Background/Purpose: Orchestration of immune checkpoints is central for the outcome of immune activation, especially in patients with rheumatoid arthritis (RA). We have previously shown that galectin-9 (Gal-9) binding to 4-1BB is central for the increased cytokine production in RA¹. Here we extend these studies, showing that Gal-3 has the opposite effect, leading to decreased 4-1BBL signaling in RA.

We hereby aim to investigate the modulatory potential of Gal-3 on 4-1BB in RA.

Methods: Gal-3 and 4-1BB were measured in plasma, synovial fluid and synovial tissue samples from RA patients (n=8). Fluorescence polarization and surface plasmon resonance (SPR) analyses were used to evaluate the binding between 4-1BB, 4-1BBL and Gal-3. Synovial fluid mononuclear cells (SFMCs) from patients with RA were cultured for 24 hours and co-incubated with either 4-1BB, 4-1BBL, Gal-3, or combinations thereof. Nanoparticle Tracking Analysis (NTA) was used to detect Gal-3 and 4-1BB complexes in healthy control (HC) plasma (n=10), RA plasma, and synovial fluid samples (n=10). Flow cytometry and Imagestream analyses were performed to detect complex-depletion to the cell membrane and evaluate the functional implications of complex-depletion. Glycan dependence was shown by addition of a Gal-3-specific inhibitor, lactose, or using PNGase F pretreatment of the cells.

Results: In RA, Gal-3 and 4-1BB levels in synovial fluid were increased by a factor 4 and 12 compared with paired plasma (p< 0.05), (p< 0.01) respectively. Cells expressing 4-1BB also co-expressed Gal-3 in the inflamed RA synovial tissue. Gal-3 was capable of binding to both soluble and membrane bound 4-1BB with a K_D=1.79 m M, without interfering with the binding of 4-1BBL. In plasma, Gal-3 was detected as part of complexes > 300 nm in size. Activated 4-1BB⁺ T cells and 4-1BB transfected HEK293 cells were both capable of depleting these complexes from plasma (p< 0.01) . After cell mediated depletion, expression levels of Gal-3 and 4-1BB complexes were increased on the cell surface compared with untreated cells (p< 0.05). T he increase in 4-1BB expression was accompanied by a 4-fold decreased in TNFa production (p< 0.01). Gal-3^high4-1BB^high T cells, were now found to be less responsive to 4-1BBL stimulation. In RA patients, complexes containing Gal-3 were dramatically reduced in both plasma and synovial fluid (p< 0.01). The level of Gal-3 complexes in RA plasma correlated inversely with disease activity (DAS28crp). Production of MCP-1 in SFMC cultures stimulated with Gal-3, followed by 4-1BBL, decreased by 50% compared with untreated cultures or cultures stimulated with either 4-1BBL or Gal-3 alone (p < 0.05).

Conclusion: Gal-3 is a carbohydrate-dependent 4-1BB binding protein, which does not block the co-binding of 4-1BBL. In plasma and synovial fluid, Gal-3 binds to s4-1BB and forms complexes which again can bind to T cells, followed by a decreased cytokine production. These observations provide evidence that galectins control the effects of 4-1BB signaling in RA by decoy binding, highlighting the interplay between galectins and an important checkpoint molecule in RA.

References:

1. Nielsen MA et al. Rheumatology (Oxford). 2016

« Back to ACR Convergence 2020

ACR Meeting Abstracts - https://acrabstracts.org/abstract/galectin-3-decreases-the-activity-of-4-1bb-by-facilitating-its-decoy-surface-binding-in-rheumatoid-arthritis/