Dataset Information

B5909cb4-a63a-4fe7-b453-171b83d30b39 - samples

ABSTRACT: The clinical relevance of immune landscape intratumoural heterogeneity (immune-ITH) and its role in tumour evolution remain largely unexplored. Here, we uncover significant spatial and phenotypic immuneÃ¢Â€Â“ITH from multiple tumour sectors and decipher its relationship with tumour evolution and disease progression in hepatocellular carcinomas (HCC). ImmuneÃ¢Â€Â“ITH is associated with tumour transcriptomic-ITH, mutational burden, and distinct immune microenvironments. Tumours with low immuneÃ¢Â€Â“ITH experience higher immunoselective pressure and escape via loss of heterozygosity in human leukocyte antigens and immunoediting. Instead, the tumours with high immune-ITH evolve to a more immunosuppressive/exhausted microenvironment. This gradient of immune pressure along with immune-ITH represents a hallmark of tumour evolution, which is closely linked to the transcriptome-immune networks that contributes to disease progression and immune inactivation. Remarkably, high immune-ITH and its transcriptomic signature are predictive for worse clinical outcome in HCC patients. This in-depth investigation of ITH provides evidence on tumour-immune co-evolution along HCC progression.

PROVIDER: EGAD00001006570 | EGA |

REPOSITORIES: EGA

ACCESS DATA

Similar Datasets

Project description:Intratumoral heterogeneity (ITH) arises from distinct subclonal expansion following genetic or epigenetic alterations, and profoundly influences how tumors response to their immune microenvironment. Tumor progression trees based on single-cell mutational profiles have made it possible to trace subclonal evolution; however, conventional tree-building methods can incorporate only a limited number of cells and mutations, restricting their application to larger single-cell data. To investigate the effect of ITH on the therapeutic response of melanoma, we have developed Trisicell (https://trisicell.rtfd.io), a computational toolkit for scalable inference of mutational ITH through assessment of single-cell genomic variant data. By applying Trisicell to genetically matched mouse melanoma datasets, we found that expressed mutations are sufficient to effectively drive subclonal evolution. On single-cell, full-length RNA sequencing data of mouse melanoma from preclinical immune checkpoint blockade (ICB) studies, the analysis showed that the subtree-seeding mutations in the trees identified distinct subclones associated with a specific developmental state and neural crest lineage markers. Using the tree to trace cell lineages, we found that neoantigens depleted by ICB were predominantly expressed in minor subclones, suggesting that post-treatment recurrence is driven by immunoediting. Moreover, these neoantigens were enriched with those derived from frameshift mutations and mutated nuclear genes. Importantly, recurrently mutated genes in ICB-responding human melanoma exhibited the same features. We next used Trisicell to analyze single-cell, full-length RNA data of brain metastases (BM) from melanoma patients treated with ICB, and discovered that relapsing BM from ICB-responding patients exhibited subclones that also expressed a higher fraction of frameshift mutations and were associated with elevated levels of infiltrated T cells. Notably, they also exhibited more mutated HLA genes and expressed high level of the novel immune checkpoint HLA-G, a putative local immunosuppressive mechanism. In summary, applying Trisicell to single-cell transcriptomic analyses allowed us to identify novel features of ICB-responding melanoma neoantigens and distinct mechanisms for adapting to immunotherapy at different sites. These results have important implications for both melanoma evolution and target identification for immunotherapies, including ICB and cancer vaccines. This is the first study to trace subclonal evolution of neoantigens at single-cell resolution, demonstrating that Trisicell and our datasets represent important resources for the field.

Dataset Information

B5909cb4-a63a-4fe7-b453-171b83d30b39 - samples

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets