The Y79 cells were purchased from the American Type Culture Collection (ATCC, Manassas, VA, USA) and Weri-RB1 cells were purchased from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China). Both cells were maintained in Roswell Park Memorial Institute (RPMI)-1640 medium (#12633012, Thermo Fisher, Waltham, MA, USA) containing 10% fetal bovine serum (FBS, #16140071, Thermo Fisher, Waltham, MA, USA) and antibiotics (#15640055, Thermo Fisher, Waltham, MA, USA). All the cells were identified by short tandem repeat profiling and validated for absence of mycoplasma.
Primers and antibodies
Primers used in this study were provided in Supplementary Table 1. The antibodies used in this study were listed in Supplementary Table 2.
Isolation of retinoblastoma stem-like cells
The Y79 and WERI-RB1 cells were first normally cultured. When the cells reached 80% confluence, the cells were washed and dissociated. The dissociated cells were then resuspended in stem cell culture medium (RPMI-1640 medium containing 20 ng ml−1 recombinant human epidermal growth factor (rhEGF, #PHG0314, Thermo Fisher, Waltham, MA, USA), 20 ng ml−1 recombinant human basic fibroblast growth factor (rh-bFGF, #PHG0261, Thermo Fisher, Waltham, MA, USA), 1× B-27 (#17504044, Thermo Fisher, Waltham, MA, USA), 5 μg ml−1 heparin (H0200000, Sigma-Aldrich, St. Louis, MO, USA), 2 mM L-glutamine (#21051040, Thermo Fisher, Waltham, MA, USA), 20 ng ml−1 recombinant human leukemia inhibitory factor (rh-LIF, # PHC9483, Thermo Fisher, Waltham, MA, USA), and antibiotics). The single-cell suspension was then prepared by repetitive pipetting. The cells were then diluted by stem cell medium to a concentration of 1 cell per 10 μl. Then, the cells were seeded into ultra-low attachment 96-well plate (#3474, Corning Inc., Corning, NY, USA) at the density of 1 cell per well (10 μl single-cell suspension solution per well), following by careful check under microscope. The cells were cultured for 25 days and the spheres in each well were checked under microscope. The cells with the strongest sphere-forming capacity were selected and cultured in ultra-low attachment dish for expansion. The stem-like properties (including sphere-forming capacity, the frequency of sphere-forming cells and the expression of stem cell makers) were subsequentially determined.
The cells were first cultured in normal culture medium for at least 24 hours. When the cells reached 80% confluence, the cells were washed and single-cell suspension solution was prepared by stem cell culture medium. The cells were then seeded into ultra-low attachment 6-well plate (#3471, Corning Inc., Corning, NY, USA) at the density of 3,000 cells per well. After 15 days of culture, the spheres in each well were counted under microscope.
Real-time quantitative reverse-transcription PCR (qRT-PCR).
Total RNA from indicated cells was isolated with Trizol reagent (#A33254, Thermo Fisher, Waltham, MA, USA). The ethanol was removed through evaporation. The residual DNA was removed by Turbo DNAse (#AM2239, Thermo Fisher, Waltham, MA, USA) according to manual, followed by DNAse inactivation. The mRNA concentration was determined by Nanodrop spectrophotometer (Thermo Fisher, Waltham, MA, USA). TaqMan RNA-to-Ct 1-Step Kit (A25605, Thermo Fisher, Waltham, MA, USA) was used for qRT-PCR assay. GAPDH was chosen as internal control. The relative mRNA levels were presented as ∆∆Ct. The details of primer are listed in supplementary materials.
The gene expression data (Tumor Retinoblastoma – Dorsman – 76 dataset) was downloaded from R2 bioinformatics website (https://hgserver1.amc.nl/cgi-bin/r2/main.cgi). The Spearman correlation analysis was used for determined the correlation between genes.
Isolation of CD44+ and ABCG2+ cells
The CD44+ and ABCG2+ cells were isolated by MACS system (magnetic activated cell sorting) according to manufacturer instruction. All reagents and products related to MACS were purchased from Miltenyi Biotec (Auburn, CA, USA). Briefly, 50 μl of cell suspension (107 cells) was mixed with 25 μl of blocking reagent and 25 μl of anti-CD44 (#130-095-194, Miltenyi Biotec, Auburn, CA, USA) or anti-ABCG2 (#130-10-7-680, Miltenyi Biotec, Auburn, CA, USA) microbeads, followed by incubation for 20 min at 4 °C. After centrifugation (1,000 rpm for 5 min), the supernatant with cells were loaded onto a MACS column, the CD44+ or ABCG2+ cells retained in the column and CD44- or ABCG2- cells that passed through column were harvested.
Lentivirus vectors pCDH-CMV-MCS-EF1-Puro and PLKO.1-puro were kindly provided by professor Hongbin Ji (Shanghai Institutes for Biological Sciences, Shanghai, China). For SOX2 overexpression, the coding sequence of SOX2 was inserted into pCDH-CMV-MCS-EF1-Puro plasmid. For SOX2, YAP and WWTR1 knockdown, shRNAs specifically against SOX2, YAP and WWTR1 were inserted into PLKO.1-puro plasmids. The shRNAs used were provided in Supplementary Table 3. The lentivirus were packaged by co-transfection with the reconstruction, VSV-G and delta R8.2 plasmids into HEK293T cells. After 24 hours, the medium containing virus particles was collected and used for infection. The stable cell lines were selected by puromycin and the efficiency of transfection was determined by western blot.
Western blot was performed according to standard protocol. Briefly, RIPA buffer (#89901, Thermo Fisher, Waltham, MA, USA) was used to prepare the cell lysis. After centrifugation, the protein samples were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and blotted onto Immobilon-P membranes (#IPVH00010, Millipore-Sigma, St. Louis, MO, USA), followed by incubation with the primary antibodies at 4 °C overnight. After incubation with corresponding horseradish peroxidase (HRP)-conjugated secondary antibody, the signals were visualized by incubation with Immun-Star HRP Substrate (#1705040, BioRad, Hercules, CA, USA). β-actin and Lamin B1 were used as internal control for total and nuclear protein, respectively. Uncropped gels were attached in Supplemented Figure 2.
the cells were harvested and resuspended in the medium containing 25 % Matrigel (A1413301, Thermo Fisher, Waltham, MA, USA), followed by inoculation into the flank of 6-8 weeks old female BALB/c nude mice (Pusheng Technology, Nanjing, China). The tumors were counted at 30-45 days after incubation. All the procedures were approved by Institutional Animal Care and Use Committee in the Affiliated People’s Hospital of Ningbo University, (Zhejiang, China).
Limiting dilution assay
The cells were first cultured in 6 cm cell culture dish. When the cell density reached 90% confluence, the cells were harvested and resuspended in sphere medium as mentioned above. The cells were then cultured in 96-well ultra-low attachment culture plate (#3474, Corning Inc., Corning, NY, USA) at a density of 20, 10, 5 cells / well. After 25 days culture, the number of wells with spheres was recorded. To determine the frequency of tumor-initiating cells in vivo, the cells were inoculated into the flank mice at the density of 50, 25, 5 cells per mice. The number of mice with tumor was counted at 45 days after inoculation. ELDA online software (http://bioinf.wehi.edu.au/software/elda/) was used to calculate the frequency of sphere-forming cells.
Chromatin immunoprecipitation quantitative PCR (ChIP-PCR)
EpiTect ChIP qPCR Assay kit (#334211, Qiagen, Valencia, CA, USA) was used for ChIP-PCR assay. Briefly, the cells were fixed for cross-linking, followed by isolation of chromatin. The chromatin DNA was subsequently fractionated to 0.8 to 1.0 kb by sonication, followed by incubation with SOX2 antibody-bound beads. The DNA samples on the bead were then reverse cross-linked and collected. After purification and quantification, the Levels of YAP and WWTR1 promoter were determined by qRT-PCR.
Luciferase reporter assay
To determine the transcriptional activity of YAP and WWTR1 promoters, the promoter fragments were cloned into PGL4.20 luciferase reporter plasmid (E6751, Promega, WI, USA). To analyze the transcriptional activity of YAP, an 8XGTIIC-luciferase reporter plasmid containing YAP/TAZ target sequence was used. The reconstructed plasmids and hRluc/TK control plasmids were co-transfected into the cells. The luciferase activity was determined by Dual-Luciferases Reporter Assay kit (E1901, Promega, WI, USA) according to the manual.
Nuclear run-on assay
The cells were cultured in 6-well plate containing complete culture medium to 70% confluence, followed incubation with medium containing 0.5 mM 5-ethynyl uridine (EU, E10345, Thermo Fisher, Waltham, MA, USA) for 1 h. The total RNA was subsequently isolated. After biotinylation with biotin azide, the labled RNA was purified by streptavidin coupled beads, followed by qRT-PCR examination.
DNA pull-down assay
The bait DNA fragments were labeled with Biotin using 3’ End DNA labeling Kit (#89818, Thermo Fisher, Waltham, MA, USA). The labeled DNA fragments were subsequently immobilized on the DynabeadsTM M-270 Streptavidin (#65306, Thermo Fisher, Waltham, MA, USA). The nuclear extracts from 2 × 107 cells were prepared, followed by incubation with bait DNA-coated beads in binding buffer. The DNA-protein complex bound by beads was then eluted and subjected to western blot examination.
Data were represented as ‘mean ± SD’ from three experiments except where indicated. The significance was determined by Student’s t-test and One-way ANOVA (*P < 0.05, **P < 0.01, ***P < 0.001).