Nucleophosmin 1 (NPM-1) is a highly conserved, ubiquitously expressed nucleolar protein that functions as a molecular chaperone shuttling protein-binding partners between the nucleolus, nucleus and cytoplasm. NPM-1 and has been assigned more than a dozen functions in the cell, including ribosome biogenesis and centrosome duplication. The NPM-1 gene maps to chromosome 5q35, a region that is the target of deletions in both de novo and therapy-associated MDS in humans. Additionally, heterozygous mutations in the NPM-1 gene have been identified in 60% of cytogenetically normal adult AMLs. Mutant NPM-1, (NPMc+), commonly results in the generation of a novel nuclear export signal (NES) leading to cellular mislocalization of NPMc+ from the nucleolus and nucleus to the cytoplasm. The role of NPMc+ in contributing to AML however remains unresolved to date. Two hypotheses to explain the role of NPMc+ in leukemogenesis have been advanced. The first purports that aberrant cytoplasmic mislocalization of NPMc+ also mislocalizes a number of NPM1-cargo proteins into the cytoplasm including for example, the tumor suppressor Arf, leading to the activation of the c-MYC oncogene, thereby contributing to leukeomogenesis. A second hypothesis states that reduction in wild type levels of NPM-1 in the nucleolus as a result of both heterozygosity as well as mislocalization into the cytoplasm following association with NPMc+, contributes to tumorigenesis. To address these hypotheses we generated a series of IL-3-dependent cell lines from the bone marrow of NPM1+/+ and NPM+/-mice. In addition, we stably transduced an NPMc+ expression vector into the NPM+/- cells resulting in the NPM+/-c+ line, thereby providing cellular models for both NPM1 haploinsufficieny as well as mutant NPM-1 associated AML. We then sought to examine the role of the master myeloid transcription factor C/EBPα in contributing to NPM-1-associated AML. C/EBPα is a single exon, bzip transcription factor that generates four isoforms derived from separate in-frame AUGs resulting in the translation of a nucleolar p50, a full length p42, a p40 and a dominant negative p30 isoform. We found that in NPM+/-c+ and OCI-AML3 cells (derived from a CN-AML patient harboring the NPMc+ mutation), only the p40 isoform of C/EBPα migrated to the cytoplasm while the full length p42 isofom remained in the nucleus. The p40 isoform of C/EBPα lacks the first 14 N-terminal amino acids when compared to the full length p42 isoform, and its function has not been well described in the literature. However, we have demonstrated that like the full length p42 isoform, the p40 isoform is also capable of transcriptional activation as measured by a reporter gene assay. Hence mislocalization of p40 could affect the overall activity of C/EBPα in NPMc+ expressing cells. We next sought to determine the contribution of NPM-1 haploinsufficieny to C/EBPα activity using our NPM+/- cells. We showed that a) expression levels of the dominant negative p30 isoform of C/EBPα are elevated in NPM+/- cells thereby blocking the activity of wildtype p42 b) C/EBPα p42 protein in the nucleus is rendered inactive due to phosphorylation by an unknown kinase at the S21position and c) expression of the downstream targets of C/EBPα (lactoferrin, MMP8, MMP9, gp91phox) are blocked in NPM+/- cells. These observations together suggest that in CN-AML where one copy of NPM-1 is wildtype and one copy is mutant (NPMc+), both haploinsufficient levels of wild type NPM-1 as well as mutant NPMc+ contribute to aberrant myeloid differentiation by inactivating C/EBPα activity. Our data thus provide new insights into the molecular mechanisms of mutant NPM-1 associated AML.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.

Sign in via your Institution