Background: Sjogren’s disease (SjD) primarily affects the salivary and lacrimal glands, resulting in tissue inflammation that is characterised by formation of tertiary lymphoid structures (TLS), and a loss of glandular function resulting in dryness of the eyes and mouth (sicca symptoms), fatigue and poor health-related quality of life. TLS are accumulations of lymphoid cells sharing similar cellular compartments, organization, and function as secondary lymphoid organs. Notably, the presence of TLS in inflamed salivary glands (SGs) is associated with active disease, increased autoantibody production, and an elevated risk of malignancies, including B-cell lymphoma. Drug development for SjD has predominantly focused on suppressing aberrant immune cells but have failed to show efficacy in clinical studies. To date, SjD lacks any approved biological therapies that directly target its underlying pathogenesis.

Objectives: To treat patients effectively, comprehensive understanding of their SG microenvironment is needed. Current profiling efforts often struggle to capture high-plex ‘omics data while preserving the spatial architecture of the tissue.

Methods: To generate a SjD patient SG atlas we utlized both 10x scRNAseq Genomics platform (on enzymatically dissociated tissue) and CosMx™ Spatial Molecular Imager (SMI) using the Human Universal Cell Characterization (UCC) Panel (on FFPE SG tissue sections). SjD patients fulfilled 2016 ACR/EULAR classification criteria.

Results: We were able to map both identified cell types from scRNAseq as well as novel populations in SjD SGs CosMx data. We unbiasedly clustered tissue architectural features (cellular niches) using cellular neighbourhood analysis to define 8 neighbourhoods. Neighbourhood 1 and 2 were mainly enriched with epithelial cells. However, other neighbourhoods were mainly associated with different immune cell populations. Interestingly, IgG plasma cells were present within both Neighbourhood 5-6 which had features of TLS-germinal centre (GC), as well as in neighbourhood 4 and 8. Neighbourhood 8 was also enriched with IgA plasma cells and was distinctly away from TLS niche scattered across the SG. The IgA neighbourhood 8 was associated with myeloid populations in contrast to other IgG plasma cells niches. The T cell populations ( CD4+CCR6+CD40L+LTb+IFNG+TNF+ and CD4+CXCL13+CD40L+ ) were mainly confined to neighbourhood 5 of the TLS niche. Intriguingly, we also detected Treg and GranzymeK+CD8 T cells in neighbourhood 5. We identified a BAFF+CCL19+CXCL13+CD40+ immunostimulatory fibroblast (immunofibroblast) population in the TLS-GC neighbourhood 5-6. However, at the periphery of TLS-GC in neighbourhood 7, we identified a pericyte cluster expressing lymphoid chemokines CCL21 and CCL19 , both implicated in TLS formation. However, gene-set profiling of this pericyte population significantly differs from immunofibroblasts by lacking TNFSF13B, CD82 , and PDGFRA expression. Notably, NFkB gene-set enrichment is exclusive to the immunofibroblast subcluster, with RELB and NFKB2 expression limited to immunofibroblasts, confirming that only immunofibroblasts, but not CCL21+CCL19+ pericytes, rely on the lymphotoxin β receptor (LTβR) signalling pathway to produce lymphoid chemokines. CCL21+CCL19+ pericytes, on the contrary expressed both TNFα and IFNγ associated genes, such as IFNGR1 , IFNGR2 , TNFRSF1A , IRF1 and SOCS1 . Cell-chat analysis indicated CCL21+CCL19+ pericytes crosstalk by HLA interaction with IL7R+IFNG+TNF+ CD8 T cells.

Conclusion: Overall, spatial mapping of SjD SGs has the potential to reveal novel cellular landscape and their interactions, aiding therapeutic and discoveries.

REFERENCES: NIL.

Acknowledgements: NIL.

Disclosure of Interests: None declared.