MUSASHI-2 duels in myelodysplastic syndromes

Myelodysplastic syndromes (MDS) are a heterogeneous set of myeloid disorders typically characterized by cytopenias in several lineages within the hematopoietic hierarchy.¹ They are clonal disorders, driven by alterations that originate from a dysregulated hematopoietic stem and progenitor cell (HSPC) compartment. Large genomic studies have uncovered numerous mutations in MDS, of which SF3B1, SRSF2, TET2, and ASXL1 have been identified as the most common. Conventional chemotherapeutic agents for high-risk MDS, including decatabine or 5-azacytidine, are broadly cytotoxic and do not specifically target the MDS HSPCs. Moreover, it is neither clear how MDS stem cells can persist and mediate disease, nor what mechanisms drive transformation to an acute myelogenous leukemia (AML).

Recent studies have focused on an RNA binding protein called MUSASHI-2 (MSI2), which controls the self-renewal program in HSPCs and is implicated in the most aggressive leukemias. Based on this work, we posited that MSI2 might play an important role in HSPC dysfunction in MDS. We first demonstrated that MSI2 expression correlates with high-risk MDS and a poor clinical outcome (Fig. 1).² We validated these findings after measuring MSI2 intracellular abundance in low-risk and high-risk MDS patients. To determine if Msi2 plays a functional role in MDS, we utilized transgenic mice that express the fusion NUP98-HOXD13 (NHD13). Despites its rarity in MDS, NHD13 drives phenotypes that recapitulate many MDS features, including transformation to AML.³ We crossed NHD13 mice with Msi2 conditional knockout mice and found that, after deletion, there was a rapid loss of donor chimerism in transplanted animals and a reversal of MDS. Conversely, bone marrow derived from an NHD13 tetracycline inducible MSI2 overexpression model was transplanted, and after 7 months the NHD13/MSI2 mice developed lethal aggressive myeloid diseases, which included MPN/MDS or AML/MDS. These diseases appeared addicted to active MSI2 induction, as transplanted disease could be cleared if mice were withdrawn from doxycycline-mediated MSI2 expression (Fig. 1).

Figure 1.

MSI2 drives more aggressive myelodysplastic syndromes. Patients were more likely to have increased MSI2 expression and poor survival in high risk vs. low risk MDS. Overexpression of MSI2 results in activated hematopoietic stem and progenitor cells (HSPC). If forced MSI2 expression is removed, the MDS or myeloid disease becomes more differentiated.

The HSPC compartment was expanded after a 5-day induction of MSI2, and this was maintained in the diseased mice. NHD13/MSI2 overexpressing mice had reduced apoptosis and increased G1 positive cells compared to the NHD13 HSPCs. Similarly, gene expression profiling of NHD13/MSI2 HSPCs before disease progression displayed a more activated hematopoietic stem cell program. Our results suggest that MSI2 plays a role in activating and maintaining the diseased HSPC and may potentiate myelodysplastic syndrome toward transformation.

Studies have shown that MDS is driven by a rare and distinct stem cell. However, the LSC in AML has been thought to be downstream of the HSC.⁴ Our data suggest that MSI2 overexpression expands and maintains a phenotypic HSPC. Even though MSI2 expression is increased in high-risk MDS compared to low-risk MDS, it is not clear if MSI2 levels increase before or during disease progression. Serial assessment of MDS patients would help to determine if MSI2 expression could be used as a predictive marker for AML progression. Additionally, it is not clear how MSI2 is epigenetically regulated in the MDS HSPC. HOX genes are commonly activated in leukemia and have been implicated in the control of MSI2 expression. Potentially mirroring what happens in patients, MSI2 levels increase when some of the NHD13 mice transform to AML. This may suggest that another pathway needs to be altered to drive high levels of MSI2 expression in the NHD13 model.

Relatedly, we found that MSI2 can also mediate the translation of HOXA9 and control the MLL epigenetic program, suggesting that this pathway is maintaining a positive feedback loop in AML.⁵ MSI2 also controls Tetraspanin-3, which was found to alter chemokine response and interactions within the myeloid leukemic microenvironment.⁶ In normal HSCs, MSI2 modulates the response to TGFβ signaling and can control translation of Tgfbr1 and Smad3.⁷ It remains to be seen whether MSI2 modulates MDS transformation through regulation of these pathways in the stem cell compartment or through a novel molecular mechanism, and if they are at play in patients with MDS. Nevertheless, this study and others suggest that MSI2 might be an attractive therapeutic target in MDS and in leukemia. One strategy is to identify patients based on MSI2 expression and determine if this can be used for predicting the therapeutic response to emerging therapies. The development of MSI2 specific inhibitors should also be pursued as a more direct approach to tackle the dysregulated stem and progenitor compartment in MDS.

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.