Global, integrated analysis of methylomes and transcriptomes from laser capture microdissected bronchial and alveolar cells in human lung

Gene regulatory analysis of highly diverse human tissues <i>in vivo</i> is essentially constrained by the challenge of performing genome-wide, integrated epigenetic and transcriptomic analysis in small selected groups of specific cell types. Here we performed genome-wide bisulfite sequencing and RNA-seq from the same small groups of bronchial and alveolar cells isolated by laser capture microdissection from flash-frozen lung tissue of 12 donors and their peripheral blood T cells. Methylation and transcriptome patterns differed between alveolar and bronchial cells, while each of these epithelia showed more differences from mesodermally-derived T cells. Differentially methylated regions (DMRs) between alveolar and bronchial cells tended to locate at regulatory regions affecting promoters of 4,350 genes. A large number of pathways enriched for these DMRs including GTPase signal transduction, cell death, and skeletal muscle. Similar patterns of transcriptome differences were observed: 4,108 differentially expressed genes (DEGs) enriched in GTPase signal transduction, inflammation, cilium assembly, and others. Prioritizing using DMR-DEG regulatory network, we highlighted genes, e.g., <i>ETS1</i>, <i>PPARG</i>, and <i>RXRG</i>, at prominent alveolar vs. bronchial cell discriminant nodes. Our results show that multi-omic analysis of small, highly specific cells is feasible and yields unique physiologic loci distinguishing human lung cell types <i>in situ</i>.