Project description:The Genomics and Transcriptomics of Human Insulinoma The common forms of diabetes - Types 1 and 2 - ultimately result from a deficiency of insulin-producing pancreatic beta cells. The Genomics and Transcriptomics of Human Insulinoma study was performed in order to identify novel approaches to inducing human pancreatic beta cells to replicate and regenerate. As a corollary, developing drugs that are able to expand human beta cell mass in people with diabetes should reverse diabetes. Unfortunately, identifying druggable pathways that can enhance human beta cell replication has been a major challenge. In 2017, there is only one class of drugs - the harmine analogues - that can induce human beta cells to replicate, and in this case, higher replication rates are desirable. Thus, identifying additional drugs and druggable pathways is a priority in diabetes research. Insulinomas are rare, benign adenomas of the pancreatic beta cell that cause excess insulin production and hypoglycemia: exactly the opposite of Types 1 and 2 diabetes. Beta cell proliferation rates in insulinomas are abnormally high. Thus, the premise for The Genomics and Transcriptomics of Human Insulinoma study is that benign human insulinomas hold the genomic and transcriptomic "recipe", and the repertoire of druggable pathways, that can be exploited to induce regeneration or replication of human beta cells in diabetes. Because, insulinomas are so rare, are almost always benign (non-malignant), and are easily resected by laparoscopic surgery, little attention has been paid to understanding the genomics or transcriptomics of insulinoma. There are at present only three published studies employing next-gen sequencing in insulinoma (<a href="https://www.ncbi.nlm.nih.gov/pubmed/?term=24326773">PMID:24326773</a>; <a href="https://www.ncbi.nlm.nih.gov/pubmed/?term=25787250">PMID:25787250</a>; and <a href="https://www.ncbi.nlm.nih.gov/pubmed/?term=25763608">PMID:25763608</a>). These studies contained 10, 7 and 8 insulinomas, respectively, and highlighted likely mutations in YY1 and MEN1. Our goal was to markedly expand the database and to add RNAseq to these earlier studies. The Genomics and Transcriptomics of Human Insulinoma study, in press in Nature Communications in 2017, reports next-gen sequencing on 38 insulinomas, by far the largest series of human insulinomas subjected to next-gen sequencing. This includes paired (genomic plus tumor) whole exome sequencing on 26 human insulinomas (22 sequenced at Mount Sinai, 4 downloaded from Cao et al, <a hre="https://www.ncbi.nlm.nih.gov/pubmed/?term=24326773">PMID:24326773</a>), and 25 sets of RNAseq from insulinomas, some of which also had paired whole exome seq, and some of which did not. The insulinoma RNAseq was compared to RNAseq from 22 sets of FACS-sorted normal human beta cells. Since insulinomas are so rare, the 38 insulinomas were collected by several investigators at several institutions over several decades, but most (22 whole exome sets, and all RNAseq) were sequenced at the Icahn School of Medicine at Mount Sinai in New York. The current dataset contains whole exome seq and RNAseq on the 11 insulinomas harvested at Mount Sinai. The four from Cao et al can be retrieved from Cao et al <a href="https://www.ncbi.nlm.nih.gov/pubmed/?term=24326773"> PMID:24326773</a>. Fastq files from the remaining 23 insulinomas will be added as the local IRBs and Institutional Certifications are acquired. Complete patient data are provided in our Nature Communications report. Going forward, our intention is to expand this series, with the goal of sequencing 100 human insulinomas. These will be added to dbGaP as they accrue. Paired-end whole exome seq (mean usable sequencing depth 79X and 105X for blood and insulinoma, respectively) was performed using an Illumina HiSeq 2500. Insulinoma and sorted normal beta cell RNAseq was performed on Ribozero and polyA paired end libraries using the Illumina HiSeq 2500. Complete sequencing and bioinformatic details are provided in our Nature Communications report. The principal findings from the study are that although each insulinoma has a different set of presumptive driver mutations, the majority converge on genes that are members of the Polycomb Complex, Trithorax Complex and other epigenetic modifying enzymes. In addition, 20% of insulinomas have copy number loss or loss of heterozygosity of all or most of chromosome 11, and the majority display abnormalities in CpG methylation and imprinting control on the imprinted Chr 11 p15.5-15.4 region that contains <a href="https://www.ncbi.nlm.nih.gov/gene/?term=INS">INS</a>, <a href="https://www.ncbi.nlm.nih.gov/gene/?term=IGF2">IGF2</a>, <a href="https://www.ncbi.nlm.nih.gov/gene/?term=CDKN1C">CDKN1C</a>, <a href="https://www.ncbi.nlm.nih.gov/gene/?term=KCNQ1">KCNQ1</a>, and other genes involved in beta cell specification and proliferation.

Project description:Background Insulinoma is the most common pancreatic neuroendocrine tumour in dogs and humans. The understanding of driving factors and critical survival genes in insulinomas is limited and overall survival is poor for canine and human malignant insulinoma. This study aimed to use single-cell RNA-sequencing to conduct a multispecies analysis of insulinoma cell lines to understand their single-cell transcriptomic landscape. Secondly, the impact of freeze-thawing on the pancreatic beta single-cell transcriptome was investigated, to determine whether cryoarchiving of primary insulinoma samples may be feasible in future studies. Methods Single-cell transcriptomic analysis was performed using fresh and cryopreserved multispecies insulinoma cell lines (canINS, CM, INS-1 and MIN6). R and Seurat were used to perform cell clustering and specific cluster marker genes were identified by the FindMarkers function. Metascape was used to identify statistically enriched pathways for specific cell clusters. Differentially expressed genes between fresh and cryopreserved single-cell transcriptome profiles, were defined as genes with a log2 fold change >0.25 and a Bonferroni-adjusted P<0.05, based on the Wilcoxon rank sum test. Results Based on the specific cell line single-cell transcriptome profiles, five or six cell clusters were constructed per cell line. All cell lines expressed neuroendocrine markers and additionally INS-1 and MIN6 displayed a gene signature indicative of mature/functional pancreatic islet/beta-cells. DEPTOR, BICC1, GHR, CCNB2, CENPA, LMO4, VANGL1, and L1CAM were identified as cross-species conserved insulinoma cluster marker genes. Little effect was found of cryopreservation and thawing on overall gene expression at the single-cell level in insulinoma cell lines: only 6 and 29 genes had a log2 fold change > 1 in cryopreserved versus fresh canINS and CM, respectively. Conclusions canINS, CM, INS-1 and MIN6 are all principally relevant as insulinoma models and the demonstrated differences in their single-cell transcriptomic profiles could aid researchers in selecting the appropriate cell lines for specific study objectives. Cross-species conserved insulinoma cluster marker genes were identified that harbour oncogenes and their involvement in insulinoma tumourigenesis should be investigated in future studies. The good comparability between cryopreserved and fresh insulinoma cells allows for inclusion of cryopreserved insulinoma patient samples in future studies, which allows for reduced assay-based variability.

Dataset Information

Genomics and Transcriptomics of Human Insulinoma

Dataset's files

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets