GGA TAT AG-3′.

Data are presented as means±SEMs. Differences between groups were evaluated by unpaired Student's t-test. p values are indicated when differences between two groups were statistically significant (<0.05).

Autoantibodies abundant in K/BxN serum can induce arthritis in normal mice by initiating innate inflammatory responses in joints (4). To determine the role of IL-17 in the innate effector phase of arthritis, we assessed the susceptibility of IL-17-deficient mice to K/BxN serum-induced arthritis. Whereas all WT mice developed severe arthritis, evident in ankle thicknesses and arthritic indices, in response to the serum, none of the IL-17^−/− mice showed any symptoms of arthritis (Fig. 1A). Accordingly, severe leukocyte infiltration and bone invasion were observed in the hind-paw joints of serum-recipient WT mice, whereas the joint architecture of IL-17^−/− mice was intact compared to that of untreated mice (Fig. 1B). IL-17-expressing cells were detected in the hind-paw synovial tissue of serum-recipient WT mice, but not in that of the serum-recipient IL-17^−/− mice (Fig. 1C). In addition, IL-17 transcripts were more abundant in the synovial tissue of WT mice after serum transfer, but were undetectable in the tissue from IL-17^−/− mice, as expected (Fig. 1D). IL-23 transcripts were similarly increased in both strains of mice in response to serum transfer.

These results indicate that IL-17 is required for the development of arthritis triggered by K/BxN serum and suggest that autoantibodies promote IL-23 expression. IL-23, in turn, activates cells that are residing in or recruited to the synovial tissue to produce IL-17, an essential player in synovitis.

Previous studies have shown that IL-17 is produced not only by CD4⁺ Th17 cells but also by other types of cells such as γδ TCR⁺ T cells, iNKT cells, NK cells, mast cells, and innate lymphoid cells with diverse phenotypes (1119202122). To identify the IL-17-producing cells that mediated the K/BxN serum-induced arthritis, we extracted synovial cells from swollen synovial tissues when clinical signs reached their peak and assayed the tissues by FACS. We observed that a small but substantial fraction of IL-17-producing cells were present in the synovial cell subpopulation with higher FSC and SSC levels (hereafter referred to as FSC^intSSC^int cells), but not in the subpopulation with low FSC and SSC levels. The IL-17-producing cells lacked lineage markers specific for B220⁺ B cells, CD3⁺ T cells, CD4⁺ Th cells, γδ TCR⁺ T cells, CD11b⁺ myeloid cells, CD11c⁺ dendritic cells, Gr-1⁺ granulocytes, NK1.1⁺ NK and NKT cells, and c-Kit⁺ mast cells (Fig. 2A). The IL-17-producing cells were identified in a fraction of Thy1^hiSca1^int cells within the FSC^intSSC^int subpopulation (Fig. 2B). Despite the existence in both strains of mice even at the steady state, the Thy1^hiSca1^int cells produced IL-17 only in WT mice, not in IL-17^−/− mice, in response to serum transfer (Fig. 2C). Therefore, these results suggest that IL-17 was produced by cells resident in the synovial tissue, rather than by cell infiltrates in response to arthritogenic stimulation, and these cells retained a phenotype distinct from all other lineages known to express IL-17 in the synovium.

Because a certain population of innate lymphoid cells residing in the intestinal mucosa was recognized to produce IL-17 in response to IL-23 (22), we examined whether IL-17 expression by Thy1^hiSca1^int synovial cells could be induced by IL-23. To this end, synovial cells extracted from normal mice were stimulated with IL-23, and their IL-17 production was assayed by RT-PCR and FACS. IL-23 treatment greatly increased the level of IL-17 transcripts, as well as the proportion of IL-17-expressing cells within Thy1^hiSca1^int synovial cells (Fig. 3A).

Because autoantibodies present in K/BxN serum were the major arthritogenic factor, we evaluated whether autoantibody-containing immune complexes could activate Thy1^hiSca1^int synovial cells to produce IL-17. For this purpose, we generated K/BxN immune complexes by incubating K/BxN serum with GPI protein. A mixture of BxN serum and GPI protein (BxN immune complexes) was used as a negative control. Treatment of synovial cells with K/BxN immune complexes resulted in significantly enhanced IL-17 production by Thy1^hiSca1^int synovial cells, whereas the effect of BxN immune complexes was marginal (Fig. 3B).

These results, taken together, demonstrate that IL-23 and K/BxN immune complexes promote IL-17 production by Thy1^hiSca1^int synovial cells in vitro, although it is not clear whether IL-23 and K/BxN immune complexes act directly on Thy1^hiSca1^int synovial cells or elicit other synovium-resident cells to activate Thy1^hiSca1^int synovial cells. Interestingly, these cells share characteristics such as the Thy1⁺Sca1⁺ phenotype and responsiveness to IL-23 with IL-17-producing innate lymphoid cells identified in inflamed intestines.

Although previous studies have identified diverse populations of innate lymphoid cells, mostly residing in barrier tissues such as the skin, intestine, and lungs (23), there have been no reports of these cells residing in synovial tissues. Here, we provide evidence of IL-17-producing synovium-resident cells. We also revealed that these cells are responsive to immune complexes and IL-23 and thereby play a prominent role in inflammatory arthritis.