Knockdown of MTCH2 promotes differentiation of leukemia. (A) TEX cells were transduced with MTCH2 shRNA or control shRNA. Eight days after transduction, gene expression was measured by RNA sequencing. MTCH2 knockdown (233 genes) and control (209 genes) genes were used as signature gene sets in gene set enrichment analysis. Changes in gene expression were compared with LSC⁺ and LSC⁻ populations of AML cells (top) or origins of relapsed AMLs (relapsed origin-committed [ROc] or relapsed origin-primitive [ROp]). (B) OCI-AML2 cells were transduced with MTCH2 shRNAs or control shRNA. Fifteen days after transduction, cells were stained for nonspecific esterase (NSE). NSE staining was quantified with ImageJ software. *P < .05; **P < .01; ***P < .001; ****P < .0001 by ANOVA. (C) mRNA levels of the Lysozyme (LYZ) in TEX cells after MTCH2 knockdown. Data represent relative mean ± SD (n = 3; control shRNA = 1.0). ****P ≤ .0001 by Student t test. (D) MTCH2 F/F and MTCH2 F/F Vav1-Cre⁺ LSK⁺ cells (stained for Ter119, CD11b, B220, Gr1, CD4, CD3, CD8a, cKit, and Sca1) were transduced with MLL-AF9. Two weeks after transduction, cells were analyzed for lineage-positive cells by staining for CD11b and Gr1 surface markers. Data represent 3 independent experiments. *P ≤ .05 by Student t test. (E) TEX AML cells were transduced with shRNA targeting MTCH2 or control sequences and then injected into sublethally irradiated NOD/SCID-GF mice. The percentage of human CD45⁺ cells in the noninjected femur was determined by flow cytometry (n = 10/group) after 5 weeks. The bar represents mean engraftment. **P = .0074 by Student t test. (F) 8227 leukemia cells were transduced with GFP lentivectors expressing MTCH2 shRNA2 or an empty control vector (Control). GFP⁺ cells were flow sorted and cultured. MTCH2 protein levels were measured by immunoblotting 5 days postsorting. The abundance of viable CD34⁺/CD38⁻ cells were also analyzed by flow cytometry. The corresponding population in the control sample was set at 100%, and the population in the MTCH2 knockdown was expressed relative to control. *P ≤ .05; **P ≤ .01 by Student t test. (G) Colony formation from GFP⁺ control and MTCH2 knockdown clones in 8227 cells. Mean ± SD colony counts are shown. *P = .0136 by Student t test. (H) 8227 cells were transduced with GFP-expressing plasmids containing control or MTCH2 shRNA2 sequences and then injected (1 × 10⁶ cells/mice) into the sublethally irradiated NOD/SCID-GF mice (n = 8 mice/group). Eight weeks later, the percentage of human GFP⁺, CD45⁺ cells in the noninjected femur was determined by flow cytometry. The transduction efficiency of the control and MTCH2 lentiviral vectors was 30% and 32%, respectively. The bar represents mean engraftment. *P = .0233 by Student t test. (I) CD34⁺-enriched primary AML cells were transduced with GFP-expressing plasmids containing control or MTCH2 shRNA2 sequences and then injected (3.2 × 10⁵ cells/mice for patient 90249 and 5 × 10⁵ cells/mice for patient 100214) into the sublethally irradiated NOD-SCID mice (n = 8 mice/group). Eight weeks later, the percentage of human GFP⁺, CD33⁺ cells in the noninjected femur was determined by flow cytometry. The transduction efficiency of the control and MTCH2 lentiviral vectors was 11.9% and 15.1% for patient 90249 and 10.1% and 8.6% for patient 100214, respectively. (J) Secondary engraftments of primary AML cells were performed in NOD-SCID mice (for patient 90249, 2 × 10⁶ cells/mice were injected and n = 5 mice/group were used; for patient 100214, 1.7 × 10⁶ cells/mice were injected and n = 6 mice/group were used). Eight weeks later, the percentage of human GFP⁺, CD33⁺ cells in the noninjected femur was determined by flow cytometry. The bar represents mean engraftment, **P ≤ .01; ****P ≤ .0001 by Student t test.