Background: Rheumatoid Arthritis (RA) is a progressive and systemic autoimmune disorder associated with chronic and destructive inflammation of the joints. In vitro approaches simulating RA synovial tissue are crucial in preclinical and translational research to expand our knowledge on RA human pathophysiology and to test new diagnostic and therapeutic applications. Recently, we developed a novel RA synovial tissue model by incorporating RA fibroblast-like-synoviocytes (RAFLS), endothelial cells (ECs), and macrophages, in which both angiogenesis and inflammatory processes can be studied.

Objectives: In the current study, we set out to examine the potential of the spheroid-based RA synovial tissue model in the presence of either “M1”-like or “M2”-like macrophages to modulate key RA inflammatory processes using the established TNF inhibitor etanercept, as well as novel small molecule inhibitors of the NF-κB intracellular signaling pathways.

Methods: Spheroids of RAFLS, ECs and either pro-inflammatory “M1”-like macrophages or anti-inflammatory “M2”-like macrophages derived from monocytes were formed, followed by growth in 3 dimensions in a collagen-based matrix. The spheroids were left unstimulated, or cultured in the presence of the pro-angiogenic factors VEGF/bFGF (GF), the pro-inflammatory cytokine Tumor-Necrosis Factor α (TNF) or RA synovial fluid (SF). Next, the 3D model was compared to investigate the impact of macrophage phenotype on different readouts. Finally, the 3D model containing the “M1”-like macrophages was used to test therapeutic compounds in the presence of GF, SF or TNF. Morphological changes were quantified using confocal imaging and digital image analysis by machine learning, while concentrations of pro-inflammatory soluble factors were measured in culture supernatants.

Results: The presence of SF significantly enhanced containment of the ECs and pro-inflammatory “M1”-like macrophages within the spheroid core compared to the unstimulated condition. Overall, the anti-inflammatory “M2”-like macrophages did not sustain within the core and migrated away to a larger extent than its “M1”-like counterpart. No significant changes were observed upon the different stimulation strategies. In the 3D model containing the “M1”-like macrophages, the addition of etanercept prior to GF stimulation decreased spheroid outgrowth by 30%, while a small molecule NF-κB-inducing kinase (NIK) inhibitor fully prevented spheroid outgrowth. Moreover, the concentrations of interleukin (IL)-6 in culture supernatants upon GF stimulation were diminished by around 30% for both compounds. Secondly, all compounds tested strongly inhibited macrophage containment within the core upon SF stimulation compared to the conditions with SF only. Lastly, stimulation with TNF triggered the spheroid core to collapse across all conditions and significantly increased IL-6 production, which could be inhibited by etanercept.

Conclusion: We demonstrated the capacity of the 3D model of synovial inflammation to replicate alterations in cellular interactions depending on macrophage phenotype. Moreover, we showed the inhibitory effects of therapeutic compounds on spheroid outgrowth, cell migration and soluble mediator production, highlighting the potential of this novel 3D model to test the effect of various therapeutic applications on RA synovial inflammation.

REFERENCES: [1] Philippon EML, van Rooijen LJE, Khodadust F, van Hamburg JP, van der Laken CJ and Tas SW (2023) A novel 3D spheroid model of rheumatoid arthritis synovial tissue incorporating fibroblasts, endothelial cells, and macrophages. Front. Immunol. 14:1188835. doi: 10.3389/fimmu.2023.1188835.

Acknowledgements: NIL.

Disclosure of Interests: Eva Philippon: None declared, Jan Piet Van Hamburg: None declared, Lisanne van Rooijen: None declared, Gary P Sims AstraZeneca, Conny J. van der Laken: None declared, Sander W. Tas: None declared