Characterizing the Cellular Constituents of Proximal Airway Disease in Granulomatosis With Polyangiitis

Abstract

Objective: Granulomatosis with polyangiitis (GPA) is a rare multisystem autoimmune vasculitis. 10-20% of patients suffer life-threatening obstruction of their proximal airways. Although progress has been made in the treatment of systemic disease, ameliorating airway disease in GPA remains an unmet need arising from limited understanding of disease pathogenesis. We sought to characterize the cellular constituents of the affected proximal airway mucosa in GPA airway scar.

Study design: Basic/translational study.

Setting: Single tertiary care center.

Methods: Using single-cell RNA sequencing, we profiled the cellular constituents of proximal airway samples from GPA and disease comparators (GPA; n = 9, idiopathic subglottic stenosis: iSGS; n = 7, post-intubation proximal stenosis: PIPS; n = 5, and control; n = 10). We report transcriptomes for subglottic epithelial, immune, endothelial, and stromal cell types and map expression of GPA risk genes to tissue types present in the proximal airway. We compared differential gene expression across immune cell populations and performed pseudotime analysis using Monocle 3.

Results: Similar to iSGS and PIPS, the subglottic mucosa of GPA patients demonstrated an abundant immune infiltrate. 71% of the established GPA risk genes (10 of 14) localized to T cells and macrophages. Differential gene expression and pseudotime analysis revealed a sub-population of CD4-/CD8- inflammatory T cells that only originated from GPA.

Conclusion: We characterized the cellular composition of GPA airway disease and demonstrated that the expression of GPA risk alleles is predominantly localized to immune cell populations. We also identified a subset of inflammatory T cells that is unique to GPA.

Keywords: airway fibrosis; granulomatosis with polyangiitis; immunology; single‐cell RNA sequencing; subglottic stenosis.

Figure 1

Schematic diagram for (A) GPA‐associated auto‐immune vasculitis with (B) subglottic and (C) bronchial sites of airway stenosis. (D) Pipeline for processing of airway biopsy specimens, cell dissociation and subsequent single‐cell RNA sequencing.

Figure 2

iSGS, PIPS, GPA, and control scRNAseq samples were plotted via UMAP dimensional reduction into (A) specific cell types and (B) parsed into disease comparators. (C) Various cell populations were compared between disease and control states.

Figure 3

Quantification of immune, epithelial, endothelial, and mesenchymal cell populations was performed and compared between disease and control states. Data are shown as cell count as percentage of total cells.

Figure 4

(A) Expression of established GPA risk alleles was analyzed across cell types. (B) Heat map expression of GPA risk alleles superimposed on scRNAseq UMAP. (C) Quantification of genetic distribution of risk genes by cell type.

Figure 5

(A) Visualization of T cells from all scRNAseq samples (B) T cells labeled by disease with identification of T cell subset unique to GPA. (C) Pseudotime analysis with start time at the CD4+ effector population.