Investigation of dormancy in acute myeloid leukaemia cells and the induction of dormancy in their response to genotoxic stress

Cancer cells can exist in a reversible state of dormancy (G0 phase of the cell cycle). Relapse of acute myeloid leukaemia (AML) is likely due to dormant cells escape frontline treatment. Dormant AML cells have been identified in the bone marrow (BM) endosteal region which is characterised by an excess of TGFβ1 and a shortage of nutrients. As the first step in this project, we developed an in vitro model of AML cell dormancy by exploiting these features. Following a preliminary investigation of four AML cell lines, the CD34+CD38- line TF1-a was selected for in-depth investigation. TF1-a showed significant inhibition of proliferation, with features of dormancy and stemness, in response to 72 hours of TGFB1+mTOR inhibitor treatment (mTOR pathway inhibition mimics major effects of nutrient scarcity) without affecting cell viability or inducing differentiation of these cells. Secondly, whole human genome gene expression profiling using Affymetrix microarray strips (HuGene2.1ST) was conducted to molecularly characterise the dormancy-induced TF1-a cells. As a result, we identified 240 genes which were significantly up-regulated by at least twofold, including genes involved in adhesion, stemness, chemoresistance, tumor suppression and genes involved in canonical cell cycle regulation. The most upregulated gene was osteopontin (17.1fold). Immunocytochemistry of BM biopsies from AML patients confirmed high levels of osteopontin in the cytoplasm of blasts near the paratrabecular BM. Osteopontin and other genes identified in this model, including well-characterised genes (e.g. CD44, CD47, CD123, ABCC3 and CDKN2B) as well as little-known ones (e.g. ITGB3, BTG2 and PTPRU), are potential therapeutic targets in AML. AML cells which are identified in the bone marrow (BM) endosteal region are likely to survive chemotherapy for several reasons including the low concentration of the drug delivered to this poorly perfused area of the BM. We hypothesised that these cells might induce dormancy features that help them escaping chemotherapy. Moreover, little is known regarding the molecular changes in those AML cells surviving genotoxic stress. The third aim of this study was to investigate the AML cells surviving genotoxic stress. Therefore, we developed and characterised an in vitro model of prolonged sub-lethal genotoxicity in AML cells by utilising the TF1-a cells treated with etoposide for 6 days. TF1-a cells survived this conditioning with significant inhibition of proliferation and limited damage and apoptosis. The molecular signature of these cells was characterized using GEP of the whole human genome and was compared to that of the dormancy-induced TF1-a cells in an aim to identify genes commonly up-regulated in both scenarios that might act as possible drivers of dormancy in AML cells residing in a sub-lethal genotoxic environment. In this context, 31 genes were significantly commonly upregulated in both scenarios including ITGB3, SLFN5, C15orf26 and GRAP2 which are candidates for in-depth investigation in AML. To sum up, in this study we developed and molecularly characterised in vitro models of dormancy and sub-lethal genotoxicity in AML cells with stemness characteristics through novel approaches that took bone marrow microenvironmental features into consideration. These features likely contribute to the resistance of the residual sub-population of AML cells that causing disease relapse. The current models helped to overcome the obstacles facing the in-depth studying of these rare AML cells due to the difficulty in obtaining them from clinical samples. Finally, the molecular findings of this study paved the way to potentially important future directions of research that could help to achieve the ultimate goal of eradicating AML cells and preventing relapse