Computational strategies for single-cell genomic profiling in human immuno-oncology

In the rapidly evolving field of immuno-oncology, advanced computational strategies for single-cell genomic profiling are critical for unraveling the complexities of immune responses and enhancing cellular immunotherapy. This dissertation presents two interrelated projects that address key challenges in immune gene expression analysis and the optimization of natural killer (NK) cell-based therapies for acute myeloid leukemia (AML). The first project, titled Allele-specific immune gene quantification and expression analysis in single-cell RNA-seq data, addresses the challenge of both allele typing and quantifying allele-specific expression of highly polymorphic immune genes from single-cell RNA sequencing (scRNA-seq) data. Immune molecules such as human leukocyte antigens (HLAs) exhibit extensive allelic diversity, which is crucial for immune function and disease susceptibility. Traditional quantification methods often struggle to capture this complexity. To overcome these limitations, I developed an alignment-free computational methodology that leverages the novel approach of donor-specific reference sequences, thereby ensuring an accurate representation of allele-specific expression and precise allele typing. In addition, I utilized an R/Bioconductor data structure, originally developed within our group as part of a master's thesis, to integrate expression data across multiple immunogenetic annotation layers. This enabled an interactive and comprehensive exploration of gene-, functional-, and allele-specific expression patterns. Validation of this approach on several scRNA-seq datasets demonstrated its robustness, revealing key insights such as the loss of HLA expression in tumor cells, differential expression of specific HLA alleles across distinct immune cell subtypes, and accurate allele typing from single-cell data. These findings highlight the potential of this methodology to inform precision immunogenomics and guide personalized immunotherapy strategies. Overall, the presented novel methodology provides a robust approach for investigating the expression of key immune mediators across human donors at a very high molecular detail. The second project, titled CRISPR/Cas9 editing of NKG2A improves the efficacy of primary CD33-directed chimeric antigen receptor natural killer cells, focuses on improving the efficacy of NK cell-based immunotherapy by genetically modifying NK cells to overcome inhibitory immune checkpoints. Although NK cells inherently possess antitumor activity and a favorable safety profile, their therapeutic potential is often compromised by immunosuppressive mechanisms within the tumor microenvironment, specifically the interaction between HLA-E and the inhibitory receptor NKG2A. In collaboration with Prof. Evelyn Ullrich’s group, dual-modified NK cells were generated by combining lentiviral CAR33 transduction with CRISPR/Cas9-mediated knockout of the NKG2A-encoding KLRC1 gene. My primary contribution to this project involved the analysis of single-cell multi-omics data generated through CITE-seq, with the goal of characterizing the transcriptional and proteomic changes associated with NK cell activation, maturation, and cytotoxic function. Using NK cells derived from two donors, these modifications produced eight distinct cell populations (non-transduced, KLRC1 knockout, CAR33-transduced, and dual-modified CAR33-KLRC1ko, each evaluated before and after co-culture with AML cells). Comprehensive single-cell and pseudo-bulk analyses were employed to characterize the transcriptional and proteomic changes associated with each NK cell modification. The analyses revealed distinct transcriptional states, with dual-modified CAR33-KLRC1ko-NK cells exhibiting an expression profile characterized by elevated levels of genes involved in activation, maturation, migration, and antigen presentation. These results indicate that NK cells with dual modifications perform better than those modified with either CAR33 transduction or KLRC1 knockout alone, supporting their enhanced cytotoxic activity against CD33+ AML cells. In summary, this dissertation integrates innovative computational methodologies with advanced cellular immunotherapy strategies to enhance our understanding of immune gene expression and improve NK cell functionality in the context of immuno-oncology. By developing robust techniques for allele-specific expression profiling and demonstrating that dual genetic modifications can potentiate NK cell-mediated cytotoxicity against AML, this work lays the foundation for more effective and personalized cancer immunotherapies.