Multiomic profiling of diabetes-resistant and -prone mouse islets for the identification of diabetes risk gene (Insel-Multi-Seq)

Abstract

While the basic pathomechanisms of T2D are largely known, universally effective treatment strategies are lacking. This underlines the need to identify and influence new signaling pathways in order to develop individualized and effective therapies. Therefore, the overarching translational goal of the project is to contribute to a better understanding of diabetes pathogenesis, laying the groundwork for more precise or new treatment options. We previously evaluated pancreatic islet transcriptomes of obese mice with different diabetes risk to unravel the genetic basis of type 2 diabetes (T2D). Both bulk and single-cell RNA sequencing technologies were used to compare transcriptomes between diabetes-resistant (B6.V.Lepob/ob; ob/ob) and diabetes-prone (New Zealand Obese; NZO) mouse islets on an established feeding regimen (carbohydrate-free high-fat diet until 18 weeks of age, diabetogenic carbohydrate-rich high-fat diet for two days). We found several genes associated with T2D pathogenesis, including a significant overlap with human T2D genes. More recently, our single-cell RNA sequencing study of ob/ob and NZO islet cells demonstrated that ob/ob mice exhibit a protective β-cell cluster with indications of reduced β-cell identity, whereas NZO β-cells progressed toward stress-related clusters with a strikingly different expression pattern, presumably leading to later β-cell loss. Following up on these results, we intended to clarify (1) which alterations in regulatory regions might be responsible for differentially expressed genes (DEGs) in specific cell types, (2) the transcriptional regulators that are related to those DEGs, (3) to which extent SNPs have a specific effect in different cell types, and (4) discover which SNPs might contribute to the stress-related β-cell clusters in NZO. To this end, islets of male ob/ob and NZO mice with or without diabetogenic diet were isolated, dispersed into single cells, and cryogenically stored. Upon thawing, single nuclei were isolated from 2-3 animals per group and cDNA libraries generated for single-cell multiomics analysis of gene expression (GEX) and chromatin accessibility (Assay for Transposase-Accessible Chromatin; ATAC). Initial quality control data indicated successful generation of GEX and ATAC libraries from all groups. These libraries were sequenced and the obtained data underwent further processing. Detailed bioinformatic analysis to identify regulatory elements and cell-type-specific (transcription) factors driving the diabetes-prone phenotype of NZO mice are still ongoing. In the meantime, we established the main bioinformatic methods for integrating single-cell data sets from humans and mice in our group and have gained some initial insights from our earlier islet single-cell RNA-Seq data, which will be utilized and compared to the multiome analyses in the future. Regarding the role of non-β islet cells in T2D pathogenesis, we planned to remove β-cells from isolated and dispersed mouse islet cell suspensions (same groups as above) by means of fluorescent zinc staining coupled with fluorescence-activate cell sorting (FACS). Unfortunately, we did not achieve an efficient enrichment of non-β-cells using the original fluorescent zinc dye or an alternative dye. Further optimizations are now underway for an antibody-based sorting strategy.