Granulocyte Macrophage-Colony Stimulating Factor (GM-CSF) is a cytokine that regulates various cellular processes including differentiation, proliferation, survival, and leukocyte activation. The receptor for GM-CSF is a dodecamer composed of the CSF2RA and CSF2RB receptor subunits. CSF2RB is also a shared common beta subunit for the IL3 and IL5 receptors, and is the predominant subunit for signaling. CSF2RA is primarily a ligand-binding subunit, with a short 54 amino acid intracellular domain, which provides specificity of signaling. We previously reported that GM-CSF signaling is inhibitory to leukemogenesis in a murine model for t(8;21) acute myeloid leukemia (AML), and aids in promoting myeloid differentiation of leukemic blasts. Interestingly, around 32-59% of t(8;21) AML patients suffer from haploinsufficiency of CSF2RA, a gene located on the pseudoautosomal region (PAR) of the sex chromosomes, due to loss of a sex chromosome (LOS). CSF2RA expression has also been reported to be lower in t(8;21) AML patients compared to non-t(8;21) patients. Although we discovered that GM-CSF signaling is inhibitory to t(8;21) leukemogenesis, we hypothesize that CSF2RA itself may act as a tumor suppressor.
Although CSF2RA confers specificity of GM-CSF signaling, very little is known about the role of its expression and signaling in the context of t(8;21) leukemogenesis. To address whether CSF2RA expression has negative effects on the leukemic potential of t(8;21) cells, we restored CSF2RA expression in the t(8;21) leukemia cell lines, Kasumi-1 and SKNO-1. Interestingly, CSF2RA expression specifically inhibited cell proliferation and induced apoptosis in the t(8;21) cell lines, but not in other non-t(8;21) myeloid leukemia cell lines. To further confirm that these effects were specific to the presence of t(8;21), we expressed CSF2RA with RUNX1-ETO, the oncofusion protein generated from t(8;21), in primary murine bone marrow cells. CSF2RA expression in RUNX1-ETO cells had similar effects as was observed in the t(8;21) cell lines. Additionally, this was specific to RUNX1-ETO expression, as control cells did not exhibit these effects. Moreover, we determined that the anti-proliferative and pro-apoptotic effects of CSF2RA expression were ligand-independent, due to the fact that GM-CSF has no cross-species reactivity between humans and mice. CSF2RA expression also reduced the stemness of RUNX1-ETO bone marrow cells and inhibited their colony forming ability. To identify which region of the receptor was mediating these pro-apoptotic effects, we generated truncation mutants and determined that the 25 amino acids at the carboxyl-terminus of the intracellular domain of CSF2RA are required.
IL3RA, the gene encoding the IL3 alpha receptor, is also located on the PAR and suffers from haploinsufficiency in the case of LOS. Therefore, IL3RA also has the potential to also serve as a tumor suppressor in t(8;21) leukemogenesis. Additionally, given that IL3RA oligomerizes with CSF2RB for ligand binding and signaling, we investigated whether IL3RA expression could elicit similar effects as with CSF2RA expression. Interestingly, IL3RA expression had no inhibitory effect on t(8;21) cells, indicating that the observed pro-apoptotic and anti-proliferative effects are unique to CSF2RA in t(8;21) cells.
Altogether, we have discovered that CSF2RA expression, which would be reduced upon LOS, inhibits the leukemic potential of t(8;21) cells by reducing proliferation and inducing apoptosis in a ligand-independent fashion. These phenotypes were especially surprising given that GM-CSF generally promotes cell proliferation and survival, and indicates that CSF2RA may have a novel role as a tumor suppressor in the pathogenesis of t(8;21) AML. Our findings provide greater insight into how LOS may serve as a critical cooperating event in t(8;21) leukemogenesis. Further mechanistic studies will aid in elucidating which signaling pathways are differentially affected upon CSF2RA expression in t(8;21) cells, and may uncover novel therapeutic targets for treating t(8;21) AML.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.