GZMK+CD8+ T cells Target A Specific Acinar Cell Type in Sjögren's Disease

Abstract

Sjögren's Disease (SjD) is a systemic autoimmune disease without a clear etiology or effective therapy. Utilizing unbiased single-cell and spatial transcriptomics to analyze human minor salivary glands in health and disease we developed a comprehensive understanding of the cellular landscape of healthy salivary glands and how that landscape changes in SjD patients. We identified novel seromucous acinar cell types and identified a population of PRR4+CST3+WFDC2- seromucous acinar cells that are particularly targeted in SjD. Notably, GZMK+CD8 T cells, enriched in SjD, exhibited a cytotoxic phenotype and were physically associated with immune-engaged epithelial cells in disease. These findings shed light on the immune response's impact on transitioning acinar cells with high levels of secretion and explain the loss of this specific cell population in SjD. This study explores the complex interplay of varied cell types in the salivary glands and their role in the pathology of Sjögren's Disease.

Figure 1

Clinical research investigations for 25 subjects included comprehensive oral, rheumatological, and ophthalmologic investigations applying American College of Rheumatology 2016 Sjögren’s Disease Classification Criteria including salivary gland biopsies on all subjects. Single-cell RNA-sequencing data from 94,227 cells was gathered across 25 patients with and without SjD disease. (a) A cartoon depicting the sample collection from subjects and the creation of scRNA-seq and spRNA-seq libraries. Histological interpretations were rendered on all subjects and patients’ glands. Additional subjects, not included in the scRNAseq analyses, were used for flow cytometry, 10X Visium spatial transcriptomics, multiplex fluorescent in situ hybridization (PhenoCycler-Fusion), and multiplex immunofluorescence microscopy. (b) The microscopic appearance of minor salivary glands from non-SjD and SjD patients. Note the multiple scattered lymphocytic foci, periductal fibrosis, and atrophy characteristic of SjD. (c) A heatmap of clinical features for the scRNA-seq patients exhibiting the intrinsic clinical heterogeneity of subject phenotypes (e.g., SjD, non-SjD sicca). (d) PCA of patients using five of the clinical features. Importantly, the second principal component of the PCA analysis divides the patients on autoantibody positivity for anti-SSA autoantibodies (SSA+). (e) Leiden clustering followed by manual annotation based on gene expression and UMAP embeddings enables granular distinctions between cell types. Cells clustered with Leiden clustering in scanpy are labeled along with manual embeddings based on cell expression and boundaries in UMAP embeddings in boxes. In total, 31 unique cell types, including seven types of seromucous acinar cells, were identified. Two PRR4- SMAC populations were combined into one “transitioning” population. (f) Cells by the diagnosis of the patient, with and without SjD. (g) Cells clustered by anti-SSA autoantibody positivity from patients.

Figure 2

One seromucous acinar cell population represents the greatest loss of epithelium in disease. (a) Expression of acinar cell type markers in the scRNAseq data including in the six distinct populations of seromucous acinar cell types and two populations of mucous acinar cells. (b) Cell trajectory for ductal and seromucous acinar cells using wishbone, starting with the cell most positive for KRT5, a marker of progenitor state. The cell trajectory starts with a small population of ductal progenitors and proceeds along the ducts to the seromucous acinar cells, ending in the CST3− populations. (c) Gene expression (after log-transformation) for marker genes along the wishbone trajectory as a fraction of maximum expression. WFDC2 expression is lost first, followed by CST3. MUC7 expression peaks in the center and LYZ and ZG16B are highest in the most differentiated SMACs. (d) Statistically significant changes in cell proportions between SjD and nonSjD. PRR4+CST3+WFDC2− SMACs were the most abundant population in non-SjD patients and presented the greatest loss in SjD. (e) Representative images showing multiplex in-situ RNA hybridization of acinar and disease marker genes in salivary glands.

Figure 3

Clinical research investigations for 25 subjects included comprehensive oral, rheumatological, and ophthalmologic investigations applying American College of Rheumatology 2016 Sjögren’s Disease Classification Criteria including salivary gland biopsies on all subjects. Single-cell RNA-sequencing data from 94,227 cells was gathered across 25 patients with and without SjD disease. (a) A cartoon depicting the sample collection from subjects and the creation of scRNA-seq and spRNA-seq libraries. Histological interpretations were rendered on all subjects and patients’ glands. Additional subjects, not included in the scRNAseq analyses, were used for flow cytometry, 10X Visium spatial transcriptomics, multiplex fluorescent in situ hybridization (PhenoCycler-Fusion), and multiplex immunofluorescence microscopy. (b) The microscopic appearance of minor salivary glands from non-SjD and SjD patients. Note the multiple scattered lymphocytic foci, periductal fibrosis, and atrophy characteristic of SjD. (c) A heatmap of clinical features for the scRNA-seq patients exhibiting the intrinsic clinical heterogeneity of subject phenotypes (e.g., SjD, non-SjD sicca). (d) PCA of patients using five of the clinical features. Importantly, the second principal component of the PCA analysis divides the patients on autoantibody positivity for anti-SSA autoantibodies (SSA+). (e) Leiden clustering followed by manual annotation based on gene expression and UMAP embeddings enables granular distinctions between cell types. Cells clustered with Leiden clustering in scanpy are labeled along with manual embeddings based on cell expression and boundaries in UMAP embeddings in boxes. In total, 31 unique cell types, including seven types of seromucous acinar cells, were identified. Two PRR4- SMAC populations were combined into one “transitioning” population. (f) Cells by the diagnosis of the patient, with and without SjD. (g) Cells clustered by anti-SSA autoantibody positivity from patients.

Figure 4

CD8 T Exhausted T cells are enriched in the immune infiltrates of SjD and exhibit an effector phenotype. (a) Highly multiplexed (4/35-plex shown) immunofluorescence microscopy shows distribution of CD4+ and CD8+ T cells, as well as, CD68+ macrophages in the glands of SjD and nonSjD. (b,c) scRNAseq UMAPs of expression using only T cells, with disease diagnosis and anti-SSA positivity visualized. T cells from anti-SSA positive individuals cluster on the periphery while SjD-positive T cells are distributed throughout. (d) Expression of key T cell genes in exhausted and effector CD8+ T cells across disease severity. (e) Single cell pathway enrichment analysis shows similar activated profiles between GZMK+ CD8+ Exhausted T cells and CD8+ Effector T cells in SjD. (f) The potential for degranulation and cytotoxicity of T cells was measured ex vivo using flow cytometry-based T-lymphocyte cytotoxicity assay. CD45+ immune cells were dissociated from patients’ salivary glands (nonSjD: n=5 patients; SjD: n=6 patients). (g) Spatial plots of segmented and phenotyped multiplex immunofluorescence data confirm alterations in the cellular arrangement of the glands, and highlight T cells at the epithelial interface in SjD. (h) CellChat Ligand-receptor analysis of scRNAseq data shows enriched signaling pathways (e.g., “MHC-I signaling”) is specific to anti-SSA+ SjD with connections to multiple cell types and the physical location of CD8+ T cells to immune-involved epithelia is shown.

Figure 5

GZMK drives cellular innate immune signaling in SjD. (a) Multiplex immunofluorescence of salivary glands sections from SjD and nonSjD hybridized with anti-GZMK, -GZMB, and -CD8. Immune infiltrates enriched with GZMK+CD8+ T cells in close proximity to acinar structures and ducts are shown. (b) Fluorescent confocal microscopy was used to measure cytosolic mitochondrial DNA after GZMK protein transfection. The transfection of recombinant GZMK and GZMB were monitored by their His tag. Image analysis demonstrates increased mtDNA in the cytoplasm. (c) Multiplex immunofluorescence microscopy shows cytosolic transfection of GZMK drives phosphorylation IRF3 (pIRF3) and nuclear translocation in pSGEC.

Figure 6

The quantification of cell types per spot according to Cell2Location. (a) Cell cooccurrence scores most different in disease. Immune cell colocalization with the PRR4+CST3+WFDC2− seromucous population is highest in SSA+ Sjögren’s patients. (b) Cooccurrence score spatial plots depict the altered cooccurrences between cell types including loss of cooccurrence of ductal cells with fibroblasts and increased cooccurrence of ductal cells with ductal progenitors, CD8+ Exhausted T cells with PRR4+CST3+WFDC2− SMACS, and high ZG16B SMACS with PRR4+CST3-WFDC2− SMACS.