RASSF2 Functions As a Tumor Suppressor in t(8;21) AML Via Promotion of Apoptosis

The t(8;21) chromosomal translocation is one of the most common chromosomal translocations associated with acute myeloid leukemia (AML), found in approximately 12% of de novo AML cases. The majority of these cases are classified as FAB-subtype M2 AML. The t(8;21) results in the stable fusion of the AML1 (RUNX1) and ETO (RUNX1T1) genes. The AML1-ETO fusion protein is composed of the N-terminal portion of AML1, which includes the DNA-binding Runt-homology domain, and nearly the full-length ETO protein. The primary accepted mechanism by which AML1-ETO promotes leukemia development is through the aberrant recruitment of transcriptional repression/activation complexes to normal AML1 target genes. Therefore, the identification of individual genes or biological pathways that are specifically disrupted in the presence of AML1-ETO will provide further molecular insight into the pathogenesis of t(8;21) AML and lead to the possibility for improved treatment for these patients.

We identified RASSF2 as a gene that is specifically downregulated in (2-4 fold) in total bone marrow of t(8;21) patients compared to non-t(8;21) FAB-subtype M2 AML patients by analyzing publicly available gene expression datasets. Similarly, using a mouse model of t(8;21) AML we found Rassf2 mRNA levels to be nearly 30-fold lower in t(8;21) leukemia cells compared to wild-type Lin-Sca-cKit+ (LK) myeloid progenitors. Gene expression analysis by RT-qPCR in leukemia cell lines confirmed that RASSF2 mRNA levels are significantly downregulated (8-10-fold) in both Kasumi-1 and SKNO-1 t(8;21) cell lines as compared to a similar non-t(8;21) HL-60 cell line and to primary human CD34+ control cells. In addition, expression of AML1-ETO in HL-60 or CD34+ cells results in a decrease in RASSF2 mRNA expression, which further suggests that RASSF2 is a target for regulation by AML1-ETO. Assessment of published ChIP-seq data shows that AML1-ETO binds the RASSF2 gene locus at two distinct regions in both primary t(8;21) AML patient samples and in the Kasumi-1 and SKNO-1 cell lines. These regions are similarly bound by several important hematopoietic transcription factors in primary human CD34+ cells, including AML1, ERG, FLI1, and TCF7L2, implicating these two regions as important for the regulation of RASSF2 expression during blood cell differentiation. Overexpression of RASSF2 in human leukemia cell lines using an MSCV-IRES-GFP (MIG) construct revealed that RASSF2 has a strong negative effect on leukemia cell proliferation and viability. The overall percentage of GFP-positive cells in MIG-RASSF2 transduced cells markedly decreased compared to MIG-control transduced cells over a period of 14 days. This effect was primarily due to significantly increased apoptosis in the RASSF2 expressing cell populations. Similarly, we found that expression of RASSF2 significantly inhibits the long-term self-renewal capability of hematopoietic cells transduced with AML1-ETO in a serial replating/colony formation assay. AML1-ETO transduced hematopoietic cells were normally capable of serial replating for more than 6 weeks. However, AML1-ETO transduced cells co-expressing RASSF2 consistently had reduced colony number and lost their ability to replate after 3-4 weeks. This was due to a dramatically increased rate of apoptosis in RASSF2 expressing cells.

RASSF2 is reported to be a tumor suppressor that is frequently downregulated at the transcriptional level by hypermethylation in primary tumor samples, but not healthy controls. Here we have identified RASSF2 as a target for repression, and demonstrated its tumor suppressive function in t(8;21) leukemia cells. Further insights into the molecular mechanisms of RASSF2 function in AML will continue to be explored.