Mutations at exon-12 of nucleophosmin (NPM) gene are the most common gene lesion in Acute Myeloid Leukemia (AML) (Falini et al., NEJM, 352:254, 2005). They result in

  1. disruption of tryptophan (W) 288 and/or 290, constituting the Nucleolar Localization Signal (NoLS), and

  2. creation of a new C-terminal Nuclear Export Signal (NES) motif, with 6 variations observed to date.

Both alterations are necessary for aberrant NPM accumulation in leukemic cell cytoplasm (NPMc+ AML,

Falini et al., Blood, 107:4514, 2006
).

In NPM mutants, there is a correlation between NES sequence and the type of NoLS disruption: the most common NES motif (LxxxVxxVxL) always associates with loss of both W288 and W290, e.g. mutant A; when retained, W288 always associates with rare NES variant sequences, such as LxxxLxxVxL in NPM mutant E. We demonstrated this correlation is explained by efficiency differences between NES motifs: NPM mutant A is not efficiently exported in cytoplasm upon artificial W288 reinsertion; the same happens when LxxxLxxVxL NES motif in mutant E is artificially replaced with LxxxVxxVxL, typical of mutant A. This suggest LxxxVxxVxL NES motif is weaker than the others, being unable to efficiently export mutant NPM to cytoplasm when artificially coupled with W288 (repesenting a partial NoLS).

Such changes at NPM C-terminus also influence NPM-Arf binding and subcellular distribution. Tumor suppressor Arf is a major NPM interactor in the nucleolus, and its deregulation would be strongly implicated in leukemogenesis. In co-transfected cells, NPM leukemic mutants partially relocates Arf from nucleolus to nucleoplasm and cytoplasm and, in turn, Arf partially relocates NPM mutants A and E from cytoplasm to nucleus in a dose-related manner. Artificial NPM mutants carrying W288 coupled with LxxxVxxVxL NES motif show higher degree of co-localization with Arf, and lower doses of Arf are required for them to be completely relocated to nucleoli as compared to NPM leukemic mutants A and E. Co-immunoprecipitation studies show that leukemic mutants A and E bind less Arf than wild-type NPM or artificial NPM mutants. This obervation suggest that NPM leukemic mutants and Arf reciprocally interact and influence each otheri n a dose-related manner, with C-terminal NPM mutated sequence playing a role in Arf binding.

Cytofluorimetric analysis of Arf-transfected NIH-3T3 cells shows G1 cell cycle arrest. Upon co-transfection, neither wild-type NPM nor leukemic or artificial NPM mutants are able to relieve Arf-induced cell cycle arrest.

In conclusion, we suggest there is a “determinism” in the combination of NES motif/NoLS disruption: NPM mutants are selected for efficient cytoplasmic export and reduced Arf binding as compared to wild-type NPM. Both mechanisms may contribute to Arf dislocation/degradation, thus having the same functional consequences as NPM silencing; moreover, it is tempting to speculate that NPM leukemic mutants could be themselves less sensitive to ARF induced inhibition. Nevertheless, in our experimental model, changes in Arf subcellular distribution didn’t result in functional alterations. To clarify wether these findings will be relevant to the pathogenesis of NPMc+ AML, future studies should address in deeper detail the relationship between NPM and Arf doses and function in specific subcellular compartments.

Disclosures: This study has been supported by AIRC and FIRC.; B. Falini has applied for a patent on clinical use of NPM mutants.

Author notes

*

Corresponding author

Sign in via your Institution