MNT: a new target for AML Open Access

Acute myeloid leukemia (AML) remains a major clinical challenge despite encouraging results from new targeted therapeutic options.^1,2 AML can arise from diverse genetic changes.^3,4 Among the most common, particularly in infant and childhood AML and in therapy-induced AML, are chromosomal translocations/inversions that involve the gene MLL (mixed lineage leukemia), also known as KMT2A (lysine methyl transferase 2A), which is located at chromosome 11q23. MLL encodes a large multidomain epigenetic regulator,^5,6 and the translocations lead to loss of its C-terminus, which contains a SET (Su(var)3-9, Enhancer-of-zeste, and Trithorax) domain responsible for H3K4 methyltransferase activity. More than 100 fusion partners have been identified, and the consequent MLL fusion proteins differ in potency. Several of the most common (eg, MLL::AF9, MLL::ENL, MLL::AF4, MLL::AF10, and MLL::ELL) constitute a transcriptional activation complex that includes p-TEFb, a cyclin T–dependent kinase that controls elongation by RNA polymerase II, and/or the histone methyltransferase DOT11. Therefore, most MLL gene rearrangements (MLL-r) lead to overexpression of MLL target genes, including HOXA9 and its co-factor MEIS1 (reviewed by Schreiner et al⁷).

Importantly, MLL fusion proteins directly activate the expression of c-MYC,^8-10 a basic helix-loop-helix leucine zipper (bHLHLZ) transcription factor that controls a host of genes involved in cell growth, cell cycle progression, metabolism, and DNA damage responses.^11,12 Counterintuitively, cells that express elevated MYC undergo apoptosis under stress conditions (eg, cytokine or nutrient deprivation),^13,14 particularly at high MYC levels.¹⁵ Although this serves as a critical restraint on MYC-driven neoplastic transformation, the safety mechanism can be overridden by high levels of proteins that enhance cell survival (eg, BCL-2¹⁶) or loss of proteins that promote cell death (eg, BH3-only protein BIM¹⁷).

MYC heterodimerizes with MAX (MYC-associated factor X), a related bHLHLZ protein, and together they bind E-box motifs (CACGTG and variants) in regulatory regions of target genes to stimulate their transcription.¹⁸ MAX also binds several MYC-related transcriptional repressors, including MNT (MAX network transcriptional repressor),¹⁹ which is widely expressed in mammalian tissues.^19,20 MNT/MAX heterodimers suppress transcription by recruiting SIN3/HDAC (switch-independent 3/histone deacetylase) co-regulatory complexes to E-box motifs.^11,21,22 

As a MYC antagonist, MNT was expected to be a tumor suppressor. Early studies that showed adenocarcinoma development after conditional Mnt deletion in mammary epithelium supported this role.²³ However, more recently, mouse genetic studies have shown that Mnt deletion inhibits B and T lymphoma development by enhancing apoptosis in lymphoid cells that overexpressed MYC.^24-26 Thus, at least in the lymphoid system, MNT actually aids MYC in oncogenesis instead of antagonizing it. It does so by suppressing MYC-driven apoptosis,^23-25 presumably by competing for critical target genes that promote apoptosis, either directly or indirectly.

Because many AMLs express high levels of MYC, which is an important negative prognostic factor,²⁷ we wondered whether AMLs, like MYC-driven lymphomas, are dependent on MNT for sustained cell survival and expansion. In this study, we explored the impact of inducing MNT loss in mouse and human AML cells that expressed MLL-fusion proteins and investigated whether MNT loss could enhance the response of these AMLs to BH3-mimetic drugs, which inhibit prosurvival members of the BCL-2 family.^28,29 

The full details are provided in the supplemental Materials and methods.

All mice were maintained at the Walter and Eliza Hall Institute (WEHI) on a C57BL/6 background. Mnt^fl/+30 and Rosa26^CreERT2/+31 (hereafter referred to as CreERT2) mice were crossed to generate Mnt^fl/fl and Mnt^fl/fl/CreERT2 genotypes. Homozygous Mnt deletion, reported as being perinatally lethal,³⁰ is fatal at around embryonic day 10 (E10) in C57BL/6 mice at WEHI.²⁵ CreERT2 transgene expression enables inducible deletion of floxed Mnt alleles by treatment with tamoxifen in vivo or 4-hydroxy tamoxifen (4-OHT) in vitro.³² 

All mice used in this study were bred, housed, and monitored in the Bioservices Facility at WEHI under the supervision of trained veterinarians and in accordance with the WEHI animal ethics committee regulations and the Australian Code for the Care and Use of Animals for Scientific purposes.

Primary AMLs (T0) were generated by MLL::AF9/GFP retrovirus infection of E14 fetal liver cells from C57BL/6-Ly5.2 mice, followed by transplantation into sublethally irradiated (7.5 Gy) C57BL/6-Ly5.1 recipient mice (Figure 1A), as described previously.^33,34 

Secondary (T1) AMLs were generated by transplanting spleen cells from primary (T0) AML mice into multiple nonirradiated C57BL/6-Ly5.1 mice of the same sex as the initial E14 embryo; recipients developed secondary (T1) AML within ∼2 weeks.

MLL::AF9 AML cell lines (CLs) were generated by culturing bone marrow cells from T0 MLL::AF9 AML mice in interleukin-3 as described.^33,34 All CLs were regularly tested for mycoplasma (Lonza MycoAlert).

Cells were plated into 96-well flat-bottom plates (5 × 10⁴ cells per well) and incubated for 24 hours in 4-OHT (0.5 μM; catalog no. H7904; Sigma-Aldrich) or vehicle (ethanol), followed by treatment with BH3-mimetic drugs or vehicle for 48 hours. The BH3 mimetics used were S63845 (MCL-1 inhibitor; catalog no. A-6044; Active Biochem), ABT-199/venetoclax (BCL-2 inhibitor; catalog no. A-1231; Active Biochem), and A-1331852 (BCL-X_L inhibitor; Guillaume Lessene, WEHI), all dissolved in dimethyl sulfoxide. If required, the pan-caspase inhibitor QVD-OPh (25 μM; catalog no. HY-12305; MedChemExpress) was added 30 minutes before treatment with 4-OHT. Live cells (annexin V–negative propidium iodide–negative) were quantified by flow cytometry.

Multiple nonirradiated, 8-week-old C57BL/6-Ly5.1 mice were transplanted via tail vein injection with bone marrow cells from a primary (T0) AML mouse (5 × 10⁵ cells per recipient) and treated on days 3, 4, and 5 by oral gavage with tamoxifen (200 mg/kg body weight; catalog no. T5648; Sigma-Aldrich) or vehicle (peanut oil; catalog no. P2144; Sigma-Aldrich). Mice were monitored daily and euthanized humanely at the ethical end point or after 195 days (experimental end point).

For AML treatment with tamoxifen with/without MCL-1 inhibitor S63845, 12 of 24 nonirradiated mice transplanted with 5 × 10⁵ bone marrow cells from T1 AML mice were treated with tamoxifen and 12 of the 24 were treated with vehicle. Subsequently, in each cohort, 6 were injected via the tail vein on days 6 to 10 with S63845 (25 mg/kg body weight; catalog no. 6044; Active Biochem) and 6 with vehicle (2% vitamin E; catalog no. 57668; in Dulbecco's phosphate-buffered saline; Sigma-Aldrich).

The human AML CLs used were MV4;11,³⁵ THP-1,³⁶ MOLM-13,³⁷ and OCI-AML3,³⁸ authenticated by STR (short tandem repeat) profiling at CellBank Australia. The cells were stably transduced simultaneously with lentiviruses that expressed either FuCas9-Cherry (Addgene Plasmid, catalog no. 70182) or single-guide RNAs (sgRNAs) in doxycycline-inducible FgH1tUTG/GFP (Addgene Plasmid, catalog no. 70183) as described.^39,40 To induce sgRNA expression, mCherry⁺GFP⁺ cells were treated with doxycycline hyclate (1 μg/ml; catalog no. D9891; Sigma-Aldrich) for 5 days.

NOD SCID Gamma (NSG) mice were injected IV with 5 × 10⁵ MV4;11 cells that expressed Cas9 and doxycycline-inducible sgRNAs that targeted human MNT (MNT2 and MNT4). Doxycycline treatment commenced on day 4 after transplantation (600 mg/kg doxycycline hyclate in chow, fed orally ad libitum; catalog no. SF08-026; Specialty Feeds). Treatment with the BCL-2 inhibitor venetoclax (ABT-199) commenced on day 10 (50 mg/kg body weight; every day, except on weekends, by oral gavage, for 4 weeks). Mice were monitored daily and euthanized humanely at the ethical end point. MV4;11 (human CD45⁺) cells were quantified in the bone marrow on day 28 by intrafemoral sampling⁴¹ and at the experimental end point.

Statistical comparisons were made using a 1-way or 2-way analysis of variance with Tukey multiple comparison test or a 1-way analysis of variance with Dunnett multiple comparison test using Prism software, version 9 (GraphPad, San Diego, CA). Unless otherwise stated, the data are shown as the mean ± standard error of the mean with P values of ≤.05 considered statistically significant. Mouse survival curves were plotted in GraphPad Prism, version 9. The curves were compared using the log-rank (Mantel-Cox) test with Bonferroni correction.

To explore the consequences of MNT loss for MLL-t AML, we first focused on the well-studied MLL::AF9 mouse model^42,43 using the CreERT2 transgene³¹ to conditionally delete floxed Mnt alleles in MLL::AF9 AML cells.

In view of the widespread expression of both Mnt²⁰ and the CreERT2 transgene,³¹ before embarking on the AML studies, we investigated whether Mnt deletion perturbs normal hemopoiesis. Cohorts of 8 week-old Mnt^fl/flCreERT2 mice were treated with tamoxifen to activate the CRE-ERT2 protein³¹ and then analyzed 4 and 8 weeks later, thereby avoiding the reported transient toxicity.⁴⁴ CreERT2-mediated Mnt deletion was highly efficient as evidenced by polymerase chain reaction and western blot analysis (supplemental Figure 1A-B). The mice remained viable and healthy and only modest hemopoietic changes were observed (supplemental Figure 1C-E; supplemental Table 1). At 4 weeks, the treated mice were mildly leukopenic mainly because of a deficit of lymphocytes and monocytes, but the white blood cell levels returned to normal by 8 weeks. The red blood cell counts were mildly reduced at both 4 and 8 weeks. The major change noted in the bone marrow at 8 weeks was an increase in long-term hemopoietic stem cells (LT-HSC; CD48^–CD150⁺) and multipotential progenitor 2 cells (MPP2; CD48⁺CD150⁺) in the Lin^- Sca1⁺ c-Kit⁺ stem cell compartment. The spleen remained essentially normal. These data indicated that MNT loss did not have a major impact on normal hemopoiesis in the likely time period of the AML study.

To generate Mnt-deletable MLL::AF9 AMLs, hemopoietic stem/progenitor cells (HSPCs) cells from E14 fetal livers of Mnt^fl/flCreERT2 mice were infected with MLL::AF9/GFP or GFP MSCV (Murine Stem Cell Virus) retroviruses and injected into sublethally irradiated (7.5 Gy) recipient mice (Figure 1A). As expected, all mice that were transplanted with MLL::AF9 virus–infected HSPCs developed typical AML and required ethical euthanasia within 45 days (supplemental Figure 2A), whereas the controls that received GFP virus–infected HSPCs remained healthy. The AML-bearing mice presented with leukocytosis, anemia, thrombocytopenia, and splenomegaly (supplemental Figure 2B-E; supplemental Table 2), and their bone marrow was replete with AML cells, which exhibited robust MYC and MNT expression (Figure 1B). These primary AMLs are designated hereafter by the number of the mouse that received transplant, genotype of the donor HSPCs, and virus (eg, 2232 Mnt^fl/flCreERT2/MLL::AF9 indicates AML that developed in mouse 2232 transplanted with Mnt^fl/flCreERT2 fetal liver cells infected with MLL::AF9/GFP virus). In all, we generated 6 independent Mnt^fl/flCreERT2/MLL::AF9 AMLs and, as controls, 5 independent CreERT2/MLL::AF9 AMLs and 2 independent Mnt^fl/flMLL::AF9 AMLs.

To determine the impact of Mnt deletion, we first performed in vitro studies. Short-term MLL::AF9 AML CLs were established by culturing bone marrow cells from primary AML mice in interleukin-3–containing medium.^33,34 Polymerase chain reaction tests indicated that Mnt deletion using 0.5 μM 4-OHT was largely complete by 24 hours (Figure 1C). Multiple independent CLs (n = 7) were treated with 4-OHT or vehicle, and cell viability was determined by flow cytometry after 24 hours (Figure 1D). The viability of control (Mnt^+/+) CreERT2/MLL::AF9 CLs (blue) decreased after exposure to 4-OHT, as previously noted,⁴⁵ probably because of off-target CRE-mediated toxicity. Nevertheless, the viability of Mnt^fl/flCreERT2/MLL::AF9 AML CLs (pink) was significantly lower again (P = .0170; Figure 1E) because of the Mnt deletion. Importantly, MNT loss did not alter the level of expression of the driver fusion protein MLL::AF9 (Figure 1F).

The BCL-2 family of prosurvival and proapoptosis proteins regulates the mitochondrial cell death pathway.^28,MLL::AF9 AML CLs express prosurvival BCL-2, MCL-1, and BCL-X_L (Anstee et al³⁴; Figure 2A), and MCL-1 is essential for the development and sustained growth of AML driven by MLL fusion genes.⁴⁵ Recently, BH3 mimetic drugs that inhibit individual prosurvival family members have been developed to promote the apoptosis of cancer cells,⁴⁶ and the BCL-2 inhibitor venetoclax is being used in combination therapy to treat AML.⁴⁷To explore whether MNT loss could sensitize the AML cells to BH3 mimetic drugs, 5 independent Mnt^fl/flCreERT2/MLL::AF9 AML CLs were treated for 24 hours with 4-OHT or vehicle (ethanol) and then for another 24 hours with S63845 (MCL-1 inhibitor), ABT-199/venetoclax (BCL-2 inhibitor), A-1331852 (BCL-X_L inhibitor), or vehicle (dimethyl sulfoxide) in the presence of 4-OHT. Cell death was significantly increased by combining Mnt loss (+4-OHT) with exposure to each of the BH3 mimetic drugs (Figure 2B). The inclusion of the pan-caspase inhibitor QVD-OPH (Quinoline-Val-Asp-Difluorophenoxymethylketone) in the medium enhanced cell survival, indicating that cell death provoked by Mnt loss alone, or in combination with BH3 mimetic drugs, was by apoptosis (Figure 2C). Combining BH3 mimetics with different specificities was more effective than using a higher concentration of either alone, and Mnt loss again augmented killing, which was dependent on caspase activity (Figure 2D). Western blot analysis of QVD-OPH–treated Mnt^fl/flCreERT2/MLL::AF9 AML cells derived from mouse 2235 (Figure 2E) suggested that the increased efficacy of BH3 mimetic drugs after Mnt deletion was because of a further elevation of the BH3-only protein BIM (Bcl-2 Interacting Mediator of cell death), a proapoptotic member of the BCL-2 family.⁴⁸ No consistent changes were observed in the level of the prosurvival proteins MCL-1, BCL-X_L, or BCL-2.

We next tested the impact of Mnt deletion in transplanted MLL::AF9 AMLs (Figure 3A). Bone marrow cells from each primary (T0) AML mouse were injected IV into a cohort of 8 nonirradiated C57BL6 mice, 4 of which received tamoxifen (200 mg/kg) and 4 of which received only vehicle (negative control). A total of 13 independent primary AMLs were transplanted in 6 different experiments, each of which included both Mnt-deletable and control Mnt nondeletable genotypes. The AML cell death induced by MNT loss would be expected to improve the survival of transplant recipients.

Regardless of the AML genotype, the median survival of transplant recipients that were treated with vehicle was 17 days (pink and blue broken lines in Figure 3A). Control CreERT2/MLL::AF9 AML recipients that were treated with tamoxifen (blue solid line) survived longer (∼24 days), probably because of the impact of tamoxifen and/or off-target CRE activity. However, of note, the survival of Mnt^fl/flCreERT2/MLL::AF9 AML recipients treated with tamoxifen (solid pink line) was considerably longer (median, 57 days; P < .0001).

Remarkably, 11 of the 22 tamoxifen-treated Mnt^fl/flCreERT2/MLL::AF9 AML recipients (2/4 transplanted with 2205, 2/3 with 2206, 4/4 with 2235, and 3/3 with 2233) seemed to be cured (supplemental Table 3). Analysis of these survivors 195 days after transplantation revealed that their blood profiles had returned to normal and that the AML cells (ie, CreERT2⁺, MntΔ⁺, MYC protein^hi) were largely undetectable in the bone marrow, as exemplified for mouse 9841 and mouse 9844 in Figure 3B-C.

The other 11 tamoxifen-treated Mnt^fl/flCreERT2/MLL::AF9 AML recipients developed typical leukemia, with a mean survival of 30 days (supplemental Table 3). When autopsied, these mice had high white blood cell counts and low platelet counts, and their bone marrow and spleen was dominated by Mnt-deleted AML cells that expressed high levels of MYC protein (eg, 9842, 9843 in Figure 3B-C). We infer that, despite experiencing some initial survival benefit, these mice were eventually overwhelmed by clone(s) bearing mutations that conferred resistance to MNT loss. We have not yet identified these lesion(s) but have ruled out loss of proapoptotic BIM and, for 2 tumors, loss of proapoptotic BAX or BAK (not shown).

Taken together, these results suggest that MNT loss very significantly disadvantages MLL::AF9 AML cells in vivo, presumably by enhancing cell death, but mutations that confer resistance can emerge.

In view of the MCL-1 dependency of MLL fusion AMLs,⁴⁵ we wondered whether combining Mnt deletion with S63845 treatment would be more efficacious than either alone. For these experiments, because stocks of T0 AML bone marrow cells had been depleted, we used bone marrow cells from secondary (T1) AML mice, generated by transplanting cryopreserved spleen cells from T0 AML mice. The treatment protocol is outlined in Figure 4A and outcomes are summarized in Figure 4B-E and supplemental Table 4.

Mnt deletion alone (tamoxifen treatment; solid pink curve) prolonged survival of the recipient mice, but the extent varied for individual T1 AMLs. The impressive effect for T1 3413 (Figure 4C; mean survival, >195 days), which derived from T0 2205 (hence designated 2205/3413), mirrored that of T0 2205 (mean survival, 127 days; supplemental Table 3). T1 2232/3384 also responded to Mnt deletion in a similar manner as T0 2232 (Figure 4D; mean survival, 30 days vs 28.5 days, respectively). However, T1 2206/3427 (Figure 4E) was far less responsive than T0 2206 (mean survival, 16.5 days vs 85 days), presumably because further mutation had occurred during the expansion in vivo before testing.

Although treatment with S63845 alone (compare solid black and broken gray curves) prolonged the survival of some T1 MLL::AF9 AML-transplanted mice (outcome for 2205/3413, 2232/3384; Figure 4C-D), it had no impact on others (T1 2201/3378 and T1 2206/3427; Figure 4B,E). Furthermore, it did not provide any further survival benefit to mice that had first been treated with tamoxifen to delete Mnt (compare pink and orange curves; Figure 4C-E).

To extend our studies to human AML, we investigated 4 patient-derived human AML CLs, 3 of which had MLL fusion genes (MV4;11; MLL::AF4,³⁵ MOLM-13; MLL::AF9,³⁷ and THP-1; MLL::AF9),³⁶ and the fourth (OCI-AML3) had NPM1 and DNMT3A mutations.³⁸ Each of these long-established lines expresses MYC and MNT, as well as varying levels of the prosurvival proteins BCL-2, BCL-X_L, and MCL-1 (Figure 5A).

To perform the CRISPR/Cas9 deletion, we used a dual lentiviral vector system^39,40 to introduce CAS9 and sgRNAs. Two independent sgRNAs that targeted human MNT (MNT2, MNT4) were tested, as well as an sgRNA for mouse Bim as a nontargeting control. After transduction, viable cells that expressed high levels of both mCherry (CAS9⁺) and GFP (Green fluorescent protein) (sgRNA⁺) were isolated by fluorescence-activated cell sorting and treated with doxycycline to induce sgRNA expression. Western blot analysis indicated reduced MNT protein levels after 3 to 5 days (Figure 5B).

To assess whether MNT loss increased the sensitivity to BH3 mimetic drugs, we treated the pooled populations of mCherry^hiGFP^hi cells with doxycycline for 3 days, followed by treatment with the BH3 mimetic drugs S63845 (MCL-1 inhibitor) or ABT-199/venetoclax (BCL-2 inhibitor) in the presence of doxycycline for a further 2 days. The BH3-mimetic concentrations used were based on predetermined 50% inhibitory concentration values. Figure 5C-F shows that MNT protein depletion enhanced the response of MV4;11, THP-1, and OCI-AML3 cells to S63845 and, in the case of MV4;11, also to ABT-199 (venetoclax). However, the sensitivity of MOLM-13 cells was minimally affected.

Next, we turned to a xenograft model to test the impact of MNT loss in vivo in comparison with the outcome of treatment with venetoclax (ABT-199), which is approved for treating certain patients with AML.⁴⁹ For this trial, MV4;11 human AML cells transduced with Cas9 and MNT2 sgRNA (Figure 5B) were additionally transduced with the MNT4 sgRNA vector. In vitro testing of mCherry^hiGFP^hi cells to be used for the xenograft showed highly effective MNT loss after 3 days of doxycycline treatment (Figure 6A). Of the 24 transplanted NSG mice, 6 were left untreated and the others (3 cohorts of 6 mice each) were treated with doxycycline alone (provided in chow throughout the experiment, starting on day 4), venetoclax alone, or doxycycline plus venetoclax.

Notably, MNT loss induced by doxycycline treatment significantly prolonged the survival of transplant recipients when compared with the untreated control mice (compare pink and pale blue curves; P < .001). Indeed, survival was better with MNT loss than with venetoclax treatment (compare pink and dark blue curves; P < .01). Although combining venetoclax with MNT loss (green curve) did not statistically improve the survival when compared with MNT loss alone (pink curve), the trend toward longer survival (median survival, 39 days vs 36 days; supplemental Table 5) suggests that this may be achievable by modifying the treatment regimen. Consistent with longer survival being a consequence of a reduction in tumor burden, flow cytometric analysis of the intrafemoral bone marrow taken on day 28 during treatment (Figure 6C) showed a reduction in transplanted MV4;11 (human CD45⁺) cells when compared with controls. The positive response for this aggressive, long-cultured AML CL suggests that an effective MNT inhibitor would be helpful for treating human MLL-r AMLs. Of note, many other AMLs in the BEAT AML data set have MNT levels at least as high as MLL-r AMLs (Figure 6D) and thus may also be dependent on MNT.

The last 2 decades have seen significant progress in identifying diverse genetic driver mutations in AML and in developing therapies directed against those mutations, such as inhibitors of FLT3 (FMS-related tyrosine kinase 3), IDH1/IDH2 (Isocitrate dehydogenase 1/2), BCL-2, and MENIN, and epigenetic modifiers.^2,47,49-54 Although targeted therapeutics may have only modest activity when used as single agents, they can significantly augment clinical efficacy when administered in combination with other drugs.⁴⁹ Treatment with the BCL-2 inhibitor venetoclax, in combination with azacytidine or decitabine^47,51,55 or with low dose cytarabine,^56,57 is now approved for treatment of newly diagnosed patients with AML who are ineligible for intensive chemotherapy, and MCL-1 inhibitors have entered clinical trials.⁵⁸ However, although venetoclax combination therapy has improved the response rates and reduced toxicities, most patients with AML still relapse within 8 to 17 months.^51,57 Hence, despite these advances, there is still an urgent need for additional therapeutic agents to improve the outcomes for patients with AML.

Here, we have established for the first time to our knowledge, the dependency of certain myeloid leukemias on the bHLHLZ transcription factor MNT. Using multiple independent primary murine MLL::AF9 AMLs, we showed in vitro that Mnt deletion provoked apoptosis and augmented the cell death triggered by the BH3 mimetics S63845 (MCL-1 inhibitor; P < .0001), ABT-199/venetoclax (BCL-2 inhibitor; P < .0001), and A-1331852 (BCL-X_L inhibitor; P < .01). In vivo, we found that Mnt deletion significantly extended the survival of mice that were transplanted with MLL-AF9 AMLs (P < .0001; median survival, 57 days vs 17 days). Most notably, 11 of the 22 tamoxifen-treated Mnt^fl/flCreERT2/MLL::AF9 transplant recipients seemed to be cured, because their blood and bone marrow profiles had returned to normal by the time the experiment was terminated (195 days). The greater efficacy of Mnt deletion in vivo when compared with in vitro may be a consequence of the tumor environment in vivo.

To extend our studies to human AML, we used doxycycline-inducible CRISPR/Cas9^39,40 to delete MNT from 4 long-established human AML CLs, namely MV4;11, which harbors an MLL::AF4 fusion gene; MOLM-13 and THP-1, which have MLL::AF9 fusion genes; and OCI-AML3, which has the NPM1 and DNMT3A mutations that are common in adult AML.⁵⁹ MNT reduction in MV4;11 and, to a lesser extent, in THP-1 and OCI-AML3 rendered these CLs more sensitive to killing by the MCL-1 inhibitor S63845 and, in the case of MV4;11, also by the BCL-2 inhibitor ABT-199/venetoclax. Importantly, we also showed that inducing MNT deletion in MV4;11 cells transplanted into immunodeficient NSG mice significantly enhanced the survival of transplant recipients.

In conclusion, this study has established, for the first time to our knowledge, that primary murine MLL::AF9 AMLs and long-established human AML CLs are MNT dependent. These findings extend those we have made previously about the MNT dependency of MYC-driven B and T lymphomas, which we have shown was a consequence of MNT-mediated suppression of MYC-driven apoptosis, at least in part, via reduced expression of proapoptotic BIM.^25,26 

The deregulated overexpression of MYC is critical for driving many human cancers^60,61. However, to date, MYC has proven to be an intractable target for the development of clinically effective inhibitors.⁶² Our findings present an entirely new avenue to explore. Rather than inhibiting MYC, we propose taking advantage of MYC’s innate capacity to induce apoptosis and to amplify that drive by inhibiting MNT (or its regulators). To progress this possibility, it will be important to investigate which other tumor types are MNT dependent.

Possible approaches to MNT inhibition include inhibiting its heterodimerization to MAX; inhibiting MNT/MAX DNA binding; inhibiting MNT from binding to its vital co-repressor SIN3; and a PROTAC (Proteolysis Targeting Chimaera) strategy for targeting the degradation of MNT protein. MNT inhibitors are likely to be more effective in combination with additional drugs, and our results suggest that they could be a useful adjunct to BH3 mimetic drug therapy.

The authors thank Peter Hurlin (Shriner’s Hospital for Children, Portland, OR) for his kind gift of floxed Mnt mice; Guillaume Lessene (Walter and Eliza Hall Institute [WEHI]) for providing A-1331852; David Huang for providing human acute myeloid leukemia cell lines; and our colleagues, Andreas Strasser and Hai Vu Nguyen, for useful discussions and comments on the manuscript. The authors are grateful to the WEHI Bioservices for help with the animal experiments and skillful mouse husbandry, in particular Michael Watters, Dan Fayle, Jaclyn Gilbert, and Giovanni Siciliano. The authors also thank Simon Monard and his team at the WEHI’s fluorescence-activated cell sorting facility.

This work was supported by grants and fellowships from the Worldwide Cancer Research (19-0232 [S.C.]), the Australian National Health and Medical Research Council (NHMRC; Program Grant 1016701 [S.C.]; Synergy Grant GNT2011139 [G.L.K.]; Investigator Grant APP2018071 [A.H.W.]; Investigator Grant APP1176175 [F.C.B.]), the Victorian Cancer Agency (ECRF Fellowship 21006 [S.T.D.]), Dyson Bequest (G.L.K.), other philanthropic support to WEHI, and operational infrastructure grants to WEHI through the Victorian State Government Operational Infrastructure Support and the Australian Government NHMRC Independent Research Institute Infrastructure Support Schemes.

Contribution: K.C.F. designed and performed most of the experiments with contributions from V.L. and S.T.D. and advice from S.C., G.L.K., and A.H.W.; V.L. and K.C.F. performed the xenograft studies of human acute myeloid leukemia (AML) MV4;11; F.C.B. and M.J. analyzed the BEAT AML data set; S.C. and K.C.F. wrote the manuscript with input from G.L.K. and A.H.W.; and S.C. conceived and led the project and secured funding.

Conflict-of-interest disclosure: Walter and Eliza Hall Institute and its employees benefit from milestone and royalty payments related to venetoclax (ABT-199). G.L.K. and A.H.W. report receiving research funding from Servier for work on the development of S63845. The remaining authors declare no competing financial interests.

Correspondence: Suzanne Cory, Walter and Eliza Hall Institute of Medical Research, 1G Royal Parade, Parkville, Melbourne, VIC 3052, Australia; email: cory@wehi.edu.au.