ATRA and the RARα Specific Agonist, NRX195183, Have Opposing Effects on the In Vitro Clonogenicity of Pre-Leukemic Murine AML1-ETO Bone Marrow Cells

Abstract 999

All-trans retinoic acid (ATRA) is used successfully to treat acute promyelocytic leukemia (APML), however, to date it has not shown promise in treating other AML subtypes. ATRA has been shown to enhance hematopoietic stem cell (HSC) self-renewal (requiring RARγ activation) but promotes differentiation of myeloid progenitors likely through RARα activation. We hypothesized that (1) the lack of success of ATRA in treating other AML subtypes may be due to the potential ability of ATRA to enhance self-renewal of the leukemic stem cell and (2) the use of a specific RARα agonist may have more promise in enhancing AML differentiation. We therefore compared the effects of pharmacological levels (1μM) of ATRA and an RARα-specific agonist, NRX195183, on bone marrow cells harvested from a Cre-inducible conditional AML1-ETO (AE) knock-in murine model. AE cells cultured for 2 weeks with ATRA showed significant reductions in the proportions of mature myeloid cells (Gr1brightCD11b+) by fluorescence activated cell sorting (FACS) (DMSO: 14.2±4.3%, ATRA: 4.0±1.6%, p=0.04, n=4). By 4 weeks of culture, ATRA-treated AE cells had increased blast and reduced maturing myeloid cell proportions (Blasts %: DMSO 70.2 ± 3.0, ATRA 95.3 ± 1.2, p=0.08; Intermediate %: DMSO 14.3 ± 2.6, ATRA 3.8 ± 1.0, p=0.01; Neutrophils %: DMSO 2.3± 1.0, ATRA 0.3 ± 0.2, p=0.07, n=6). Furthermore, ATRA potentiated the clonogenicity of the AE cells after 5 weeks of treatment in vitro (Mean±SEM for colony #/ 5×10⁴ cells: DMSO 505.8±337.0, ATRA 4394±388.9, p=0.001; n=6). In contrast, AE cells cultured for 2 weeks with NRX195183 showed significant increases in the proportions of mature myeloid cells by FACS (DMSO: 15.8±3.5%, NRX195183 26.7±3.0%, p=0.03; n=5). By 4 weeks of culture, NRX195183-treated AE cells had decreased blast and increased maturing myeloid cell proportions (Blasts %: DMSO 82.4±3.0, NRX195183 58.8±9.1, p=0.03; Intermediate %: DMSO 14.5±2.5, NRX195183 29.0±6.8, p=0.07; Neutrophils %: DMSO 1.6±0.8, NRX195183 8.2±4.7 p=ns; DMSO n=8, NRX195183 n=5). Moreover, NRX195183 reduced the clonogenicity of the AE cells after 5 weeks of treatment in vitro (Mean±SEM for colony #/ 5×10⁴ cells DMSO 554.8±252.6, NRX195183 82.6±61.6, p=0.05; n=8). Short-term in vivo transplants of fetal liver cells overexpressing the truncated AE gene, AE9a, into sublethally irradiated recipients revealed similar findings in the NRX195183-treated mice with a decrease in blasts and an increase in mature neutrophils in the peripheral blood on morphological analysis after 4 weeks of treatment (Blasts x10⁶/ml: DMSO 3.1±1.0, NRX195183 0.9±0.3, p=0.08; Neutrophils x10⁶/ml: DMSO 0.5±0.1, NRX195183 0.8±0.1, p=0.04; DMSO n=16, NRX195183 n=11). Taken together, these findings support a model whereby ATRA promotes self-renewal of leukemic blasts whilst NRX195183 has the opposing effect. To understand the mechanism by which ATRA promotes self-renewal in AE cells, we performed genome-wide gene expression analyses on the ATRA- versus control-treated AE cells. This revealed 16 differentially upregulated genes after 24 hours of treatment. Using Ingenuity Pathway Analysis, the top scoring network in the ATRA-treated AE cells was cell-to-cell signalling and interaction (p=1.1E^-7-2.4E^-3), lipid metabolism (p=2.3E^-7-2.0E^-3) and small molecule biochemistry (p=2.3E^-7-2.1E^-3); SERPINE1 and BMP2 were the genes with the highest connectivity within the network interacting with molecules known for their roles in tumorigenesis, including AKT, NF-kβ complex and TGFβ1. SERPINE1 upregulation has been shown to be RARα-mediated whilst BMP2 has been shown to be a RARγ-regulated gene. Interestingly, the specific RARγ agonist, NRX204723, had no effect on the clonogenic potential of these AE progenitors thus raising the hypothesis that both RARα and RARγ activation are required to promote self-renewal of the AE progenitors. Further studies using both RARα/RARγ agonists are warranted to assess if the ATRA effects on AE cells are phenocopied. Collectively, these findings reveal the contrasting roles of specific RARα activation in promoting loss of self-renewal ability and enhancing differentiation in the AE cells whilst ATRA promotes clonogenicity of these cells. This has potential significant implications in AML treatment as specific RARα agonists may be beneficial in improving the efficacy of current treatment modalities to achieve sustained remission in other AML subtypes.