In vitro modelling of high-risk ALL fusions uncovers genomic alterations and non-canonical signalling pathways as a mode of TKI-resistance – implications for targeted therapy

Abstract

Treatment-resistant acute lymphoblastic leukaemia (ALL) remains a significant clinical issue. Genomic profiling has identified a subtype of high-risk ALL termed Philadelphia-chromosome-like (Ph-like) ALL that is associated with accumulation of kinase-activating fusion genes and poor outcome. Studies have demonstrated the kinases encoded by Ph-like ALL fusion genes are targetable in vitro and in vivo by tyrosine kinase inhibitors (TKI), though it is known through CML and Ph+ ALL studies, that TKI-resistance is likely, leading to relapse and/or progression. Therefore, there is an urgent need to identify drivers of potential TKI-resistance in the Ph-like ALL setting. This study aims to model TKI-resistance in high-risk Ph+ and Ph-like ALL in order to understand the underlying mechanisms and inform future therapeutic strategies that may overcome resistance, thereby improving patient outcomes. In this thesis, Ba/F3 cells expressing Ph-like ALL fusion genes were generated by retroviral transduction of fusions identified in primary Ph-like ALL samples, while Ph+ ALL was modelled using the immortalised cell line SUP-B15. TKI-resistance was generated through long-term exposure (~6-12 months) of cells to increasing concentration of TKI (imatinib, dasatinib, ponatinib or ruxolitinib – depending on fusion kinase). TKI-resistant Ba/F3 Ph-like ALL cells were found to acquire the clinically-reported ABL1-T315I, ABL1-G250E/T315I and JAK2-Y931C kinase domain mutations associated with drug resistance. Interestingly, a novel CSF1RL785M mutation and a previously unreported ABL1-L302H/T315I compound mutation were identified, with these mutations predicted to alter drug binding. Importantly, next-generation transcriptomic sequencing uncovered the association of non-canonical signalling pathways in TKI-resistant high-risk ALL cells. In ruxolitinib-resistant JAK2-Y931C Ba/F3 cells, mRNA-seq pathway analysis revealed the involvement of glycolytic metabolism and G-protein coupled receptor signal transduction pathways as potential modes of TKI-resistance. Transcriptomic analysis also uncovered a KRAS and NRAS mutation in dasatinib and ponatinibresistant SUP-B15 Ph+ ALL cells respectively, resulting in upregulation of mitogenactivated protein kinase (MAPK) cell survival signalling cascades. Together these findings represent attractive targets for future studies into parallel treatment modalities that may inhibit the development of, or overcome established treatment resistance in ALL. Overall this thesis has demonstrated that TKI-resistance in high-risk Ph-like and Ph+ ALL can be attributed to clinically-relevant point-mutations and the involvement of key alternative signalling mechanisms which override TKI sensitivity. This knowledge is critical towards potentially informing future management of patients who relapse on first-line TKI therapies, as well as the prospective development of combinatorial treatments that target critical fusion kinase and parallel signalling pathways likely contributing to the emergence of treatment resistance, in high-risk ALL and other diseases of similar treatment-resistant modalities.