Nucleophosmin (NPM1) mutations occur in 50–60% of adult acute myeloid leukemia (AML) with normal karyotype.1, 2 About 40 NPM1 mutations2 have been so far identified, all clustering in exon-12. In spite of molecular heterogeneity, all mutations cause common changes at the C terminus of NPM mutants, i.e. loss of tryptophans 288 and 290 (or 290 alone) and creation of a new nuclear export signal (NES) motif.2 As a consequence, NPM mutants aberrantly accumulates in the cytoplasm of leukaemic cells;3, 4 hence, the term NPMc+ (cytoplasmic-positive) AML.1, 2 Here, we report on the identification and functional characterization of a cytoplasmic nucleophosmin mutant generated by a novel exon-11 NPM1 mutation in a patient with AML.
Samples from 98 AML patients at diagnosis (Supplementary Table) were analyzed for NPM1 mutations at exons 11 and 12. Denaturing high-pressure liquid chromatography (DHPLC) screening (Supplementary Material 1) identified in one 73-year-old male patient a new sequence variant (named Vi2), with eight nucleotides inserted at position 902 in the middle of exon-11 (Figure 1a and b). Variant Vi2 was confirmed by allele-specific oligonucleotide polymerase chain reaction (ASO-PCR) (Supplementary Material 2). Using the normal primer for amplification, definite bands of 232 bp were detected on agarose gel for the patient and healthy controls; no amplicons were observed in cDNA from normal controls when amplified with allele-specific oligonucleotide (Figure 1c). Nucleotide insertion led to a stop codon at amino acid 275 (Met274Stop). The predicted truncated protein consisted of 274 amino acid residues compared with 294 in wild-type NPM (NPMwt) (Figure 1a).
Aberrant cytoplasmic expression is a distinguishing feature of NPM leukaemic mutants generated by exon-12 NPM1 mutations.1, 2, 3, 4 To assess whether the exon-11 NPMVi2 mutant is also cytoplasmic, we used pEGFP-C1-NPMwt1 as template to generate the pEGFP-C1-NPMmVi2 construct in which the exon-11 NPM-Vi2 mutation is expressed in frame with enhanced green fluorescent protein (eGFP). Plasmids were generated using the QuikChange MultiSite-Directed Mutagenesis Kit (Stratagene, La Jolla, CA, USA), with primers designed on the following sequence: pEGFP-C1-NPMmVi2, 5'-GTGGAAGCCAAATTCAGGCGCCTATCAATTATGTGAAG. Confocal microscope analysis of NIH-3T3 cells transfected with the pEGFP-C1-NPMmVi2 construct (Supplementary Materials 3 and 4) showed that the NPMVi2 mutant localizes exclusively in the cytoplasm (Figure 2, top left).
Cytoplasmic accumulation of exon-12 NPM mutants is dictated by the concerted action of C-terminal mutated tryptophans and the creation of a NES motif.3 Notably, both alterations were also present in the NPMVi2 mutant. In fact, the NPMVi2 mutation resulted in loss of the whole exon-12 of the NPM1 gene (encoding for the C-terminal aromatic region and the tryptophan residues 288 and 290 responsible for nucleolar binding). Moreover, the NESbase version 1.0 program5 identified a putative NES motif with the VxxxFxxLxI sequence (Figure 1a) at the C terminus of the predicted NPMVi2 mutant protein. Therefore, the NPMVi2 mutant has not only lost the ability to target nucleoli, but also gained enforced nuclear export capabilities. Accordingly, the images of pEGFP-NPMVi2-transfected cells are the same as those of NPM mutant A (Figure 2, middle left) with which mutant NPMVi2 shares common features. In contrast, they differ from the less frequent mutants retaining tryptophan 288 (e.g. NPM mutant E), which are capable of partial nucleolar localization (Figure 2, bottom left).
To prove that the enforced nuclear export capabilities of NPMVi2 mutant are NES-dependent, we transfected NIH-3T3 cells with pEGFP-NPMVi2, incubated them with the Crm1 inhibitor leptomycin B and found that 100% of transfected cells showed exclusive nucleoplasmic localization (Figure 2, top right). This pattern is similar to that observed with NPM mutant A (Figure 2, middle right), but differ from the leptomycin B-induced nucleoplasmic and nucleolar relocation of NPM mutant E (Figure 2, bottom right). These findings prove that export of NPMVi2 mutant is NES-dependent and that the mutant cannot bind to nucleoli, despite artificially raised levels in nucleoplasm.
Since the NPMVi2 mutant harbours two physiological N-terminal NES sequences (residues 42–49 and 94–102, respectively2), the specific role of the new NES motif at the C terminus of the mutant cannot be established conclusively in Leptomycin B-based experiments. Therefore, using pEGFP-C1-NPMwt1 as template, we generated a mutant carrying an altered VxxxFxxLxI C-terminal NES sequence of NPMVi2 (Figure 3). In this construct (pEGFP-C1-NPMmVi2-no-NES), the C-terminal putative NES motif of NPMVi2 mutant (VxxxFxxLxI) was disrupted by substitution of phenylalanine (F) and leucine (L) with two guanine (G) residues (F → G and L → G). Plasmids were generated with primers designed on the following sequence: pEGFP-C1-NPMmVi2-no-NES, 5'-GTGGAAGCCAAAGGCAGGCGCGGATCAATTATGTGAAG.
Cells transfected with this eGFP-tagged plasmid showed exclusive nucleoplasmic positivity for the recombinant protein (Figure 3, right), demonstrating that VxxxFxxLxI is a functional C-terminal NES and that NPMVi2 mutant export abilities depend upon it.
Cell transfection experiments show that NPMVi2 localizes in cytoplasm, but neither provide information as to whether this also occurs in the patient's AML cells, nor provide information on which haemopoietic cell lineages are involved by the mutation. To address these issues, we immunostained paraffin sections from the patient's bone marrow biopsy using monoclonal antibodies that recognize fixative-resistant epitopes of NPM1 (Supplementary Material 5). The bone marrow was hypercellular and infiltrated by myeloid blasts (positive for myeloperoxidase and CD68 macrophage-restricted), atypical megakaryocytes and erythroid precursors (Figure 4, left). Leukaemic cells were CD34-negative (not shown) and exhibited cytoplasmic (in addition to nuclear) NPM expression (Figure 4, right); C23/nucleolin was not detected because of antigenic denaturation. Notably, the NPMVi2 mutant was detected in the cytoplasm of myeloid blasts, megakaryocytes and immature erythroid precursors (Figure 4, right), indicating that multilineage involvement, a common feature of NPMc+ AML,6 may occur with mutations on NPM1 exons other than exon-12.
Our sequence analysis and functional studies reveal that, like NPM leukaemic mutants generated by exon-12 NPM1 mutations,3 cytoplasmic accumulation of NPMVi2 mutant is dictated by the concerted action of tryptophan(s) changes and a new NES motif at the C terminus of the protein. The tryptophan residues 288 and 290 at the NPMVi2 C terminus are missing due to truncation of the last 20 amino acids. This alteration is expected to have the same functional effect on nucleo-cytoplasmic traffic as tryptophan(s) replacement by exon-12 NPM1 mutations. In fact, tryptophan residues are essential for nucleolar binding7 since their absence (or replacement) may favour nuclear export, by reducing mutant binding to nucleoli and increasing affinity for Crm1.3 However, since B23.2, the physiologically truncated NPM isoform present in low amounts in tissues, lacks both tryptophan residues but localizes in nucleoplasm,8 tryptophan loss alone cannot accomplish NPMVi2 mutant delocalization into cytoplasm. The additional force ensuring nuclear export and cytoplasmic accumulation is the new NES (VxxxFxxLxI) at the NPMVi2 mutant C terminus. The VxxxFxxLxI NES sequence is that of a typical Rev-type NES, which is defined by a short stretch of closely spaced leucine or other hydrophobic residues, that is, isoleucine, methionine, valine or phenylalanine.5 Although functionally active, the VxxxFxxLxI NES is slightly different from those identified at C terminus of exon-12 NPM mutants.1, 2, 3, 4 This points to heterogeneity of C terminal NES motifs deputed to ensure nuclear export of NPM mutants.
Thus, we have shown that alterations at C-terminal NPMVi2 mutant mimic those of exon-12 NPM mutants. Since mutations at different exons of NPM1 are all associated with aberrant cytoplasmic NPM expression, they seem to be designed to export mutants into cytoplasm, further supporting the view that cytoplasmic NPM dislocation is a critical step in leukaemogenesis. NPM mutants bind and delocalize endogenous NPM and also the ARF protein into cytoplasm,8 interfering with a proper p53 response. Furthermore, an imbalance in the nuclear–cytoplasmic NPM ratio could alter other functions of NPMwt.8 However, the exact role of NPM mutants in leukaemogenesis still remains unknown.
Our patient also shared several biological and clinical features with NPMc+ AML carrying NPM1 exon-12 mutations, i.e. normal karyotype, M4-FAB morphology, CD34 negativity and involvement of myeloid, erythroid and megakaryocytic series.1, 2 He harboured a NPM1 mutation without FLT3-ITD, a molecular combination that has been associated with a relatively good prognosis in AML with normal karyotype.2 However, the prognostic value of this association cannot be established in our patient. In fact, because of age and cardiac problems, he received only supportive therapy and died of infectious complications five months after diagnosis.
Our findings raise the question about the incidence of NPM1 exon-11 mutations among AML patients. In a previous study correlating immunohistochemistry with molecular analysis of NPM1 mutations, cytoplasmic NPM could predict with 100% accuracy the presence of NPM1 mutations, and all 200 NPM cytoplasmic-positive AMLs harboured NPM1 mutations at exon-12.4 Thus, the mutation described in this study (and other mutations that may putatively involve NPM1 exons other than exon-12) must be extremely rare and clinically of scarce importance. Nevertheless, it is necessary to establish guidelines for detecting these rare NPM1 mutation variants. If using cDNA as template, DHPLC allows the analysis of amplicons including more than one exon at time. For this reason, DHPLC assay is a good approach suitable to identify genetic aberrations occurring outside the specific mutational hot spot. Only the samples showing an abnormal chromatogram are subsequently sequenced. As an alternative, immunohistochemistry that detects, through cytoplasmic dislocation on NPM, ‘all types’ of NPM1 mutations,3, 4 could serve as first step for directing further molecular studies, i.e. restricting analysis for NPM1 exons other than exon-12 to AML cases that may turn out to be NPM cytoplasmic-positive in the absence of exon-12 NPM1 mutations.
Supported by the Associazione Vicentina Leucemie e Linfomi (AVILL), Associazione Italiana per la Ricerca sul Cancro (AIRC) and the Federazione Monte dei Paschi di Siena. EA was recipient of a grant from ‘Fondazione Progetto Ematologia’ – Vicenza. NB is a recipient of a fellowship from FIRC (Federazione Italiana per la Ricerca sul Cancro). We are grateful to Dr A Montaldi for cytogenetic analysis, Dr S Roberti for statistical analysis, Dr GA Boyd for editorial assistance and Mrs I Frasson for excellent secretarial assistance. B Falini applied for a patent of the clinical use of NPM mutants.
About this article
Annals of Hematology (2013)