IL-23 is critical for induction of arthritis, osteoclast formation, and maintenance of bone mass

Abstract

The role of IL-23 in the development of arthritis and bone metabolism was studied using systemic IL-23 exposure in adult mice via hydrodynamic delivery of IL-23 minicircle DNA in vivo and in mice genetically deficient in IL-23. Systemic IL-23 exposure induced chronic arthritis, severe bone loss, and myelopoiesis in the bone marrow and spleen, which resulted in increased osteoclast differentiation and systemic bone loss. The effect of IL-23 was partly dependent on CD4(+) T cells, IL-17A, and TNF, but could not be reproduced by overexpression of IL-17A in vivo. A key role in the IL-23-induced arthritis was made by the expansion and activity of myeloid cells. Bone marrow macrophages derived from IL-23p19(-/-) mice showed a slower maturation into osteoclasts with reduced tartrate-resistant acid phosphatase-positive cells and dentine resorption capacity in in vitro osteoclastogenesis assays. This correlated with fewer multinucleated osteoclast-like cells and more trabecular bone volume and number in 26-wk-old male IL-23p19(-/-) mice compared with control animals. Collectively, our data suggest that systemic IL-23 exposure induces the expansion of a myeloid lineage osteoclast precursor, and targeting IL-23 pathway may combat inflammation-driven bone destruction as observed in rheumatoid arthritis and other autoimmune arthritides.

FIGURE 1

Systemic IL-23 induces joint inflammation and extensive bone loss in vivo. A, H&E-stained decalcified sections of the knee joint showing extensive subchondral, metaphyseal, and epiphyseal bone loss, hyperplastic bone marrow, and pannus formation 30 d after IL-23 MC injection compared with GFP MC control. Scale bars, 500 μm (top panels) and 250 μm (bottom panels). Representative images of three independent experiments. B, Clinical disease severity score after IL-23 MC injection (average of four independent experiments). C, Three-dimensional rendered images from the paw and the anterior and side knee views 60 d after IL-23 MC injection (representative images from three independent experiments). D, Day 90 whole-body scan of control mice compared with the IL-23 MC-injected mice showing severe loss of bone mineral density in IL-23 MC-injected mice. E, Serum cytokine levels 28 d after IL-23 MC injection compared with GFP MC control mice. Cytokines presented were statistically different from control using serum data pooled from three independent experiments. Dotted line shows sensitivity of the assay. p < 0.05, Mann–Whitney nonparametric test.

FIGURE 2

Systemic IL-23 induces myelopoiesis, serum TRAP5b, and IL-17A. Flow cytometry of (A) bone marrow and (B) spleen 11 d after IL-23 MC injection compared with control GFP MC mice. C, Serum TRAP5b 11 d after IL-23 MC injection compared with GFP MC control (p < 0.05, Mann–Whitney nonparametric test; data pooled from four experiments). D, Serum IL-23, IL-17A, and TNF levels 11 d after IL-23 MC injection compared with GFP MC control (p < 0.05, Mann–Whitney nonparametric test; data pooled from three experiments).

FIGURE 3

Systemic IL-23–induced arthritis and elevated serum TRAP5b is inhibited by depletion of phagocytic cells. A, IL-23 MC induced arthritis after depletion of CD4 T cells or phagocytic cells, and neutralization of IL-17A or TNF. GFP MC and IL-17A MC controls also are presented (data pooled from three experiments). B, Serum TRAP5b 11 d after IL-23 MC injection (ANOVA).

FIGURE 4

IL-23p19^−/− osteoclasts have impaired differentiation and function. A, Flow cytometry of bone marrow isolated from IL-23p19^−/− or control mice cultured for 24 h in the presence of M-CSF. B, BMM isolated from IL-23p19^−/− and control mice cultured for 6 d in the presence of M-CSF+RANKL form TRAP⁺ cells, which are also multinucleated (TRAP stain). Scale bars, 200 μm. C, Scanning electron photomicrographs of BMM cultures showing mature osteoclast resorbing activity (i.e., resorbed dentine has a rough, lighter color appearance). Scale bars, 50 μm. Representative data of three experiments done in triplicate.

FIGURE 5

Twenty-six–wk-old male IL-23–deficient mice have bone abnormalities caused by osteoclast dysregulation. A, Representative high-resolution μCT analysis of 26-wk-old male mouse’s trabecular bone in the distal femur. B, Von Kossa-stained sections of 26-wk-old male mouse’s distal femurs. Original magnification ×2. C, Toluidine blue counterstained von Kossa-stained sections of 26-wk-old male mouse’s distal femurs. Original magnification ×20. Black arrows indicate osteoclasts. D, BV/TV, Tb.N, OcS/BS, and OcN/OcS (NOc/OcS) measured in the distal femur. *p < 0.05.