Background: Type-17 CD8+ tissue resident memory (T RM ) cells are enriched in the skin in psoriasis and in the joints in psoriatic arthritis (PsA) pointing to a key role for these cells in driving inflammation in both sites. We hypothesised that PsA is an antigen driven, tissue resident, IL-17+ CD8+ T cell mediated disease and that skin and synovial tissue inflammation is linked in terms of CD8+ T cell clonality and phenotype.
Objectives: The objectives of this study were 1) To determine whether synovial fluid (SF) is an appropriate proxy for investigating the inflammatory T cell infiltrate in synovial tissue (ST) in inflammatory arthritis; and 2) To compare the T cell receptor (TCR) repertoire and phenotype of T cells from paired samples of skin and ST and/or SF from patients with PsA.
Methods: Memory (CD45RA-CD27±and CD45RA+CD27-) T cells were sorted from paired samples of blood, skin epidermis, synovial fluid (SF) and synovial tissue (ST) from patients with PsA (n=7) and from blood, SF and ST from patients with RA (n=1) for 5’ single cell RNA and TCR sequencing.
Results: Comparison of SF and ST T cells from 6 patients (n=5 PsA, n=1 RA) revealed striking similarities in the TCR repertoires in SF and ST, particularly for CD8+ T cells. Furthermore, once artefact arising from collagenase digestion had been accounted for, the frequencies of T cell subsets were similar in SF and ST. These data indicate that SF is an appropriate proxy to investigate the T cell infiltrate in ST in inflammatory arthritis. Comparison of T cells from paired skin and ST and/or SF from 6 patients with PsA revealed an enrichment of type-17 CD8+ TRM cells in both sites. However, there were phenotypic differences between the two sites including a stronger IL-17 signature in the skin compared to increased expression of GZMK (encoding granzyme K) in the joint. Importantly, several T cell clones were shared between the skin and the joint. Across the six patients 155 CD8+ T-cell clones were shared between the two sites, comprising 1,071 CD8+ T cells and taking up a median of 13% of the skin and 8% of the joint CD8+ TCR repertoire. In the CD4+ T cell compartment there were 46 skin-joint shared clones, comprising 130 cells and taking up 6% of the skin and 1% of the joint TCR repertoire. CD8+ skin-joint shared clones tended to have a similar phenotype at both sites, characterised by increased expression of genes associated with a cytotoxic, tissue-resident phenotype.
Conclusion: Whilst there were phenotypic differences between T cells from the skin and the joint in PsA, T cells expressing identical TCRs and with similar phenotypes were present in both sites. Skin-joint shared clones had increased expression of cytotoxic genes suggesting that they were more activated and pro-inflammatory than their non-shared counterparts. Collectively these data indicate that in PsA the T cell infiltrate in the skin and joint is linked in terms of T cell clonality and raises the possibility that T cells migrate between the skin and the joint to propagate inflammation across both sites.
REFERENCES: NIL.
Acknowledgements: Supported by MRC and NIHR-BRC at KCL/GSTT.
Disclosure of Interests: Lucy Durham is currently funded by an MRC-AstraZeneca Industry Partnership fellowship, Frances Humby: None declared, Nora Ng: None declared, Sarah Ryan: None declared, Rosamond Nuamah: None declared, Kathy Fung: None declared, Athul Menon: None declared, Pawan Dhami: None declared, Bruce Kirkham: None declared, Leonie Taams has received research grant funding from Sanofi.