Preview

PDF, English Download (11MB) | Terms of use

Abstract

Chronic lymphocytic leukaemia (CLL) is a malignancy ofbmature B lymphocytes, characterised bybthe accumulation of malignant cells within the blood, bone marrow and lymph nodes. The disease follows a heterogeneous course driven by multiple factors, including genetic aberrations, signalling via the B cell receptor (BCR) and interactions with non neoplastic cells in the tumour microenvironment. Many individual disease drivers have been well-characterised, and CLL represents a success story in linking advances in molecular understanding to improvements in clinical approaches. Building on this, multi-omics approaches promise to further our understanding of disease biology through the integration of multiple datatypes. There remains a clinical need to generate systematic understanding of the role that genetic aberrations and signalling pathways collectively play in CLL.

In this thesis, I investigate the impact of cell-intrinsic and extrinsic factors on CLL biology and drug response. I employ a dataset consisting of viability data of CLL primary samples treated with drugs and microenvironmental stimuli, along with molecular profiles of the same samples. The viability assay was performed using 12 drugs each co-applied with 17 individual microenvironmental stimuli (n=192) and combined with molecular profiles covering point mutations, copy number variations, DNA methylation and mRNA expression. The dataset is complemented with patient clinical data and immunohistochemistry (IHC) stains of lymph node biopsies. I apply statistical inference and regression modelling to reveal several key findings.

Firstly, CLL samples demonstrate heterogeneous responses to the panel of microenvironments stimuli and Interleukin-4 (IL4) and Toll-like Receptor (TLR) signalling have the strongest impact. The response profiles delineate four patient subgroups that show differential disease dynamics and molecular profiles. Secondly, a systematic survey of genetic determinants of microenvironments response identifies trisomy 12 as a key modulator. The data suggest that the transcription factors Spi-B and PU.1 may mediate this effect. Thirdly, I generate a map of interactions between microenvironmental signalling pathways and drugs and identify novel drug-resistance pathways including the effect of IFN on ibrutinib toxicity. I demonstrate how these drug - microenvironment interactions can be further modulated by molecular features, and identify context-dependent drug resistance mechanism that occur in specific genetic backgrounds. For example, TLR activity induced resistance to fludarabine in IGHV-U and trisomy 12 CLLs, but not in IGHV-M. Finally, the in vivo relevance of these findings is established within CLL-infiltrated lymph nodes, which show higher levels of IL4 signalling compared to non-neoplastic samples (p<0.001). Elevated IL4 signalling in CLL-infiltrated lymph nodes correlates with poorer outcomes (p=0.038).

The presented dataset and analysis are available online (https://github.com/Huber-group-EMBL/CLLCytokineScreen2021) and through an interactive shiny app (https://www.imbi.uni-heidelberg.de/dietrichlab/CLLMicroenvironment/). Collectively, these results demonstrate the impact of genetic aberrations and microenvironmental stimuli on drug response and propose novel pathogenic mechanisms and causes of drug resistance.