Lung tuft single cell transcriptomics

While the lung bears significant regenerative capacity, severe viral pneumonia can chronically impair lung function by triggering dysplastic remodeling. The connection between these enduring changes and chronic disease remains poorly understood.  We recently described the emergence of tuft cells within Krt5+ dysplastic regions after influenza injury. Using bulk and single cell transcriptomics, we characterized and delineated multiple distinct tuft cell populations that arise following influenza clearance. Distinct from intestinal tuft cells which rely on Type 2 immune signals for their expansion, neither IL-25 nor IL-4Ra signaling are required to drive tuft cell development in dysplastic/injured lungs. Furthermore, tuft cells were also observed upon bleomycin injury, suggesting that their development may be a general response to severe lung injury. While intestinal tuft cells promote growth and differentiation of surrounding epithelial cells, in the lungs of tuft cell deficient mice, Krt5+ dysplasia still occurs, goblet cell production is unchanged, and there remains no appreciable contribution of Krt5+ cells into more regionally appropriate alveolar Type 2 cells. Together, these findings highlight unexpected differences in signals necessary for lung tuft cell amplification and establish a framework for future elucidation of tuft cell functions in pulmonary health and disease. Post-influenza tuft cells were sorted from day 28-post influenza lung. Single-Cell RNASeq was performed using the Chromium System (10x Genomics) and the Chromium Single Cell 3’ Reagent Kits v2 (10x Genomics) at the Children’s Hospital of Philadelphia Center for Applied Genomics. As with bulk RNA-Seq, ~3000 EpCam+ Trpm5-GFP+ cells were sorted into PBS + 0.1% BSA from mice at day 28 post-influenza and loaded onto the 10x Chromium system. After sequencing, initial data processing was be performed using Cellranger (v.3.1.0). Cellranger mkfastq was used to generate demultiplexed FASTQ files from the raw sequencing data. Next, Cellranger count was used to align sequencing reads to the mouse reference genome (GRCm38) and generate single cell gene barcode matrices. Post processing and secondary analysis was performed using the Seurat package (v.4.0). First, variable features across single cells in the dataset will be identified by mean expression and dispersion. Identified variable features was then be used to perform a PCA. The dimensionally-reduced data was used to cluster cells and visualize using a UMAP plot. Contaminating non-tuft cells were removed by sub-setting data, requiring counts for Trpm5 > 1. Raw reads and CellRanger output data is deposited at GEO.