Transcriptomic heterogeneity of non-beta islet cells is associated with type 2 diabetes development in mouse models.

AIMS/HYPOTHESIS: The aim of this work was to understand the role of non-beta cells in pancreatic islets at early stages of type 2 diabetes pathogenesis.

METHODS

Specific clustering was employed to single-cell transcriptome data from islet cells of obese mouse strains differing in their diabetes susceptibility (diabetes-resistant B6.V.Lep [OB] and diabetes-susceptible New Zealand Obese [NZO] mice) on a diabetogenic diet.

RESULTS

Refined clustering analysis revealed several heterogeneous subpopulations for alpha cells, delta cells and macrophages, of which 133 mapped to human diabetes genes identified by genome-wide association studies. Importantly, a similar non-beta cell heterogeneity was found in a dataset of human islets from donors at different stages of type 2 diabetes. The predominant alpha cell cluster in NZO mice displayed signs of cellular stress and lower mitochondrial capacity (97 differentially expressed genes [DEGs]), whereas delta cells from these mice exhibited higher expression levels of maturation marker genes (Hhex and Sst) but lower somatostatin secretion than OB mice (184 DEGs). Furthermore, a cluster of macrophages was almost twice as abundant in islets of OB mice, and displayed extensive cell-cell communication with beta cells of OB mice. Treatment of beta cells with IL-15, predicted to be released by macrophages, activated signal transducer and activator of transcription (STAT3), which may mediate anti-apoptotic effects. Similar to mice, humans without diabetes possess a greater number of macrophages than those with prediabetes (39 mmol/mol [5.7%] < HbA < 46 mmol/mol [6.4%]) and diabetes.

CONCLUSIONS/INTERPRETATION: Our study indicates that the transcriptional heterogeneity of non-beta cells has an impact on intra-islet crosstalk and participates in beta cell (dys)function.

DATA AVAILABILITY

scRNA-seq data from the previous study are available in gene expression omnibus under gene accession number GSE159211 ( https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE159211 ).