DNA hypermethylation-induced miR-182 silence targets BCL2 and HOXA9 to facilitate the self-renewal of leukemia stem cell, accelerate acute myeloid leukemia progression, and determine the sensitivity of BCL2 inhibitor venetoclax

Rationale: microRNAs (miRNAs) are frequently deregulated and play important roles in the pathogenesis and progression of acute myeloid leukemia (AML). miR-182 functions as an onco-miRNA or tumor suppressor miRNA in the context of different cancers. However, whether miR-182 affects the self-renewal of leukemia stem cells (LSCs) and normal hematopoietic stem progenitor cells (HSPCs) is unknown. Methods: Bisulfite sequencing was used to analyze the methylation status at pri-miR-182 promoter. Lineage-negative HSPCs were isolated from miR-182 knockout (182KO) and wild-type (182WT) mice to construct MLL-AF9-transformed AML model. The effects of miR-182 depletion on the overall survival and function of LSC were analyzed in this mouse model in vivo. Results: miR-182-5p (miR-182) expression was lower in AML blasts than normal controls (NCs) with hypermethylation observed at putative pri-miR-182 promoter in AML blasts but unmethylation in NCs. Overexpression of miR-182 inhibited proliferation, reduced colony formation, and induced apoptosis in leukemic cells. In addition, depletion of miR-182 accelerated the development and shortened the overall survival (OS) in MLL-AF9-transformed murine AML through increasing LSC frequency and self-renewal ability. Consistently, overexpression of miR-182 attenuated AML development and extended the OS in the murine AML model. Most importantly, miR-182 was likely dispensable for normal hematopoiesis. Mechanistically, we identified BCL2 and HOXA9 as two key targets of miR-182 in this context. Most importantly, AML patients with miR-182 unmethylation had high expression of miR-182 followed by low protein expression of BCL2 and resistance to BCL2 inhibitor venetoclax (Ven) in vitro. Conclusions: Our results suggest that miR-182 is a potential therapeutic target for AML patients through attenuating the self-renewal of LSC but not HSPC. miR-182 promoter methylation could determine the sensitivity of Ven treatment and provide a potential biomarker for it.


Competitive reconstitution assays
Donor cells (Ly45.2, 5×10 5 ) from 182WT or 182KO mice and equal WT competitors (Ly45.1, 5×10 5 ) were mixed and xenografted into lethally irradiated recipients (Ly45.1). PB was collected monthly after transplantation for four times. The ratio of CD45.2/CD45.1 was analyzed by flow cytometry (CytoFLEX LX, Beckman-Coulter, Brea, CA, USA). C57BL/6 mice congenic for the CD45 locus Sequencing was performed on Illumina HiSeq 4000 sequencing platform after double-stranded cDNA samples were verified with an Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA). We used Illumina/Solexa Pipeline Image analysis to analyze base calling and error estimation. To reconstruct the transcriptome, the trimmed reads were mapped to the corresponding reference genome by HISAT2 and StringTie. To analyze GSEA, the negative logarithm of the adjusted p-value was first multiplied by the sign of the fold change for each gene to generate preranked gene lists. The enrichment score for each gene set was calculated by input.

Cell cycle
Leukemic cells (2×10 6 ) were incubated in PBS buffer with propidium iodide (0.04 mg/mL, Invitrogen) and RNase (100 mg/mL) at room temperature for 30 min. Cells were collected by flow cytometry and FlowJo software v10.0 (Ashland) was performed to produce histograms to assess number of cells in each phase of the cell cycle.

Construction of plasmids
Human and murine pre-miR-182 and its flanking regions were cloned into retroviral vector pMSCVpuro (Clontech, Palo Alto, CA, USA) to construct the plasmids expressing human and murine miR-182. 3'-UTR of HOXA9 and BCL2 including miR-182-binding sites were amplified by PCR and cloned into vector psiCHECK-2 (Promega). The mutations on miR-182-binding sites in 3'-UTR of HOXA9 and BCL2 were generated by the site-directed mutagenesis kit (Stratagene, Santa Clara, CA, USA). To construct the plasmids expressing HOXA9 and BCL2, we amplified coding sequence (CDS) of HOXA9 and BCL2 and inserted them into vector pLVX-puro. All of the primer sequences are shown in Table S2. DNA sequencing was performed to confirm the sequence.

Lentiviral and retroviral production and transduction
For the production of lentivirus and retrovirus, HEK293T cells (410 6 ) were plated in a 10 cm dish one day before transfection. Lentiviral vector LVX-puro (7.5 g) together with packaging plasmids

Flow cytometry analysis
To measure GFP + cells in the murine AML model, peripheral blood (PB) cells were collected and lysed with lysis buffer (BD PharMingen, San Diego, CA, USA). BM cells were isolated by crushing bones from mice and stained with various primary antibodies in staining buffer for 30 min.

Hematological and histological analysis
Peripheral blood (PB) counts were measured by a hemacytometer (MINDRAY Medical International Co., Ltd, Shenzhen, China). Murine BM cytospins and PB smears were stained by Wright-Giemsa stain following standard protocols for morphological analysis [4].
Paraformaldehyde-fixed paraffin-embedded sections of murine spleen and liver tissues were subjected to H&E staining by standard protocols.
Producing miR-182 knockout (182KO) mice The precursor sequence of miR-182 was replaced by the neomycin resistance gene under the control of the phosphoglycerate kinase 1 promoter in a targeting construct. The targeting vector was linearized and transfected into embryonic stem (ES) cells. Recombinant ES clones were selected and injected into blastocysts to generate male chimeras. Heterozygous and homozygous mice were obtained after five generations of backcrossing and were confirmed using genotyping and northern blotting [5]. Genotype of 182KO mice was performed according to previous report [5]. Briefly, DNA was extracted from the tail of 182WT and KO mice and PCR was performed by standard procedure. Primers of PCR for genotype were indicated in Table S2.    The numbers of response mice mean that the recipient mice develop full-blown leukemia and die within six months after transplantation.