Single-cell transcriptomic and spatial landscapes of the developing human pancreas

Current differentiation protocols have not been successful in reproducibly generating fully functional human beta cells in vitro, partly due to incomplete understanding of human pancreas development. Here, we present detailed transcriptomic analysis of the various cell types of the developing human pancreas, including their spatial gene patterns. We integrated single-cell RNA sequencing with spatial transcriptomics at multiple developmental time points and revealed distinct temporal-spatial gene cascades. Cell trajectory inference identified endocrine progenitor populations and branch-specific genes as the progenitors differentiate toward alpha or beta cells. Spatial differentiation trajectories indicated that Schwann cells are spatially co-located with endocrine progenitors, and cell-cell connectivity analysis predicted that they may interact via L1CAM-EPHB2 signaling. Our integrated approach enabled us to identify heterogeneity and multiple lineage dynamics within the mesenchyme, showing that it contributed to the exocrine acinar cell state. Finally, we have generated an interactive web resource for investigating human pancreas development for the research community.

Abstract 5710: Identification of convergent gene repression mechanisms through integrative genomic and DNA methylation analysis in TRACERx

Introduction: Lung TRACERx is a prominent study employing multi-region and longitudinal multi-omics sequencing to unravel the evolutionary trajectories of lung cancer. Aberrant DNA methylation patterns have been widely described in nearly all human cancers, yet their interplay with DNA mutations in lung cancer is not well understood. Incorporating the contribution of epigenetic modifications to cancer evolution trajectories within TRACERx could improve our understanding of the intricate relationship between genetic and epigenetic changes in non-small cell lung cancer (NSCLC) evolution.

Methods: Multi-region sampling from 38 TRACERx patients including 112 tumor regions and 37 matched normal adjacent tissue samples was performed. Reduced representation bisulfite sequencing (RRBS) was performed to assess DNA methylation and the CAMDAC (Larose-Cadieux et al, 2020) was applied to estimate purified tumor methylation rates and correct for copy number changes. Whole exome sequencing and somatic copy number alterations (SCNAs) were inferred using the ASCAT tool (Van Loo et al, 2010) and Methsig (Pan et al, 2021) was performed to discover new methylation driver genes.

Results: Using multi-region sequencing, we identified ubiquitous hypermethylation of 29 known driver genes in both lung adenocarcinoma (LUAD) and squamous cell lung cancer (LUSC), together with an additional 9 and 27 genes exclusive to LUSC and LUAD, respectively. We also identified 13 and 7 driver genes non-ubiquitously hypermethylated exclusively in LUSC and LUAD, respectively. Using a differential methylation based approach, we describe a method to determine the extent of intra-tumor methylation heterogeneity akin to established ITH scores based on genomics data. In addition, we report the identification of novel subtype-specific methylation driver genes enriched in HOX family members which are related to cancer progression. Through integration of DNA methylation and genomic sequencing data, we identify parallel mechanisms contributing towards ubiquitous tumor suppressor gene alterations. At the patient level, multiple driver genes such as NSD1, GATA3 and MGA were subject to repression by both copy number loss and DNA hypermethylation. Finally, we describe dosage-compensation of genes such as the Notch ligands JAG2 and DLK1 that are proximal to amplified oncogenes and hypermethylated during tumor evolution.

Conclusion: We describe the contribution of DNA methylation and genomic alterations to altering the landscape of NSCLC. Leveraging DNA methylation, we can determine the extent of convergent repression mechanisms in different regions of the same tumor, assess DNA methylation heterogeneity, and discover DNA methylation-based driver genes in NSCLC.

Citation Format: Francisco Gimeno-Valiente, Carla Castignani, Elizabeth Larose-Cadieux, Kezhong Chen, Nana Mensah, Olga Chervova, Thomas Watkins, Pawan Dhami, Heli Vaikkinen, Andrew Feber, TRACERx Consortium, Jonas Demeulemeester, Miljana Tanic, Stephan Beck, Peter Van Loo, Charles Swanton, Nnennaya Kanu. Identification of convergent gene repression mechanisms through integrative genomic and DNA methylation analysis in TRACERx [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 5710.

Systematic Evaluation of the Immune Environment of Small Intestinal Neuroendocrine Tumors

PURPOSE

The immune tumor microenvironment and the potential therapeutic opportunities for immunotherapy in small intestinal neuroendocrine tumors (siNET) have not been fully defined.

EXPERIMENTAL DESIGN

Herein, we studied 40 patients with primary and synchronous metastatic siNETs, and matched blood and normal tissue obtained during surgery. We interrogated the immune checkpoint landscape using multi-parametric flow cytometry. In addition, matched FFPE tissue was obtained for multi-parametric IHC to determine the relative abundance and distribution of T-cell infiltrate. Tumor mutational burden (TMB) was also assessed and correlated with immune infiltration.

RESULTS

Effector tumor-infiltrating lymphocytes (TIL) had a higher expression of PD-1 in the tumor microenvironment compared with the periphery. In addition, CD8+ TILs had a significantly higher co-expression of PD-1/ICOS and PD-1/CTLA-4 (cytotoxic T lymphocyte antigen-4) and higher levels of PD-1 expression compared with normal tissue. IHC revealed that the majority of cases have ≤10% intra-tumoral T cells but a higher number of peri-tumoral T cells, demonstrating an "exclusion" phenotype. Finally, we confirmed that siNETs have a low TMB compared with other tumor types in the TCGA database but did not find a correlation between TMB and CD8/Treg ratio.

CONCLUSIONS

Taken together, these results suggest that a combination therapy approach will be required to enhance the immune response, using PD-1 as a checkpoint immunomodulator backbone in combination with other checkpoint targeting molecules (CTLA-4 or ICOS), or with drugs targeting other pathways to recruit "excluded" T cells into the tumor microenvironment to treat patients with siNETs.

Comparison and imputation-aided integration of five commercial platforms for targeted DNA methylome analysis

Targeted bisulfite sequencing (TBS) has become the method of choice for the cost-effective, targeted analysis of the human methylome at base-pair resolution. In this study, we benchmarked five commercially available TBS platforms-three hybridization capture-based (Agilent, Roche and Illumina) and two reduced-representation-based (Diagenode and NuGen)-across 11 samples. Two samples were also compared with whole-genome DNA methylation sequencing with the Illumina and Oxford Nanopore platforms. We assessed workflow complexity, on/off-target performance, coverage, accuracy and reproducibility. Although all platforms produced robust and reproducible data, major differences in the number and identity of the CpG sites covered make it difficult to compare datasets generated on different platforms. To overcome this limitation, we applied imputation and show that it improves interoperability from an average of 10.35% (0.8 million) to 97% (7.6 million) common CpG sites. Our study provides guidance on which TBS platform to use for different methylome features and offers an imputation-based harmonization solution that allows comparative, integrative analysis.

Whole-genome sequencing of single circulating tumor cells from neuroendocrine neoplasms

Single-cell profiling of circulating tumor cells (CTCs) as part of a minimally invasive liquid biopsy presents an opportunity to characterize and monitor tumor heterogeneity and evolution in individual patients. In this study, we aimed to compare single-cell copy number variation (CNV) data with tissue and define the degree of intra- and inter-patient genomic heterogeneity. We performed next-generation sequencing (NGS) whole-genome CNV analysis of 125 single CTCs derived from seven patients with neuroendocrine neoplasms (NEN) alongside matched white blood cells (WBC), formalin-fixed paraffin-embedded (FFPE), and fresh frozen (FF) samples. CTC CNV profiling demonstrated recurrent chromosomal alterations in previously reported NEN copy number hotspots, including the prognostically relevant loss of chromosome 18. Unsupervised hierarchical clustering revealed CTCs with distinct clonal lineages as well as significant intra- and inter-patient genomic heterogeneity, including subclonal alterations not detectable by bulk analysis and previously unreported in NEN. Notably, we also demonstrated the presence of genomically distinct CTCs according to the enrichment strategy utilized (EpCAM-dependent vs size-based). This work has significant implications for the identification of therapeutic targets, tracking of evolutionary change, and the implementation of CTC-biomarkers in cancer.

Application of long read sequencing to determine expressed antigen diversity in Trypanosoma brucei infections

Antigenic variation is employed by many pathogens to evade the host immune response, and Trypanosoma brucei has evolved a complex system to achieve this phenotype, involving sequential use of variant surface glycoprotein (VSG) genes encoded from a large repertoire of ~2,000 genes. T. brucei express multiple, sometimes closely related, VSGs in a population at any one time, and the ability to resolve and analyse this diversity has been limited. We applied long read sequencing (PacBio) to VSG amplicons generated from blood extracted from batches of mice sacrificed at time points (days 3, 6, 10 and 12) post-infection with T. brucei TREU927. The data showed that long read sequencing is reliable for resolving variant differences between VSGs, and demonstrated that there is significant expressed diversity (449 VSGs detected across 20 mice) and across the timeframe of study there was a clear semi-reproducible pattern of expressed diversity (median of 27 VSGs per sample at day 3 post infection (p.i.), 82 VSGs at day 6 p.i., 187 VSGs at day 10 p.i. and 132 VSGs by day 12 p.i.). There was also consistent detection of one VSG dominating expression across replicates at days 3 and 6, and emergence of a second dominant VSG across replicates by day 12. The innovative application of ecological diversity analysis to VSG reads enabled characterisation of hierarchical VSG expression in the dataset, and resulted in a novel method for analysing such patterns of variation. Additionally, the long read approach allowed detection of mosaic VSG expression from very few reads–the earliest in infection that such events have been detected. Therefore, our results indicate that long read analysis is a reliable tool for resolving diverse gene expression profiles, and provides novel insights into the complexity and nature of VSG expression in trypanosomes, revealing significantly higher diversity than previously shown and the ability to identify mosaic gene formation early during the infection process.

Antigenic variation is a system whereby pathogens switch identity of a protein that is exposed to the host adaptive immune response as a way of remaining one step ahead and avoiding being detected. African trypanosomes have evolved a spectacularly elaborate system of antigenic variation, with variants being used from a library of ~2,000 genes. Our ability to understand how this rich repository is used has been hampered by the resolution of available technologies to discriminate between what can be closely related gene variants. We have applied a long read sequencing technology, which generates sequence information for the whole length of the antigen gene variants, thereby avoiding having to try and piece together antigen sequences from lots of small fragments, the pitfall of standard sequencing. Applying this technology to material taken at specific time points from batches of mice infected with trypanosomes reveals that the diversity of variants is much higher than previously suspected, and that there is a clear semi-predictable pattern in the gene expression. Additionally, using this technology we have been able to detect the presence of ‘mosaic’ genes, which are created by stitching together fragments from several donor genes in the library, much earlier in infection than has been shown previously. Therefore, we shed new light on the complexity of antigenic variation and show that long read sequencing will be a very useful tool in analysing and understanding the expression patterns of closely related genes, and how pathogens use them to cause persistent infections and disease.

Undifferentiated Sarcomas Develop through Distinct Evolutionary Pathways

Undifferentiated sarcomas (USARCs) of adults are diverse, rare, and aggressive soft tissue cancers. Recent sequencing efforts have confirmed that USARCs exhibit one of the highest burdens of structural aberrations across human cancer. Here, we sought to unravel the molecular basis of the structural complexity in USARCs by integrating DNA sequencing, ploidy analysis, gene expression, and methylation profiling. We identified whole genome duplication as a prevalent and pernicious force in USARC tumorigenesis. Using mathematical deconvolution strategies to unravel the complex copy-number profiles and mutational timing models we infer distinct evolutionary pathways of these rare cancers. In addition, 15% of tumors exhibited raised mutational burdens that correlated with gene expression signatures of immune infiltration, and good prognosis.