Silencing of the Long Noncoding RNA MYCNOS1 Suppresses Activity of MYCN-Amplified Retinoblastoma Without RB1 Mutation

Purpose MYCNOS (MYCN opposite strand) is co-amplified with MYCN in pediatric cancers, including retinoblastoma. MYCNOS encodes several RNA variants whose functions have not been elucidated in retinoblastoma. Thus, we attempted to identify MYCNOS variants in retinoblastoma and aimed to decipher the role of MYCNOS variant 1 (MYCNOS1) on the activity of MYCN-amplified retinoblastoma. Methods The profiles of MYCNOS variants and MYCN status were determined in 17 retinoblastoma tissues, cell lines, retinas, and retinal organoids. A functional study of MYCNOS1 expression was conducted in patient-derived tumor cells and in retinoblastoma cell lines via short hairpin RNA-mediated gene silencing. We carried out MYCN expression, cell viability, cell cycle, apoptosis, soft agar colony formation, and transwell assays to examine the role of MYCNOS1 in MYCN and cell behaviors. We analyzed a transcriptome of MYCN-amplified retinoblastoma cells deficient for MYCNOS1 and, finally, tested the responses of these cells to chemotherapeutic agents. Results Expression of MYCNOS1 was associated with the expression and copy number of MYCN. Knockdown of MYCNOS1 caused instability of the MYCN protein, leading to cell cycle arrest and impaired proliferation and chemotaxis-directed migration in MYCN-amplified retinoblastoma cells in which RB1 was intact. MYCNOS1 expression was associated with gene signatures of photoreceptor cells and epithelial–mesenchymal transition. MYCNOS1 silencing enhanced the response of retinoblastoma cells to topotecan but not carboplatin. Conclusions MYCNOS1 supports progression of retinoblastoma. Inhibition of MYCNOS1 expression may be necessary to suppress MYCN activity when treating MYCN-amplified cancers without RB1 mutation.

Genomic DNA was extracted from primary retinoblastoma, blood, and cells using the DNeasy Blood and Tissue Kit (Qiagen, Valencia, CA, USA). MYCN copy number was determined by digital droplet PCR (ddPCR). A 20-µL reaction contained 1× QX200 ddPCR EvaGreen (BioRad Laboratories, Hercules, CA, USA), 50 nM forward primer (I2_MYCN_F) and 100 nM reverse primer (I2_MYCN_R) for MYCN amplicon, 150 nM forward primer (RPP30_F) and 100 nM reverse primer (RPP30_R) for reference gene amplicon (RPP30), and 3 ng of HaeIII-digested DNA template. The droplets were generated by QX200 Droplet Generator (BioRad Laboratories), followed by PCR. Thermal cycling condition consisted of an activation period (95°C for 5 minutes) followed by 40 cycles of two-step thermal profile (95°C for 30 s followed by 60°C for 1 minute), a dye-stabilization step (4°C for 5 minutes, then 90°C for 5 minutes), and finally a 4°C indefinite hold. A 2°C/s ramp rate was applied for all thermal cycling steps. A signal in droplets (16,000-18,000 droplets) was read using the QX200 Droplet Reader (BioRad Laboratories), and copy number was analyzed by QuantaSoft analysis software (BioRad Laboratories). The assay was repeated using RLBP as a reference gene, and DNA from healthy controls was included in assays. Primer sequences are listed in Supplementary Table 2.
Whole-genome analysis of tumor and paired blood DNA was performed for patient RB170.
Germline variants were called using Isaac Variant Caller 8 followed by SnpEff 9 for annotation and effect prediction for single-nucleotide variants and small indels, and Manta 10 for structural variants. Copy number alterations and loss of heterozygosity were determined in blood and tumor DNA by using the CGH/SNP array (Infinium CytoSNP-850K array, Illumina Inc.) in accordance with the manufacturer's instructions; the results were analyzed and visualized using Nexus copy number software (BioDiscovery, El Segundo, CA, USA).
Multiplex ligation-dependent probe amplification (MLPA) and Sanger sequencing were conducted according to the previous work 11 to determine genetic abnormalities in the RB1 gene.

RNA Expression Analysis
RNA was extracted from primary retinoblastoma and cells using TRI Reagent (Molecular Research Center Inc., Cincinnati, OH, USA). cDNA was synthesized from 1 µg of total RNA using random primer and ImProm-II Reverse Transcriptase (Promega, Madison, WI, USA).
RT-PCR reactions were conducted to amplify MYCNOS (variants 1-5), MYCN, and coding RB1 regions. The RT-PCR products of RB1 were purified using a gel extraction kit (Qiagen) for sequencing. RT-qPCR was performed using iTaq Universal SYBR Green Supermix consisted of initial denaturation (95°C for 20 s) followed by 40 cycles of two-step thermal profile (95°C for 5 s and 60°C for 30 s), followed by melting curve analysis. Thermal cycling condition of RT-PCR for RB1 consisted of initial denaturation (98°C for 30 s) followed by 30 cycles of three-step thermal profile (98°C for 10

Western blotting
Cells were lysed using radioimmunoprecipitation assay (RIPA) lysis and extraction buffer containing halt protease inhibitor cocktail (Thermo Fisher Scientific, Waltham, MA, USA) in accordance with the manufacturer's instructions. Lysates were mixed with Laemmli buffer and freshly added β-mercaptoethanol and boiled at 95°C for 15 minutes. Total proteins were separated on 10% SDS polyacrylamide gel and transferred to the PVDF membrane, which was then incubated in 5% non-fat skim milk and 0.2% Tween in Tris-buffered saline. (1:5000, #12620; Cell Signaling Technology). Actin protein was used as a loading control.
Protein signals were developed using chemiluminescent HRP substrate, in accordance with the manufacturer's instructions, and recorded using the Chemidoc MP Imaging System (Bio-Rad Laboratories).

Protein Stability Assay
Retinoblastoma cells were treated with 10 µg/mL cycloheximide for 0, 15, 30, 60, and 120 minutes before cell lysates were collected. A total protein of 40 µg was used for immunoblotting for MYCN, and the levels of actin were used as a loading control.

Histology, Immunofluorescence, and Imaging
Histology, immunofluorescence, including antibody dilutions, and imaging were conducted according to a previous report 12 . Fluorescent images were acquired by confocal laser scanning microscopy, and Z-stacking was performed with NIS-Element AR (Nikon, Tokyo, Japan).

Soft Agar Colony Formation Assay
Colony formation assays were performed using initial seeding of 0.5 and 1 × 10 3 cells for Y79 and 1 and 5 × 10 5 cells for RB170 mixed with 0.36% low-melting temperature agarose solution in growth medium in 6-well plates. The cell-agarose solution mixtures were plated on the solidified bottom layer of 0.75% low-melting temperature agarose in growth medium and grown at 37°C under 5% CO2 for 3 weeks. Green fluorescent protein (GFP)-positive colonies were visualized, and photographs were recorded using the Operretta high-content imaging system (Perkin Elmer, Waltham, MA, USA). In photographs, a group of GFPpositive colonies larger than 1 × 10 3 µm 2 (50 cells) was considered a colony, and the size and number of colonies were analyzed using Columbus Image Data Storage and Analysis System (Perkin Elmer). Colonies were stained with 0.05% crystal violet-40% methanol in PBS and photographed by Chemidoc Imaging System (BioRad Laboratories).

Cell Viability Assay and Drug Test
Cell proliferation analysis was performed by measurement of ATP production using the The data were analyzed using FlowJo software (BD Biosciences).

Live Cell Imaging
In total, 5 × 10 4 cells were seeded on a Hi-Q4 culture dish secured in a chamber maintained at 37°C under a humidified atmosphere of 5% CO2. Growth medium contained 5% Matrigel solution to facilitate cell adhesion. Cell were monitored using a Biostation IMQ (Nikon) equipped with a camera for video recording. Time-lapse video was conducted for 120 hours, and images were captured every 3 hours. Cell motility including velocity magnitude and speed was analyzed using the image analysis software CL-Quant Ver. 2.0 (Nikon).

RNA-sequencing
RNA was extracted from cultures by using TriPure isolation reagent (Roche Applied Science, Penzberg, Germany). The quality and quantity of RNA were determined by RNA6000 assay

Processing RNA-seq Reads:
Low-quality reads were purged using Trimmomatic (v0.36) 13 . Quantification of the transcriptome was carried out using Kallisto (v0.43) 14 using annotation from Ensembl version 84 at the transcript level. Gene level counts were generated using the tximport R package 15 , and gene count level data were normalized using the TMM (trimmed mean of Mvalues) method followed by CPM (counts per million) computation.

Differential Expression Analysis:
Differential expression analysis between distinct sample groups was performed using exactTest function as implemented in the edgeR R package 16 . Genes were selected based on a log2 fold-change value ≥1 and adjusted p-value ≤0.01.

Statistical Analysis
Pearson correlation test was conducted between MYCN copy number and expression of MYCN or MYCNOS1, and MYCN and MYCNOS1 expression. Student's t-test was used to test the difference in stability of MYCN protein and drug response following MYCNOS silencing. One-way ANOVA followed by Tukey's multiple comparison test was used to test the difference in expression of RNA and protein, cell cycle phase, caspase 3-positive cells, colony size and number, and migration following MYCNOS silencing. The F-test was used for analysis of proliferation curves. When p-value was < 0.05, the results were concluded to be statistically significant. Statistical analyses were conducted using R-packages.

Data Availability
Raw data of RNA-seq and whole genome sequencing data have been deposited in Gene Expression Omnibus (GEO) and Sequence Read Archive (SRA) through accession numbers GSE161449 and PRJNA678350, respectively.

Supplementary Video
Time-lapse video microscopy over the course of 120 hours with images captured every 3 hours.