Background
Acute myeloid leukemia (AML) is characterized by the dysregulated proliferation and impaired differentiation of myeloid precursor cells. Many of these leukemias harbor genetic translocations, which determine both the molecular mechanistics and the prognosis of the disease [
1]. The t(6;9)(p23;q34) chromosomal translocation is found in 1% of AML, where it is associated with young age and poor prognosis [
2]. The translocation occurs between specific introns in the gene
DEK on chromosome 6 and the gene
NUP214 on chromosome 9, creating the fusion gene
DEK-NUP214 (formerly
DEK-CAN). The original reading frames are preserved, yielding an invariable fusion protein comprising almost the entire chromatin remodeling protein DEK and the carboxy-terminal two thirds of the nucleoporin NUP214 [
3].
Despite extensive characterization of many other fusion genes, the role of DEK-NUP214 is still poorly understood. We have previously shown that DEK-NUP214 promotes the activating phosphorylation of the eukaryotic translation initiation factor 4E (eIF4E) on Ser
209 and increases the protein synthesis of myeloid cells [
4]. However, the cause as well as the cellular effects thereof remain to be explored. Recently, DEK-NUP214 has been shown to induce leukemia in a murine model, but only from long-term repopulating stem cells and with long latency, emphasizing the need for cooperating mutations [
5]. A striking feature of leukemias with the
DEK-NUP214 fusion gene is the concomitant internal tandem duplication (ITD) in the tyrosine kinase FLT3. The
FLT3-ITD genotype is more than three times as common in leukemias with t(6;9)(p23;q34) as in other AML [
2,
6]. This suggests a classic oncogenic cooperation between a pro-proliferative FLT3-ITD and a differentiation-blocking DEK-NUP214. However, recent studies have identified a role for FLT3-ITD also in inhibition of myeloid differentiation [
7]. And contrary to many fusion proteins observed in AML, DEK-NUP214 does not seem to inhibit differentiation, at least not when expressed in the monocytic cell line U937 [
8]. This raises the possibility of a role for DEK-NUP214 in proliferation.
The mechanistic target of rapamycin (mTOR) is a central node in the regulation of both proliferation and translation [
9]. The mTOR protein is found in two complexes. Activated by growth factor signaling, the mTOR complex 2 (mTORC2) phosphorylates Akt at Thr
450 and Ser
473, in turn activating mTOR complex 1 (mTORC1) [
10]. mTORC1 initiates translation by phosphorylation of its downstream targets, such as the p70 S6 kinase [
11]. Although mTORC1 regulates the translation of most mRNAs, some transcripts are particularly sensitive. These include many oncogenes such as c-myc and cyclin D1. Activation of the mTORC1 pathway thus promotes cellular growth and proliferation [
12,
13].
In addition to its role in translation, mTORC1 also affects cellular metabolism by promoting the more energy-efficient oxidative phosphorylation over glycolysis. This role is independent of the translational regulation machinery and rather seems to involve phosphorylation of mitochondrial enzymes [
14,
15]. Due to its multiple roles in carcinogenesis and its common overactivation in cancer, mTOR has become an attractive target for cancer therapy and there are currently several inhibitors in clinical trials [
16]. Recently, the FDA approved the highly specific mTORC1 inhibitors everolimus (RAD001) and temserolimus (CCI-779) for the treatment of advanced renal cell carcinoma and everolimus is currently in clinical trial for acute myeloid leukemia [
17‐
19].
In this study, we show that overexpression of DEK-NUP214 in the myeloid cell line U937 leads to increased levels of mTOR and activation of the mTOR target p70S6K. This translates into higher protein synthesis and a metabolic shift from glycolysis to oxidative phosphorylation. Accordingly, cells expressing DEK-NUP214 proliferate faster than their normal counterparts. Treatment with the mTORC1 inhibitor everolimus selectively reverses the DEK-NUP214-induced proliferation, suggesting that the effect is mTOR-dependent and that patients with t(6;9)(p23;q34) may be suitable for treatment with mTOR inhibitors.
Methods
Cell culture
The cell lines U937 and PL-21 (ATCC, Manassas, VA, USA) and stable clones derived thereof were cultured in RPMI 1640 medium (Life Technologies, Carlsbad, CA, USA) supplemented with 10% fetal bovine serum (Life Technologies). Stable clones expressing either the DEK-NUP214 fusion gene [
4], DEK-NUP214 deletion mutants [
4] or the corresponding empty pcDNA3 vector, were generated by electroporation followed by incubation for 48 h and subsequent seeding of 10 000 cells per well in 100 μl medium. After two weeks of selection by culture in growth medium supplemented with 0.5 mg/ml geneticin (Life Technologies), clones were selected and expanded.
Proliferation experiments
For proliferation experiments, cells were seeded in fresh culture medium at a density of 0.5 × 106 cells/ml and when indicated treated with daily additions of the mTORC1 inhibitor everolimus (Sigma-Aldrich, St. Louis, MO, USA). Cell counting was performed with the Countess Automated Cell Counter (Life Technologies) and viability was determined on the basis of trypan blue dye exclusion (Life Technologies).
Protein expression
Protein expression was analyzed by western blot one day after seeding, as described above. Cells were washed in PBS (PAA Laboratories, Pasching, Austria), resuspended and frozen in sample buffer containing 0.1 M Tris–HCl pH 6.8, 0.2 M β-mercaptoethanol, 14% glycerol (v/v), 3% SDS (w/v), 0.01% bromophenol blue (w/v), Complete protease inhibitor cocktail (Roche Diagnostics GmbH, Mannheim, Germany) and PhosStop protease inhibitor cocktail (Roche Diagnostics GmbH). Samples were sonicated in a UP50H ultrasonic homogenizer (Dr. Hielscher GmbH, Teltow, Germany), boiled for 5 minutes and centrifuged at 14 000 × g for 5 minutes. Lysates corresponding to 500 000 cells were run on tris-glycine gels (Life Technologies) and transferred by an SV20-SDB semi-dry blotter (Sigma-Aldrich) to Hybond ECL membrane (GE Healthcare, Uppsala, Sweden). Membranes were blocked with 5% bovine serum albumin (Sigma-Aldrich) and incubated with one of the following antibodies according to the manufacturers’ recommendations: anti-α-tubulin (Sigma-Aldrich), anti-GAPDH (Santa Cruz Biotechnology, Santa Cruz, CA, USA), anti-phospho-mTOR-Ser2448, anti-mTOR, anti-phospho-Akt-Ser473, anti-phospho-Akt-Thr308 or anti-phospho-p70-S6K-Thr389 (Cell Signaling Technology, Danvers, MA, USA). HRP-conjugated anti-mouse or anti-rabbit were used as secondary antibodies (Bio-Rad Laboratories, Hercules, CA, USA) and detected with the EZ-ECL kit (Biological Industries, Kibbutz Beit Haemek, Israel). Quantification was performed using the Molecular Imager FX (Bio-Rad Laboratories) with the Quantity One 4.2.2 software (Bio-Rad Laboratories).
Gene expression analysis
Gene expression was determined by quantitative real-time PCR. RNA was extracted using the RNeasy Mini Kit (Qiagen GmbH, Hilden, Germany) and reverse transcription was performed with the High Capacity cDNA Reverse Transcription Kit (Life Technologies). Expression levels were assayed with the TaqMan Gene Expression Assay and primer-probe pairs for the detection of glyceraldehyde-3-phosphate dehydrogenase (GAPDH; Hs99999905_m1), mechanistic target of rapamycin (mTOR; Hs00234522_m1) or DEK-NUP214 (custom made) (Life Technologies). The amplification reaction was performed using the StepOnePlus Real-Time PCR System (Life Technologies). The expression of DEK-NUP214 and mTOR was calculated relative to the expression of GAPDH using the comparative C
T method, as previously described [
20]. cDNA from a patient with the t(6;9)(p23;q34) chromosomal translocation was kindly provided by professor Bertil Johansson at the Department of Clinical Genetics, Lund University.
Global translation assay
The translation rates of the stable clones were assessed by radioactive labeling of newly synthesized proteins. Cells were seeded in fresh culture medium at a density of 0.5 × 106 cells/ml. At indicated time points, EXPRESS35S Protein Labeling Mix containing radioactively labeled methionine and cysteine (PerkinElmer, Waltham, MA, USA), was added to cell cultures to a final concentration of 50 μCi/ml. After incubation for 2 h, 100 000 viable cells of each clone were sorted by a FACSAria cell sorter (BD Bioscience, San José, CA, USA), washed in PBS and lysed in radioimmunoprecipitation buffer (30 nM HEPES, pH 7.3, 1% Triton-X (v/v), 1% sodium deoxycholate (w/v), 0.1% SDS (w/v), 0.15 M NaCl) containing the Complete Protease Inhibitor Cocktail (Roche Diagnostics GmbH, Mannheim, Germany). Proteins were precipitated by addition of trichloroacetic acid to a final concentration of 9%. The precipitate was washed twice in acetone, suspended in 50 μl 0.1 M Tris–HCl, pH 8.6, and added to 5 ml of scintillation fluid (Beckman Coulter, Fullerton, CA, USA). The radioactivity of the samples was measured by a Wallac Guardian 1414 liquid scintillation counter (PerkinElmer). Values were corrected for background by subtracting the values from samples incubated with the EXPRESS35S Protein Labeling Mix on ice.
Cells were seeded in fresh culture medium at a density of 0.5 × 106 cells/ml. At indicated time points, cell suspension was taken out and centrifuged at 145 × g for 5 minutes. Supernatant was collected and stored at −80°C to prevent degradation of lactate. The glucose concentration was measured by applying 10 μl of supernatant to the Glucose Assay Kit II (BioVision, Mountain View, CA, USA). After dilution of the supernatant 1:50 in lactate assay buffer, the lactate concentration was determined by applying 10 μl to the Lactate Assay Kit II (BioVision). Absorbance was measured at 450 nm with a Labsystems Multiskan Plus Plate Reader (Thermo Fisher Scientific, Waltham, MA, USA).
Statistical analysis
Statistical testing was performed using the two-tailed t test, where the averages of the three DEK-NUP214 clones from each experiment were tested against the averages of the three control clones from the same experiments. Stars represent conventional significance levels; single stars indicate p < 0.05, double stars indicate p < 0.01 and triple stars indicate p < 0.001.
Discussion
This study is the first to demonstrate that the expression of the fusion gene DEK-NUP214 leads to increased cellular proliferation. We show that this is dependent on upregulation of the signal transduction protein mTOR with subsequent effects on protein synthesis and glucose metabolism. We proceed to show that the proliferative effect can be overcome by inhibition of mTORC1 with everolimus, suggesting that patients with the DEK-NUP214 fusion gene may benefit from treatment with mTOR inhibitors.
The biology of DEK-NUP214 is notoriously elusive. Although the genetic translocation was characterized almost two decades ago, only a few reports have studied its role in leukemogenesis and none has been able to show whether the contribution is on the level of proliferation or differentiation. We find in this study that DEK-NUP214 increases the proliferation of myeloid cells. This is a property shared by several fusion proteins, the most similar being SET-NUP214, which contains the same portion of NUP214 [
23]. But also other nucleoporin fusions such as NUP98-HOXA9 and NUP98-HHEX show similar pro-proliferative properties both in culture and in vivo [
24‐
26]. In some aspects, this finding is in contrast with a previous study of the NUP214 gene, which also included one DEK-NUP214 clone. This clone displayed equal or slightly lower proliferation as compared to wild-type cells [
8]. We cannot with certainty determine the reason for this discrepancy but it may be the result of different expression levels of the fusion gene. Interestingly, Boer et al. selected the clone with the highest inducible expression of DEK-NUP214 for their proliferation experiment. As with some other oncogenes, DEK-NUP214 may promote proliferation at low or moderate levels and inhibit proliferation when highly expressed. Such a disadvantageous effect of high gene expression could also explain the low expression levels of DEK-NUP214 in cells with stable expression of the gene; both our clones and cells from patients with the t(6;9)(p23;q34) translocation [
27].
In characterizing the proliferative effect, we find that DEK-NUP214 promotes signaling through the mTOR pathway. We demonstrate that DEK-NUP214 increases the level of mTOR protein, without affecting any of the upstream regulators AMPK, GSK3 or Akt. Despite extensive characterization of mTOR activation, surprisingly little is known about the regulation of its expression. β-catenin is known to influence the transcription of mTOR [
28] but since this was unaffected by DEK-NUP214, we suggest another mode of regulation. The mechanism remains to be investigated and may involve miRNA-mediated inhibition of translation, subcellular relocalization or covalent modification, but most likely involves the stabilization of mTOR by protein-protein interaction, since this has been described for several other proteins such as Raptor [
29], C/EBPδ [
30], Tti1 [
31] and Tel2 [
31]. We also see an increase in the level of mTOR protein phosphorylated on Ser
2448. This phosphorylation is mediated by p70S6K in a feedback loop, whose effect on the activity of mTOR is not yet understood [
32,
33]. The increase in mTOR-p-Ser
2448 may arise from the observed activation of p70S6K or may reflect the increased availability of mTOR protein in cells expressing DEK-NUP214. By examining the phosphorylation of their substrates, we can conclude that in this context, the increased level of mTOR confers increased activity of mTORC1 but not mTORC2. The reason for this may be that the availability of the other factors of the complexes makes mTORC1 more susceptible to an mTOR increase or that the mTORC1 substrates are more sensitive to changes in mTOR complex activity.
To address the functional relevance of the increased mTOR signaling, we analyzed the cellular translation rate. The first day after seeding, nutrients and growth factors are readily available and the conditions for translation are highly favorable. The rate of translation is subsequently very high. Hence, it is not very surprising that the expression of DEK-NUP214 does not markedly enhance the translation rate at this time point. However, three days after seeding, the control cells have decreased their rate of protein synthesis by two thirds whereas the cells expressing DEK-NUP214 sustain a 68% higher translation rate than the control cells. Due to the rapid growth and proliferation of cancer cells, they require extensive protein synthesis also when nutrients and growth factors are scarce [
34]. This key feature is crucial for malignant transformation and could be a mechanism by which DEK-NUP214 contributes to leukemogenesis.
A more recently discovered function of mTOR is in glucose metabolism. Most cancer cells initially rely heavily on aerobic glycolysis, a phenomenon known as the Warburg effect [
35]. However, as proliferation increases, so does the energy demand. A second metabolic shift can serve to reestablish the more energy-efficient oxidative phosphorylation, while also providing metabolites for macromolecule anabolism [
36]. Dysregulation of the mTOR pathway has been proposed as such an event, as overactivation of mTORC1 leads to a shift from glycolysis to oxidative phosphorylation [
15]. Our findings confirm this notion. Cultures of cells expressing DEK-NUP214 produce less lactate but consume as much glucose as cultures of control cells, indicating such a shift. Given the higher proliferation rate and thus higher number of cells in the DEK-NUP214 cultures, the glucose consumption per cell is somewhat lower than for the control cells. However, this decrease alone cannot explain the decrease in lactate levels, since a reduction in glucose consumption that only offsets the effect of increased proliferation on total glucose levels would consequently also only offset the effect of proliferation on total lactate levels. What we observe here is a reduction in total lactate levels, thus indicating a metabolic shift from glycolysis to oxidative phosphorylation.
mTOR has attracted widespread attention as a target for cancer therapy and several variants of the original mTOR inhibitor rapamycin are being evaluated in clinical trials, both for solid tumors and leukemias [
18]. One of these is everolimus, which employs the same mechanism of action as rapamycin and has been approved by the FDA for the treatment of renal cell carcinoma [
17]. Our results show that treatment with everolimus completely ablates the proliferative phenotype induced by DEK-NUP214. Strikingly, it does so at concentrations that do not affect the control cells. This may be because the higher proliferation rate of the DEK-NUP214 cells produces higher demands and thus increased dependence on mTORC1 signaling. Compensatory pathways may thus be able to sustain the proliferation rate of the control cells but not the increase caused by the expression of DEK-NUP214. These results demonstrate that the increased proliferation by DEK-NUP214 is indeed dependent on mTORC1. Furthermore, it suggests that patients with leukemias harboring the t(6;9)(p23;q34) may benefit from treatment with the novel mTOR inhibitors that are becoming increasingly available.
Acknowledgements
The authors wish to thank Falko Apel for technical assistance. This work was supported by grants from the Medical Faculty at Lund University (ALF), the Swedish Cancer Society, the Swedish Research Council, the Swedish Childhood Cancer Foundation, the Gunnar Nilsson Cancer Foundation, the Österlund Foundation, the Siv-Inger and Per-Erik Andersson Memorial Fund and the Åke Olsson Foundation.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
CS, MA, AL and UG designed experiments. CS, JP and AL performed experiments. CS, MA, JP, AL and UG analyzed data. CS and UG wrote the paper. All authors read and approved the manuscript.