Session 10: Ground Breaking Technologies

Background: T-cell receptors (TCRs) used by T cells in the islets and antigens targeted by such T-cells may be utilized for T-cell biomarkers and therapeutic purposes for type 1 diabetes (T1D). Thus, we aimed to determine islet-specific TCR repertoires and their antigen specificity.
Methods: We isolated 1,478 CD4 and 1,123 CD8 T-cells from islets of six organ donors having T1D, and identified TCR alpha and beta chain sequences of each single cell using Illumina-MiSEQ. To seek antigen specificity, TCR transductants were analyzed for the response to candidate antigens.
Results: We identified 1,637 alpha and 1,967 beta in-frame TCR sequences. A large number of T-cells in each donor (26%-71%) had identical clonotypes, suggesting that these cells expanded upon antigen stimulation. Furthermore, 6 TCRs were detected in the islets of multiple donors. Notably, donors having the identical clonotypes shared at least one HLA class-I or II alleles when the clonotypes were found from CD8 or CD4 T-cells, respectively. Next, we analyzed antigen specificity for frequently detected TCRs. Preproinsulin 2-14 and 15-24 presented by HLA-A2 were recognized by TCRs that were found from CD8 T-cells in the islets of multiple donors. Regarding TCRs derived from CD4 T-cells, we detected reactivity to insulin B-chain 9-23 presented by DQ8 and DP4, and C:peptide 19-35 presented by DQ8-trans. Remarkably, further analysis determined that ZnT8 270-279 presented by DP4 is an epitope for one of the “public” TCRs detected in the islets of multiple T1D donors.
Conclusions: “Public” TCR clonotypes that may be utilizable for surrogate T1D T-cell biomarkers exist. We confirmed common reactivity to epitopes within preproinsulin and ZnT8 by CD4 and CD8 T-cells in the islets. Particularly, specificity to epitopes presented by HLA-DP4, which is expressed by approximately half of humans in the world and therefore can be beneficial for a large number of patients, suggests utilizing DP4-specific antigens as diagnostic and therapeutic tools for T1D. 

Backgroud: A significant gap in the study of the onset of the autoimmune response for Type 1 Diabetes (T1D) and its progression is the lack of biomarkers that can be used to accurately predict and monitor these processes. Therefore, two goal of The Environmental Determinants of Diabetes in the Young (TEDDY) study are to identify new biomarkers and obtain a mechanistic understanding of the autoimmune response leading to T1D. Methods: As a first step towards achieving these goals, we have performed integrative machine learning of multiple omics datasets (genomics, metabolomics and lipidomics), as well as associated meta-data, on a nested case-control cohort from TEDDY, a prospective study of children higher genetic risk of developing diabetes. We applied a pipeline that first identified the best machine learning classification algorithm per data type. We subsequently performed feature selection in the context of a Bayesian integration across the multiple data sources, cross-validation, and optimization via simulated annealing generating a likelihood that each feature is important to the panel predicting seroconversion. A hold-out set of 25% of the total number of available case-control pairs was used to validate the model. Results: We performed predictive modeling with the goal of identifying the features that separate cases from controls in increments of 3, 6, 9 and 12 months prior to seroconversion. The number of case-control pairs with adequate data across the multiple omics and demographic data ranged from 208 to 336 for the various time points. The matched cross-validated training sets returned an average area under the Receiver Operating Characteristic curve of over 0.71 and the feature sets included dozens of disparate markers that were relatively uniform across the various data sources. This proof-of-principle demonstrates the ability to identify candidate biomarker panels for prediction of time to seroconversion.

The majority of disease associated variants in type 1 diabetes (T1D) are located in non-coding DNA and are enriched in areas of open or active chromatin. Functional annotation of the genome has revolutionised our ability to understand the “grammar” of the non-coding genome and integration of such data with GWAS summary statistics showed strong enrichment of T1D-associated SNPs in enhancers in lymphocytes, thymus tissue and haematopoietic stem cells. The practice of nominating candidacy to the closest or most biologically relevant candidate gene in a disease-associated region ignores the growing evidence that enhancers can act over long distances and upon multiple genes. 
We used promoter capture Hi-C (PCHi-C), a method that can indicate which regions physically interact with gene promoters genome-wide, in combination with total RNA-sequencing, ChIP-seq, Immunochip and GWAS summary statistics to explore principles underlying induction of gene expression and to prioritise genes in 17 primary human haematopoietic cell types including activated CD4+ T cells.
Activation of CD4+ T cells induced changes in gene transcription that correlated with acquisition of promoter interacting regions (PIRs) and expression of enhancer RNAs and revealed the complexity underlying gene regulation where, for example, enhancers can interact with multiple gene promoters, “skip” multiple gene promoters and switch target promoters upon activation or differentiation. Using blockshifter, a method developed to examine the enrichment of GWAS summary statistics between tissue specific PIRs, we found T1D associated SNPs were most strongly enriched in activated CD4+ T cells when compared to non-activated CD4+ T cells and two non-lymphoid  haematopoietic cell types. Integration of PCHi-C data from 17 different blood cell types using a novel Bayesian method and a conservative statistical threshold identified 97 novel protein coding and 39 non-coding transcripts in 29 of the 58 T1D regions, 4 of which had no previously nominated candidate genes.  
 

Background: Type 1 diabetes (T1D) is characterised by autoimmune pancreatic beta cell destruction, resulting in insulin deficiency and hyperglycaemia. By the time of diagnosis, the autoimmune response has already destroyed many of the insulin-producing pancreatic beta cells, making it challenging to investigate the early phase of disease. This study used novel single-cell transcriptomic technology in a spontaneous model of T1D, the non-obese diabetic (NOD) mouse, to investigate the transcriptional profile of the variety of cells present in pancreatic islets prior to the onset of hyperglycaemia.
Methods: Pancreatic islets were purified from 10-12-week-old female NOD mice (n=10) prior to the onset of hyperglycaemia and from control 10-12 week-old B6 mice (n=2). After hand-picking and counting islets, a single-cell suspension was prepared from whole islet tissue. Bead-based 10X Genomics technology was used to prepare single-cell RNA-seq libraries, capturing 500-1000 islet cells per mouse. High-throughput sequencing was undertaken using HiSeq4000 at c.100k reads per cell. 
Results: The transcriptome of multiple pancreatic and infiltrating immune cell populations during the early stages of beta cell destruction was established, with detection of >2,000 transcripts per cell. Cells were grouped into distinct clusters based on their transcriptome, using a variety of single-cell analysis software including CellRanger. In NOD mouse islets, the transcriptome of approximately 1 in 3 cells was consistent with an immunological origin, including activated CD8+ lymphocytes, macrophages, dendritic cells and plasma cells. We also identified a subset of atypical pancreatic beta cells expressing immune response genes. This work, which allowed analysis of both the immune response and the beta cell response to the immune attack, demonstrates the potential of single-cell transcriptomics to reveal important new insights into the nature of early T1D.

High parameter single cell analysis has driven deep understanding of immune processes. Using a next-generation single-cell “mass cytometry” platform we quantify surface and cytokine or drug responsive indices of kinase target with 45 or more parameter analyses (e.g. 45 antibodies, viability, nucleic acid content, and relative cell size). Similarly, we have developed two advanced technologies termed MIBI and CODEX that enable deep phenotyping of solid tissue in both fresh frozen and FFPE formats (50 – 100 markers). Collectively, the systems allows for subcellular analysis from the 70nm resolution scale to whole tissue in 3D.
I will present evidence of deep internal order in immune functionality demonstrating that differentiation and immune activities have evolved with a definable “shape”. Further, specific cellular neighborhoods of immune cells are now definable with unique abilities to affect cellular phenotypes—and these neighborhoods alter in various cancer disease states. In addition to cancer, these shapes and neighborhoods are altered during immune action and “imprinted” during, and after, pathogen attack, traumatic injury, or auto-immune disease. Hierarchies of functionally defined trans-cellular modules are observed that can be used for mechanistic and clinical insights in cancer and immune therapies.