Advertisement

The efficiency of murine MLL-ENL–driven leukemia initiation changes with age and peaks during neonatal development

Theresa Okeyo-Owuor, Yanan Li, Riddhi M. Patel, Wei Yang, Emily B. Casey, Andrew S. Cluster, Shaina N. Porter, David Bryder and Jeffrey A. Magee

Data supplements

Article Figures & Data

Figures

  • Figure 1.

    MLL-ENL initiates AML most efficiently when it is induced in neonates. (A) Overview of the strategy for inducing MLL-ENL at E10.5, P0, and 3 weeks and 8 weeks after birth. (B) Kaplan-Meier survival curves for mice after MLL-ENL induction at the indicated ages. **P < .01, ***P < .0001 relative to induction at 8 weeks after birth. #P < .0001 for P0 relative to E10.5 induction. Survival curves were compared with the log-rank test. Group sizes are indicated. (C) Representative flow plot for AML specimens from Tet-OFF-MLL-ENL mice. (D) Representative AML cytospin. Scale bar indicates 100 μM. (E) Expression of the MLL-ENL transcript by RNA-seq after induction at E10.5. Reads from HPCs and pGMs were mapped to the human MLL-ENL sequence for the indicated induction ages and genotypes. Reads per kilobase of transcript per million mapped reads values from each sample were normalized to one allele of endogenous Mll1.

  • Figure 2.

    Transient suppression of MLL-ENL in neonates prevents AML in most mice. (A) Overview of experimental design. (B) Kaplan-Meier survival curves for Tet-OFF-MLL-ENL mice after neonatal suppression of the transgene. The P value is indicated and reflects a comparison of mice that received a 4-week doxycycline (Dox) pulse, beginning at P0, to mice that did not receive doxycycline. Group sizes are indicated. Survival curves were compared by the log-rank test. (C-F) Spleen weights, white blood cell (WBC) counts, and hemoglobin (Hgb) and platelet counts for mice that survived for >1 year after transient MLL-ENL suppression. All mice were euthanized ∼12 months after birth. (G) Representative bone marrow cytospins for mice that developed AML, in the absence of doxycycline, and mice that survived for >1 year after transient MLL-ENL suppression. Scale bars indicate 100 μM. For panels C-F, CTL indicates control and ME indicates Tet-OFF-MLL-ENL genotypes.

  • Figure 3.

    MLL-ENL depletes HSC and HPC numbers when it is induced in fetal, but not adult, mice. (A-B) HSC and HPC numbers in P0 liver (n = 15-19) or 10-week-old bone marrow (n = 10-12) after fetal MLL-ENL induction or in 6-month-old bone marrow (n = 6 per genotype) or 14-month-old bone marrow (n = 10-11) after adult MLL-ENL induction. Bone marrow cell numbers reflect 2 hindlimbs (tibias and femurs). (C) Twenty-four-hour HSC and HPC BrdU incorporation percentages were measured in E18.5 livers (n = 5-7) or 10-week-old bone marrow (n = 7 per genotype) after fetal MLL-ENL induction. For all panels, error bars indicate standard deviations. **P < .01, ***P < .001 for Tet-OFF-MLL-ENL relative to controls. Groups were compared by the 2-tailed Student t test.

  • Figure 4.

    MLL-ENL induces changes in gene expression more efficiently in neonatal HPCs than in older or more committed progenitors. (A) Volcano plots show genes that are significantly induced (red) or significantly repressed (green) in HPCs, pGMs, or GMPs at P0, or in 10-week-old HPCs, after fetal MLL-ENL induction. The MLL-ENL transcript was mapped, and expression is indicated as a blue data point. RNA-seq data were analyzed, and adjusted (adj.) P values were calculated, as indicated in "Materials and methods." Genes that changed with an adjusted P < .05 were considered significantly differentially expressed. (B) Volcano plots showing a lack of differential gene expression after adult MLL-ENL induction. The panels indicate gene expression 6-month-old or 14-month-old mice (4 months or 12 months after doxycycline withdrawal). MLL-ENL was expressed at these ages as indicated by the blue dots. (C) Venn diagrams comparing the number of genes for the indicated cell types and ages that showed significantly increased expression and a fold change >2. Data reflect fetal MLL-ENL induction. FDR, false discovery rate. (D-G) Reads per kilobase of transcript per million mapped reads (RPKM) values for Hoxa9, Hoxa10, Mecom, and Igf1 at the indicated ages after fetal MLL-ENL induction. Error bars indicate standard deviation. ***Adjusted P < .001. For all groups, n = 3 to 4 independent biological replicates.

  • Figure 5.

    Increased MLL-ENL expression partially overcomes the barrier to target gene activation and repression in adult progenitors. (A) MLL-ENL transcript levels in Col1a1TetO_MLL-ENL heterozygous or homozygous mice. Transcript expression was normalized to endogenous Mll1 in each sample, as in Figure 1E. (B-E) HSC, HPC, pGM, and GMP numbers in 10-week-old, homozygous Tet-OFF-MLL-ENL mice after fetal induction. Error bars indicate standard deviations. n = 5 to 8, *P < .05, **P < .01; groups were compared with the 2-tailed Student t test. (F-H) Volcano plots show genes that are significantly induced (red) or significantly repressed (green) in HPCs, pGMs, or GMPs in 10-week-old mice that are homozygous for the Col1a1TetO_MLL-ENL allele. The MLL-ENL transcript was mapped, and expression is indicated as a blue data point. Adjusted P values were calculated as indicated in "Materials and methods." (I-J) Numbers of unique genes that were significantly induced or repressed with a fold change >2 in P0 mice with a heterozygous Col1a1TetO_MLL-ENL allele, as compared with adult mice with a homozygous Col1a1TetO_MLL-ENL allele. Changes in gene expression were far more extensive in neonatal HPCs than at other time points or in other cell types, despite the presence of only a single MLL-ENL transgene. For all groups, n = 4 independent biological replicates. For panels B-E, Ctl indicates control and ME indicates Tet-OFF-MLL-ENL genotypes.

  • Figure 6.

    Transplanted HPCs and pGMs give rise to AML after adult MLL-ENL induction. (A) Overview of experiment timeline. (B) Percentage of donor (CD45.2) peripheral blood myeloid cells (CD11b+Gr1+), B cells (B220+), and T cells (CD3+) at 4 weeks after transplantation. *P < .05, **P < .01; groups were compared with the 2-tailed Student t test. (C) Kaplan-Meier survival curves for mice transplanted with control or MLL-ENL–expressing bone marrow. (D) Survival curves for a second cohort of MLL-ENL–expressing mice transplanted with 300 000 whole bone marrow cells or 3000 HPCs or pGMs. Group sizes are indicated. Survival curves were compared with the log-rank test.

  • Figure 7.

    Lin28b suppresses MLL-ENL–driven AML. (A-C) Lin28b, Hmga2 and Igf2bp2 were hyper-activated in MLL-ENL expressing neonatal, but not adult, HPCs. ***Adjusted P < .001. (D) Survival of Tet-OFF-MLL-ENL and Tet-OFF-MLL-ENL/Lin28b mice after neonatal transgene induction. Curves were compared by the log-rank test. (E-G) Spleen weight and HSC and HPC numbers in surviving Tet-OFF-MLL-ENL/Lin28b after they were euthanized at >1 year old. (H-I) Representative spleen histology and CD3 expression in control mice or Tet-OFF-MLL-ENL/Lin28b mice that became moribund. Scale bars indicate 100 μM (large panels) or 40 μM (insets). (J) MLL-ENL initiates AML most efficiently when it is induced during the neonatal stage of development. Lin28b might protect against AML initiation in utero. At later stages of development, MLL-ENL fails to activate target genes as efficiently as it does in neonates.