| dc.description.abstract | Rheumatoid arthritis (RA) and Psoriatic arthritis (PsA) encompass two of the most prevalent forms of inflammatory arthritis (IA), an umbrella term for the vast range of autoimmune driven arthropathies characterised by highly dysregulated and destructive immune and stromal activity within the joint, resulting in debilitating effects on patient quality of life. Several features are common to both pathologies including the extensive synovial inflammation which precedes bone and cartilage erosion in addition to the widespread systemic manifestations observed. However, although not fully understood currently, it is becoming increasingly recognised that there are a number of distinguishing characteristics in relation to circulating biomarkers in addition to disease pathogenesis at the clinical, immunological, cellular, and molecular levels. At the cellular level, a critical perpetuator of disease in both arthropathies is the fibroblast-like synoviocyte (FLS), a key stromal cell that assumes a central role in maintaining the integrity of the joint under homeostasis. However, as of yet, differences in their transcriptional profile and regulation in RA vs PsA remain unexplored. Therefore, the aims of this thesis are to identify phenotypic and functional characteristics that define distinct FLS populations and regulators in RA vs PsA, in addition to identifying circulatory miRNA as cellular biomarkers that can distinguish RA from PsA and to evaluate the potential implications for disease pathogenesis.
In Chapter 2, following global scRNA-seq analysis of all cells within the synovium and through receptor-ligand interactions analysis, the existence of a population of THY1+FAP+ FLS enriched in RA specifically was identified, whose pathogenic activity was being driven by the combined effects of IL-1� and TGF-�, derived from macrophages and T cells respectively, within the inflamed joint microenvironment. It was demonstrated that IL-1� and TGF-� synergistically/additively interact to promote the expression of several proinflammatory cytokines and chemokines, adhesive and angiogenic molecules, and matrix degrading MMPs, whilst altering the cellular bioenergetics in favour of glycolysis. Moreover, this was paralleled by changes in endoplasmic reticulum stress and mitochondrial fission/fusion properties, further supporting mitochondrial dysfunction and invasive capacity in RA FLS. Finally, it was demonstrated that addition of Takinib, the TAK1 inhibitor, could suppress some of these synergy-induced effects both in RA FLS and RA ex vivo synovial explants, highlighting the central role of this downstream molecule in coordinating the synergistic interactions of IL-1� and TGF-� in RA FLS. Cumulatively, the data suggests that the pathogenic nature of specific populations of FLS enriched in RA is in part driven by synergistic interactions between T cell and macrophage-derived cytokines, thus, blocking specific immune-stromal cell interactions may offer new avenues of therapeutic intervention in RA and PsA.
Whilst emphasis has been placed on stratifying phenotypically and functionally distinct subsets of FLS in RA, analysis of FLS subsets in PsA exclusively, in addition to direct comparisons of FLS populations in RA vs PsA remain largely uninvestigated. Therefore, expanding on Chapter 3, for the first time, an enrichment of THY1+ sub-lining FLS in RA and conversely, an enrichment of THY1- lining FLS in PsA was identified. Moreover, following cell sorting, it was revealed that THY1+ FLS are functionally distinct from THY1- FLS with heightened angiogenic and chemokine markers observed in the THY1+ population, and greater MMPs in the THY1- FLS. This correlated with the histological analysis demonstrating significantly elevated immune cell infiltration and lymphoid aggregate presence in RA whilst lining layer thickness was increased in PsA. Further flow cytometry stratification of FLS subsets identified six distinct FLS populations in RA and PsA based on surface marker expression of THY1, in addition to CD34, CD55, and FAP, with two of these populations enriched in RA: THY1+CD34+CD55-FAP+ and THY1+CD34-CD55-FAP+ and conversely, two others enriched in PsA: THY1-CD34-CD55+FAP+ and THY1+CD34-CD55+FAP+. In parallel, reanalysis of scRNA-seq data on FLS exclusively revealed sixteen distinct FLS clusters, with enrichment of three sub-lining clusters in RA compared to PsA: THY1+POSTN+, CXCL12+CHI3L2+, and CXCL12+APOE+. scRNA-seq analysis focused only on synovial FLS and ex vivo experiments highlighted differences in their functional properties with immune/inflammatory responses associated with RA dominant populations in contrast to high matrix degradative gene expression in PsA dominant populations. Finally, whilst passage 0 RA and PsA FLS maintained similar profiles to ex vivo FLS, once expanded to passage 3, the FLS started to lose specific phenotypic characteristics, suggesting that once removed from the joint microenvironment FLS subset stability appears transient with convergence towards common phenotypes.
Lastly, it has been widely accepted that the earlier the diagnosis the superior the outcomes in rheumatic disease, thus identifying biomarkers that can provide a reliable and non-invasive method of achieving this are important. In this thesis, seven miRNAs were successfully identified that were significantly increased in RA serum compared to PsA and importantly also healthy controls (HC). Further sub analysis revealed that these miRNAs were not affected by the ACPA status of the patient while the differences between RA and PsA were not affected by differences in disease activity measures. Moreover, further PCA and biplot analysis determined that three miRNAs in particular were more skewed towards RA in contrast to PsA, specifically miR-22-3p, miR-223-3p, and miR-29b-3p. Downstream analysis using DIANA and STRING software identified several pathways of interest as being primary targets of these three specific miRNA all of which assume important roles in driving aspects of RA pathogenesis such as angiogenesis, cell death, and bone metabolism. These included the FoxO signalling pathway, the Hippo signalling pathway, and importantly the P13K-AKT pathway which had the greatest number of target genes at 87. Thus this data highlights the valuable nature of using circulating miRNA not only as diagnostic biomarkers that can distinguish RA from PsA, but additionally to deepen understanding of their differential pathogenesis.
Cumulatively, the data collated in this thesis suggests there is considerable heterogeneity between RA and PsA, not only at the cellular but additionally the biomarker level. It provides a deeper insight into the roles of FLS subsets with unique functions in driving differential pathogenic mechanisms in RA and PsA, whilst highlighting the impact of immune-stromal communication in augmenting this activity. Moreover, the data indicates that serum miRNA may indeed provide a method of distinguishing RA from PsA patients. Taken together, this knowledge can facilitate earlier and more rapid diagnosis whilst additionally, improving understanding of disease pathogenesis and consequently, the implementation of more targeted therapeutic approaches in RA and PsA. | en |
