miR-133 regulates Evi1 expression in AML cells as a potential therapeutic target

The Ecotropic viral integration site 1 (Evi1) is a zinc finger transcription factor, which is located on chromosome 3q26, over-expression in some acute myeloid leukemia (AML) and myelodysplastic syndrome (MDS). Elevated Evi1 expression in AML is associated with unfavorable prognosis. Therefore, Evi1 is one of the strong candidate in molecular target therapy for the leukemia. MicroRNAs (miRNAs) are small non-coding RNAs, vital to many cell functions that negatively regulate gene expression by translation or inducing sequence-specific degradation of target mRNAs. As a novel biologics, miRNAs is a promising therapeutic target due to its low toxicity and low cost. We screened miRNAs which down-regulate Evi1. miR-133 was identified to directly bind to Evi1 to regulate it. miR-133 increases drug sensitivity specifically in Evi1 expressing leukemic cells, but not in Evi1-non-expressing cells The results suggest that miR-133 can be promising therapeutic target for the Evi1 dysregulated poor prognostic leukemia.


Results
We screened for miRNAs that potentially target Evi1 to suppress its expression using computational prediction and luciferase assays. In silico prediction of Evi1 targets using the miRanda software revealed that 42 miRNAs potentially bind to the Evi1 3′ UTR. (Table 1) Next, we examined whether they suppress the translation of a luciferase reporter containing the 3′ UTR of the human Evi1 mRNA. The precursors for 42 miRNAs were available in our miRNA precursor library. Pre-miR ™ Precursor Molecules for these 42 miRNAs were co-transfected into NIH3T3 cells with a luciferase reporter vector containing the 3′ UTR region of the human Evi1 mRNA. (Figure 1a) Two miRNAs reproducibly downregulated luciferase activity: miR-133 and miR-466. (Figure 1b) To examine whether the endogenous expression of Evi1 is affected by miR-133 and miR-466, we overexpressed miR-133 in HEL cell lines, which express high levels of Evi1. We found that miR-133 suppressed endogenous Evi1 expression in the HEL cell lines. (Figure 1c) Overexpression of miR-466 did not affect endogenous Evi1 expression, but suppression of miR-466 activity by miR-466 TUD 13 resulted in upregulation of endogenous Evi1 expression. (Figure 1d) Since miR-133 and miR-466 both regulated Evi1, we further analyzed their functions in leukemia cell lines. Evi1 overexpression is associated with poor prognosis and shorter survival in AML, because AML shows strong drug resistance. We examined whether miR-133 and miR-466 increase sensitivity to Adriamycin (ADR), one of the key drugs in chemotherapy for AML. HEL and K562 cells derive from AML patients with high Evi1expression, while HL60, U937 and THP1 cells are from ones without Evi1 expression. (Figure 1e,f) We compared drug sensitivity between the two groups of cell lines. Ectopic expression of miR-133 in HEL cells sensitized the cells to ADR. HEL cells were transduced with a retroviral expression vector, MDH, expressing miR-466, miR-133, or no miRNA (control). Transduced HEL and K562 cells were sorted for those expressing GFP (a marker gene in the MDH vector) and were treated with ADR for 44 h. Annexin V staining was used to measure apoptosis. Approximately 3-7% of cells expressing miR-133 were Annexin V-positive (apoptotic) in the absence of ADR treatment, similar to sorted control MDH HEL, K562, HL60, U937 and THP1 cells. (Figure 2a,b) Treatment with ADR for 48 h dose-dependently increased the number of Annexin V-positive miR-133-overexpressing HEL and K562 cells, but not HL60, U937 and THP1 cells, compared to control cells. (Figure 2a,b) These results clearly show that miR-133 induces apoptosis in Evi1-overexpressing cells, but not in cells without Evi1 expression. miR-466 did not show any effect on ADR sensitivity.
That miR-133 increases drug sensitivity in Evi1-high-expressing AML 1 cells was confirmed by caspase activation in these cells. ADR dose-dependently increased cleaved caspase-3 levels in miR-133-overexpressing cells compared with control cells. (Figure 2c) This indicates that miR-133 promoted apoptosis in the presence of ADR in Evi1-high-expressing HEL leukemic cells. (Figure 3) Discussion miR-1-2 and miR-133a-1 are clustered together at the same locus on chromosome 18 14 suggesting that their transcription might be regulated by similar mechanisms. It was previously reported that transcription of these two miRNAs was directly regulated by Evi1, which acts as a transcription factor for them 15 .
Both miRNAs are upregulated by overexpression of Evi1, while only miR-1, and not miR-133, increased cell proliferation 16 . The function and significance of miR-133, which is transcriptionally upregulated by Evi1, needed to be explored. In this study, we demonstrated that miR-1 and miR-133 might act antagonistically, at least in Evi1-overexpressing leukemic cells. Similarly, a previous study showed that miR-1 and miR-133, which are preferentially expressed in cardiac and skeletal muscle and have been shown to regulate differentiation and proliferation of cells in these tissues, produce opposing effects: miR-1 is pro-apoptotic in cardiac cell apoptosis whereas miR-133 is anti-apoptotic. This suggests that the relative levels of miR-1 and miR-133 are more important than their let-7b miR-466a-3 miR-669f-3p   Recently, an in silico study showed that a SNP in the predicted miR-133 binding site in the 3′ UTR of Evi1 predicted worse prognosis in AML. This suggests that in patients miR-133 may play a critical tumor suppressive role whose abrogation results in a worse prognosis 17 . Our functional assay clearly showed that miR-133 is a tumor suppressor for Evi1-overexpressing leukemic cells. The restoration of miR-133 in gastric cancer suppresses cell proliferation and induces apoptosis, indicating that miR-133 is a promising therapeutic target, consistent with our study 18,19 . Accordingly, regulation of miR-133 processing, chemically modified mimics of miR-133, and drug delivery systems should be further studied to better understand the function of miR-133 in Evi1-overexpressing leukemia and its therapeutic potential. Target molecules of Evi1 that induce drug resistance include ITGA6, GPR5, and ANG1 [20][21][22] .
In summary, we identified miR-133 as a miRNA that regulates Evi1, whose overexpression is associated with a poor prognosis in AML. Since miR-133 modulates dysregulated excess Evi1 expression but not normal expression, it could be a promising therapeutic target in Evi1-overexpressing AML patients.

Materials and Methods
Prediction of miRNAs using a computational target prediction system. To detect candidate miR-NAs targeting Evi1, we first evaluated a series of miRNA precursors. To narrow the screened miRNAs to fewer than 100 miRNAs, we used a computational target prediction system (miRanda) containing updated sequences for all known miRNAs. Cutoff scores for selection of candidate miRNAs were < − 20.0 for energy and > 120 for binding 28 . Cell culture. Five cell lines (HEL, K562, U937, HL-60, and THP1) were maintained in RPMI 1640 medium (Wako, Japan) supplemented with 10% (v/v) fetal bovine serum (FBS), 50 U/mL penicillin, and 50 mg/mL streptomycin in a 10 cm dish (Corning, Inc., Corning, NY, USA). Cells were passaged twice per week.
Quantitative PCR for genes. For target gene detection, RT-PCR was performed using the High Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Inc., CA, USA) and qPCR was carried out with the Fast SYBR Green Master mix. All real-time qPCR was conducted using the StepOnePlus real-time PCR system (Applied Biosystems). Threshold cycle (CT) values were calibrated to β -actin and analyzed by the 2 −ΔΔCT method. Sequences of specific primers are listed in Supplementary Table 1. Western blotting. For Western blot analyses, cells were harvested by centrifugation and washed twice with phosphate-buffered saline (PBS). Cells (1.0 × 10 5 ) were lysed in radioimmunoprecipitation assay (RIPA) buffer for 5 min on ice. Cell lysates were centrifuged to remove debris. Protein samples were separated electrophoretically on a 5-20% SDS-polyacrylamide gel and blotted onto PVDF membranes (Bio-Rad Laboratories, Tokyo, Japan). The blots were blocked with 2% low-fat dry milk in TBST (20 mM Tris-HCl, pH 7.5, 150 mM NaCl containing 0.1% Tween 20 [Sigma, MO, USA]) for 1 h at room temperature. The blocked membrane was incubated with anti-Evi1 (CST#2593) (1:2000) or anti-β actin (1:5000) for 2 h, followed by incubation with anti-rabbit IgG (CST#7074) (1:2000) secondary antibody for 1 h. Drug sensitivity assay. Five cell lines (HEL, K562, U937, HL-60, and THP1) were seeded in 24-well plates with 2.0 × 10 5 cells per well in growth medium. Adriamycin was added at specific concentrations and incubated for 48 h, before being analyzed by FACS with immunostaining for APC-Annexin V (BioLegend, Japan).