Fig 2: AML1-ETO occupies the miR-29a/b-1 locus and transcriptionally down regulates the miR-29a/b-1 primary transcriptA. Top schematic shows the miR-29a/b-1 locus on human chromosome 7 (Accession Number EU154353) [45]. Runx binding sites are indicated. RNAs originating from the locus are shown; orange blocks indicate exons, while the orange line shows intronic regions. Genomic tracks from ChIP-seq dataset from Trombly et al [35] were re-analyzed for occupancy of the miR-29a/b-1 locus by AML1-ETO, NCoR, or the activating H3K4me3 and suppressive H3K27me3 histone modifications. B. Actively proliferating Kasumi-1 cells were subjected to chromatin immunoprecipitation with two separate antibodies against ETO (black bars), a rabbit polyclonal antibody against carboxy-terminus of RUNX1, which is not present in AML1-ETO (white bar), and an IgG isotype control (gray bar). Immuno-enriched chromatin was analyzed by quantitative PCR using primers encompassing Runx binding sites in the regulatory region of the miR-29a/b-1 locus. Bar graph shows immuno-enriched chromatin as a percentage of input. C. Western blot analysis of SKNO-1 cells infected with a non-silencing short hairpin RNA (shNS), or an shRNA selectively targeting AML1-ETO (shA/E). Significant knock down of the AML1-ETO protein was reproducibly obtained 5 days post-infection. GAPDH was used as protein loading control. D. Total cellular RNA from SKNO-1 cells infected with shNS and shA/E was analyzed for expression of the pri-miR-29a/b-1, as well as of CDKN1A, a known AML1-ETO target gene. Bar graph represents an average of three independent experiments, and error bars indicate standard deviation (SD). Student's t-test shows highly significant changes in the expression of pri-miR-29b-1 and CDKN1A (p > 0.01) upon AML1-ETO knockdown.

Fig 3: AML1-ETO epigenetically activa SETDB2 expression.Notes: (A) Schematic diagrams of the CpG islands along the SETDB2 promoter. Numbers indicate the nucleotides relative to SETDB2 (+1 nt). Vertical lines indicate CpG dinucleotides. Lower panels: A series of constructs and their mutants. (B) Luciferase reporter activities of human 293T cells transiently co-transfected for 48 h with luciferase reporter constructs containing the wild-type sequence of SETDB2 promoter or its mutant counterparts, together with AML1-ETO or mock cDNA. (C) After ChIP using the indicated antibodies or IgG, qRT–PCR was performed to evaluate the specificity of protein binding in the region containing the predicted AML1-binding site. (D) After Kasumi-1 cells were treated with C646 (30 nM) for 24 h, qRT–PCR and Western blot were performed to quantify SETDB2 levels. Expression values are shown as mean ± SEM. *P<0.05.

Fig 4: Impact of shRNA-mediated knockdown of SETDB2 on AML1-ETO -positive AML cell lines.Notes: (A) After SKNO-1 and Kasumi-1 cells were infected with shRNA lentivirus for 48 h, qRT–PCR and Western blot were performed to quantify SETDB2 levels. (B) Effect of SETDB2 knockdown compared to scramble control on colony formation. (C) Effect of SETDB2 knockdown compared to scramble control on proliferation. Expression values are shown as mean ± SEM. *P<0.05.

Fig 5: AML1-ETO and miR-29b-1 form a regulatory circuit that modulates leukemic phenotypeOur results establish a regulatory circuit between the AML1-ETO oncogene and the miR-29b-1 tumor suppressor; AML1-ETO transcriptionally regulates miR-29b-1, and miR-29b-1 translationally inhibits AML1-ETO. MicroRNA-29b-1 has been implicated in hematological malignancies and is a direct transcriptional target of the hematopoietic C/EBPa transcription factor [29]. Our results suggest that AML1-ETO regulates the expression of miR-29b-1 directly by binding to the promoter regulatory region of the miR-29a/b-1 locus, as well as indirectly through down-regulation of C/EBPa, which activates the miR-29a/b-1 locus [29]. It has been shown that miR-29b-1 directly inhibits the expression of Myc, a key regulator of cell cycle and cell growth, Akt2, an upstream sensor in the DNA repair/damage pathways and a regulator of cell apoptosis, and Cyclin D2 (CCND2), which mediates cell cycle control [30]. The current study provides evidence of a complex interplay between AML1-ETO and miR-29b-1 that leads to partial apoptosis, decreased BrdU incorporation, and release from AML1-ETO-mediated myeloid differentiation block. Our study and others together implicate miR-29b-1 as a key modulator of leukemic phenotype in acute myeloid leukemia. Green arrows indicate an activating event, while red symbols show an inhibitory event.