Explorations in cancer genomics and tumor immunology through analysis of next-generation sequencing data

Abstract

Cancer is an evolutionary disease. From the time we are a fertilized egg, our cells are subject to processes that can alter DNA sequence and epigenetics leading to random changes in cell fitness. Certain phenotypes can lead to a positively selected fitness advantage, to the detriment of the broader individual. Alterations allowing bypass of cellular checkpoints, genomic instability, metastasis, immune evasion, and many others are considered hallmarks of cancer. Analysis of large next generation sequencing datasets has enabled unprecedented insight into the genetic underpinnings of these traits, and how they can be exploited for improved cancer screening and treatment. Here we use next generation sequencing to uncover mechanisms of mutagenesis in cancer, interrogate the interplay of early pre-cancers with the immune system, and characterize the molecular underpinnings of response to a leading cellular therapy. Genomic instability is considered one of the hallmarks of cancer, with additional sources of DNA damage or deficiencies in DNA repair often being present. Understanding these mutational processes gives us mechanistic insight into how cancer arises, and can present tumor vulnerabilities that may be exploited by therapeutics. As part of the Pan-cancer Analysis of Whole Genomes Consortium, we used a cohort of 4,645 whole genome sequences from patients across tumor types to discover common underlying mutational processes using Bayesian non-negative matrix factorization. In total, we found a set of 49 single-base substitution, 11 doublet base, and 17 small insertion and deletion signatures encompassing the major mutational processes active in cancer. We explored the mechanisms of these signatures, and published them on the COSMIC website as a standard reference for use by the community. We then turn to examine a tumor-extrinsic factor in cancer disease progression: the immune system. In particular, the precise role of the immune system in controlling or supporting early pre-cancerous lesions is not well understood. To this end, we leveraged single-cell RNA sequencing technologies to study the immune microenvironment in the bone marrow of 9 healthy and 22 individuals with varying stages of multiple myeloma (MM) and its well defined precursor states. We found changes in immune composition even at the earliest stages of disease. NK cells with distinct chemokine receptor expression, non-classical monocytes, regulatory and cytotoxic T cells became enriched over the course of progression. We observed further transcriptional and proteomic changes, including a loss of MHC class II presentation in the CD14+ monocytes of patients. Our results provide an initial map of immune changes at play over the evolution of premalignant MM and demonstrate the importance of the immune system in early pre-cancerous development. Finally, we turn our attention to how the immune system can be leveraged to provide treatments to patients. Chimeric Antigen Receptor (CAR)-T cell therapy has revolutionized the treatment of hematologic malignancies, with approximately half of patients with refractory large B-cell lymphomas achieving durable responses from CD19-targeting CAR-T treatment. However, failure mechanisms are identified in only a fraction of cases. We performed single-cell transcriptome sequencing of 106 pre- and post-treatment peripheral blood samples and infusion products collected from 32 individuals with high-grade B cell lymphoma treated with two leading CD19 CAR-T products. We characterize a distinct phenotype of CAR-T cells in responders and non-responders, discover differences in CD8 expansion dynamics between different commercial products, and identify a potential role for CAR-T-regs in mediating relapse. In summary, we leveraged a variety of genomic and transcriptomic sequencing technologies to study different aspects of tumor evolution and its interactions with immune cell populations. In addition to shedding light on basic biological mechanisms, in many cases this understanding can help advance patient risk assessment, illuminate potential treatment options, or advance existing ones. As additional techniques, such as proteomics and spatial sequencing, continue to be developed we, and the field as a whole, can further resolve the complex picture of tumor, immune, and stromal evolution.