Single-cell proteomics of pancreatic islet cells reveals type 1 diabetes and donor-specific features

Abstract

Type 1 diabetes mellitus (T1DM) is the most common severe chronic disease in children and adolescents and requires life-long exogenous insulin treatment. Investigating key differences between the pancreas of healthy and T1DM patients is key to help uncover new treatments. Here we performed a single-cell proteomic (SCP) analysis of purified pancreatic islets from several healthy and T1DM donors at a depth of > 4000 protein quantified. We employed tools designed for single-cell mRNA sequencing (scRNAseq) for cell type annotation highlighting discrepancies between mRNA and protein markers. Subsequently, we defined a subset of markers useful for future proteomics analyses of pancreatic islets. This marker list was employed to refine the annotations of our combined dataset, which we exploited with a machine learning approach to successfully transfer annotations into a new islet dataset. Lastly, we showcase the applicability and value of SCP in a clinical context by providing novel insights into cell type-specific differences between healthy and T1DM islets as well as between healthy islets of different donors. Our study is the first SCP analysis of pancreatic islets from multiple donors and T1DM islets, performed at this level of analytical depth. This provides methodological and biological insight as well as a reference dataset for future SCP studies and for any researcher interested in T1DM and in pancreatic islet physiology.