Background

Several pediatric cancers feature amplification at the chromosomal region 2p24 including alveolar rhabdomyosarcoma (ARMS), neuroblastoma (NB), medulloblastoma, Wilms’ tumor, and retinoblastoma [1,2,3]. The minimum common region of amplification at 2p24 in rhabdomyosarcoma (RMS) and NB has been found to consistently include the oncogene MYCN and amplification of MYCN is used clinically as a prognostic marker in NB [3,4,5,6,7]. Amplification or overexpression of MYCN leads to dysregulation of proliferation, differentiation and the cell cycle in NB [8] and contributes to cell growth in ARMS [9]. MYCN is also capable of positive auto-regulation as well as auto-suppression in NB, potentially fine-tuning MYCN levels [10,11,12]. In ARMS, MYCN transcription is driven by PAX3-FOXO1, the protein product of a fusion between the PAX3 and FOXO1 genes that has prognostic significance in these tumors [9, 13]. RMS and NB are a major cause of cancer related death in children with a five-year survival rate of around 50% for high-risk NB cases, including those with MYCN amplification, and 40% for PAX3-FOXO1 positive RMS cases [13, 14].

MYCN opposite-strand (MYCNOS, N-CYM, MYCN-AS1, NYCM, CYMN) is produced by antisense transcription across exon 1 and intron 1 of MYCN that has been shown to be highly expressed in MYCN-amplified NB and small cell lung cancer [15, 16]. Two alternative transcripts denoted MYCNOS-01 and MYCNOS-02 (Additional file 1: Figure S1A) are fully sequence-verified [15]. There is an emerging body of evidence for roles of MYCNOS-02 through an encoded protein (NCYM) that promotes NB tumorigenesis, in particular via its regulation of MYCN expression, and also its role as a long non-coding RNA (lncRNA) [17,18,19,20,21,22]. NCYM has been shown to mediate expression of MYCN protein by both direct interaction and also indirectly via inhibition of GSK3β, leading to decreased MYCN phosphorylation and a concomitant increase in MYCN protein stability [17]. This was associated with increased tumor growth and metastasis [17]. The same study also concluded that MYCN positively drives the promoter of MYCNOS-02 in an E-box-dependent manner [17]. As well as increasing MYCN protein, NCYM has been found to increase MYCN cleavage to produce the anti-apoptotic protein Myc-nick in NB [22]. NCYM has also been shown to promote aggressiveness in NB by increasing OCT4 expression via its stabilization of MYCN [20].

LncRNAs are commonly defined as transcripts of over 200 nucleotides in length that in general do not code for a protein [23]. MYCNOS-02 lncRNA is able to regulate the usage of two MYCN promoters and therefore expression of different MYCN transcripts via interaction with binding partners such as G3BP1. This in turn results in expression of different isoforms of the MYCN protein [18]. MYCNOS-02 lncRNA has also been found to recruit CTCF to the MYCN promoter to increase recruitment of activating chromatin marks and thus increase MYCN expression [19]. This positive regulation of MYCN suppressed differentiation and increased growth, invasion and metastasis in NB [19]. Additionally, a recent study has shown MYCNOS-02 lncRNA can interact with the RNA-binding protein NonO to indirectly increase MYCN transcript levels post-transcriptionally [21]. Overall, these studies show that both the MYCNOS-02 encoded protein and lncRNA play a role in growth, invasion and metastasis of NB cells [17, 19, 20].

Unlike MYCNOS-02, a functional role for the MYCNOS-01 transcript has not yet been investigated, despite original annotation of its sequence being consistent with a lncRNA [15]. In this study we therefore investigated the role of MYCNOS-01 as a lncRNA in RMS and NB. We demonstrate that MYCNOS-01 post-transcriptionally regulates MYCN protein levels without affecting MYCN mRNA levels, whilst MYCN regulates MYCNOS-01 transcription. We show that silencing of MYCNOS-01 in RMS and NB cell lines with MYCN amplification reduces cell viability, similar to the effects of MYCN reduction. Thus, we conclude that regulation of MYCN by MYCNOS-01 contributes to the reduction in cell growth in RMS and NB cell lines after MYCNOS-01 silencing.

Methods

Translation and Kozak sequence prediction tools

Translation prediction for the MYCNOS-01 transcript sequence was carried out using the ExPASy translate tool (http://web.expasy.org/translate/) [24] and Kozak sequence prediction was carried out using ATGpr (http://atgpr.dbcls.jp/) [25].

Cell culture

Human ARMS cell line RMS-01 was available directly from the authors [26] and the RH30 cell line was a gift from Peter Houghton (St Jude Children’s Research Hospital, Memphis, Tennessee). The human NB cell lines KELLY and SY5Y were obtained from ECACC (cat. No. 92110411) and ATCC (cat. No. CRL-2266) respectively. RMS-01 and RH30 were cultured in DMEM (Thermo Fisher Scientific, MA, USA) and KELLY and SY5Y were cultured in RPMI-1640 medium (Thermo Fisher Scientific, MA, USA) supplemented with 10% Foetal Bovine Serum (FBS), 2 mM L-glutamine and 1% penicillin/streptomycin. The MYCN overexpressing and matched empty vector expressing RH30 lines were generated as previously described in Tonelli et al., (2012) [9] and cultured in DMEM supplemented with 400 μg/ml geneticin. Cells were maintained at 37 °C and 5% CO2. Data from short tandem repeat testing of the cell lines using the GenePrint 10 system (Promega, WI, USA) were compared with records for these cell lines in a repository database or our own archival records. This was consistent with the origin of these cell lines.

Analyses of expression profiling data

Data uploaded to R2 Genomics Analysis and Visualisation Platform (http://r2.amc.nl) were used for analyses. These included 101 RMS samples (ITCC) [6], a set of 19 RMS cell lines (Versteeg) [27], 88 NB samples (Versteeg) and 24 NB cell lines (Versteeg) that had been previously profiled using the Affymetrix GeneChip with the HGU133 Plus2 array. Probe sets could distinguish MYCNOS-01 and MYCNOS-02 transcripts: probe set 216188_at detects MYCNOS-01, set 207028_at detects MYCNOS-02 and set 209757_s_at was used to detect MYCN.

qRT-PCR

MYCNOS-01 and MYCNOS-02 expression data was available from primary sample biopsies from RMS patients. Samples and details of RNA extraction were previously described [6, 28] with appropriate approvals for investigation. RNA was isolated from cell lines using the RNeasy mini kit (Qiagen, Hilden, Germany) according to the manufacturer’s instructions. Cell line cDNA was synthesised using the High Capacity cDNA Reverse Transcriptase kit (Applied Biosystems, CA, USA) and patient sample cDNA was synthesised using SuperScript II reverse transcriptase (Invitrogen, CA, USA) following the manufacturers’ protocol. Samples were run for qRT-PCR on the ViiA™ 7 Real-Time PCR System (Applied Biosystems, CA, USA). The following Taqman® probe and primer sets were used: MYCNOS-01 Hs01032821_m1, MYCNOS-02 Hs01040745_m1, MYCN Hs_00232074_m1. Human ACTB (Beta Actin) endogenous control (Applied Biosystems, CA, USA) was used to normalise gene expression. Each sample was run in triplicate. Analysis of MYCN expression and copy number in patient samples is described in [6].

siRNA transfection

Oligonucleotides for specific silencing of MYCNOS-01, MYCNOS-02 and MYCN were transfected into cells using Lipofectamine RNAimax (Invitrogen, CA, USA) according to the manufacturer’s instructions. All siRNAs were obtained from GE Dharmacon (CO, USA). The sequence from 5′ to 3′ for the three siRNAs against MYCNOS-01 were as follows: siMYCNOS-01 1 GGGACAAGAGCACAGUUUCUU, siMYCNOS-01 2 GGUAAGUUAAGGUACAGCCUU, siMYCNOS-01 3 GGAGUAUUUGUUUAGUGCUUU. The sequences for the three siRNAs against MYCNOS-02 were GAAAGAAGGGUAGUCCGAAUU for siMYCNOS-02 1, GACCGAUGCUUCUAACCCAUU for siMYCNOS-02 2, CCGCUUUGACUGCGUGUUGUU for siMYCNOS-02 3. For knockdown of MYCN a pool of three siRNAs was used with sequences GAAGAAAUCGACGUGGUCA, CCAAGGCUGUCACCACAUU, AAUUGAACACGCUCGGACU, as previously described [9]. The control siRNA used was the ON-TARGETplus non-targeting control pool (GE Dharmacon, CO, USA). Samples were analyzed by qRT-PCR, Western blot, flow cytometry or phenotypic assays at time-points indicated in the relevant figures.

Western blotting

Protein lysates were prepared using Cell Lysis Buffer (Cell Signaling Technology, MA, USA) and their concentration measured by the Pierce™ BCA protein assay kit (Thermo Fisher Scientific, MA, USA). Protein samples were resolved by SDS-PAGE and transferred onto PVDF membranes. Blots were incubated with the following primary antibodies: MYCN SC-791 (1:200, Santa Cruz, TX, USA), PARP 9542 (1:1000, Cell Signaling Technology, MA, USA), Phospho-C-Myc (Thr58/Ser62) 04–217 (1:4000, Merck Millipore, MA, USA), GAPDH MAB374 (1:10000, Merck Millipore, MA, USA). Blots were then incubated with rabbit (sc-2313, Santa Cruz, TX, USA) or mouse (A9044, Sigma-Aldrich, MO, USA) horseradish peroxidase-conjugated secondary antibody diluted to 1:4000 depending on primary antibody species. Blots were developed using the ECL™ Prime Western Blotting System (GE Healthcare, IL, USA) on the Chemidoc Touch Imaging System (Bio-Rad, CA, USA). Densitometry was performed using Bio-Rad Image Lab 5.2.1 (Bio-Rad, CA, USA).

Immunofluorescence staining

Cells cultured in chamber slides were fixed with 2% paraformaldehyde for 15 min at room temperature and permeabilised with 0.1% Triton X-100. Samples were blocked in PBS with 10% goat serum and 1% BSA for 1 h. Samples were incubated with primary MYCN antibody SC-53993 (1:500, Santa Cruz, TX, USA) overnight at 4 °C followed by secondary antibody Alexa Fluor 555 goat anti-mouse (1:400, Invitrogen, CA, USA) for 30 min at room temperature. Cells were counterstained with DAPI. Fluorescent images were captured using a Zeiss Axioplan 2 microscope (Oberkochen, Germany) using a 16× objective and a standard exposure time optimised for control treated cells. The sum of the intensity for MYCN staining was measured using Image J software and made relative to the number of cells in that field of view, indicated by DAPI.

Plasmid production and transfection

Full-length MYCNOS-01 transcript (RefSeq NR_110230) was cloned into the pcDNA5/TO vector (Invitrogen, CA, USA) and the construct verified by Sanger sequencing (Eurofins Genomics, Ebersberg, Germany). For plasmid transfection, Lipofectamine 2000 (Invitrogen, CA, USA) was used following the manufacturer’s instructions. Samples were analyzed by qRT-PCR and Western blot at time-points indicated in the relevant figures.

Proteasome inhibition

For protein stability experiments, cells were transfected for a total of 48 h and treated for the final 4 h with either 10 μM MG132 (Sigma-Aldrich, MO, USA) in DMSO or DMSO control. Protein was then extracted from cells for analysis by Western blot.

Cell viability assay

Cells were transfected as six repeats in a 96-well plate to assess the effects of gene knockdown. Cell viability was assessed by the MTS method using the CellTiter 96® Aqueous One Solution Cell Proliferation Assay (Promega, WI, USA). Fresh media plus 20 μl assay reagent were added at the indicated time-point. After 2.5 h incubation at 37 °C and 5% CO2 the absorbance of each well was measured at 492 nm on a FLUOstar Optima plate reader (BMG Labtech, Ortenberg, Germany).

Apoptosis assay

Cells were transfected in quadruplicate in a 96-well plate and apoptosis measured by evaluating the activation of caspase 3/7 at the indicated time-point by replacement of 50 μl of media with 50 μl of Caspase-Glo® 3/7 Assay (Promega, WI, USA). After 1 h incubation protected from light at room temperature, the samples were transferred to a white-walled 96-well plate and luminescence of each well was read on a FLUOstar Optima plate reader (BMG Labtech, Ortenberg, Germany). The caspase signal intensity was normalised by the absorbance measurement from the corresponding MTS assay.

Flow cytometry

Cells were fixed in 70% ethanol at -20 °C overnight, washed with PBS and resuspended in 1 ml PBS containing 100 μg/ml RNase A and 40 μg/ml Propidium Iodide. Samples were incubated for 30 min at 37 °C then analyzed on a BD™ LSRII Flow Cytometer (BD Biosciences, CA, USA).

Statistical analysis

Graphs represent means ± standard deviation from multiple independent experiments as stated in figure legends. Statistical significance was measured by unpaired two-tailed Student t-test or by one-way analysis of variance (ANOVA) with post hoc Dunnett’s test for multiple comparisons. For linear correlation studies of gene expression Pearson’s coefficient (R) was calculated between each pair of variables to indicate the strength of the linear association. p < 0.05 was considered significant and indicated by a single asterisk, p < 0.01 is indicated by a double asterisk, and p < 0.001 is indicated by a triple asterisk.

Results

MYCNOS-01 is predicted to be non-protein coding and correlations between MYCNOS and MYCN transcript levels in RMS and NB patients and cell lines

Previous data showed protein encoding potential for MYCNOS-02 transcripts [15]. In contrast, translation and Kozak sequence analysis for MYCNOS-01 predicted transcripts to be non-protein coding since the probability any start codon present in the MYCNOS-01 sequence is an initiation codon is very low (Additional file 1: Figure S1B and C). The relationship between MYCN and MYCNOS-01 transcript levels was then investigated by mining publically available expression profiling data for RMS and NB samples from patients and cell lines. A significant correlation between levels of MYCNOS-01 and MYCN expression and MYCNOS-02 and MYCN expression was identified considering all RMS samples (Fig. 1a and b) and cell lines (Fig. 1c and d). There was also a significant positive correlation between both MYCNOS transcripts and MYCN transcript expression in NB patient samples (Fig. 1e and f) and cell lines (Fig. 1g and h) that is consistent with the literature for MYCNOS-02 [17, 19]. For NB patient samples and cell lines, levels of MYCNOS-01 and 02 were higher in cases with MYCN amplification (Fig. 1e-h) versus cases without. Amplification at 2p24 in RMS is also associated with high MYCNOS-01 and MYCNOS-02 expression levels, as indicated in data mined for the cell lines (Fig. 1c and d) and qRT-PCR analyses of a limited number of cases with known MYCN amplification status as well as the cell lines (Additional file 2: Figure S2A–D). MYCN and MYCNOS-01 transcript levels in RMS and NB showed no significant correlations when MYCN amplified cases were excluded (RMS p = 0.72, R = 0.078; NB p = 0.42, R = 0.097).

Fig. 1
figure 1

Correlation of MYCNOS transcripts and MYCN transcript expression in RMS and NB. a Correlation between MYCNOS-01 and MYCN measured by Affymetrix microarray in 101 RMS patient samples (R = 0.501, p < 0.0001) and (b) correlation between MYCNOS-02 and MYCN in these patient samples (R = 0.483, p < 0.0001). c Correlation between MYCNOS-01 and MYCN measured by Affymetrix microarray in 19 RMS cell lines (R = 0.523, p = 0.022) and (d) correlation between MYCNOS-02 and MYCN (R = 0.672, p = 0.0016). e Correlation between MYCNOS-01 and MYCN measured by Affymetrix microarray in 88 NB patient samples (R = 0.696, p < 0.0001) and (f) correlation between MYCNOS-02 and MYCN in these patient samples (R = 0.759, p < 0.0001). g Correlation between MYCNOS-01 and MYCN measured by Affymetrix microarray in 24 NB cell lines (R = 0.782, p < 0.0001) and (h) correlation between MYCNOS-02 and MYCN (R = 0.877, p < 0.0001). Data derived from R2 genomics analysis and visualisation platform. MYCN-amplified cases or cell lines are depicted with circles

Full size image

MYCNOS-01 regulates MYCN protein but not transcript levels in MYCN-amplified RMS and NB cells

To determine whether MYCNOS-01 regulates MYCN transcript levels, which may be consistent with the correlations in their expression levels in NB and RMS derived samples, we performed siRNA-mediated silencing of MYCNOS-01 in cell lines. RMS cell lines tested had either high (RMS-01) or intermediate (RH30) MYCN expression levels, and NB cell lines used had high (KELLY) and low (SY5Y) levels of MYCN [7,8,9]. The relative MYCNOS and MYCN transcript levels for the four cell lines used are indicated in Additional file 2: Figure S2E.

In high MYCN-expressing lines, no consistent effect on MYCN transcript levels was observed after silencing of MYCNOS-01 for 72 h (Fig. 2a-d). To determine whether MYCNOS-01 is regulating MYCN post-transcriptionally, we also assessed MYCN protein levels by Western blotting. MYCN protein was decreased in cells transfected with MYCNOS-01 siRNAs compared to non-targeting control in high MYCN-expressing cell lines RMS-01 and KELLY (Fig. 2e, f). This is supported by the decrease in MYCN immunofluorescence signal after MYCNOS-01 depletion compared to control in these cell lines (Additional file 3: Figure S3A-F). Furthermore, MYCNOS-01 overexpression has no effect on MYCN transcript levels but causes a slight increase in MYCN protein; the effect observed is limited since RMS-01 and KELLY cells express very high basal levels of MYCN (Additional file 4: Figure S4). Experiments on RMS-01 cells including a proteasome inhibitor showed no increase in phosphorylated MYCN with MYCNOS-01 knockdown, indicating that MYCN protein stability was not affected (Additional file 5: Figure S5). However, MYCNOS-01 silencing had no strong effect on either MYCN transcript or protein levels in the ARMS cell line RH30 with intermediate MYCN expression (Fig. 3a-c). Due to the low expression levels of MYCN protein in the NB cell line SY5Y, no bands were visible by Western blot but there was similarly no MYCNOS-01-mediated effect on MYCN transcript expression (Fig. 3d and e).

Fig. 2
figure 2

MYCN transcript and protein expression after MYCNOS-01 knockdown in high MYCN-expressing RMS and NB. qRT-PCR detecting MYCNOS-01 transcript level after MYCNOS-01 knockdown for 72 h with three siRNAs in (a) RMS-01 and (c) KELLY. Corresponding MYCN transcript level shown in (b) RMS-01 and (d) KELLY. Expression relative to NT control. (e) and (f) MYCN protein levels after MYCNOS-01 knockdown for 72 h and 96 h with three siRNAs in RMS-01 and KELLY respectively. GAPDH used as loading control. Densitometry values shown above each blot normalised to GAPDH and relative to NT control. Data representative of three repeats. NT = non-targeting control. Relative expression of transcripts in these cell lines are indicated in Additional file 2: Figure S2E

Full size image
Fig. 3
figure 3

MYCN transcript and protein expression after MYCNOS-01 knockdown in intermediate and low MYCN-expressing RMS and NB. qRT-PCR detecting MYCNOS-01 transcript level after MYCNOS-01 knockdown for 72 h with three siRNAs in (a) RH30 and (d) SY5Y. Corresponding MYCN transcript level shown in (b) RH30 and (e) SY5Y. Expression relative to NT control. (c) MYCN protein levels after MYCNOS-01 knockdown for 72 h and 96 h with three siRNAs in RH30. GAPDH used as loading control. Densitometry values shown above each blot normalised to GAPDH and relative to NT control. Data representative of three repeats. NT = non-targeting control. Relative expression of transcripts in these cell lines are indicated in Additional file 2: Figure S2E

Full size image

As a role for MYCNOS-02 in RMS has not been previously evaluated, we also performed siRNA-mediated silencing of MYCNOS-02 to determine the effects on MYCN mRNA and protein expression in RMS-01 and KELLY cells (Additional file 6: Figure S6). Similar levels of MYCNOS-02 silencing were observed with all 3 siRNAs, but there was no consistent effect on MYCN expression at the RNA or protein level.

Overall, these results demonstrate that MYCNOS-01 reduces MYCN protein levels in MYCN-amplified RMS and NB cell lines but not through altering MYCN phosphorylation.

MYCN regulates MYCNOS-01 transcript levels

Silencing MYCN in RMS and NB caused a significant increase in MYCNOS-01 expression in both high MYCN-expressing (RMS-01, KELLY) and intermediate MYCN-expressing (RH30) RMS and NB cell lines (Fig. 4a-i). Conversely, overexpressing MYCN in RH30 decreased MYCNOS-01 expression (Fig. 4j-l). MYCN was found to similarly regulate MYCNOS-02 in RMS and NB with MYCN amplification (Additional file 7: Figure S7). Overall, these results indicate MYCN negatively regulates MYCNOS-01 expression whilst MYCNOS-01 positively regulates MYCN protein levels.

Fig. 4
figure 4

Effect of MYCN knockdown on MYCNOS-01 transcript expression in RMS and NB. a qRT-PCR and (b) Western blot for MYCN, and (c) qRT-PCR for MYCNOS-01 after treatment with three pooled MYCN siRNAs for 72 h in RMS-01. d-f in KELLY as for (a)-(c). g-i in RH30 as for (a)-(c). j qRT-PCR and (k) Western bot for MYCN, and (l) qRT-PCR for MYCNOS-01 in RH30 stably transfected with empty vector (EV) or HA-MYCN expressing vector. Graphs and Western blots representative of three repeats. NT = non-targeting control

Full size image

Decreasing MYCNOS-01 levels results in decreased cell viability in MYCN-amplified RMS and NB cells

We next investigated the phenotypic effects of MYCNOS-01 and MYCNOS-02 depletion on RMS (RMS-01, RH30) and NB (KELLY, SY5Y) cells. Decreasing MYCNOS-01 expression resulted in a significant decrease in cell viability compared to negative control with all MYCNOS-01 siRNAs in MYCN-amplified RMS-01 and KELLY cells (Fig. 5a and b), similar to the reduction in cell viability that is seen with siRNA-mediated MYCN silencing (Fig. 5a and b). Silencing MYCNOS-01 did not reduce MYCN protein lower than direct MYCN silencing (Fig. 5c), but showed a similar or greater effect on RMS cell viability (Fig. 5a) raising the possibility that MYCNOS-01 may have targets in addition to MYCN in RMS. Decreasing MYCNOS-02 expression also significantly decreased cell viability in these cell lines (Additional file 8: Figure S8A and B). In contrast, silencing of MYCNOS-01 did not significantly affect cell viability in intermediate (RH30) and low (SY5Y) MYCN expressing cells overall (Fig. 6a and b). However, although the effect was less for RH30 compared to RMS-01, MYCN knockdown did significantly decrease cell growth as we have previously reported [9] (Fig. 6a). MYCN reduction in SY5Y had no effect, although levels of MYCN are very low (Fig. 6b). There was no significant increase in caspase 3/7 activation or PARP cleavage in MYCNOS-01 or MYCN siRNA treated cells compared to non-targeting control indicating apoptosis was not induced (Additional file 9: Figure S9). However, MYCNOS-02 knockdown promoted apoptosis in RMS-01 and KELLY cells (Additional file 10: Figure S10). Decreasing MYCNOS-01 expression had no effect on cell cycle progression but silencing MYCN caused a G1 arrest in both RMS-01 and KELLY (Additional file 11: Figure S11). Based on all our results, we propose a feedback model for MYCNOS-01 and MYCN regulation in RMS and NB (Fig. 7).

Fig. 5
figure 5

Knockdown of MYCNOS-01 inhibits cell viability in high MYCN-expressing RMS and NB. (ai) MTS assay showing cell viability of RMS-01 up to 144 h after transfection with three MYCNOS-01 siRNAs or three pooled MYCN siRNAs relative to NT siRNA. (bi) KELLY as in (ai). Corresponding qRT-PCR for MYCNOS-01 (ii) and MYCN (iii) at 24 h shown below line graph. c MYCN protein levels after MYCNOS-01 or MYCN knockdown for 72 h in RMS-01 and (d) in KELLY. Densitometry values shown above each blot normalised to GAPDH and relative to NT control. Data representative of three repeats. Statistical analysis relative to non-targeting control. NT = non-targeting

Full size image
Fig. 6
figure 6

Knockdown of MYCNOS-01 does not inhibit cell viability in intermediate or low MYCN-expressing RMS and NB. (ai) MTS assay showing cell viability of RH30 up to 96 h after transfection with three MYCNOS-01 siRNAs or three pooled MYCN siRNAs relative to NT siRNA. (bi) SY5Y as in (ai). Corresponding qRT-PCR for MYCNOS-01 (ii) and MYCN (iii) at 24 h shown below line graph. Data representative of three repeats. Statistical analysis relative to non-targeting control. NT = non-targeting

Full size image
Fig. 7
figure 7

Schematic of proposed model for regulation between MYCN and MYCNOS-01. MYCN and MYCNOS-01 expression are both driven by genomic amplification, although MYCN is expressed at higher levels than MYCNOS-01. MYCNOS-01 positively regulates MYCN protein levels and MYCN negatively regulates MYCNOS-01 transcript levels as a form of negative feedback on MYCN. MYCNOS-01-mediated regulation of MYCN has effects on cell viability but there is also likely a MYCN-independent signaling pathway

Full size image

Discussion

In this study, we have shown that MYCNOS-01 and MYCNOS-02 can play important roles in RMS as well as NB cell growth, at least in part via their regulation of MYCN. Roles for MYCNOS-01 in RMS and NB and MYCNOS-02 in RMS have not been previously explored whilst our data for the effect of MYCNOS-02 on growth of NB cells is consistent with previous findings [17,18,19,20,21,22]. Regulation of MYCN by MYCNOS-01 and MYCNOS-02 was readily apparent in MYCN-amplified RMS and NB, which express these transcripts at high levels, presumably as a result of their co-amplification at the genomic level. In contrast, effects of MYCNOS transcripts on MYCN protein levels in RMS and NB without high level MYCN amplification were either less marked or not seen. The positive regulation of MYCN by MYCNOS-01 and MYCNOS-02 likely contribute to the cell growth of RMS and NB. This is consistent with the phenotypic dose dependent effects and dependencies of RMS and NB cells on MYCN levels seen in this and previous studies [9, 29, 30].

In addition to the positive regulation of MYCN by MYCNOS-01, MYCN also negatively regulates MYCNOS-01 transcription, summarized in Fig. 7. As MYCN protein has been found to be recruited to its own intron 1 [10], in line with the transcription start site for MYCNOS-01 on the opposite strand, this binding activity could be involved in MYCN-mediated regulation of MYCNOS-01. Previous studies have also identified indirect MYCN negative feedback mechanisms involving trans-acting factors [12]. This is likely a mechanism that fine-tunes MYCN expression levels.

An increasing number of lncRNAs have now been characterized and many have been linked to cancer progression [31]. Here we have shown that MYCNOS-01 can act as a cis-antisense lncRNA on its sense partner MYCN. Although MYCN transcript expression was not regulated by MYCNOS-01, we have identified a post-transcriptional role for this lncRNA in regulating MYCN protein levels. This is consistent with the lack of correlation we found between the two transcripts in non-amplified lines.

There are several examples of lncRNAs that regulate protein partners post-transcriptionally without affecting transcript expression [32,33,34,35]. For example, the lncRNA PVT1 has been shown to be required for increasing MYC protein stability and high expression levels in 8q24-amplified cancers [32]. In our study, we found no evidence that MYCNOS-01 altered the stability of MYCN protein via Thr58/Ser62 phosphorylation. Another example of post-transcriptional regulation is the lncRNA treRNA, which has been shown to play a role in tumor invasion and metastasis in breast cancer, and which regulates the translation of E-cadherin mRNA in these cells via redistribution of CDH1 to low molecular weight polysomes to suppress translation [34]. Potentially MYCNOS-01 could have a similar mechanism to regulate translation efficiency of MYCN and thus its protein expression. In addition, the lncRNA BACE1-AS regulates translation of BACE1 by masking the binding site for miR-485-5p, thus preventing miRNA-induced translational repression and mRNA decay [35]. Another possibility therefore is that MYCNOS-01 interacts with an miRNA that targets MYCN for degradation, therefore increasing MYCN protein expression by sequestering away a negative regulating factor. The above examples indicate possible post-transcriptional mechanisms that could be explored for MYCNOS-01-mediated regulation of MYCN. Further investigations are required to determine molecular interactions with MYCNOS-01 and how these regulate MYCN protein levels.

Both MYCNOS-01 and MYCNOS-02 shown in this study in RMS and NB, and MYCNOS-02 shown previously in NB, regulate MYCN protein levels [17, 19, 20]. However, data on whether MYCNOS-02 is able to regulate MYCN at the transcriptional level is conflicting. One report for MYCNOS-02 indicates silencing of MYCNOS-02 does not affect MYCN expression at the transcriptional level [17]. However, other studies suggest that MYCNOS-02 can affect MYCN transcript expression in NB [18, 19, 21]. MYCNOS-02 has been found to interact with CTCF to affect chromatin remodeling at the MYCN promoter therefore MYCN transcript expression [19]. However, in silico prediction techniques suggest CTCF interacts with a region of MYCNOS-02 that does not overlap with the sequence of MYCNOS-01, supporting the possibility that the two transcripts could have different binding partners that are involved in MYCN regulation. It is possible for two overlapping lncRNAs from the same locus to have different characteristics and function. For example, the CCAT1 locus encodes CCAT1-L that is located in the nucleus and positively regulates MYC transcription and CCAT1-S that is mainly located in the cytoplasm with no effect on MYC transcript levels [36,37,38].

Previous studies of MYCNOS-02 have identified its role in NB tumor growth and metastasis. In vitro, MYCNOS-02 can suppress differentiation and promote metastasis, invasion and cell proliferation partially due to its indirect regulation of MYCN [19]. Our study found MYCNOS-01 also plays a role in MYCN-amplified RMS and NB cell viability, although no effect was seen on cell cycle progression. Often a decrease in proliferation occurs with concomitant cell cycle arrest, however it is possible for these two effects to be separated. For example, one study found decreasing the tumor suppressors RPL5 or RPL11 resulted in a reduction in ribosome content and translation capacity, causing cells to progress at a lower rate through all stages of the cell cycle thus resulting in decreased proliferation without cell cycle arrest [39].

MYCNOS-01 was also found to play a role in cell growth via regulation of MYCN; silencing of MYCNOS-01 resulted in a reduction in MYCN protein levels. However, MYCNOS-01 knockdown did produce a slightly different phenotype to MYCN knockdown due to differences in MYCN protein levels achieved. We have previously identified that using different molecular tools to diminish MYCN can affect the strength of phenotype detected depending on the magnitude and endurance of MYCN reduction [9]. MYCNOS-01 knockdown did not decrease MYCN protein sufficiently to produce a G1 arrest, in contrast to direct MYCN knockdown. However, MYCNOS-01 reduction affecting other protein targets and signaling pathways that contribute to the phenotype observed cannot be excluded. Further defining how MYCNOS-01 regulates MYCN, and possibly other proteins, may lead to new approaches to perturb the clinically aggressive phenotype of RMS and NB tumors.

Conclusions

MYCNOS-01 can be added to the growing list of lncRNAs involved in tumorigenesis; MYCNOS-01 positively regulates MYCN and MYCN negatively regulates MYCNOS-01, potentially to fine-tune MYCN protein levels. MYCNOS-01 affects cell growth of MYCN-amplified RMS and NB and could also play a role in other MYCN-driven cancers. Although directly targeting lncRNAs and MYCN is challenging, future therapeutic strategies could disrupt specific lncRNA-protein interactions to reduce MYCN levels.