Mir142 loss unlocks IDH2R140-dependent leukemogenesis through antagonistic regulation of HOX genes

AML is a genetically heterogeneous disease and understanding how different co-occurring mutations cooperate to drive leukemogenesis will be crucial for improving diagnostic and therapeutic options for patients. MIR142 mutations have been recurrently detected in IDH-mutated AML samples. Here, we have used a mouse model to investigate the interaction between these two mutations and demonstrate a striking synergy between Mir142 loss-of-function and IDH2R140Q, with only recipients of double mutant cells succumbing to leukemia. Transcriptomic analysis of the non-leukemic single and leukemic double mutant progenitors, isolated from these mice, suggested a novel mechanism of cooperation whereby Mir142 loss-of-function counteracts aberrant silencing of Hoxa cluster genes by IDH2R140Q. Our analysis suggests that IDH2R140Q is an incoherent oncogene, with both positive and negative impacts on leukemogenesis, which requires the action of cooperating mutations to alleviate repression of Hoxa genes in order to advance to leukemia. This model, therefore, provides a compelling rationale for understanding how different mutations cooperate to drive leukemogenesis and the context-dependent effects of oncogenic mutations.


Plasmid generation
To generate the lentiviral SFFV-IDH2 R140Q -IRES-GFP vector, cDNA encoding mutant IDH2 R140Q was inserted into a unique BamHI site in the multiple cloning site of the CSI (SFFV-IRES-GFP) lentiviral plasmid. Lentiviral vectors expressing wildtype or variant forms of MIR142, were generated by cloning the miRNA into the CSI plasmid downstream of the GFP. gBlock oligonucleotides encoding wildtype or mutant MIR142 pre-miR sequences (mut55 A>G, mut57 U>C and mut58 G>C) with additional flanking sequences, were synthesized by Integrated DNA Technologies (Leuven, Belgium) and cloned into NotI and SbfI sites in the CSI plasmid. To generate the miR-142-3p luciferase reporter, an oligonucleotide containing three copies of the canonical bulged miR-142-3p target sequence was cloned into the pmirGLO Dual-Luciferase miRNA Target Expression Vector (Promega, E1330).

Relative quantification of expression by qRT-PCR
RNA was extracted from cells using TRI reagent (Sigma Aldrich, 93289) according to the manufacturer's protocol and DNase treated. For miRNA quantification, reverse transcription was performed using the TaqMan MicroRNA Reverse transcription kit (Thermo Fisher Scientific, 4366596) followed by qPCR using TaqMan MicroRNA Assays (Thermo Fisher Scientific, 4427975) according to the manufacturer's protocol. Assays used for analysis were miR-142-3p (Assay ID: 000464) and U6 endogenous control (Assay ID: 001973). For quantification of Ash1l expression, reverse transcription was performed using the SuperScript III First-Strand Synthesis System (Thermo Fisher Scientific, 18080051) and TaqMan Gene Expression Assays (Thermo Fisher Scientific, 4331182). Assays used for analysis were Ash1l (Assay ID: Mm00467322_m1) and Actb endogenous control (Mm02619580_g1). For quantification of IDH2 R140Q expression, the SuperScript III First-Strand Synthesis System (Thermo Fisher Scientific, 18080051) and SYBR Green PCR mastermix (Thermo Fisher Scientific,4309155) were used, with primers to mouse Gapdh and the following transgene-specific primers: IDH2CSI_F2 ctggatgggaaccaagacct IDH2CSI_R2 gggatccacgcgtctactg

Quantification of 2-hydroxyglutarate by mass spectrometry
Washed cell pellets were extracted by adding 300µl of 80% methanol and placed on ice for 30 minutes. After centrifugation for 5 minutes (10,000g, 4 o C), supernatant was transferred and dried under Savant speed vac (Thermo Fisher Scientific). The dried extracts were reconstituted in 100 µl of 10% acetonitrile (+0.1% formic acid) and 10 µl were injected directly into the LC-MS/MS system. Separation was achieved on an UPLC system (Accela system, Thermo Scientific) equipped with an ACE Ultracore 2.5 µm, Super C18, 100 × 2.1 mm column (Hichrom, UK). Isocratic elution was employed using a mobile phase of 0.1% formic acid in water (70%) and 0.1% formic acid in acetonitrile (30%) for 6 minutes, all at a flow rate of 250 µl/min. Triple-stagequadrupole mass spectrometry (TSQ Vantage, Thermo Fisher Scientific) equipped with a heated electrospray ion source was used for mass detection. Samples were analyzed in the Multiple Reaction Monitoring (MRM) negative ion ionization mode was employed at a spray voltage of 3000V. Nitrogen was used as sheath and auxiliary gas at a flow rate of 50 and 20 arbitrary units, respectively. Argon was used as collision gas with pressure of 1.5 mTorr. The optimum transitional daughter ions mass and collision energy for 2-HG was: -ve m/z 147.0 ® 129.1 (collision energy 12 4 V). Data acquisition and chromatography analysis was carried out using Xcalibur chromatography software version 2.2 from Thermo Fisher Scientific. The method was found to be linear between 0.1 and 10µg/ml (R 2 for weighted 1/x 2 linear regression > 0.97) and recovery was 83.4 ± 8.3%. Intra-day and Inter-day imprecision were < 2.9% and 10.2%, respectively and inaccuracy was < 9.0% for all quality control levels.

Processing of mouse tissues
For bone marrow extraction, leg and hip bones were crushed in a pestle and mortar and cell suspensions filtered using 40 µm filters. Spleen and liver were diced and passed through a 70 µm filter. Cell suspensions were then centrifuged and resuspended in 1x RBC lysis buffer (BioLegend, San Diego, CA, USA, 420301) for 10-15 minutes on ice. Cells were centrifuged, resuspended in PBS containing 2% heat-inactivated FBS and 2mM EDTA (FACS buffer), filtered, and stained with the appropriate antibody panels. All centrifugation steps were carried out at 400g, 4°C for 5 minutes.

Peripheral blood analysis
Peripheral blood was collected in EDTA-coated tubes and blood counts were obtained using a Sysmex XP-300™ Automated Hematology Analyzer. Red blood cells were lysed using 1xRBC lysis buffer (BioLegend, 420301) for 10-15 minutes at room temperature. Cells were centrifuged, resuspended in FACS buffer, filtered and stained with appropriate antibody panels. All centrifugation steps were carried out at 400g, 4°C for 5 minutes.

Lentiviral shRNA for Ash1l knockdown
Lentiviral constructs expressing shRNAs targeting Ash1l (ULTRA-3233179) or a nontargeting control (ULTRA-NT#4), were obtained from TransOMIC technologies. In the pZip-SFFV-turboRFP-Puro vector, the shRNA is expressed from an optimised UltramiR backbone and co-expressed with RFP. Ash1l knockdown was validated in the HPC5 murine hematopoietic progenitor cell line. FACS-purified KLS cells were then co-transduced with the shRNA pZip vector and the SFFV-IDH2 R140Q -IRES-GFP vector. Double transduced GFP+RFP+ cells were sorted by FACS, and CFC assays performed as described above.