Background
We originally identified human
NBPF1 (Neuroblastoma BreakPoint Family, member 1) in a neuroblastoma (NB) patient on the basis of its disruption in a
de novo, constitutional translocation between chromosomes 1p36.2 and 17q11.2 [
1-
3].
NBPF1 is a member of a gene family with intricate genomic structure [
4]. The
NBPF members are primarily located on duplicated regions of chromosome 1, and analysis of the predicted protein sequences showed that several pairs of exon types encode a protein domain called the NBPF/DUF1220 repeat [
4,
5]. The number of encoded NBPF/DUF1220 repeats varies from 4 to 52 copies, depending on the gene member and the NBPF1 protein has 7 repeats [
6]. The copy number of the NBPF/DUF1220 repeat is much larger in humans than in other primates, which suggests an important role in human evolution [
4,
5]. The
NBPF genes are likely involved in cancer and in brain and developmental disorders (reviewed in [
7]). This has been ascribed to their location in unstable high-identity duplication blocks, which leads to recurrent chromosomal rearrangements. One tumor type of particular interest is NB.
NB tumors are derived from the sympathetic nervous system and account for approximately 15% of cancer deaths in children [
8]. However, a fascinating feature of NB is its remarkable biological heterogeneity, which is evident in the broad variety of clinical courses of the disease. While some patients experience spontaneous regression or differentiation of the tumor, others are affected by rapid and fatal tumor progression despite increasingly intensive treatment strategies [
9].
Evidence for the involvement of
NBPF genes in NB comes from the abovementioned disruption of
NBPF1 in a NB patient, and from the association of NB with copy number variation of an
NBPF1 paralog [
10]. Interestingly, the 1p36 region is frequently deleted not only in NB, but also in other human cancer types, including those of neural, epithelial and hematopoietic origin, indicating that the same tumor suppressor genes might be involved in a broad range of human cancers [
11].
Based on these findings, we hypothesized that NBPF1 acts as a tumor suppressor. Previously, we reported that expression of
NBPF1 mRNA is significantly decreased in NB cell lines with loss of heterozygosity for 1p36 compared to cell lines with a normal 1p36 locus [
3]. This decreased expression is a hallmark of tumor suppression activity. Moreover, NBPF1-expressing colon cancer cells formed significantly fewer colonies in soft agar than control cell lines, indicating that NBPF1 might be important for suppression of anchorage-independent growth [
3].
In this study, we show that NBPF is expressed in the non-proliferative suprabasal layers of squamous stratified epithelia of human skin and cervix. Moreover, we show that forced expression of NBPF1 in the human HEK293T cell line resulted in a p53-dependent G1 cell cycle arrest that was accompanied by upregulation of the cyclin-dependent kinase inhibitor p21
CIP1/WAF1. Additionally, overexpression of
NBPF1 in two p53-mutant NB cell lines resulted in G1 cell cycle arrest and concomitant
CDKN1A induction. However, G1 cell cycle arrest and
CDKN1A upregulation were not observed in a colon cancer cell line in which NBPF1 expression was induced, despite the clear NBPF1-dependent inhibition of anchorage-independent clonal growth in this cell line [
3]. This demonstrates that NBPF1 exerts cell-specific tumor suppressive effects. In conclusion, this study advances the understanding of the role of NBPF1 as a tumor suppressor.
Methods
Reagents
Doxycycline hydrochloride (dox; Sigma, Bornem, Belgium) was used at a final concentration of 2 μg/ml to induce expression of NBPF1-IRES-EGFP or EGFP in DLD1Tr21/NBPF1 and DLD1Tr21/Mock cells, respectively. Doxorubicine hydrochloride (Doxo; Sigma) was used at a final concentration of 350 nM during 24 h. Aphidicolin (Sigma) was used at a final concentration of 2 μg/ml.
Cell culture and transfections
Human HEK293tsA1609neo (in short HEK293T) and its derivative HEK_shRNAp53 (in which the expression of p53 has been silenced by an shRNA) were cultured in DMEM medium supplemented with 10% fetal calf serum (FCS), 2 mM glutamine and 0.4 mM sodium pyruvate. SH-SY5Y neuroblastoma cells (ATCC, CRL-2266) were cultured in DMEM medium supplemented with 15% FCS, 2 mM glutamine and 0.4 mM sodium pyruvate. The SK-N-AS (ATCC, CRL-2137) and NLF [
12] neuroblastoma cell lines were cultured in RPMI medium supplemented with 10 % FCS, 2 mM glutamine and 0.4 mM sodium pyruvate. HEK293T cells were transfected using the calcium phosphate precipitation method, and neuroblastoma cells were transfected with Fugene HD (Roche Applied Science, Vilvoorde, Belgium). DLD1Tr21/Mock and DLD1Tr21/NBPF1 cells have been described and were grown in standard medium composed of RPMI with 10% FCS [
3].
Plasmid constructions
To construct the pdEGFP-NBPF1 plasmid for expression of EGFP-fused NBPF1 under the control of a CMV promoter, full-length NBPF1 cDNA was transferred to the pdEGFP vector by Gateway technology (Life Technologies Europe, Ghent, Belgium). The vector expressing EGFP-luciferase under the control of an SV40 promoter (pGL3-EGFP-luciferase) was a kind gift from Dr. Eric Raspé. To generate DLD1Tr21/NBPF1 cells, we cloned cDNA of NBPF1-IRES-EGFP, fused to an amino-terminal Flag tag, into the pcDNA4/TO vector (Life Technologies Europe). This plasmid expresses flag-tagged NBPF1, under the control of a doxycycline-inducible promoter [
3].
Immunofluorescence
Cells on glass coverslips were rinsed briefly with PBS and fixed with 4% paraformaldehyde. They were incubated for 1 h with primary antibody diluted in blocking solution (0.4% gelatin). Antibodies used were goat anti-NBPF sc-82241 (Santa Cruz Biotechnology, Heidelberg, Germany, dilution 1/200) and mouse anti-p53 DO-1 (Santa Cruz Biotechnology, dilution 1/250). Cells were then washed in PBS and incubated for 1 h with secondary antibodies coupled to Alexa fluorophores (Life Technologies Europe) diluted in blocking solution. Coverslips were mounted in Vectashield containing DAPI (Vector Laboratories, Peter Borough, United Kingdom) to prevent photobleaching.
SDS-PAGE and immunoblotting
EGFP-positive and EGFP-negative populations of HEK293T, HEK293T_shRNAp53, SH-SY5Y, NLF and SK-N-AS cells were isolated at 48 h after transfection by FACS (Epics Altra, Beckman Coulter, California, USA) and used for protein extraction with Trizol (Life Technologies Europe). For subsequent SDS-PAGE analysis we used a 10 % gel and loaded a total of 40 μg protein per lane. After blotting to PVDF membranes (Millipore, Billerica, MA, USA), blocking with 5 % non-fat dry milk occurred. The membrane was washed 3 times with 5% w/v BSA, 1x TBS, 0.1% Tween® 20 and then incubated overnight with primary anti-p21CIP1/WAF1 antibody (rabbit anti-p21CIP1/WAF1 12D1, Cell Signaling Technology, Beverly, MA, USA; dilution 1/1000) in the same buffer at 4 °C. After several washings, the membrane was incubated for 1 h with secondary horseradish peroxidase (HRP)-conjugated antibodies and proteins were detected by the enhanced chemiluminescence (ECL) detection system (GE Healthcare Europe, Diegem, Belgium). Anti-vinculin (mouse anti-vinculin, Sigma, dilution 1/1000) or anti-actin (mouse anti-actin, MP Biomedicals, dilution 1/10.000) detection was used as a loading control and anti-EGFP (rabbit anti-GFP A11122, Life Technologies, dilution 1/1000) or anti-NBPF detection (sc-82241, dilution 1/1000) was used to confirm transfection efficiency or in cell sorting experiments.
Immunohistochemistry on paraffin-embedded tissue sections
Formalin-fixed, paraffin-embedded tissue specimens were retrieved from the archives of the Pathology Department at the University Hospital of Liège (protocol approved by the Liège University Hospital Ethics Committee). No other ethical approval was required for this study. Human and mouse tissues were fixed in 4% paraformaldehyde, dehydrated and paraffin-embedded. For staining, paraffin was removed and sections were rehydrated. Endogenous peroxidase was blocked with a 1/100 dilution of H2O2 in methanol. Antigen retrieval was carried out in a pressure cooker (2100 Retriever, PickCell Laboratories, Amsterdam, The Netherlands) with the sections soaked in citrate retriever buffer. Aspecific binding sites were blocked with 3 % horse serum in PBS + 1% BSA. Incubation with primary antibody (sc-82241, diluted 1/200 in PBS/1 % BSA) was done overnight at 4°C. Incubation with biotinylated secondary antibody (Dako, Glostrup, Denmark, dilution 1/300) was carried out for 30 min at room temperature. This was followed by a 20-min incubation with StreptABComplex/HRP (Dako) according to the manufacturer’s protocol. After washing, the sections were treated with DAB (BioGenex, Fremont, CA, USA) and counterstained with hematoxylin following standard procedures. Finally, the sections were embedded in Pertex (HistoLab, Gothenburg, Sweden).
Time-lapse microscopy
HEK293T cells were transfected in an eight-well LabTek II chamber glass slide (Sigma) with plasmids encoding either EGFP-NBPF1 or EGFP-luciferase. After 6 h, the transfection medium was replaced with fresh medium and in one set of experiments aphidicolin was added after 24 h. At 48 h after transfection, EGFP-positive cells were selected for time-lapse recording using a cell observer spinning disk system (Zeiss, Zaventem, Belgium). This system includes an observer Z.1 microscope equipped with a Yokogawa disk CSU-X1. Images were taken with a pln Apo 40x/1.4 oil DIC III objective and a Rolera em-c
2 Camera. The setup includes an incubation chamber, maintained at 37°C in a 5% CO
2 atmosphere. DIC and fluorescence images were obtained every 10 min for 15 h. Images were processed into movies using Fiji (Just ImageJ;
http://pacific.mpi-cbg.de/).
Cell cycle analysis
Cells were harvested 48 h after transfection and fixed for 1 h in ice-cold ethanol. Fixed cells were treated with RNase A (Sigma, 1 mg/ml) and stained with propidium iodide (Sigma, 50 μg/ml). The cellular DNA content of EGFP-positive and EGFP-negative populations was evaluated using flow cytometry (FACSVerse; Becton Dickinson, San Jose, California, USA). We applied software-based compensation (Flowjo) to the collected data. Compensation matrix was defined with reference to single color controls. Gating strategy: 1/ side scatter versus forward scatter to exclude debris; 2/ FSC-height versus FSC-area for doublet discrimination; 3/ EGFP (FITC) versus forward scatter to identify EGFP-negative and EGFP-positive subpopulations. Cells with very strong EGFP expression were not taken into account for cell cycle analysis because time lapse experiments had shown that cells with such strong EGFP-NBPF1 expression were dying, and so we considered these EGFP levels as not physiological. All experiments were performed at least three times, and one representative figure is shown.
Real-time qRT-PCR
EGFP-positive and EGFP-negative populations of HEK293T and HEK293T_shRNAp53 cells were isolated at 48 h after transfection by FACS (Epics Altra, Beckman Coulter, California, USA) and used for RNA extraction and real-time qRT-PCR analysis. RNA was prepared from the different cell populations with Trizol (Life Technologies Europe). cDNA was prepared with the iScript cDNA synthesis kit according to the manufacturer's instructions (Bio-Rad). qRT-PCR mixes contained 20 ng template cDNA, SensiFast SYBR No-ROX kit (GC Biotech, Alphen aan den Rijn, The Netherlands) and 300 nM forward and reverse primers. Reactions were performed on a LightCycler 480 (Roche) using the following protocol: incubation at 95°C for 5 min, followed by 40 cycles: 95°C for 10 s, 60°C for 30 s, and 72 °C for 1 s. Primers used are listed in Table
1. Gene expression levels were normalized using the geometric mean of the most stable reference genes [
13].
Table 1
List of primer sequences used for quantitative RT-PCR of the genes indicated
ANAP2 | GACAGGAGTTGTTAGTGGCCT | TCCAGGACCGAGTGTAGCC |
ATM | ACGTTACATGAGCCAGCAAAT | GAAAATGAGGTGGATTAGGAGCA |
ATR | TCTCAGCCAACCTCCGTGAT | GCAGAAGTCTCGTTATGATCCAA |
CCND1 | GTGCTGCGAAGTGGAAACC | ATCCAGGTGGCG ACGATCT |
CCNE1 | GAGCCAGCCTTGGGACAATAA | GCACGTTGAGTTTG GGTAAACC |
CDC34 | GACGAGGGCGATCTATACAACT | GAGTATGGGTAGTCGATGGGG |
CDK2 | GCTAGCAGACTTTGGACTAGCCAG | AGCTCGGTACCACAGGGTCA |
CDK4 | ATGTTGTCCGGCTGATGGA | CACCAGGGTTACCTTGATCTCC |
CDK6 | GCGCCTATGGGAAGGTGTTC | TTGGGGTGCTCGAAGGTCT |
CDKN1A | CGCTAATGGCGGGCTG | CGGTGACAAAGTCGAAGTTCC |
CDKN1B | CTGCAACCGACGATTCTTCTACT | GGGCGTCTGCTCCACAGA |
CUL1 | ACCAGTCAAACCAAGCACGAG | GTCTGCCCCTTTTTCGACTTAG |
CUL2 | ACGACAATAAAAGCCGTGGTC | GGTTCAGGATAGGCCACACATA |
CUL3 | TGTGGAGAACGTCTACAATTTGG | TGCGCCTCTGTCTACGACT |
SKP2 | ACCTCCAGGAGATTCCAGACC | CCCAGGTTTGAGAGCAGTTCC |
TP53 | CCTCAGCATCTTATCCGAGTGG | TGGATGGTGGTACAGTCAGAGC |
NBPF1 | GCGAGGCTGCCCGAGCTTCT | GACTTCGCGTAACTTCCCATTCA |
EGFP | GAGCTGAAGGGCATCGACTT | TCTGCTTGTCGGCCATGAT |
Proteome analysis
DLD1Tr21/Mock and DLD1Tr21/NBPF1 cells were grown for 4 days in their standard medium, containing 2 μg/ml dox. Cell pellets were re-suspended in 0.5 ml of cell lysis buffer (50 mM sodium phosphate, pH 7.5, 100 mM NaCl, 0.8% (w/v) CHAPS, and Complete Protease Inhibitor cocktail (Roche)). The insoluble fraction was removed by centrifugation (15 min at 16.000 g at 4°C). Samples were desalted on a NAP5 column in 20 mM tri-ethylammonium bicarbonate (pH 7.0). Endoproteinase Lys-C digestion was carried out at 37°C overnight using a ratio of 1/100 (w/w) endoproteinase Lys-C/protein. Peptides from the DLD1Tr21/Mock cell lysate were post-metabolically labeled with light NHS-12C3-propionate, whereas the peptides from DLD1Tr21/NBPF1 cell lysate were labeled with heavy NHS-13C3-propionate. Equal amounts of the peptide mixtures were combined, concentrated by vacuum-drying and re-dissolved in 1% acetic acid. The peptide mixture was then separated by RP-HPLC chromatography and 20 fractions were collected for further LC-MS/MS analysis. LC-MS/MS analysis was done on a LTQ-Orbitrap mass spectrometer. We acquired 158,141 MS/MS spectra, from which we identified 13,315 unique peptides (Swiss-Prot database, restriction to human proteins) and 3613 proteins (Mascot software used at 99% confidence). Mascot Distiller was then used to quantify the proteins and only proteins with a minimum of two quantifiable peptides in the experiment were included in the Ingenuity Pathway Analysis (IPA, Ingenuity Systems, Redwood City, CA). Keratins and ribosomal proteins were considered non-specific and were excluded from IPA analysis.
Discussion
We previously reported evidence for a tumor suppressor role of NBPF1 [
3]. However, the molecular mechanisms responsible for its anti-tumor effect were not elucidated. This paper shows that NBPF1 can exert its tumor suppressive effect in different ways: by blocking cell division, inducing cell death, or by modulating a cancer-associated proteome.
Live-cell imaging of the embryonic kidney cell line HEK293T, transfected with an EGFP-NBPF1 expressing construct, showed a temporally fluctuating pattern of EGFP-NBPF1 expression in comparison with control transfected cells continuously expressing EGFP-luciferase. In addition, we did not observe any EGFP-NBPF1 transfected cell that completed a cell cycle, whereas control EGFP-luciferase transfected cells and mock transfected cells divided frequently. Analysis of the different stages of the cell cycle in this cell line showed that overexpression of NBPF1 resulted in a clear G1-arrest in comparison with the controls. In the HEK293T cells, this arrest could be explained by the specific induction of p21CIP1/WAF1, seen at both the mRNA and protein levels.
The best known activator of p21
CIP1/WAF1 is the tumor suppressor p53 [
19]. Activation of p21 by p53 has indeed been shown to play a key role in the ability of p53 to induce cell cycle arrest [
20]. Therefore, we performed experiments to determine if NBPF1 was acting via p53 induction. First, we performed immunofluorescence for p53 in HEK293T cells and showed an increased nuclear accumulation of p53 upon overexpression of NBPF1, whereas knock-down of p53 did not result in a G1 cell cycle arrest and abolished NBPF1-induced
CDKN1A expression. Still, the p53 status in HEK293T cells is controversial as these cells express adenoviral proteins that inhibit p53 activity. To clarify this controversy, we stimulated HEK293T and DLD1Tr21 cells with an inducer of DNA double-strand breaks, doxorubicin, which results in p53-dependent upregulation of p21 [
21]. Treatment with doxorubicin induced p21 in the HEK293T cells, but not in the DLD1Tr21 cell line with mutated
TP53 (data not shown). Therefore, we concluded that the activity of p53 is only attenuated in HEK293T cells, but not completely inhibited. This leads to transcriptional activation of
CDKN1A upon doxorubicin treatment, and therefore, the induction of p21 upon overexpression of NBPF1 in HEK293T cells can be considered to be p53-dependent. Moreover, stable knockdown of p53 in HEK293T cells completely abrogated the NBPF1-induced cell cycle block and
CDKN1A upregulation, thereby confirming that NBPF1-induced cell cycle arrest is p53-dependent in HEK293T cells.
Nuclear translocation of p53 indicates its activation [
22], and NBPF1 seems to act in HEK293T cells via increasing the nuclear expression of p53. However, the mechanism by which NBPF1 can modulate p53 remains unknown. As NBPF1 proteins are located in the cytoplasm, whereas active p53 is in the nucleus, a direct interaction between NBPF1 and p53 seems unlikely. However, MDM2, the main inhibitor of p53 activity, is located in the cytoplasm, and NBPF1 might interact with MDM2, thereby preventing the negative effects of MDM2 on p53. To test this hypothesis, we performed co-immunoprecipitation experiments after overexpression of both NBPF1 and MDM2 in HEK293T cells. Although we could demonstrate a minor interaction between NBPF1 and MDM2 (data not shown), we believe that this interaction is of insufficient extent to explain the observed NBPF1-driven nuclear accumulation of p53. On the other hand, it is possible that NBPF1 promotes the nuclear localization of p53 via a posttranslational modification, such as phosphorylation or acetylation [
23], but additional experiments are required to elucidate this.
The human
NBPF1 gene was originally identified in a NB patient [
1-
3], and we reported previously that expression of
NBPF1 mRNA is significantly decreased in NB cell lines with loss of heterozygosity for 1p36 compared to cell lines with a normal 1p36 locus [
3]. Therefore, we investigated the role of NBPF1 in several NB cell lines, both with mutant or wild-type
TP53 alleles. Overexpression of NBPF1 in NLF and SK-N-AS resulted in a G1 cell cycle arrest and in
CDKN1A upregulation. NLF and SK-N-AS bear loss-of-function mutations of
TP53, namely, a missense mutation in the
TP53 exon 6 (607G > A) of NLF cells, and a homozygous deletion of
TP53 exons 10 and 11 of SK-N-AS cells [
12]. These results indicate that the G1 cell cycle arrest is p53-independent or that these mutant p53 proteins can still induce p21. Stable knockdown of p53 in these NB cell lines is required to confirm that NBPF1-induced cell cycle arrest is p53-independent in these NB cell lines. Surprisingly, overexpression of NBPF1 in the SH-SY5Y cell line, expressing wild-type p53, did not result in a G1 cell cycle arrest, but led to the appearance of a sub-G1 peak, indicative of induced cell death. Therefore, we hypothesize that the growth-inhibitory effect upon NBPF1 overexpression is p21-dependent, and that overexpression of NBPF1 in the SH-SY5Y cell line does not result in G1 arrest due to lack of p21 induction.
The observed arrest of the cell cycle by overexpression of NBPF1 is in line with the apparent expression of NBPF family members in the non-proliferating suprabasal layers of squamous stratified tissues, whereas the basal layers with proliferating cells are essentially NBPF-negative. We demonstrated this by immunohistochemistry using a polyclonal antibody after having clearly demonstrated its specificity for NBPF proteins. This commercial antibody might therefore be used in future studies on NBPF expression in human tumor samples.
We could not confirm NBPF1-dependent
CDKN1A upregulation in DLD1Tr21/NBPF1 colon cancer cells. Our proteome analysis of DLD1 cells with induced NBPF1 expression indicates that the absence of a cell cycle arrest in these cells might be due to the upregulation of several cell-cycle promoting genes, such as
EEF1A2 and
MCM3 [
24,
25]. Whether these genes are directly upregulated by NBPF1 or whether their upregulation represents compensatory effects counteracting NBPF1 function in DLD1 cells needs further investigation. In any case, these data indicate that the proposed tumor suppressive properties of NBPF1 might be very versatile and cell-specific.
Despite the lack of
CDKN1A upregulation in DLD1 cells with inducible NBPF1 expression, these cells show a marked decrease of clonal growth in a soft agar assay upon NBPF1 induction [
3]. Anchorage-independent growth of cells is one of the hallmarks of cellular transformation and is regarded as a critical step during metastasis [
26]. Intriguingly, VCP (valosin-containing protein), which is associated with anti-apoptotic functions and metastasis via activation of the NF-κB signaling pathway [
27], was significantly downregulated upon NBPF1 expression in DLD1 cells. Moreover, NBPF1 induction decreased the expression of S100P, the expression of which is associated with drug resistance, metastasis, and poor clinical outcome in many different cancers such as colon, breast and prostate cancer [
28]. Also, overexpression of S100P in PC3 prostate cancer cells promotes anchorage-independent growth in soft agar [
29].
Both overexpression and knockdown strategies are commonly used to scrutinize gene functions. Thus, knockdown of
NBPF1 would have been a complementary approach to the experiments described here. We used several approaches and numerous attempts to knock down the expression of the
NBPF1 gene or the
NBPF gene family in human cells (Table
2). On the one hand, we aimed at silencing specifically the
NBPF1 gene. On the other hand, we tried to silence the
NBPF gene family as a whole because downregulation of only
NBPF1 might lead to compensation by
NBPF paralogs having overlapping functions and therefore obscure the phenotype. However, none of the many siRNAs and shRNAs tested showed an acceptable knockdown level for silencing endogenous NBPF, which has prevented us so far from proceeding with functional studies.
Table 2
Overview of shRNAs and siRNAs used in this study with the aim to silence the expression level of NBPF family members
shRNA_19-mer | 1 | CCUCUUCUGCCACAAACGUUU | X | | 1 |
2 | GGAAUGUGCCAUCACUUGU | X | | 5 |
3 | GGAAGUCCCUGAGGACUCA | X | | 6 |
4 | GAAGAAGAUCAACGAAGAA | | X | 12 |
5 | CUUACUUUGAUGGGAACAA | | X | 14 |
shRNA_29-mer | 6 | UACCUCUUCUGCCACAAACGUCAGCAUGGU | X | | 1 |
7 | AUGUCAGGAUGCUGUAAACAUUCUCCCAG | | X | 5 |
8 | AACAUCAACAUCACCUUUGAGGAAGACAA | | X | 5 |
9 | UGACUCCAACCAGCCUCACAAGAACAUCA | X | | 5 |
10 | GGAGAAGACGGUGUACCAGCCUCACAAGA | X | | 5 |
11 | ACCAGUCUUACAGCGGCACAUUUCACUCA | | X | 8 |
12 | CCUGACUCAUUCCAGCACUACAGAAGUGU | X | | 14 |
13 | AGCUGGCAGAGAACAAACAGCAGUUCAGA | X | | 1 |
14 | AUGCUGAGGAAUGAGCGACAGUUCAAGGA | X | | 2 |
15 | UAAGGGAGAAGUUACGGGAAGGGAGAGAU | X | | 3 |
16 | CUUCCUGGCCAACCAGCAGAACAAAUACA | X | | 1 |
siRNA pool | 17 | GAAUCGUACUGCUAAGAAU | | X | 5' UTR |
ACAGAGACAGACAAAGAUA | X | | 5' UTR |
GAUGGCCAAUCUUUCCUAA | X | | 5' UTR |
GCAAGUCCAGGUCAUACUG | X | | 5' UTR |
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made.
The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
VA and KS acquired the data. PB, GVI and DG performed flow cytometry and FACS. EP performed time lapse microscopy. JV and KG performed proteomic analysis. VA, KV, GB and FvR contributed to conception and design. VA, KV, GB and FvR drafted the manuscript. All authors read and approved the final manuscript.