AML cell lines engineered to express either WT IDH1 or mutant IDH1–R132H and producing the 2-HG oncometabolite. (A) Western blot analysis of total and mutant form of IDH1 protein in diverse HL60 clones (left) or MOLM14 (right) infected with retroviruses encoding GFP alone (CTL) or GFP and the indicated IDH1 variants. One representative experiment of three experiments is shown. (B and C) LC-MS analysis of intracellular (R)-2-HG concentration (B) or intracellular α-KG concentration (C) in different clones of HL60 cell lines expressing either GFP control (CTL, clone 14), WT (clone 2), or mutant (R132H, clone 5 and 11) form of IDH1 in HL60 (left) or MOLM14 (right) cell lines. **, P ≤ 0.01; ***, P ≤ 0.001, statistical significance of the observed differences (Mann-Whitney test). n = 5 per condition.

IDH1 mutation primes AML cells to the ATRA differentiation pathway through the activation of granulomonocytic differentiation-associated genes. (A) Gene ontology enrichment analysis of down- (122) and up-regulated (206) genes from RNA expression profile of the differentially expressed genes between WT IDH1 HL60 clones and mutant IDH1–R132H HL60 clones by 1.5-fold or more (n = 3 clones for each cell type). (B) Gene to TF associations within mutant IDH1–R132H target genes using data mining algorithm (Genomatix). (C) mRNA expression of C/EBPα in WT (gray; n = 323) or mutant (black; n = 68) primary AML patients from the AML patient cohort GSE6891 () and in WT (gray) or R132H (black) HL60 (dots) or MOLM14 (square). (D) Protein level of CEBPα in mutant IDH1 HL60 (R132H, clone 5 and 11) and MOLM14 (R132H) cells or of WT IDH1 HL60 (WT, clone 4 and 7) and MOLM14 (WT) cells by Western blot (n = 3). (E) qChIP experiments showing the relative recruitment of histone H3 trimethylation at lysine 4 (H3K4me3) on CEBPa promoter in mutant IDH1–R132H HL60 and MOLM14 cells versus WT IDH1–WT HL60 and MOLM14 cells, as indicated. The precipitated DNA fragments were subjected to real-time PCR analysis with two different pairs of primers amplifying the CEBPα promoter. Results were represented as the relative ratio between the mean value of immunoprecipitated chromatin with the indicated antibodies and the one obtained with a control irrelevant antibody. ChIP values are presented as percentage of total input. Data represent the mean of three independent experiments. Error bars represent SEM. (F) Among TFs identified in this study, up- (red) or down-regulation (blue) of CEBPα downstream target genes in mutant IDH1–R132H HL60 cells compared with WT IDH1 HL60 cells. (G) Expression of two CEBPα-target genes (RXRα and EMILIN) were analyzed in mutant IDH1 HL60 (R132H, clone 5 and 11) and MOLM14 (R132H) cells or in WT IDH1 HL60 (WT, clone 4 and 7) and MOLM14 (WT) cells by RT-qPCR. (H) qChIP experiments showing the relative recruitment of CEBPα on the RXRa or EMILIN locus in mutant IDH1–R132H HL60 and MOLM14 cells versus WT IDH1–WT HL60 and MOLM14 cells, as indicated. Results were presented as the relative ratio between the mean value of immunoprecipitated chromatin with the indicated antibodies and the one obtained with a control irrelevant antibody. Data represent the mean of three independent experiments. Error bars represent SEM. (I) Expression of the myeloid differentiation marker CD11b in HL60 and MOLM14 expressing mutant IDH1–R132H (clone 5 or 11) or WT IDH1 protein were assessed by flow cytometry (mean ± SD; n = 6). (J) Genes to small molecule associations within mutant IDH1–R132H target genes using data mining algorithm (genomatix). *, P ≤ 0.05; **, P ≤ 0.01; and ***, P ≤ 0.001 indicate statistical significance of the observed differences (Mann-Whitney test).

IDH1–R132H mutation sensitizes to ATRA-induced differentiation of primary patient samples in vitro. (A–C) ATRA treatment induces granulocytic differentiation of 11 primary patient samples carrying mutant IDH1–R132H (R132) compared with 10 WT IDH1 (WT) and 1 PML-RARα fusion protein (APL). (A) Primary WT IDH1, mutant IDH1–R132H AML, or APL patient samples were treated with ATRA at 1 µM for the indicated time, cytospun, and stained with May-Grünwald Giemsa. Cells were classified in five categories (from less to more differentiated; right). Bar, 10 µm. 100 cells per condition were counted three times for statistical analysis (Mann-Whitney test). (B) Intracellular amounts of 2-HG and α-KG in 12 primary patient samples expressing WT IDH1 (WT) or mutant IDH1–R132H (R132H), determined by Liquid chromatography–mass spectrometry (LC/MS; method 2). The amounts were normalized using a labeled standard, [¹³C]-succinate, spiked with the same volume in each sample (y axis). A log₁₀ scale was used. **, P < 0.05; Mann-Whitney test. (C) Flow cytometry analysis of CD11b, CD15, and CD14 expression was performed in primary WT IDH1, mutant IDH1–R132H AML, or APL patient samples treated with 1 µM ATRA or vehicle for 12 d (mean ± SEM). *, P ≤ 0.05; **, P ≤ 0.01; and ***, P ≤ 0.001 indicate statistical significance of the observed differences (Wilcoxon matched pairs signed rank test).

IDH1–R132H mutation sensitizes to ATRA-induced differentiation of AML cell lines in vitro. (A–D) Analysis of ATRA-induced differentiation in HL60 and MOLM14 infected with retroviruses encoding GFP alone (CLT) or GFP and the indicated IDH1 variants. (A) Quantification of the CD11b expression measured by flow cytometry after 3-d treatment with 1 µM ATRA or solvent (ratio of mean fluorescence intensity in ATRA-treated to vehicle-treated cells ± SEM; n = 5). Mann-Whitney test. (B) Quantification of the CD11b expression measured by flow cytometry after 3-d treatment with 1 µM ATRA or solvent (ratio of mean fluorescence intensity in ATRA-treated to vehicle-treated cells ± SEM; n = 5) in different clones (CTL, clones 14–15; WT clones 4–7; R132H, clones 5–11; Mann-Whitney test). (C) Dose-dependent increase of CD11b expression induced after 3-d treatment with 1 µM ATRA or solvent (ratio of mean fluorescence intensity in ATRA-treated to vehicle-treated cells ± SEM; n = 5) in HL60 cells parental, control (Cl 14), WT (Cl 2), or with IDH1–R132H mutations (clone 5 or 11). Two-way ANOVA. (D) Measurement of NBT reduction after 12-d treatment with 1 µM ATRA or control. The data correspond to the mean of three independent experiments (Mann-Whitney test). (E) Venn diagram of the overlapping genes enriched in mutant IDH1 HL60 cells (R132H) versus WT IDH1 HL60 cells both treated with 1 µM ATRA for 24 h with a FDR <0.05, considered as differentially expressed mRNA irrespective of the fold-change. Gene to TF associations specifically within 109 mutant IDH1–R132H target genes in response to ATRA using data mining algorithm (Genomatix). Similar screening and analysis was performed in WT IDH1 HL60 cells with exogenous 2-HG (WT+2-HG) versus WT IDH1 HL60 cells both treated with 1 µM ATRA for 24 h. (F and G) NB4, IDH1–WT HL60 clone, or MOLM14 were treated with or without 100 or 50 µM, respectively, 2-HG for 3 d with 1 µM ATRA or control. (F) Intracellular amounts of 2-HG in parental NB4 cell line or HL60 cell lines expressing WT IDH1 (WT) with or without treatment with 1-octyl-(R)-2-HG determined by LC/MS. The amounts were normalized using a labeled standard, [¹³C]-succinate, spiked with the same volume in each sample (y axis). Each condition represents a triplicate set of measurements. **, P < 0.01; Mann-Whitney test. (G) Quantification of the CD11b expression measured by flow cytometry (ratio of mean fluorescence intensity in ATRA-treated to vehicle-treated cells ± SEM; n = 5; Mann-Whitney test). (H and I) Control, WT, or R132H (clone 5 or clone 11) HL60 cells (M) or Control, WT, or R132H MOLM14 cells were pretreated with 2 µM AGI 5198 or solvent for 15 d followed by 3-d treatment for HL60 or 24-h treatment for MOLM14 with or without 1 µM ATRA ± 2 µM AGI 5198 or solvent. Mann-Whitney test. (H) Intracellular amounts of 2-HG and α-KG in HL60 cell lines or MOLM14 cell lines expressing WT IDH1 (WT) or mutant IDH1–R132H (R132H) after or without treatment with IDH1–specific inhibitor AGI-5198 determined by LC/MS (method 2). The amounts were normalized using a labeled standard, [¹³C]-succinate, spiked with the same volume in each sample (y axis). Each condition represents a triplicate. ** indicates high statistical relevance (P < 0.01; Mann-Whitney test). (I) CD11b expression was then measured by flow cytometry (ratio of mean fluorescence intensity in ATRA-treated to vehicle-treated cells ± SEM; n = 5). *, P ≤ 0.05; **, P ≤ 0.01; and ***, P ≤ 0.001, statistical significance of the observed differences (Mann-Whitney test).

ATRA preferentially reduces cell viability of mutant IDH1 AML cells through apoptotic cell death in vitro. (A) HL60 or MOLM14 cell lines expressing GFP alone (CTL) or GFP and the indicated IDH1 variants were cultured with or without 1 µM ATRA for 10 d. Viable cells were counted by Trypan blue dye exclusion (n = 3). Two-way ANOVA. (B) ATRA treatment induces granulocytic differentiation of 11 primary patient samples carrying mutant IDH1–R132H (R132) compared with 10 WT IDH1 (WT) and 1 PML-RARα fusion protein (APL). Wilcoxon matched pairs signed rank test. (C) After 5-d treatment with 1 µM ATRA or solvent, the percentage of Annexin V–positive cells was measured by flow cytometry (mean ± SEM, n = 5; Mann-Whitney test). (D) HL60 or MOLM14 cells were treated with 1 µM ATRA or solvent for 5 d. BCL2 protein expression and cleavage of poly (ADP-ribose) polymerase or caspase-3 were analyzed by Western blotting. One representative experiment of three experiments is shown. (E and F) HL60 or MOLM14 were plated in methylcellulose in the presence or absence of 1 µM ATRA for 7 d. Representative photographs taken at the end of culture are shown. (F) L-CFU were counted (mean ± SEM; n = 3). *, P ≤ 0.05; **, P ≤ 0.01; and ***, P ≤ 0.001, statistical significance of the observed differences (Mann-Whitney test). ns, not significant.

In vivo antileukemic activity of ATRA in cell line model and primary patient blasts carrying IDH1 R132H mutation. (A) Schematic representation of the in vivo protocol used for B–F. Immunodeficient NOD-scid IL2rγ^null (NSG) female mice were xenografted with IDH1 R132 or WT primary AML patient samples. When AML disease was established, mice were treated with subcutaneous (B–F) ATRA pellet at 2.5 mg (B and C) or 7.5 mg (D–F) pellet or control for 21 d. (B) Human cells were sorted from mice bone marrow, cytospun, and stained with May-Grünwald Giemsa. A representative picture is shown. (C) Cells were classified in 5 categories (from less to more differentiated in accordance of their morphological changes). 100 cells per condition were counted three times (Mann-Whitney test). Percentage of human (hCD45⁺mCD45⁻hCD33⁺hCD44⁺) CD11b⁺ cells was measured by flow cytometry analysis in bone marrow (D), spleen (E), and peripheral blood (F) of primary patient samples-xenografted mice. (Patient 1, n = 5 untreated and 5 treated mice; patient 2, n = 6 untreated and 6 treated mice; patient 3, n = 9 untreated and 9 treated mice). (G) Schematic representation of the in vivo protocol used for H–M. (H–M) Immunodeficient NOD-scid IL2rγ^null (NSG) female mice were xenografted with MOLM14 transduced with IDH1 R132H (H–J) or WT (K–M). When AML disease was established, mice were treated with subcutaneous ATRA pellet (five mice per group) or control (four mice per group). Blast differentiation and mice overall survival with or without ATRA (7.5 or 15 mg subcutaneous pellet) were monitored in vivo. After 6-d treatment, the percentage of human (hCD45⁺mCD45⁻hCD33⁺hCD44⁺) CD11b⁺ cells was assessed by flow cytometry (H and K) and the total cell tumor burden in bone marrow (hCD45⁺mCD45⁻hCD33⁺hCD44⁺Annexin-V^neg) cells was analyzed by flow cytometry in mutant IDH1–R132H (I) or WT (L) IDH1–WT MOLM14-xenografted mice. Mann-Whitney test. Survival of mutant IDH1–R132H (J) or WT (M) IDH1–WT MOLM14-xenografted mice treated with vehicle (CTL, n = 7) or ATRA (7.5 mg subcutaneous pellet, n = 7). *, P ≤ 0.05; **, P ≤ 0.01; and ***, P ≤ 0.001 indicate statistical significance of the observed differences (Log-rank [Mantel-Cox] test).