NG2 at the core: tackling MLLr B-ALL resistance

In this issue of Blood, Lopez-Millan et al¹ identified a novel pathway involved in the poor response to glucocorticoids in MLL (KMT2A) rearranged B-precursor acute lymphoblastic leukemia (MLLr B-ALL). The authors previously reported that neuron-glial antigen-2 (NG2) is highly expressed in MLLr B-ALL and associated with a poor prognosis. Extending this work, they found that NG2 is an epigenetically regulated direct target gene of the leukemic MLL-AF4 fusion protein. The upregulated NG2 protein interacts with FLT3 on the cell surface, triggering a ligand-independent FLT3 signaling cascade and inducing AP1-mediated inhibition of the glucocorticoid receptor (GR), which ultimately causes glucocorticoid resistance (see figure).

Glucocorticoids like prednisone and dexamethasone are among the most effective classes of drugs against acute lymphoblastic leukemia (ALL).² However, glucocorticoid resistance is a major barrier to cure. These drugs function by binding to GR in the cytoplasm, which then translocates to the nucleus to act as a transcription factor (TF). Whole-genome sequencing has identified many genetic and epigenetic mutations in glucocorticoid-resistant ALL, suggesting that multiple mechanisms may be involved in this resistance. Mutations in genes such as BTG1 or PTEN can inhibit GR expression or nuclear translocation. Mutations in epigenetic regulators like KMT2D, CREBBP, HDAC7, and NSD2 lead to epigenetic aberration of glucocorticoid-targeting genes. TFs such as AP-1 or NF-ĸB can directly interact with the GR to inhibit its activity, thus conferring glucocorticoid resistance. Moreover, the epigenetic landscape, particularly DNA methylation or chromatin accessibility, plays a crucial role in differential glucocorticoid responses among various cell types, potentially interrupting GR binding sites and impacting resistance.³ 

MLLr B-ALL, common in infants, is associated with poor prognosis due to treatment resistance. NG2, a membrane chondroitin sulfate proteoglycan, is aberrantly expressed in various malignancies and associated with chemoresistance. NG2’s expression on tumor cells and limited presence in healthy tissues make it a promising target for cancer therapy, including chimeric antigen receptor T-cell immunotherapy. Previous reports by the authors indicated that inhibiting NG2 increases MLLr B-ALL cells’ sensitivity to chemotherapeutic agents in patient-derived xenograft models.⁴ However, the molecular mechanisms of NG2-mediated chemoresistance in MLLr B-ALL remained unclear. In this article, the authors identified novel mechanisms of glucocorticoid resistance in MLLr B-ALL, involving NG2 gene epigenetics and NG2-mediated activation of the FLT3/AP-1 pathway. The findings presented in this article are compelling for 2 reasons.

First, the authors identified MLL-specific hypomethylated cis-regulatory elements of NG2 in MLLr B-ALL. The authors performed whole-genome DNA methylation analysis, which revealed 5 differentially methylated regions near the NG2 gene in MLLr B-ALL. An upstream hypomethylated enhancer was identified and verified using a dCas9-DNMT3A-based epigenetic editing system, which precisely methylated the enhancer, resulting in NG2 reduction. Low DNA methylation at the NG2 locus was linked to enriched intragenic MLL fusion protein binding in MLLr B-ALL, as shown by chromatin immunoprecipitation followed by sequencing (ChIP-seq) studies. Thus, although it remains to be proven whether MLL directly regulates DNA methylation at the NG2 locus, the study suggests that the MLL-specific methylation landscape (cis-regulation) and MLL fusion protein binding (trans-regulation) are crucial for NG2 gene upregulation in MLLr B-ALL. Abnormal epigenetic DNA methylation is common in cancer, but the mechanisms are diverse and complex. The MLL protein is a histone methyltransferase, with rearrangements occurring early in hematopoiesis. MLLr B-ALL often exhibits hybrid myeloid-lymphoid features and can relapse as acute myeloid leukemia.⁵ These phenotypes suggest that the epigenetic landscape of MLLr B-ALL differs from other B-ALL subtypes. Glucocorticoid-based therapies rarely succeed in treating myeloid malignancies, and the molecular basis for this remains elusive. Therefore, MLL-specific activation of NG2 identified in this article aligns with existing understandings of MLLr B-ALL’s evolution and may offer therapeutic utility based on MLL-specific epigenetic landscapes.

Second, the authors found a novel interaction between NG2 and FLT3 on the cell surface accounting for glucocorticoid resistance in MLLr B-ALL. The authors performed NG2 immunoprecipitation and mass spectrometry, identifying proteins interacting with NG2. Using western blot, in silico modeling, and single-cell analysis, they discovered a novel interaction between NG2 and FLT3 in MLLr B-ALL. FLT3 signaling is vital in glucocorticoid actions, with its downstream target AP-1 known to act as a negative cofactor of GR, leading to glucocorticoid resistance.^6,NG2 knockout experiments and fluorescence-activated cell sorting of primary leukemia cells from patients with MLLr B-ALL demonstrated reduced phosphorylated FLT3 in the NG2-negative population, indicating an upstream mechanism of GR inhibition. The study also showed dysregulation of GR-activated downstream pathways, including B-cell receptor/phosphatidylinositol 3-kinase/AKT/mTOR pathways⁷ and BCL2/BIM-involved apoptotic pathways.⁸ Therefore, FLT3/AP-1-mediated downregulation of GR and inhibition of its downstream pathways contribute to glucocorticoid resistance in MLLr B-ALL. Above all, NG2 plays a pivotal role in this resistance, with MLL-specific activation of NG2 and its direct interaction with FLT3 triggering this complex system.

Various approaches have been explored to overcome glucocorticoid resistance in lymphoblastic malignancies, including targeting FLT3, PI3K/AKT, and mTOR pathways. Compounds targeting MLL-mediated pathways, such as DOT1L and Menin-MLL inhibitors, have been developed for MLLr leukemia.^9,10 Based on this study’s findings, the authors propose using FLT3 inhibitors to treat MLLr B-ALL to restore its glucocorticoid sensitivity. Overall, this study provides a novel molecular basis for using FLT3 inhibitors, and the NG2-FLT3 interaction site may offer a new target for drug development to improve the efficacy of FLT3 inhibitors in treating MLLr B-ALL.

In conclusion, this study provides insight into the MLL-specific methylation landscape and its associated NG2 upregulation in MLLr B-ALL, revealing the critical role of the NG2-FLT3 interaction in glucocorticoid resistance. Notably, the study raises further questions about MLL fusion protein’s role in modulating DNA methylation and chromatin conformation at the NG2 locus. Additionally, the precise protein structure of the FLT3-NG2 binding has yet to be elucidated. Nevertheless, restoring glucocorticoid sensitivity is essential for improving clinical outcomes in high-risk ALL. The compelling results from this study warrant further research to further define the NG2-centered regulatory network and develop novel treatment strategies targeting NG2 and its cofactors.

Conflict-of-interest disclosure: D.J. declares no competing financial interests.