The majority of patients with adult acute myeloid leukaemia (AML) that present with an apparently normal karyotype (NK-AML) are grouped together in the “intermediate” risk category and constitute 40-45% of all adult AML patients. Mutations in the FMS-like tyrosine kinase 3 (FLT3) receptor whether it is internal tandem duplication (ITD) of its juxtamembrane domain or point mutations in its kinase domain are one of the most common mutations in NK-AML. The presence of FLT3-ITD mutation in NK-AML results in a more aggressive disease, resistance to therapy and poor survival. Acquired copy neutral loss of heterozygosity (CN-LOH) also referred to as uniparental disomy (UPD) is a common phenomenon of myeloid malignancies where an oncogenic allele is duplicated on the other chromosome. The use of single nucleotide phenotype analysis (SNP-A) karyotyping detects CN-LOH in 20% and 40% of newly diagnosed and relapsed AML respectively. CN-LOH at 6p or at the FLT3 locus 13q associated with a FLT3-ITD mutation in NK-AML results in an even more aggressive disease compared to NK-AML + FLT3-ITD without CN-LOH. It is has been proposed that CN-LOH is the result of a homologous recombination (HR) DNA repair event. However, the underlying mechanisms that confer CN-LOH have yet to be determined.

To elucidate the mechanisms that produce CN-LOH in NK-AML we determined whether oncogenes such as FLT3-ITD have the propensity to generate CN-LOH through production of excessive DNA double strand breaks (DSB) and up-regulation of HR DNA repair. Primary NK-AML (n=17) were initially characterised by SNP-A and DNA sequencing. 12/17 primary NK-AML had FLT3/ITD mutations with a clone size ≥ 25%, but were wild type (WT) for RAS. From this cohort 6/12 possessed CN-LOH with 3/12 having CN-LOH at 13q. 5/17 primary NK-AML were WT for FLT3 and RAS, but had nucleophosmin (NPM) and DNA methyltransferase 1 (DNMT1) mutations. From this cohort, 1/5 primary NK-AML + WT FLT3 had CN-LOH.

We showed that primary AML + FLT3/ITD cohort, the MOLM 13 cell line that possesses a FLT3/ITD mutation and WT FLT3 KG-1 cells transfected with a FLT3-ITD expression vector demonstrate increased ROS compared to the WT FLT3 primary NK-AML cohort, the WT FLT3 cell line, U937 and KG-1 transfected with empty vector. Immunofluorescence studies showed that primary AML with FLT3/ITD, MOLM 13 and FLT3-ITD transfected KG-1 cells have increased mobilization of phospho-γH2AX (a DNA DSB intermediate) and RAD51 (a HR factor) to sites of DNA DSB in response to mitomycin C (MMC) induced DNA damage compared to non-mutated FLT3 primary AML cells, cell lines and KG-1 transfected with empty vector (phospho-γH2AX, 29% vs. 21%, p<0.005 and RAD51, 25% VS 19%, p<0.01). Moreover, we show that HR is augmented in FLT3-ITD transfected cells and MOLM-13 with and without exogenous DNA damage using DR-GFP gene reporter assays. Increased HR activity has been associated with increased resistance to genotoxic agents. Using soft agar clonogenic assays, MOLM 13 and FLT3-ITD transfected KG-1 showed considerably increased resistance to MMC and the radiomimetic agent, bleomycin compared to U937 cells and KG-1 + empty vector. Finally, to evaluate FLT3-ITD induced genomic instability we measured sister chromatid exchanges (SCE) that result from HR mediated chromosome crossing over. FLT3/ITD primary AML, MOLM-13 and FLT3-ITD transfected KG-1 cells show a significant increase in SCE compared to WT FLT3 primary AML cells, U937 cells and KG-1 transfected with empty vector, (10 vs. 6 SCE per metaphase, p<0.001). Importantly, primary AML with CN-LOH at 13q + FLT3-ITD mutation gives significantly higher HR activity and SCE than non-13q CN-LOH primary AML + FLT3-ITD mutation (p<0.05). Moreover, treatment of FLT3/ITD primary AML, MOLM-13 and FLT3-ITD transfected KG-1 cells with the FLT3 inhibitor, CEP701 and the anti-oxidant agent, N-acetyl cysteine restored ROS levels, HR activity, sensitivity to MMC and SCE to levels observed for WT FLT3. These findings not only provide a strong proof of principle that FLT3-ITD derived DNA damage is the trigger for CN-LOH in NK-AML, but further our understanding of leukaemic progression and relapse.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.