Abstract
Clofarabine is a purine nucleoside analogue which has been incorporated into several therapeutic trial protocols for the treatment of leukaemias including acute myeloid leukaemia (AML). The aims of the study are to ascertain mechanisms of clofarabine action in AML using cell lines and presentation samples from patients. We measured apoptosis by TdT assay, cytochrome C release and flow cytometric assays of the mitochondrial membrane potential probe DiOC6. To study the effects of clofarabine on DNA synthesis and DNA double strand breaks, we used bromodeoxyuridine (BrdU) and H2AX assays respectively. Equitoxic doses were established that caused approximately 20% cell death in the AML cell lines HL-60 (0.3 μM), KG1 (1 μM) and MV4.11 (1 μM) after 6 hours of continuous exposure. At these doses, clofarabine induced apoptotic DNA nicks, as measured by the TdT assay, and Cytochrome C release in all three cell lines. However, clofarabine-induced mitochondrial hyperpolarisation and depolarisation were found to be cell-type specific, occurring in HL-60 but not MV4.11 or KG1 cells, i.e. mitochondrial membrane depolarisation is not an essential part of the mechanism of clofarabine-induced apoptosis. Following clofarabine treatment, DNA synthesis was reduced by 91% within 1 hour in KG1 cells and by 80% in HL-60, but by only 9% in MV4.11. Out of the three cell lines, MV4.11 cells, which have a FLT3 internal tandem duplication (ITD), were the only ones to completely repair DNA double strand breaks following clofarabine removal. In 12 samples from patients with AML, the proportion of cells which incorporated BrdU in a 45 minute assay - a measure of the rate of cell cycling - differed considerably between samples, from 0.8% to 23%, median 13%. Cell death induced by clofarabine was correlated with the cycling rate of untreated cells (P=0.016). All 12 samples showed inhibition of DNA synthesis within 60 minutes of clofarabine treatment (range 20%–92% inhibition, median 54%). Mitochondrial membrane hyperpolarisation and depolarisation were not observed in patient cells. However, DNA double strand breaks were induced by clofarabine in patient cells. Paradoxically, i.e. in contrast to the results seen with the MV4.11 cell line, toxicity was greatest in samples with a FLT3 mutation (P=0.007). In conclusion, the FLT3 mutant MV4.11 cell line effectively repairs clofarabine-induced double strand breaks. Cell death in patient cells cultured with clofarabine is correlated with the presence of a FLT3 mutation. As we have previously established that AML samples with FLT3 mutations have upregulated DNA repair activity, this paradox might be explained by cycles of attempted DNA repair frustrated by renewed clofarabine incoporation into DNA, thus increasing the toxicity of the drug.
FBH1 co-operates with MUS81 in inducing DNA double-strand breaks and cell death following replication stress
