Mutated NPM1 in combination with overexpression of Meis1 or Hoxa9 is not sufficient to induce acute myeloid leukemia

Background Acute myeloid leukemia (AML) carrying nucleophosmin 1 (NPM1) mutations (NPMc+) is regarded as a separate entity of myeloid neoplasms due to its distinct biological and clinical features. However, NPMc+ alone displays low leukemogenic activity and cooperating events appear crucial for AML to develop. Dysregulation of homeobox genes, such as HOXA9 and MEIS1, is a common transcriptional signature of NPMc+ AML. Furthermore, the pathogenic role for NPMc+ in AML remains incompletely understood. Aim To elucidate if NPMc+ collaborates with Meis1 or Hoxa9 in the evolvement of AML. Methods Murine bone marrow cells were genetically engineered to express mutated NPM1 variant A in combination with overexpression of Meis1 or Hoxa9. The capacity of the transduced cells to transform in vitro and to cause leukemia in vivo was then assessed. Findings and conclusion There was no synergy between NPMc+ and Meis1 or Hoxa9 in causing leukemogenic transformation of murine bone marrow cells, or in inducing AML in a transplantation model. Hence, overexpression of Meis1 or Hoxa9 in combination with NPMc+ expression was not sufficient to generate an NPMc+ AML mouse model.


Background
Nucleophosmin (NPM1) is found mutated in about 30 % of acute myeloid leukemia (AML) cases, making it one of the most commonly mutated genes in AML [1,2]. Wildtype NPM1 has been ascribed many biological functions, including a role in the biogenesis of ribosomes, genomic stability and transport of small basic proteins to the nucleus [3]. The NPM1 mutations occurring in AML disrupts the nucleolar localization signal of NPM1 and generates a nuclear export signal in its place, resulting in aberrant cytoplasmic localization of mutated NPM1 (NPMc + ) [4]. Since 2008, the World Health Organization regards AML with mutated NPM1 as a provisional AML entity due to its unique characteristics [5]. Despite the recognition of mutated NPM1 as a founder genetic lesion [6] and a putative target for novel therapy [7], the mechanism by which NPMc + contributes to leukemogenesis has only partly been explored. Animal models of NPMc + AML, aiming at understanding the NPM1-driven leukemogenesis, have thus far revealed that NPMc + alone displays low leukemogenic activity [8,9] and that additional cooperative mutations, such as FLT3-ITD [10] or N-ras mutations [9] are required for AML to develop.
Gene expression data from both NPM1-mutated adult and pediatric AML cases have identified an association between NPMc + and dysregulated expression of homeobox (HOX) and TALE genes, including Hoxa9 and Meis1, which are known to be involved in hematopoietic development [11][12][13][14]. In accordance, Ogawara et al. reported that NPMc + expression in c-Kit + bone marrow (BM) cells results in increased expression of Hoxa9 [15]. Hoxa9 and Meis1 are primarily expressed at the level of hematopoietic stem and progenitor cells during normal

Transduction of C57BL/6 J bone marrow cells
All mice experiments were approved by the Animal Ethics Research Committee in Gothenburg. C57BL/6 J mice were obtained from Charles River Laboratories (Sulzfeld, Germany). Transduction of C57BL/6 J BM with viral vectors was performed as previously described [22]. In brief, murine BM cells harvested from C57BL/6 J mice 4 days after 150 µg/g 5-fluorouracil (

Methylcellulose colony forming assay
Colony forming unit (CFU) assay was used to evaluate the proliferative capacity of cells after transduction with above mentioned genes. 100 to 5000 cells were seeded in methylcellulose medium (Methocult M3434, StemCell Technologies SARL, Grenoble, France). After 7 days, the number of colonies was counted and the following day cells were replated.

May-Grünwald-Giemsa staining
Using a Shandon CytoSpin 2 (Axel Johnson Instrument AB, Stockholm, Sweden), transduced BM cells were sprayed onto glass slides and allowed to air-dry. Slides were May-Grünwald-Giemsa stained and images taken on a Nikon Eclipse 90i microscope.

Statistics
For statistical analyses, Student´s unpaired t test was performed. All statistical analyses were calculated using GraphPad Prism Version 6.0.

Results
To clarify whether NPMc + and Meis1 or Hoxa9 cooperate in induction of AML, 5-FU-treated BM cells from C57BL/6 J mice were transduced with aforementioned genes as outlined in Fig. 1a. Successfully transduced cells were FACS-sorted based on GFP and/or YFP expression and sorted cells were utilized in the further experiments. Quantitative PCR (qPCR) analysis confirmed increased mRNA expression levels of NPMc + , Meis1 and Hoxa9 in all transduced cells (Fig. 1b-d). The presence of the correct sequence of mutated NPM1 in the transduced NPMc + BM cells was verified by sequencing. No enhancement of Hoxa9 expression was observed in the NPMc + liquid cultures (Fig. 1d). Next, the transfected cells were analyzed by CFU assays to investigate a potential synergy between NPMc + and Meis1 or Hoxa9 in leukemogenic transformation. Cells were cultured in methylcellulose medium for 7 days before enumeration of colonies. Cells transduced with NPMc + , Meis1 or Hoxa9 all showed enhanced serial colony-forming activity, compared with control cells transduced only with the selection markers, neo, GFP or YFP, where virtually no colonies were formed after the first replating (Fig. 2a). The colony forming capacity was highest in cells overexpressing Hoxa9 but no significant differences were observed in colony formation of single transduced cells and cells transduced to express both NPMc + and Meis1 or NPMc + and Hoxa9 (Fig. 2a). Similarly, it was noticed that only cells overexpressing Hoxa9 (in the presence or absence of NPMc + ) became immortalized and survived long-term in vitro cultures.
In support of the necessity of increased Hoxa9 levels for improved proliferative capacity, May-Grünewald-Giemsa staining of cells revealed that Hoxa9 expression was required for cells to maintain a blast-like cell-morphology ( Fig. 2d and f ). In the absence of increased Hoxa9 expression levels, cells differentiated and mainly displayed the morphology of monocytes, macrophages and neutrophils (Fig. 2b, c and e).
Colony-forming capability in vitro often mirrors the potential of cells to engraft in mice [23]. In accordance, transplantation with NPMc + , Meis1 or Meis1-NPMc + cells (in combination with life-sparing BM cells) to lethally irradiated C57BL/6 J mice did not lead to longterm engraftment, as the percentage of transfected cells in the blood of the mice was consistently below 1.5 % (Fig. 2g). In contrast, transplantation with Hoxa9 and Hoxa9-NPMc + cells led to long-term engraftment of leukemic cells, albeit at a low level (Fig. 2g). Three out of five mice transplanted with Hoxa9 cells developed late onset leukemia, with the first mice progressing into leukemia displaying disease symptoms, as determined by increased white blood cell counts along with an increased frequency of GFP + /YFP + leukemic cells in blood, approximately 16 weeks after transplantation (Fig. 2g). None of the mice transplanted with Hoxa9-NPMc + cells showed signs of disease progression within the 20-week observation period (Fig. 2g).

Discussion
Our in vitro and in vivo results consistently demonstrate a lack of synergy between NPMc + and Meis1 or Hoxa9 in terms of transforming murine bone marrow cells and causing AML in an experimental transplantation model. Although NPMc + has been associated with enhanced Hoxa9 expression in AML patients [11] and in transduced murine BM cells [15], we did not observe enhanced expression in our NPMc + liquid cultures (Fig. 1d). The reason for absent Hoxa9 induction and for the low leukemogenic capacity of the NPMc + cells utilized in our study might be that normal levels of wild-type NPM1 was expressed by the transduced BM cells. Hence, NPM1 mutations in patients might cause disease both by causing a reduced expression of wild-type NPM1 (expression from only one allele), and by the oncogenic activity of NPMc + . NPM1 mutations are always heterozygous and a complete knock out of the protein results in embryonic lethality in mice [24]. However, mice genetically engineered to express only one wild-type NPM1 allele display a higher susceptibility to hematological malignancies, including myeloid leukemia, indicating that loss of NPM1 function is a mechanism of pathogenicity [25].
Since NPM1 mutations in AML always result in cytoplasmic localization of the protein, it is conceivable that NPMc + plays a crucial role in pathogenicity. NPMc + has been ascribed oncogenic functions [26][27][28] and transgenic mice expressing NPMc + within the myeloid compartment develop a myeloproliferative disease, albeit no AML [8]. To mimic the situation in human AML, conditional knock-in models, referred to as Npm1 cA/+ mice, have been developed where NPMc + is expressed in mice that display reduced expression-levels of wild-type NPM1. In this setting, one-third of the Npm1 cA/+ mice developed late onset AML, indicating a more aggressive course of disease [9].
Furthermore, when NPMc + was introduced to BM cells from transgenic NPM1 +/− mice, the increase in Hoxa9 expression was much greater than when NPMc + was introduced to wild type BM cells [15]. However, in this previous study, even NPMc + expressing NPM +/− bone marrow cells failed to cause AML when transplanted into irradiated mice, highlighting the importance of cooperative genes [15]. Thus, even in conditions of reduced wild type NPM1 expression, the selection of genes to cooperate with NPMc + in triggering AML is crucial. We and others have previously shown that Hoxa9 and Meis1 act in synergy to cause leukemia in transplantation models [17,29]. Since NPMc + should trigger an enhanced Hoxa9 expression on a NPM1 +/− background [15], it is conceivable that NPMc + should cooperates with Meis1 in causing AML if NPM1 +/− bone marrow cells would be utilized, which merits further studies. Cells overexpressing Hoxa9 showed an increased colony forming and disease causing capacity compared with cells not overexpressing Hoxa9. However, our results indicate, if anything, a reduced disease causing potential of Hoxa9-NPMc + cells compared with Hoxa9 overexpressing cells alone, which do not incite additional studies of synergy between NPMc + and Hoxa9.
Authors' contributions HGW, LP and AM designed the study, HGW, TN and AJ performed experiments and analyzed the results, HGW, LP and AM wrote the manuscript. All authors read and approved the final manuscript.