Background Psoriasis [Ps] is a common chronic inflammatory skin condition affecting about 1–2% of the world population; approximately 20–30% of patients have psoriatic arthritis [PsA]. PsA can affect peripheral and axial joints, eyes, ileum, colon and skin. Genetic studies reveal common candidate genes including IL23R, IL12B, STAT3, and CARD9, all associated with IL-23 signalling. We demonstrated increased frequencies of Th17 cells in patients with Ps and PsA and the clinical importance of IL-23/Th17 is revealed by the efficacy of biologics targeting this pathway. Whilst IL-17 was thought to be pathogenic in inflammatory conditions such as experimental autoimmune encephalomyelitis, the redundancy of IL-17 but essential requirement for GM-CSF release by Th17 cells was shown. GM-CSF+ T cells were identified in cerebrospinal fluid from patients with multiple sclerosis but a role in the aetio-pathogenesis of PsA has not been established.
Objectives 1) Are GM-CSF+ T cells enriched in peripheral blood and synovial fluid obtained from patients with PsA?
2) Is GM-CSF co-produced with IFN-g?
3) Do CD4+GM-CSF+ T cells express surface markers characteristic of Th17?
4) Is GM-CSF release augmented by exogenous IL-23 and/or IL-12?
Methods PBMC or synovial fluid derived from patients with PsA, or healthy donor (HD) PBMC, were stimulated with anti-CD3/anti-CD28 +/− recombinant IL-23 or IL-12. Supernatants were harvested and analysed by ELISA for IL-17, IFN-γ or GM-CSF. Alternatively, cells were activated with phorbol myristate acetate (PMA) and ionomycin, and evaluated for the expression of the Th17-associated cell surface markers CCR6, CD161 & IL-23R plus IL-17A, IFN-g and GM-CSF by flow cytometry.
Results GM-CSF release from PsA PBMC was higher than that of HD, whereas IFN-g levels were higher in HD. This was most significant on comparing the ratio of IFN-g to GM-CSF. On examining the co-expression of GM-CSF and IFN-g in CD4+ T cells, PsA patients had higher proportions of GM-CSF+ single positive and fewer GM-CSF+IFN-g+ relative to HD. There was also marked enrichment of CD4+GM-CSF+ cells in synovial fluid. On evaluating the expression of the Th17-associated markers CCR6, CD161 and IL-23R, we demonstrated that whereas almost all CD4+IL-17+ cells expressed CCR6, approximately 50% of CD4+GM-CSF+ cells expressed this chemokine receptor. Surprisingly, there were fewer PsA patient CD4+GM-CSF+ peripheral blood T cells co-expressing CCR6, CD161 and IL-23R relative to HD. Finally, IL-23 has been shown to render Th17 cells more pathogenic, hence we examined the effect of this cytokine on GM-CSF release. We showed that whilst IL-17 release was enhanced by exogenous IL-23, GM-CSF release was significantly downregulated with little effect on IFN-g. In contrast, IL-12 increased IFN-g and GM-CSF, and reduced IL-17.
Conclusions IL-23 and Th17 cells are therapeutic targets in PsA, and mouse models of inflammation suggest that pathogenic Th17 cells release GM-CSF. We reveal that GM-CSF release is elevated in patients with PsA, and GM-CSF+ cells are enriched in diseased joints. However, only a proportion of GM-CSF+ cells exhibited features of Th17. Furthermore, IL-23 downmodulated GM-CSF whilst enhancing IL-17. Given that GM-CSF was mainly produced by cells not co-expressing IFN-g in PsA patients [in contrast to healthy donors], this subset may be pathogenic in PsA and warrants further investigation.
Disclosure of Interest None declared