Supplementary information
Essential role of METTL3 mediated m6A modification in glioma stem-like cells maintenance and radioresistance
Visvanathan et al., 2017
Page no.
Supplementary Figure legends2
Supplementary methods8
Supplementary Tables23
Supplementary Figure legends
Supplementary Figure 1
A. Expression of neural progenitor marker Nestin and astrocyte lineage specific marker GFAP (Glial fibrillary associated protein) in GB1 was showed by immunofluorescence. Scale bar=10µm. B. Stem cell surface marker SSEA1 (Stage specific embryonic antigen 1) and neural progenitor marker Nestin distribution in GB1 by flow cytometry. C. Expression of four reprogramming factors (iGSC factors) SOX2, SALL2, OLIG2 and POU3F2 in GB1 compared to DGC derived from GB1 was analyzed by Real time PCR and plotted as log2 ratio. D. Down regulation of METTL3 mRNA post pooled siRNA transfections in MGG4, MGG6, MGG8, and GB1 was measured by Real time PCR plotted as log2 ratio. E. Down regulation of METTL3 protein in MGG4, GB1 and MGG8 post 72 hours of pooled siRNA transfections by SDS-PAGE.F.The sphere formation efficiency post inhibition of METTL3 using pooled siRNA in spheres MGG4, MGG6, MGG8 and GB1.Representative images at 10x of the sphere lines after 7 days were showed for siNT and siMETTL3. The number of spheres formed after 7 days were counted using Image J software and the reduction in siMETTL3 condition compared tosiNT is depicted as percentage. Scale bar=100µm.G. Limited dilution assay with 10, 20, 50 and 100 cells of siNT and pooled siMETTL3 transfected MGG8. The number of wells with no sphere is plotted as percentage. H.Stem cell marker SSEA1 expression after METTL3 silencing of MGG8 cells was showed by flow cytometry. I and J. Expression of four iGSC factors were plotted in siMETTL3 condition in GB1 and in MGG8 respectively. RNA levels were analysed by Real time PCR and normalized to siNT condition. ddCt values were plotted as log2 ratio.
Supplementary Figure 2
A. Silencing of METTL3 expression measured by SDS-PAGE post 72hours of transfection with individual shRNA clones in MGG8 and GB1. B.Live cells post shMETTL3 silencing was measured by Alamar blue assay and plotted as percentage. shNT condition was depicted as 100%. C. Induction of apoptosis post METTL3 inhibition by pooled siRNA was showed by Annexin/PI positivity in GB1 by flow cytometry. Percentages of early and late apoptosis events were shown. D. Expression of four iGSC factors in GBM cell line derived LN229 spheres and in adherent cell line was determined by Real time PCR and plotted as log2 ratio. dCt values were normalized to cell line condition. E.Expression of four iGSC factors in GBM cell line derived U87 spheres and in adherent cell line was determined by Real time PCR and plotted as log2 ratio. dCt values were normalized to cell line condition.F. Expression of four iGSC factors in LN229 derived sphere cells post pooled siMETTL3 transfection compared to siNT was measures by Real time PCR and plotted as log2 ratio. Values are normalized to siNT
Supplementary Figure 3
A. Genome tracks depicting the METTL3 PAR-CLIP Seq peaks on SOX2 and absence of binding sites on OLIG2, POU3F2, SALL2 was showed (Liu et al., 2013). BCorrelation of SOX2 and METTL3 transcript is analyzed in. TCGA Agilent dataset. C. Correlation of SOX2 and METTL3 transcript is analyzed in. GSE22866 dataset.. D. Correlation of SOX2 and METTL3 transcript is analyzed in. GSE7696 dataset. E. Correlation of SOX2 and METTL3 transcript is analyzed in.Rembrandt dataset. F. Correlation of SOX2 and METTL3 transcript is analyzed in in-house cohort of GBM(n=46).G. Correlation of SOX2 and METTL3 protein level is analyzed in in-house GBM samples by IHC (n=25).H.SOX2 and METTL3 protein expression is analyzed in adjacent tumor sections by IHC from in-house GBM samples. Representative images of three cases are shown. I.Human SOX2 3’UTR depicting putative binding region near stop codon was (highlighted in green) identified by aligning Mouse Sox2 peaks (Meyer et al., 2013) reported in mouse brain (gene tracks provided in supplementary info). Consensus sequence GGACH required for METTL3 binding are depicted in bold underlined letters and three sites site#1, site #2 and site #3 locate within predicted binding region.
Supplementary Figure 4
A. Tertiary structure predicted for segment of 500bp from human SOX2 3’UTR by RNAcomposer for Wt, Mut A, Mut B, and Mut C. Three sites of modified A were denoted by site #1= blue (A8), site #2= red (A244) and site #3= magenta (A259) B. Tertiary structural alignment of Wt (yellow) and individual mutant (sky blue) by Chimera tool. Site A, site B and site C present in wt (wild type) or mutant (m) were shown by site #1= red (A8), site #2= green (A244) and site #3= magenta (A259). RMSD (root mean square deviation) in Å between two structures and pruned bases during structural alignment were mentioned below. C. Local secondary structure alterations caused within locality of 300bp by Mut A, Mut B and Mut C point mutations were analyzed by RNAsnp web tool. P-value difference for secondary structural alteration caused by point mutation in Mut A, Mut B and Mut C was calculated. Bars (blue and yellow) marked above the region show length of segment with maximum secondary structure changes within the full length 3’UTR of SOX2 (1.1kb) and position of mutation was showed as red mark. P value variation measured between WtVs each mutant was depicted according to color code mentioned below. Base pair probability matrix of Wt with mutants was shown for each Wt/mutant comparison. Upper and lower triangle of the matrix represents the base pair probabilities of wild-type (green) and mutant (red) respectively. Size of dots corresponds to affinity of base pairing at the position. Minimum free energy alteration (MFE) of each mutants Vs wild type were given.
Supplementary Figure 5
A. Scatter plot of HuR transcript levels across normal Vs GBM from different datasets (TCGA Agilent and TCGA Affymetrix) were showed in log2 ratio.Values are normalized to control brain B. HuR transcript level comparison between GSC Vs DGC RNA-Seq data1. The log2 value was normalized with the level in DGC. C.In vitro binding assay was performed using Nickel-NTA bead immobilized His-tag HuR protein and METTL3 over expressing LN229 cell lysate.Elutions were analysed for METTL3 by SDS-PAGE.D.Enrichment plot from GSEA analysis showing the running enrichment scores for DNA repair related geneset.Input genes are differentially expressed genespostMETTL3 silencing reported in GSE37001 and genes were pre-ranked according to regulation post METTL3 silencing. E. SOX2 ChIP-Seqgenes reported in MGG85was used as input geneset for GSEA and the enrichment plot depicting the running enrichment score for homologous recombination repair process. F.Schematic of DR-GFP/I-Sce system. p value was analyzed by Student’s T test. *** P value <0.001, ** p value <0.01, * p value <0.05. Error bars represents standard deviation.
Supplementary Figure 6
A. Scatter plot of METTL3 transcript levels in GBM compared to normal brain in different datasets TCGA Affymetrix, TCGA Agilent, GSE22866, and GSE7696 were represented as log2 ratio. B. METTL3 RNA levels in lower grades of glioma were compared to control brain samples and GBM derived from Rembrandt dataset. Values were showed as log2 ratio and normalized to control brain. C. METTL3 mRNA levels were validated from in-house GBM samples (n=17) and control samples (n=4) by Real time PCR. Values were plotted as log2 ratioand normalized to control brain. D. Representative images of IHC staining for METTL3 in GBM patient derived sections (n=40) and control samples (n=8) from in-house tissue microarray. GBM samples compared to non-tumor samples show intense cytoplasmic staining. Scale bar=20µmE. Scatter plot of METTL3 cytoplasmic raw staining intensity between control and GBM samples from in-house tissue microarrayand values are normalized to control brain F. METTL3 transcript levels were compared among 2 control samples, SVG (SV40 transformed astroglial cells), IHA (immortalized human astrocytes) and established GBM cell lines using Real time PCR. Values were plotted as ratioand normalized to control brain. G. Cell proliferation was measured by tryphan blue exclusion assay post 48, 72, 96 and 120 hours of siNT /siMETTL3 transfection in LN229 cells and plotted. H. Active caspase-3 levels were measured at 3 different time points post transfection using FITC-DEVD-FMK in LN229/ siNT and LN229/siMETTTL3 cells and plotted as median intensity at different time intervals. I.siNT and siMETTL3 transfected LN229 cells were treated with varying concentrations of Temozolomide. At 48 hours after drug addition, the proportion of live cells was quantified by MTT assay.Viable cells from untreated condition were considered as 1. P value was analyzed by Mann whitneyT test. *** P value <0.001, ** p value <0.01, * p value <0.05. Error bars represents standard deviation.
Supplementary Figure 7
Normalized RNA expression of methylases-METTL3, METTL14 and demethylases FTO, ALKBH5 inGBM samples were compared to control brain in different datasets was plotted as log2 ratio A. TCGA Agilent B. TCGA AffymetrixC. TCGA RNA-SeqD. Rembrandt E. GSC7696 F. GSE22866. P value was analyzed by Mann whitney T test. *** P value <0.001, ** p value <0.01, * p value <0.05.
Supplementary Figure 8
Normalized RNA expression of pan cancertissue samples (n=28) were compared to control tissues inTCGA RNA-Seq dataset was plotted as log2 ratio for methylases -A.METTL3, B. METTL14 and demethylasesC.FTO, D. ALKBH5.P value was analyzed by Mann whitney T test. Red stars indicate significant upregulation in respective cancers and green stars indicate significant down regulation in respective cancers. Cancers with no significant differences are left blank. *** P value <0.001, ** p value <0.01, * p value <0.05.
BLCA-Bladder urothelial carcinoma, BRCA-Breast invasive carcinoma, CESC-Cervical and endocervical cancers, CHOL-Cholangiocarcinoma, COAD-Colon adenocarcinoma, COADREAD-Colorectal adenocarcinoma, ESCA-Esophageal carcinoma, GBM-Glioblastoma multiforme, GBMLGG-Glioma, HNSC-Head and Neck squamous cell carcinoma, KICH-Kidney Chromophobe, KIPAN-Pan-kidney cohort (KICH+KIRC+KIRP), KIRC-Kidney renal clear cell carcinoma, KIRP-Kidney renal papillary cell carcinoma, LGG-Brain Lower Grade Glioma, LIHC-Liver hepatocellular carcinoma, LUAD-Lung adenocarcinoma, LUSC-Lung squamous cell carcinoma, PAAD-Pancreatic adenocarcinoma, PCPG-Pheochromocytoma and Paraganglioma, PRAD-Prostate adenocarcinoma, READ-Rectum adenocarcinoma, SARC-Sarcoma, STAD-Stomach adenocarcinoma, STES-Stomach and Esophageal carcinoma, THCA-Thyroid carcinoma, THYM-Thymoma, UCEC-Uterine Corpus Endometrial Carcinoma
Supplementary methods
Tumor samples
Tumor samples were collected from National Institute of Mental Health and Neurosciences (NIMHANS) and Sri SatyaSai Institute of Higher Medical Sciences (SSSIHMS), Bangalore, India (A cohort of 17 were utilized for RT-PCR and cohort of 40 were utilized for IHC). Sample of the anterior temporal cortex resected from epilepsy patients served as control brain sample (n=12). Samples were collected according to patients’ written consent and experiments were carried out according to the Institute ethics guidelines.
Cell lines:
The GBM cell lines, LN229, U251, U87, A172 and U373, were obtained from Sigma Aldrich, Saint Louis, Missouri, USA. The immortalized human astrocyte cell lines, NHA-hTERT-E6/E7 and SVG were obtained from Dr. Russell Pieper’s laboratory, University of California, San Francisco and Dr. AbhijitGuha’s laboratory, University of Toronto, Canada respectively. The immortalized mouse astrocytic cell line, IMA2.1, was a kind gift from Dr. Stefan Schildknecht, University of Konstanz, Germany. All cells were cultured in Dulbecco’s Modified Eagles’ Medium (DMEM) containing 10% Fetal Bovine Serum (FBS) and grown at 37°C in 5% CO2.
Establishing neurosphere lines from primary GBM tissue
Briefly the freshly resected tumor collected from operating room was transferred in HEM (HEPES buffered MEM) and trypsinized using 0.05%Trypsin-EDTA. The suspension was passed through 40µM filter (Catalog #BD352340, BD Falcon, U.S.A). The cells were pelleted down and re-suspended in Neurobasal medium at 50000 cells/ml density.
Differentiation of GSCs
For differentiation of GSC, neurospheres were chemically triturated and plated onto 10 % poly-L-lysine (Catalog #P8920, Sigma, U.S.A) coated plates in DMEM medium supplemented with 10% serum for 14-21 days. A half-volume medium change was performed every 3 days.
Transfection of GSCs, neurosphere assay and sphere diameter measurement
GSCs transfections were carried out in single cell suspension state for both siRNA and plasmids. 100 nMconcentration of either control non-targeting siRNA (siNT) or gene-specific siRNA (On-target plus SMART Pool, Dharmacon, U.K), as indicated, was transfected using Dharmafect I (Dharmacon, U.K) according to the manufacturer's instructions. Control vector or shNT with gene specific plasmids were tranfected using Lipofectamine 2000 (Life Technologies, U.S.A) according to the manufacturer's instructions. Post 48 hours of transfection, cells were harvested and confirmed for gene manipulation by qRT-PCR.After 48 hours of transfection, the aggregates formed were dissociated into single cells, counted and equal numbers of cells were plated at a density of 2 cells/µl in 24-well or in 6-well plate. Number of spheres was counted after 7 days of plating. Fresh medium was replenished every 2-3 days. Sphere diameter measurements were analysed with ImageJ software. Number of spheres above 50µM diameter were counted and plotted for total number of spheres.
RNA stability assay
Actinomycin D (Catalog #A9415, Sigma, U.S.A) at 5ug/ml was added to GSCs 24 hours after plating. After 2, 4, and 6 hours incubation, cells were collected and RNA was isolated for Real time PCR. The half life (t1∕2) of mRNA was calculated using ln2∕ slope and GAPDH was used for normalization.
Viability assay in neurospheres
1000 Cells were plated in 96 well plate (100µl volume) and post 72 hours of transfection, 10µl of Alamar blue (Invitrogen) was added. After 12 hours of incubation at 37ºC fluorescence readings were recorded in Tecan Infinite 200 florimeter.
Radiation of spheres
GSCs were transfected with METTL3 shRNA or NT shRNA and after 48 hours single cell suspension was treated with gamma irradiation (indicated doses) at a dosage rate of 2.06Gy/min in BI2000 cabinet (BRIT) with Co60 (675 Ci) source capacity. Cells were replated at equal number and allowed to recover for 5 days in 24‐well plates. For ᵧ-H2AX western blot analysis, we have incubated the cells for 4 hours post irradiation. Sphere formation assays were scored after 7 days of replating.
Plasmids and siRNAs
The over expression constructs used were METTL3 (#RC200869, Origene, U.S.A), pSIN-EF2-SOX2 3’UTR (gift from Dr. James Thomson #16577, Addgene, U.S.A) and 6x His-pET22-NH-HuR (gifted by Dr. C. Durga Rao, Indian Institute of Science, India). shRNA for METTL3 (TRCN0000289814, TRCN0000289743, TRCN0000289812, TRCN0000289813, TRCN0000289742) and shHuR (TRCN0000017277) were from MISSION shRNA TRC library for whole genome. SOX2 3’UTR reporter luciferase construct was a kind gift from Dr.Kenneth.S. Kosik, (UCSB, U.S.A). pDR GFP, pCBA-SceI was a gift from Maria Jasin (Addgene plasmid (#26475,#26477).Control siRNA and siRNA used against METTL3, SOX2 were purchased from GE HealthCare DharmaconInc (On-TARGET plus Human siRNA SMART pool).
RNA isolation and real-time Quantitative RT-PCR analysis
Total RNA was isolated using TRI reagent (Sigma, U.S.A.) and 2 µg of RNA was reverse transcribed with the High capacity cDNA reverse transcription kit (Life technologies, USA) according to the manufacturer’s protocol. qRT-PCR was performed using the ABI PRISM 7900 HT Sequence Detection System (Life technologies, USA). Expression of the genes of interest was analyzed using ATP5G rRNA and GAPDH as internal control genes and the ΔΔCt method. Real time primer information is provided as table.
FP-METTL3 / ACCTATGCTGACCATTACCAAGRP-METTL3 / CTGTTGGTTCAGAAGGCTCTC
FP-ELAVL1 / ATG AAG ACC ACA TGG CCG AAG ACT
RP-ELAVL1 / AGT TCA CAA AGC CAT AGC CCA AGC
FP-SOX2 / AACCCCAAGATGCACAACTC
RP-SOX2 / GCTTAGCCTCGTCGATGAAC
FP-OLIG2 / CCAGAGCCCGATGACCTTTT
RP-OLIG2 / AGGACGACTTGAAGCCACTG
FP-POU3F2 / TGACGATCTCCACGCAGTAG
RP-POU3F2 / GGCAGAAAGCTGTCCAAGTC
FP-SALL2 / TAATCTCGGACTGCGAAGGT
RP-SALL2 / TAGAACATGCGTTCTGGTGG
FP-ATP5G / CCAGACGGGAGTTCCAGAC
RP-ATP5G / GACGGGTTCCTGGCATAGC
FP-GAPDH / TTGTCAAGCTCATTTCCTGG
RP-GAPDH / TGATGGTACATGACAAGGTGC
FP-SOX1 / CAGTACAGCCCCATCTCCAAC
RP-SOX1 / GCGGGCAAGTACATGCTGA
FP-SOX5 / GAACAACAGGTGCTTGATGGG
RP-SOX5 / GCCCTCGGGATTCCCTATAAAT
FP-SOX21 / GCTCGCCAATCCCGAGAAG
RP-SOX21 / ATCTCTGCCATTTTGGAGCCC
FP-RFX4 / CCCGGTCCAAACTCGGAAC
RP-RFX4 / TGGCTCTTATTACAGTGTCCAGT
FP-KLF15 / TTCTCGTCGCCAAAATGCC
RP-KLF15 / CCTGGGACAATAGGAAGTCCAA
FP-POU3F3 / GCGGCTTCTAACCCCTACC
RP-POU3F3 / CCCCTGCATGAAGTCGCTC
FP-HES6 / GGCCGTGAGGATGAGGACGG
RP-HES6 / GCCAGCAGCAGCCGCA
FP-HEY2 / CCTAACAGAAGTTGCGCGGTA
RP-HEY2 / GAGGCGACAAGGGGTTGAC
FP-SOX8 / GATCCGGGGTGTCTGTCCGC
RP-SOX8 / TCATCACAGCCAGCCCTCGC
FP-ASCL1 / AAGCTCTGCCAAGATGGAGA
RP-ASCL1 / GGCAAAGAAACAGGCTGC
FP-OLIG1 / CGGACGCCAAGGAGGAGC
RP-OLIG1 / TATCTTGGAGAGCTTGCGGCC
Western blot
For western blot analysis, RIPA buffer was used to isolate protein from the neurosphere lines and tissues were quantified by Bradford’s reagent. The following antibodies have been used in these studies: Anti-METTL3 (Catalog #A301-567A, Bethyl laboratories, U.S.A, Catalog #H00056339-B01P, Abnova, Taiwan, ), Anti-SOX2 (Catalog #3579, Cell signaling, U.S.A), Anti-ᵧ-H2AX (Ser 139) (Catalog #9718, Cell signaling, U.S.A), Anti-Actin (Catalog #A5316, Sigma, U.S.A), Anti- Tubulin (Catalog #CP06, Calbiochem, U.S.A)
m6A RNA immuno precipitation
5µl rabbit anti-m6A antibody (Catalog #E1610S NEB, U.S.A) was coupled to Protein Aagaroase beads (Sigma, U.S.A) in 200 µl of 1 M IP Buffer (1 M NaCl, 10 mM sodium phosphate, 0.05% Triton-X) for 4 hours at 4°C. Beads were then washed 3 times in 200 µl of 140 mM IP Buffer (140 mMNaCl, 10 mM sodium phosphate, 0.05% Triton-X). RNA was denatured by keeping at 75°C for 5 mins, cooled on ice for 2-3 min, and bound to antibody-coupled beads in 200 µl of 140 mM IP Buffer (O/N at 4°C). Beads were treated with 100 µl Elution Buffer (5 mMTris-HCL pH 7.5, 1 mM EDTA pH 8.0, 0.05% SDS, 4.2 µl Proteinase K (20 mg/ml) for 2 hours at 50°C, and RNA was recovered with phenol:chloroform extraction followed by ethanol precipitation.
RNA immuno precipitation
RIP for anti HuR/ DDK tag was performed using 4ul of Anti-HuR (Catalog #H00001994-D01, Abnova, Taiwan) or anti-DDK tag (Catalog #2368, Cell signaling,U.S.A) as described earlier (Keene, J. et at., 2006 Nature Protocols).
m6ANorthwestern blot
RNA samples were quantified using gel and equal amount of RNA were spotted onto a nylon membrane (GE Healthcare, U.K). The membrane was then UV cross-linked and blocked for 1 h in 5% non-fat dry milk in 0.1% PBST (0.1% Tween-20 in 1x PBS, pH 7.4). Rabbit anti-m6A antibody (Catalog #E1610S, NEB, U.S.A) or was diluted 1:1000 in 0.1% PBST and incubated on the membranes for 12hrs (4°C). Following extensive washing with 0.1% PBST, HRP-conjugated anti-rabbit IgG was diluted 1:5000 in Blocking Buffer and added to the membranes for 1 h at 25°C. Membranes were washed again in 0.1% PBST and developed with enhanced chemiluminescence (ECL, GE Healthcare, U.K).
In vitro transcription
RNAs were transcribed in vitro from different linearized plasmid constructs under T7 in runoff transcription reactions. Luciferase SOX2 3’UTR vector was linearized with XhoI. The linear DNA samples were electrophoresed on agarose gels, extracted using a gel elution kit Catalog #28704(Qiagen, U.S.A), and used as was used as templates for synthesis of unlabeled or biotin labeled probes. The reaction was carried out using RiboMAX™ Large Scale RNA Production System (T7) (Catalog #P1300, Promega, U.S.A) and biotin-11-UTP (Catalog #AM8451, Invitrogen, U.S.A) as per manufacturer’s instructions. Briefly 2.5 µg of linear template DNA was added and the reaction was incubated at 37◦C for 1.5 hours. After alcohol precipitation, the RNA was resuspended in 20 µl of nuclease-free water. Primers used for amplifying the region from SOX2-3’UTR luciferase construct with T7 promoter overhang are provided as table.