Despite a steady increase in the number of targeted therapies over the past decades, the 5-year overall survival (OS) rate remains poor in acute myeloid leukemia (AML). Currently, high expression of the oncogenic Meningioma 1 (MN1) is associated with poor OS in AMLs with normal karyotype, and we correlated its expression with a more immature cellular composition in AML samples. Hence, we hypothesized that supplementation of AML cells that usually fail to engraft in xenograft models, such as acute promyelocytic leukemia (APL), with MN1 would promote in vivo expansion in mice allowing the study of potential therapeutic vulnerabilities.

The in vitro experiments were performed in human AML cell lines, primary AML cells (n=39), and isolated healthy CD34+ cells. MN1-overexpression (OE) was lentivirally performed. Bioenergetic profiles were evaluated by a Seahorse analyzer. Gene expression was measured by RNA-sequencing/qPCR. Protein levels were measured by immunofluorescence microscopy. Viability and surface protein expression was evaluated by flow cytometry. For in vivo studies, eight-weeks old female NSGS mice were transplanted with MN1-OE primary APL blast cells, and mice were followed for 12 weeks. Cells from the mouse organs were isolated, analyzed, and sorted for human APL blasts (CD45+CD117+CD33+) for ex vivo studies.

We show that genetic MN1-OE in APL cells resulted in increased cell survival upon cytotoxic therapy, and decreased sensitivity to ATRA-induced terminal differentiation in short and long term cultures (3/8 days). In healthy CD34+ cells, we observed that MN1-OE was able to increase cell proliferation and stemness properties in CMP-sorted (CD34+CD38+CD123+CD45RA-) but not in GMP-sorted cells (CD34+CD38+CD123+CD45RA+), showing in human models the results from Heuser et al., 2011 in murine MN1-model. Co-expression of MN1+PML-RARα in healthy CD34+ cells resulted in superior expansion with retention of the CD34 marker, in comparison with the single hits. At the molecular level, APL cells with MN1high displayed transcriptional programs associated with "HYPOXIA"/"HSC-UP", whereas MN1low levels correlated with "OXPHOS-UP"/"GMP-UP". Despite no evident differences in Seahorse analysis, MN1-OE cells exhibited increased levels of the glycolytic enzyme PDK1. Functionally, MN1-OE cells were more sensitive to glycolytic inhibitors 2,2-dichloroacetophenone (a PDK1 inhibitor) and 2-deoxyglucose, revealing glycolysis as a metabolic vulnerability in MN1high AMLs, which was further validated in primary AML samples. These data are in line with our previous report indicating that more immature AML subtypes typically display a more glycolytic metabolic profile (Erdem et al, 2022). MN1 is an intrinsically disordered protein containing a long polyQ-stretch that is required for transformation (Riedel et al, 2021). We questioned whether the expression of chaperone proteins would be necessary for the appropriate functioning of MN1. We observed that expression of the DNAJB6 isoform 2 (DNAJB6b) correlated with MN1 expression across a large panel of genetic subtypes of AML, including more immature AML subtypes, such as the TP53-mutant AML but also in APL. shRNA-induced knockdown in AML cells of DNAJB6b resulted in cytosolic retention of MN1, associated with reduced cell survival. In vivo, the patient-derived xenograft model showed engraftment of APL samples upon MN1-OE, with the colonization of secondary organs and splenomegaly, which was not observed in the control. Flow cytometry analysis revealed that MN1-engrafted mice displayed increased levels of APL blasts in comparison with control, which had a large proportion of mature myeloid cells (CD33+CD11b+). Ex vivo screen of engrafted cells revealed increased resistance to ATRA-induced differentiation in MN1-OE sorted blasts. We provide further evidence to support the role of MN1 as an oncogenic hit allowing in vivo studies of difficult-to-engraft AML subtypes. Functionally, high MN1 expression resulted in decreased differentiation, retention of stem-like properties, and metabolic shift towards a more glycolytic state. We demonstrated that DNAJB6b controls the cellular localization of MN1, with its downregulation associated with the loss of MN1 supportive functions. Finally, our findings suggest that targeting glycolysis can be leveraged as a therapeutically actionable mechanism for driving cell death in MN1high AMLs.

Traina:Novartis Pharmaceuticals Corporation: Consultancy. Quek:Servier: Research Funding; Bristol Myers Squibb: Research Funding; Jazz Pharmaceuticals: Speakers Bureau.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution