Cigarette Smoke or Cigarette Condensate Exposure Enhances Growth of FLT3-ITD AML Models and Alters DNA Methylation and Leukemic Gene Expression

Current and former smoker AML patients have worse survival outcomes compared to never smokers. This worsened prognosis of smokers with AML can also be seen in patients who carry activating mutations of the Fms-like tyrosine kinase 3 (FLT3) and are treated with regimens that include newly approved kinase inhibitors. While the impact of genetic mutations on survival in AML have been studied, and some have been therapeutically targeted, the role of cigarette smoking or cigarette smoke exposure (which is potentially modifiable) on leukemia progression or treatment response is understudied.

In order to elucidate molecular effects of cigarette smoke exposure (CSE) that contribute to the poor prognosis of AML patients, we developed a cigarette smoke exposure model for mice to mimic the current and former smoking habits of AML patients. NOD-SCID mice were exposed to CSE in a smoking robot for 2 hours, 5 days/week, for 2 weeks or to air alone as a control. Mice were then injected with luciferase-tagged human AML cell lines, and leukemic burden was monitored through non-invasive bioluminescent imaging. Control "non-smoking" mice were only subject to AML cell injection. Enhanced early leukemic-burden was observed two distinct FLT3-ITD AML models, MOLM13 and MOLM14, within one week post AML introduction (p-value <0.0001 and <0.001 respectively). Although the latter model showed slightly longer latency of disease with increased leukemic burden apparent 24 days post leukemic introduction (p-value <0.05). In order to address if the early increase in leukemic burden may have arisen from extrinsic factors in the tumor microenvironment, we utilized non-leukemia bearing immunocompetent mice exposed to CSE using the 2 week exposure scheme and saw enhanced myeloid progenitor growth, indicating evidence of microenvironment priming of myeloid cells by CSE. One month of CSE increased the MPP1 and MPP2 populations in the bone marrow of NOD-SCID mice. C57BL/6J mice had increased myeloid and hematopoietic stem cell populations after a month of CSE (p-value <0.05). We also modeled the effect of smoking cessation upon leukemia engraftment by halting smoke exposure compared to mice that continued smoking. Cessation significantly slowed leukemic growth in MOLM13 bearing mice (N=10, p-value<0.01).

Cigarette smoke exposure globally alters DNA methylation in blood cells and these changes can persist for decades. Independent of mutations, DNA methylation patterns in AML patients have prognostic significance. To understand how CSE accelerated leukemic growth in vivo, DNA methylation was evaluated using reduced representative bisulfite sequencing. More than two hundred significant alterations in DNA methylation across the promoter region of genes were found AML cells from spleen samples of CSE MOLM13-bearing mice as compared to non-smoking mice. Among the genes with the most significantly altered DNA methylation were GATA-2, an important protein for hematopoietic differentiation, and aryl-hydrocarbon receptor repressor (AHRR), a gene whose hypomethylation is a hallmark of cigarette smoke exposure.

To identify the impact of cigarette smoke exposure on the leukemia cells in the absence of the tumor microenvironment we treated AML cells directly using a cigarette smoke condensate (CSC) that contains the chemicals in cigarette smoke used in the previously described CSE model. MOLM13 cells either treated with DMSO or 10ug/ml CSC every passage for two weeks were injected into NOD-SCID mice. This model resulted in enhanced leukemic burden 3, 10, and 17 days after leukemic introduction (p-value <0.0001, <0.0001, and <0.001) indicating strong pro-leukemic effects of CSC.

Evaluation of in vitro CSC treated AML cells was conducted to identify causes for the enhanced leukemic burden. While CSC treatment yielded no changes in proliferation or survival of the cells over the course of two months, within one week there was increased expression of DNMT1 in several cells lines. Increased basal and maximal oxygen consumption, and modulation of the antioxidant gene, HO-1, was also observed along with modulation of AHRR and GATA-2, reinforcing roles for methylation data gained from in vivo CSE experiments. Discovering the mechanisms promoting AML progression from cigarette smoke exposure will lead to improved, tailored treatment for AML patients with smoking histories and our further studies of these gene changes will aid in that endeavor.