Allelic Complexity of KMT2A Partial Tandem Duplications in Acute Myeloid Leukemia and Myelodysplastic Syndromes

KMT2A partial tandem duplication (KMT2A-PTD) at 11q23.3 is associated with adverse risk in AML and MDS, is a potential therapeutic target, and is an attractive marker of measurable residual disease. High initial KMT2A-PTD RNA levels have been linked to poor prognosis, but mechanisms regulating KMT2A-PTD expression are not well understood. While it has been reported that KMT2A-PTD affects only a single allele, it has been theorized but not proven that duplications or genomic gains of a monoallelic KMT2A-PTD may occur, thereby potentially driving high expression and disease progression. Copy neutral loss of heterozygosity (CN-LOH) of 11q has also been described and is known to be associated with mutations in CBL but has not been reported to involve KMT2A-PTD. In this study, we identified 94 patients with KMT2A-PTDs using targeted DNA next-generation sequencing (NGS) and found that 16% (15/94) had complex secondary events, including CN-LOH and selective gain involving the KMT2A-PTD allele. High copy numbers indicating complexity were significantly enriched in AML versus MDS and correlated with higher RNA expression. Moreover, in serial samples, complexity was associated with relapse and secondary transformation. Taken together, we provide approaches to integrate quantitative and allelic assessment of KMT2A-PTDs into targeted DNA NGS and demonstrate that secondary genetic events occur in KMT2A-PTD by multiple mechanisms that may be linked to myeloid disease progression by driving increased expression from the affected allele.

Targeting BRD4 in acute myeloid leukemia with partial tandem duplication of the MLL gene

Not available.

Allogenic Stem Cell Transplantation Abrogates Negative Impact on Outcome of AML Patients with KMT2A Partial Tandem Duplication

Recently, a new subset of acute myeloid leukemia (AML) presenting a direct partial tandem duplication (PTD) of the KMT2A gene was described. The consequences of this alteration in terms of outcome and response to treatment remain unclear. We analyzed retrospectively a cohort of KMT2A-PTD-mutated patients with newly diagnosed AML. With a median follow-up of 3.6 years, the median overall survival was 12.1 months. KMT2A-PTD-mutated patients were highly enriched in mutations affecting epigenetic actors and the RTK/RAS signaling pathway. Integrating KMT2A-PTD in ELN classification abrogates its predictive value on survival suggesting that this mutation may overcome other genomic marker effects. In patients receiving intensive chemotherapy, hematopoietic stem cell transplantation (HSCT) significantly improved the outcome compared to non-transplanted patients. In the multivariate analysis, only HSCT at any time in complete remission (HR = 2.35; p = 0.034) and FLT3-ITD status (HR = 0.29; p = 0.014) were independent variables associated with overall survival, whereas age was not. In conclusion, our results emphasize that KMT2A-PTD should be considered as a potential adverse prognostic factor. However, as KMT2A-PTD-mutated patients are usually considered an intermediate risk group, upfront HSCT should be considered in first CR due to the high relapse rate observed in this subset of patients.

Mll-Partial Tandem Duplication (Mll-PTD) Causes Pseudohypoxia and Hypermethylation of the Epigenome through Suppression of Mitochondrial Complex II

The MLL-partial tandem duplication (MLL-PTD),characterized by the internal duplication of exons 3-9 or 3-11 in the MLL gene, produces an elongated protein, and is considered as a gain-of-function mutation. The MLL-PTD is primarily found in elderly patients with myelodysplastic syndromes and acute myeloid leukemiaas well as healthy individuals.Previously we showed that Mll-PTD knock-in (MllPTD/WT) mice presented enhanced self-renewal of hematopoietic stem cells (HSCs) and partially blocked differentiation of hematopoietic stem/progenitor cells (HSPCs). Interestingly, Mll-PTD increased the protein level of HIF1A in HSPCs, which is critical for enhanced self-renewal of HSCs. In the current study, we investigated the mechanisms for HIF1A activation by Mll-PTD. In normoxia, HIF1A is hydroxylated by prolyl hydroxylases (PHD), resulting in rapid protein degradation via ubiquitination. PHD is one of the well-known enzymes whose activity is dependent on the cellular level of α-ketoglutarate (α-KG), one of the metabolites in the tricarboxylic acid (TCA) cycle. Accumulation of subsequent metabolites of α-KG, such as succinate, fumarate, and malate, inhibits activity of α-KG-dependent enzymes. Indeed, mitochondrial dysfunction is known to result in accumulation of TCA cycle intermediates, leading to activation of HIF signaling. Thus, we first examined if Mll-PTD induces the alteration of mitochondrial functions. Interestingly, cellular respiration and activity of mitochondrial complexes (I, II, and III) were significantly decreased in HSPCs of MllPTD/WT mice, while the copy number of mitochondrial DNA was not altered. These results indicate that suppression of mitochondrial activity is not due to the decrease of the total mitochondria. We also examined mRNA expression levels of several major TCA cycle enzymes, and found that succinate dehydrogenase (Sdh) complex (Sdha, Sdhb, and Sdhd) was significantly downregulated in MllPTD/WT HSPCs. SDH is a critical TCA cycle enzyme which converts succinate to fumarate. Inactivation of SDH is known to result in impairment of mitochondrial biogenesis, a blockade of the TCA cycle, and accumulation of TCA cycle metabolites. We next quantified metabolites in glycolysis and TCA cycle in the plasma from WT control and MllPTD/WT mice. NMR analysis revealed that succinate, fumarate, and malate were increased in the plasma of MllPTD/WT mice. Especially, the ratios of fumarate and malate to α-KG were both significantly increased in MllPTD/WT compared to WT control. Indeed, post-α-KG metabolites increased HIF1A protein in human cord blood CD34+cells in vitro, indicating that higher levels of succinate, fumarate, and malate to α-KG levels stabilize HIF1A. We also confirmed that knockdown of Sdh increased the HIF1A protein level in murine cell line in normoxia. These results indicate that downregulation of Sdh in MllPTD/WT is one of the mechanisms for suppression of mitochondrial activity, leading to pseudohypoxia and HIF1A activation. Besides PHD, TET and histone lysine demethylases are also α-KG-dependent enzymes. We found that in MllPTD/WT HSPCs, the 5-methylcitosine (5-mC) level was increased in genomic DNA, and trimethylation levels at H3K4, H3K9, H3K36 and H3K79 were also increased. Collectively, these results suggest that metabolic pseudohypoxia due to lower mitochondrial activity not only activates HIF1A signaling but also induces hypermethylation in DNA and histones, through suppression of α-KG-dependent PHD and demethylases. In summary, we demonstrate that through suppression of mitochondrial complex II, Mll-PTD causes pseudohypoxia and hypermethylation of the epigenome, which may contribute to expansion of premalignant clones and accumulation of additional mutations in those cells. Interestingly, it has been proposed that IDH mutations are involved in tumorigenesis in leukemias and brain tumors through a similar mechanism. Moreover, loss-of-function mutations of the TCA cycle enzymes, SDH complex, and fumarate hydratase, are frequently found in various solid tumors associated with pseudohypoxia and hypermethylation phenotypes. Further investigations of the impact of metabolic-rewiring-mediated pseudohypoxia/hypermethylation on tumorigenesis may lead to the development of novel therapeutic strategies to prevent the onset and/or the progression of various types of malignant diseases. Disclosures No relevant conflicts of interest to declare.