Development and survival of MYC-driven lymphomas require the MYC antagonist MNT to curb MYC-induced apoptosis

Abstract

Deregulated overexpression of MYC is implicated in the development and malignant progression of most (∼70%) human tumors. MYC drives cell growth and proliferation, but also, at high levels, promotes apoptosis. Here, we report that the proliferative capacity of MYC-driven normal and neoplastic B lymphoid cells depends on MNT, a MYC-related transcriptional repressor. Our genetic data establish that MNT synergizes with MYC by suppressing MYC-driven apoptosis, and that it does so primarily by reducing the level of pro-apoptotic BIM. In Eμ-Myc mice, which model the MYC/IGH chromosome translocation in Burkitt's lymphoma, homozygous Mnt deletion greatly reduced lymphoma incidence by enhancing apoptosis and markedly decreasing premalignant B lymphoid cell populations. Strikingly, by inducing Mnt deletion within transplanted fully malignant Eμ-Myc lymphoma cells, we significantly extended transplant recipient survival. The dependency of lymphomas on MNT for survival suggests that drugs inhibiting MNT could significantly boost therapy of MYC-driven tumors by enhancing intrinsic MYC-driven apoptosis.

Graphical abstract

Figure 1.

Lymphoid-specific loss of MNT greatly diminishes lymphoma development and induces lymphopenia in Eμ-Myc mice. (A) Kaplan-Meier survival curves showing reduced lymphoma development in Mnt^fl/fl Eμ-Myc/Rag1Cre mice (blue; median survival, 463 days) than Mnt^fl/+ Eμ-Myc/Rag1Cre mice (lime green; median survival, 138 days) and control Eμ-Myc mice (red; median survival, 86 days) and Eμ-Myc/Rag1Cre mice (purple; median survival, 96 days). **P ≤ .01; ****P ≤ .0001. Log-rank test. Killed mice showing no malignancy on autopsy were censored (black mark). X-axis was arbitrarily terminated at 400 days, but monitoring continued. (B) PCR analysis shows efficient deletion of floxed Mnt alleles by Rag1Cre in cells sorted from bone marrow of individual 4-week-old mice. WT and floxed Mnt alleles both produce 147-bp fragments, deleted Mnt allele (MntΔ), a 386-bp fragment. Lanes 1,2: pro-B and pre-B cells, respectively, from Mnt^fl/fl Eμ-Myc/Rag1Cre mouse #360; lanes 3,4: pre-B cells from control Mnt^+/+ Rag1Cre mice (#403, #404); lanes 5,6 pre-B cells from Mnt^fl/fl Rag1Cre mice (#413, #414); C, control DNA for Mnt PCRs. (C-D) Lymphoid-specific MNT loss induces lymphopenia. Flow cytometric quantification of B lymphoid subpopulations in bone marrow (C) and spleen (D) of 4-week-old WT (light green), Eμ-Myc (red), Eμ-Myc/Rag1Cre (purple), and Mnt^fl/fl Eμ-Myc/Rag1Cre (blue) mice. Supplemental Figure 3A exemplifies sorting strategy. Bar graphs show mean ± SD; *P ≤ .05; **P ≤ .01; ***P ≤ .001; ****P ≤ .0001. (E) Lymphocyte count in blood of 4-week-old mice of indicated genotypes, determined in an Advia hematology analyzer. Mean ± SD; *P ≤ .05; **P ≤ .01; ***P ≤ .001; ****P ≤ .0001.

Figure 2.

Mnt deletion increases apoptosis but does not affect MYC protein level, cell cycling, or senescence in premalignant Eμ-Myc pro-B and pre-B cells. (A) MNT loss does not change MYC level. Pro-B and pre-B cells were sorted from bone marrow of 4-week-old mice, permeabilized and stained with MYC antibody (blue) or isotype-matched control (red). (Bottom) Mean intracellular MYC fluorescence (MFI) ± SD in WT (light green), Eμ-Myc (red), and Mnt^fl/fl Eμ-Myc/Rag1Cre (blue) pro-B and pre-B cells; n = 3 or 4; mean ± SD; **P ≤ .01; ns = not significant. (B) Cell cycle analysis. DNA content of 4′,6-diamidino-2-phenylindole-stained cells shows that the proportion of cycling (S-G2-M) pre-B cells in Mnt^fl/fl Eμ-Myc/Rag1Cre (blue) is comparable to that in Eμ-Myc (red) and Eμ-Myc/Rag1Cre (purple) mice. n = 3-6. Mean ± SD; *P ≤ .05; **P ≤ .01. (C) MNT loss does not significantly affect H3K9 trimethylation. Representative H3K9me3 staining of sorted permeabilized pro-B and pre-B cells from 4-week WT (light green), Eμ-Myc (red), and Mnt^fl/fl Eμ-Myc/Rag1Cre (blue) mice (left panels) and MFI for 3 mice of each genotype in 3 independent experiments (right panels). (D) MNT loss markedly elevates apoptosis. Annexin-V+ pro-B and pre-B cells were 2- to 3-fold more frequent in bone marrow of Eμ-Myc (red) than WT (light green) mice and ∼3-fold higher in Mnt^fl/fl Eμ-Myc/Rag1Cre (blue) than in Eμ-Myc mice. Top panels show typical fluorescence-activated cell sorting plots, and bottom panel shows mean percentage annexin-V⁺ cells; n = 6 to 9 individual mice; mean ± SD; **P ≤ .01; ***P ≤ .001.

Figure 3.

In situ analysis of cell proliferation and apoptosis. (A) Tilescan sections of whole tibia. Bones were harvested from 4-week old Eμ-Myc (n = 5), Eμ-Myc/Rag1Cre (n = 4), Mnt^fl/fl Eμ-Myc/Rag1Cre (n = 3), and WT (n = 4) mice, stained and imaged by combined 2-photon/confocal imaging of cell populations in situ. Red = CD19; green = Ki67; magenta = cleaved caspase-3; gray = second harmonic generation (bone). To better visualize CC3-positive cells in Figure 3B, the data are presented using an inverse black and white LUT (Lookup Table). Scale bar, 500 μm. (B) Zoomed areas from original tilescans in panel A (white boxes) with individual channels for CD19, Ki67, and cleaved caspase-3. A merged image (bottom panel) from each area is shown relative to bone signal (gray). Scale bar, 50 μm. (C-E) Quantification of cells in tibia positive for CD19, Ki67, and cleaved caspase-3. Color images were separated into single binary channels and then thresholded, and the number of positive pixels for each whole bone section quantified for CD19 (C), Ki67 (D) and cleaved caspase-3 (E). Each data point represents a whole quantified bone from an individual mouse in each genotype (n = 3-5). Bar graphs show mean ± SD; significance was determined by analysis of variance *P ≤ .05; **P ≤ .01).

Figure 4.

Loss of MNT impedes normal B lymphoid development. (A-B) Deletion of floxed Mnt by Rag1Cre provokes lymphopenia in mice lacking Eμ-Myc transgene. Quantification of B lymphoid subpopulations from bone marrow (A) and spleen (B) of 4-week-old WT (light green), Rag1Cre (green), and Mnt^fl/fl Rag1Cre (light blue) mice. Supplemental Figure 5A exemplifies sorting strategy. n indicates number of mice. Mean ± SD; **P ≤ .01; ***P ≤ .001; ****P ≤ .0001. (C-D) Activation of CreERT2 by tamoxifen also provokes B lymphopenia in mice lacking Eμ-Myc transgene. Quantification of indicated B lymphoid populations from bone marrow (C) and spleen (D). Young (6-week-old) WT (light green), CreERT2 (gray), and Mnt^fl/fl CreERT2 (lavender) mice were treated for 3 days with tamoxifen by oral gavage and analyzed 4 weeks later. n indicates number of mice. Bar graphs show mean ± SD; *P ≤ .05; **P ≤ .01.

Figure 5.

Loss of MNT increases apoptosis of pro-B cells in IL-7 cultures. (A) Expression of MYC, MNT, and MCL-1 protein in CD19⁺IgM⁻ cells sorted by FACS from bone marrow of 6-week-old WT mice, before and during culture in IL-7, and 24 hours after IL-7 removal. (B-D) MNT-deficient pro-B cells exhibit increased apoptosis, elevated BIM, and decreased MCL-1. Pro-B cells were obtained by culturing CD19⁺ bone marrow cells (isolated using microbeads) in IL-7 (5 ng/ml) for 4 days. (B) Annexin-V staining of Rag1Cre and Mnt^fl/fl Rag1Cre pro-B cells; profiles are typical of those from ≥3 mice of each genotype (see panel E). (C) Flow cytometric analysis of intracellular BIM using Bim^−/− cells (gray shaded) as a negative control. Bar graphs compare percentage BIM-positive cells and mean fluorescence intensity (MFI) from WT (light green) vs Mnt^fl/fl Rag1Cre (light blue) mice, determined in 3 independent experiments; MFI of BIM-null cells was subtracted from that of either WT or Mntf^l/fl Rag1Cre cells stained in same experiment. (D) Typical western blot of pro-B cells from WT, Mnt^fl/fl Rag1Cre and control Bim^−/− mice. For each blot, BIM and MCL-1 levels were quantified relative to ACTIN and normalized to WT value. Bar graphs show mean fold-change ± SD **P ≤ .01; ***P ≤ .001. (E) Bim heterozygosity significantly reduces apoptosis of MNT-null pro-B cells. Cells were stained with annexin-V after 4 days in IL-7. Bar graphs show mean ± SD; *P ≤ 0.05; **P ≤ .01; ****P ≤ .0001.

Figure 6.

Bim heterozygosity rescues B lymphopoiesis, reduces apoptosis, and accelerates lymphomagenesis in MNT-deficient Eμ-Myc mice. (A) BIM and MCL-1 expression is elevated in MNT-deficient CD19⁺ cells. Cells were fluorescence-activated cell sorted from spleens of 4-week-old mice. (B-C) Bim heterozygosity largely restores B lymphopoiesis in young Mnt^fl/fl Rag1Cre mice. Flow cytometric enumeration of B lymphoid cells in bone marrow (B) and spleen (C) of 6-week-old mice. Data include mice in Figure 4 plus additional mice. Bar graphs show mean ± SD; *P ≤ .05; **P ≤ .01; ****P ≤ .0001. (D-E) Bim heterozygosity partially restores B lymphopoiesis in young MNT-deficient Eμ-Myc mice. Enumeration of B lymphoid cell populations in (D) bone marrow and (E) spleen of the indicated genotypes. Data for controls include certain mice in Figure 1C-D. Mean ± SD; *P ≤ .05; **P ≤ .01; ****P ≤ .001; ****P ≤ .0001. (F) Bim heterozygosity ameliorates enhanced apoptosis of pro-B and pre-B cells in the bone marrow of young MNT-deficient Eμ-Myc mice. Data include mice from Fig. 1C. Mean ± SD; *P ≤ .05; ns = not significant. (G) Bim heterozygosity accelerates lymphomagenesis in MNT-deficient Eμ-Myc mice. Kaplan-Meier survival curves showing enhanced morbidity of Bim^+/− Mnt^fl/fl Eμ-Myc/Rag1Cre (orange) mice compared with Mnt^fl/fl Eμ-Myc/Rag1Cre (blue) mice, and of Bim^+/− Mnt^fl/+ Eμ-Myc/Rag1Cre (mustard) mice compared with Mnt^fl/+ Eμ-Myc/Rag1Cre (lime green) mice. Survival curves for Eμ-Myc/Rag1Cre, Mnt^fl/+ Eμ-Myc/Rag1Cre, and Mnt^fl/fl Eμ-Myc/Rag1Cre mice are those shown in Figure 1A. Log-rank test; *P ≤ .05; ****P ≤ .0001.

Figure 7.

MNT loss induced in vivo in Eμ-Myc lymphomas extends survival of transplant recipients. (A) Schematic of experimental design. Mnt^fl/fl Eμ-Myc/CreERT2 and Eμ-Myc/CreERT2 mice were aged until they developed tumors (∼100 days). Tumor cells (Ly5.2+) were harvested and either cultured to establish cell lines for in vitro treatment with 4-OH tamoxifen (4-OHT) (see panel B) or injected intravenously into syngeneic nonirradiated C57BL/6-Ly5.1 mice for in vivo treatment with tamoxifen (see panel C). (B) 4-OH tamoxifen-induced Mnt loss enhances apoptosis of Eμ-Myc lymphoma cells in vitro. Short-term cell lines established from 2 independent Mnt^fl/fl Eµ-Myc/CreERT2 lymphomas (#1129 and #1271) and a control Eµ-Myc/CreERT2 lymphoma (#1194) were treated for 24 hours with or without 0.5 μM 4-OH tamoxifen and percentage annexin-V-positive cells determined by flow cytometry. Results are from 4 (#1194, #1129) or 3 (#1271) independent experiments; mean ± SD **P ≤ .01; **P ≤ .001. Immunoblot shows MNT and ACTIN protein in cells treated with tamoxifen or vehicle. (C) Tamoxifen-induced Mnt deletion significantly extends survival of mice transplanted with Mnt^fl/fl Eμ-Myc/CreERT2 lymphomas. Kaplan-Meier survival curves of mice transplanted with 14 independent Mnt^fl/fl Eμ-Myc/CreERT2 or 9 control Eμ-Myc/CreERT2 lymphomas and subsequently treated with either tamoxifen or vehicle alone. Each lymphoma was transplanted into 6 nonirradiated C57BL/6 recipients (2-4 × 10⁶ cells/mouse), 3 of which were treated by oral gavage with tamoxifen and 3 with vehicle alone, for 2 successive days, starting on day 5; n indicates number of independent lymphomas transplanted. Significance was determined using log-rank test. See also Supplemental Figure 6A. (D) Induced Mnt deletion reduces viability of p53 wt and p53 mutant Mnt^fl/fl Eμ-Myc/CreERT2 lymphoma cell lines. Cell lines established from Mnt^fl/fl Eµ-Myc/CreERT2 and control Mnt^+/+ Eµ-Myc/CreERT2 lymphomas were incubated with 0.5 μM 4-OH tamoxifen to delete Mnt, and cell viability was determined by flow cytometry (supplemental Figure 8). Viability of 4-OHT-treated cells at 48 and 72 hours is expressed relative to that of cells incubated in parallel without 4-OHT. P53 status, determined by treatment with nutlin3a, is indicated.