Acute myelogenous leukemia (AML) is an aggressive malignancy characterized by the accumulation of myeloid blasts in the bone marrow and blood as a result of multiple serially acquired genetic mutations. The most common recurrent somatic mutation in AML affects the gene NPM1 and occurs in 25-30% of patients. This mutation results in the aberrant cytoplasmic localization of the protein and is termed NPM1c (Falini et al. 2005, Cancer Genome Atlas Research Network 2013). NPM1c is considered to be a driver of AML, however mice carrying the most common variant, NPM1cA, only develop disease after a long latency (median 18 months), suggesting that other cooperating mutations are required for AML development (Vassiliou et al. 2011).

Heterozygous mutations in one of four members of the cohesin complex (RAD21, SMC3, STAG1, and SMC1A) are commonly found in patients with AML. Different cohesin gene mutations are mutually exclusive and are frequently found in combination with NPM1 mutations (Cancer Genome Atlas Research Network 2013). We and others have shown that cohesin mutations result in enhanced hematopoietic stem and progenitor cell (HSPC) self-renewal and that aberrant expression of HOXA7 and HOXA9 contribute to this phenotype (Mazumdar et al. 2015, Viny et al. 2015, Mullenders et al. 2015, Galeev et al. 2016, Fisher et al. 2017). To explore the relationship between cohesin haploinsufficiency and Npm1cA in the development of AML, we crossed the inducible Npm1cA-flox/+ and Smc3flox/+ mouse models (Vassiliou et al. 2011 and Viny et al. 2015). Our pilot studies indicate that approximately 75% of Npm1cA/+; Smc3Δ/- mice develop AML, with a shorter latency observed compared to Npm1cA/+ only mice (8 months vs. 18 months). Surprisingly, while single-mutant Npm1cA/+ and Smc3 Δ/- mice both show increased HSPC self-renewal and HoxA gene overexpression, minimal synergy was observed in Npm1cA/+; Smc3 Δ/- mice with respect HoxA7 and HoxA9 expression. In an effort to uncover changes in gene transcription that may contribute to the accelerated development of AML observed in double-mutant mice, we performed RNA sequencing on bone marrow from WT, Npm1cA/+, Smc3 Δ/-, and Npm1cA/+; Smc3 Δ/- mice (three from each genotype). Interestingly, additive changes in gene expression were not seen in the Npm1cA/+; Smc3 Δ/- mice. Instead, upregulation of a set of unique genes was observed. Such genes include factors known to play a role in hematopoiesis or myeloid cell differentiation, including Fli1 and Csf3r. We hypothesize that Smc3 haploinsufficiency alters gene expression in Npm1cA/+ mice, resulting in the upregulation of a unique set of genes that contribute to enhanced AML development. Our work provides important insights into how combined mutations can promote AML development by altering the transcriptome. In addition, it provides a tractable model system to explore the mechanism by which NPM1 mutations cause leukemia.


Levine: Roche: Research Funding; Celgene: Research Funding; Qiagen: Equity Ownership; Qiagen: Equity Ownership; Celgene: Research Funding; Roche: Research Funding.

Author notes


Asterisk with author names denotes non-ASH members.