Single-cell transcriptomics of the human endocrine pancreas

Human pancreatic islets consist of multiple endocrine cell types. To facilitate the detection of rare cellular states and uncover population heterogeneity, we performed single-cell RNA sequencing (RNA-seq) on islets from multiple deceased organ donors, including children, healthy adults, and individuals with type 1 or type 2 diabetes. We developed a robust computational biology framework for cell type annotation. Using this framework, we show that a- and β-Cells from children exhibit less well-defined gene signatures than those in adults. Remarkably, a- and β-Cells from donors with type 2 diabetes have expression profiles with features seen in children, indicating a partial dedifferentiation process. We also examined a naturally proliferating α-cell from a healthy adult, for which pathway analysis indicated activation of the cell cycle and repression of checkpoint control pathways. Importantly, this replicating α-cell exhibited activated Sonic hedgehog signaling, a pathway not previously known to contribute to human a-cell proliferation. Our study highlights the power of single-cell RNA-seq and provides a stepping stone for future explorations of cellular heterogeneity in pancreatic endocrine cells.