Cigarette Smoke or Cigarette Condensate Exposure Accelerates Growth of FLT3-ITD AML Models, Induces Oxidative Stress, and Alters DNA Methylation

Acute myeloid leukemia (AML) patients with any history of cigarette smoking have worse survival outcomes compared to patients that have never smoked. The molecular basis of cigarette smoking or cigarette smoke exposure (CSE) impacting AML progression or treatment response is unknown. Altered DNA methylation from smoking persists decades after quitting and has been followed in peripheral blood mononuclear cells (PBMCs) but have not been evaluated in the context of AML. Smoking also causes oxidative stress in PBMCs, but this has yet to be studied in AML patients with a history of smoking.

To model how smoking worsens AML progression, we created a CSE model using a cigarette smoking robot where NOD-SCID mice received 2 hours of CSE 5 days/week or air alone. After 2 weeks of CSE, 100,000 luciferase-tagged, human AML cells were injected via tail-vein and leukemic burden was monitored by bioluminescent imaging. Two cell lines, MOLM13 and MOLM14 carrying the oncogenic fms-like tyrosine kinase 3 internal tandem duplication (FLT3-ITD) mutation, when introduced into CSE mice had enhanced leukemic progression within one week (p-value <0.0001 and <0.001 respectively). MOLM14-bearing mice showed increased leukemic burden 24 days post injection (p-value<0.05). DNA methylation was evaluated by bisulfite sequencing of splenocytes from CSE mice, which was mapped to the human (AML) or mouse (host) genome. Over 200 significant changes in DNA methylation of gene promoters was seen, including hypomethylation of aryl-hydrocarbon receptor repressor (AHRR), an established hallmark of smoking in humans. Indicating that our CSE model recapitulates DNA methylation from smoking in humans. Additionally, GATA-2, a critical protein for hematopoietic differentiation known to be frequently mutated in AML, was also amongst the top hypomethylated genes. We quarried TCGA to understand the implications of altered DNA methylation in AML patients. In AML patients, low GATA-2 methylation showed decreased survival compared to those with high GATA-2 methylation (N=42/group, p-value: 0.000138). The discovery of GATA-2 methylation in smoking models and its attribution as a poor prognostic indicator in AML is novel, which underscores a need to understand the role of GATA-2 methylation in AML.

Reactive oxygen species (ROS) have been implicated in AML progression, drug resistance, and are elevated in otherwise healthy smokers. No significant differences were seen in intracellular ROS in spleen or PBMCs of CSE mice; however, we found more than a two-fold increase of heme oxygenase 1 (HO-1) (p-value:0.0505), a protein involved in antioxidant responses and AML treatment resistance. There was also increased BCL-2 protein expression and a decrease in AHRR expression (p-value: 0.0098) in CSE mouse samples. This suggests that CSE causes oxidative stress and increases pro-leukemic signaling. To address if CSE impacted treatment efficacy, we treated mice with daunorubicin (2 mg/kg, twice weekly via tail-vein) once evidence of engraftment was detected. We found a trend towards increased leukemic burden compared to non-smoking mice which approached statistical significance.

To study the direct impact of CSE on AML, without exposure to the tumor environment, AML cells were treated in vitro with a commercially available cigarette smoke condensate (CSC) that contains the chemicals from cigarette smoke. To mimic the in vivo CSE, MOLM13 cells received two weeks of CSC treatment before being injected into mice. At 3, 10, and 17 days post injection, mice with CSC-treated AML had enhanced leukemic burden (p-value <0.0001, <0.0001, and <0.001). This model revealed that chemicals in cigarette smoke can directly promote AML. On day 14 of CSC treatment, mirroring when the cells were injected into mice, both FLT3-ITD cell lines had increased ROS levels and or glutathione as measured by flow cytometry; this is indicative of CSC-induced oxidative stress.

Cumulatively, these data define novel changes in DNA methylation and redox from smoking or tobacco products with strong potential to drive AML progression and therapy resistance. Future studies will determine if blocking these redox or methylation events can reverse the accelerated AML growth and will aid in the creation of a tailored treatment strategy for AML patients with any history of smoking.