%0 Generic
%A Aybey, Bogac
%C Heidelberg
%D 2025
%F heidok:36647
%R 10.11588/heidok.00036647
%T Comprehensive characterization of gene expression response to type-I and type-II interferon in healthy and diseased conditions
%U https://archiv.ub.uni-heidelberg.de/volltextserver/36647/
%X Interferons (IFNs) are critical regulators of the immune system, with special importance for   viral infections, autoimmune diseases, and cancer. However, there are challenges in distinguishing the effects of different IFN types and understanding their cell-type-specific responses. Single-cell sequencing can address these gaps but requires robust tools for cell-type classification and precise analysis of IFN-mediated effects. To better characterize IFN-  and cell-type-specific gene expression responses, novel bioinformatics approaches are needed.   One of the main challenges in single-cell analysis is cell-type classification. Most approaches   use expression information of all expressed genes to provide cell type labels. However, they often introduce bias into statistical procedures when testing the same genes for differential   expression that were already used for cell typing. In the first chapter of my thesis, I addressed   this issue by using only small sets of robust immune cell-type-specific genes for cell typing   within a random forest model. While most studies have generated such gene expression   signatures (GESs) from single gene expression datasets, I developed a novel GES discovery   workflow based on similarities in gene expression across seven single-cell cancer datasets.   Compared to existing algorithms and published GESs, my approach showed superior or   comparable performance, significantly improved the classification of myeloid cells and   enhanced downstream analysis of peripheral blood mononuclear cell (PBMC) datasets.   Thereby, I establish an unbiased method for classifying immune cells and statistical   investigation of expression differences between cell types.  In the second part, I dissected distinct effects of IFN-I and IFN-II across different cellular,   experimental, and disease contexts. Published IFN GESs have been purely generated from   only single gene expression datasets within specific cellular contexts and primarily comprise   genes induced by IFN-I. I used five different bulk tissue RNA-sequencing datasets of IFN   stimulation and applied a novel meta-analysis workflow to resolve GESs with specificity for   IFN-I and IFN-II response. My IFN GESs had greater functional relevance to IFN-type-specific   response and higher coherence than most published signatures. My IFN-II GES detected IFN-  II response of not only myeloid cells but also B cells, hematopoietic cells, and naïve T cells.   Further, I demonstrated the relevance of IFN-I GES in disease severity of lupus nephritis, and   IFN-II GES as predictive biomarker for immune checkpoint inhibitor response. I provide a more   precise distinction between IFN-I and IFN-II responses at the cell type level, as well as their   relevance to disease progression and therapy outcomes.   In the final part, I characterized immune cell-type-specific responses to IFNs as   comprehensive and objective studies of these responses are still lacking. Previous studies do   not provide cellular or temporal resolution, nor compare responses across different IFNs within   a single study. To address this gap, I applied the tool sets established in Chapter 1 and 2 on   a novel temporal CITE-Seq dataset of IFN-I and IFN-II stimulation of human PBMCs. I showed   that all immune cell types exhibited transient responses to IFN-I stimulation while only myeloid   and B cells responded to IFN-II, with distinct dynamic patterns. Furthermore, I identified five   unique temporal gene groups specific to monocyte responses to IFN-I or IFN-II. Those groups   consist of genes that play key roles in distinct immunological pathways. My findings enable a   more detailed characterization of IFN-mediated responses in distinct immune cell populations   compared to those provided by published datasets or GESs.   In this thesis, I introduce novel bioinformatics tools that address key challenges in immune   cell classification and the study of IFN responses. These methods advance the resolution of   immune profiling in single-cell data and provide more precise insights into IFN-mediated immune cell type responses in both health and disease. My approaches overcome the   limitations of existing workflows, offering new insights into IFN biology and its relevance to   disease mechanisms, as well as potential applications in biomarker research and therapy.