Interstitial lung diseases (ILDs) include a broad range of lung disorders characterised by the progressive scarring of interstitium in the lungs resulting in stiff lungs, coughing, and breathing difficulties. Late-stage ILD patients develop pulmonary fibrosis (PF) and fibrosis progression varies greatly among patients. Single-cell RNA sequencing (scRNA-seq) studies have been performed to characterise the cell compositions in healthy and fibrotic lungs to investigate lung fibrosis. With spatial transcriptomics technologies available, studying the molecular mechanisms of ILDs and fibrosis progression within the spatial contexts of tissues can provide new insights.
We measured gene expression for 343 genes across 28 lung tissue samples from 6 healthy and 13 ILD donors in different disease states using the 10x Xenium platform. The Xenium platform is an imaging-based single-molecule resolution method with which we observed 210 million spatially detected transcripts along with pathology (H&E) staining images of our 28 samples. We recovered more cell types compared to scRNA-seq studies, including improved recovery of fibroblasts and endothelial cells which are difficult to capture in scRNA-seq assays.
To characterise the cellular and tissue structures across tissue slices and across samples, we obtained computationally segmented spatial niches from both cell segmentation-free and cell-based analyses. We were also able to map computationally discovered niches to clinician-annotated pathology features on pathology images leading to more comprehensive understanding of PF. Specifically, we identified epithelium-depleted alveolar structures, macrophage accumulation within air spaces, and detachment of aberrant KRT5-/KRT17+ cells from the local basement membrane. To further leverage the available spatial information and study functional units in lungs, we segmented alveoli from all samples and inferred their ordering across a spectrum of disease severity to investigate the plasticity of dysregulated epithelium. Taken together, these findings using integrated approaches advance our understanding of the molecular programs regulating lung function and PF development in spatial contexts.