METTL4-mediated nuclear N6-deoxyadenosine methylation promotes metastasis through activating multiple metastasis-inducing targets

Cell culture, oxygen deprivation, and cellular stress conditions

The human head and neck cancer (FADU) and renal pelvis transitional cancer (BFTC909) cell lines were obtained from the Bioresource Collection and Research Center (BCRC, Hsinchu, Taiwan). The human lung cancer (H1299) cell line was obtained from ATCC. The embryonic kidney 293T (HEK293T) cell line was described [38]. KTCC28M, a primary cell line generated from human upper tract urothelial carcinoma cells, was obtained from Dr. See-Tong Pang (Chang Gung Memorial Hospital at Linkou, Taoyuan, Taiwan). Use of KTCC28M cells was approved by the Institutional Review Board of Chang Gung Memorial Hospital (201303642A3C501). Cell lines were cultured in Dulbecco’s modified Eagle’s medium (DMEM) or RPMI medium supplemented with 10% FBS and 1% PS. Culture conditions were maintained at 37°C in a humidified incubator containing 5% CO2. The cell lines showed negative results for mycoplasma contamination by mycoplasma DNA PCR and ultra-performance liquid chromatography with electrospray ionization tandem mass spectrometry (UPLC-ESI–MS/MS) assays. Oxygen deprivation was carried out by incubating the cells under 1% O2, 5% CO2, and 94% N2 for 18 h. For serum starvation, cells were washed twice with sterile phosphate-buffered saline (PBS) and then incubated in DMEM with 1% FBS for 24 h. For heat shock treatment, tissue-culture plates with growing cells were incubated in a water bath at 42°C for 2 h. To induce oxidative stress, cells were treated with 500 μM hydrogen peroxide at 37°C and 5% CO2 for 12 h. To induce acidosis, cells were treated with DMEM (pH 6.4) with 10% FBS at 37°C and 5% CO2 for 6 h. Actinomycin D (2 μg/ml) in dimethyl sulfoxide was used to study transcriptional regulation.

Plasmids and stable transfection

The pHA-HIF-1α, pHA-HIF-1α (ΔODD), and pHA-HIF-1α (LCLL) expression constructs were obtained from Dr. L.E. Huang (University of Utah, USA) [47]. The Flag epitope-tagged human METTL4 cDNA cloned in pcDNA3 vector (pcDNA3-Flag-METTL4) was obtained from Prof. Chuan He (University of Chicago, USA). The Flag epitope-tagged Drosophila Jumu cDNA cloned in pEF vector (pEF-Flag-Jumu) was obtained from Dr. Dahua Chen (Chinese Academy of Sciences, Beijing, China) [45]. The human lncRNA RP11-390F4.3 sequence was synthesized by Biotools (Taiwan) and cloned into pcDNA3.1 (+) vector (Life Technologies). The human lncRNA NEAT1 expression vector was purchased from Addgene (USA). The oligonucleotides used to generate various plasmids and the methods to construct these plasmids were shown (Additional file 2: Tables S7 and S8). The generation of stable clones by calcium phosphate transfection or Lipofectamine was described [38, 48]. All the stable clones were established by transfection of plasmids as designated by the name of the clones.

Establishment of METTL4 and RP11-390F4.3 knock-in cell lines

The knock-in cell lines were generated based on the Prime-Editing system [39] to make mutations at the catalytic site (287–290, DPPW to APAW) of METTL4 and the 6mA modification site on the RP11-390F4.3 promoter (−231~−224, ACTTCCGA to CCGGCCGA). The all-in-one PE3 vector, pPE2.pPuro, was obtained from the National RNAi Core Facility (Academia Sinica, Taiwan). Briefly, the PE3 gRNA and pegRNA targeting METTL4 or RP11-390F4.3, and RT-PBS containing the mutations were cloned into pPE2.pPuro to construct the pLAS-PE3-METTL4 and pLAS-PE3-RP11-390F4.3 plasmids (Additional file 2: Table S9). The plasmids were transfected respectively into FADU or BFTC909 cells and treated with 1.5 μg/ml puromycin for 3 days. Pure lines established from single cells were selected, and the genotypes were determined by PCR and DNA sequencing (the primers used are listed in Additional file 2: Table S9).

Protein extraction, co-immunoprecipitation, Western blot analysis, RNA extraction, and quantitative real-time PCR

The extraction of total proteins from cells, co-immunoprecipitation assay, and Western blot analyses were performed as previously described [38, 48, 49]. The NE-PER™ Nuclear and Cytoplasmic Extraction Kit (Thermo Fisher Scientific) was used to extract the cytoplasmic and nuclear proteins. The characteristics of the antibodies used are listed in Additional file 2: Table S10. Total RNA purification, cDNA synthesis, and quantitative real-time PCR were performed as previously described [49]. The cytoplasmic/nuclear RNAs were purified from FADU or BFTC909 cells using the SurePrep Nuclear or Cytoplasmic RNA Purification Kit (Thermo Fisher Scientific) in accordance with the manufacturer’s protocol. The sequences of primers used in the real-time PCR experiment are listed in Additional file 2: Table S11.

Immunofluorescence staining

Immunofluorescence staining followed by the acquisition of fluorescence images using confocal microscopy was performed as described previously [38, 49]. For METTL4 immunostaining, FADU, BFTC909, and KTCC28M cells were washed and fixed with 4% paraformaldehyde for 15 min at room temperature. Fixed cells were washed with PBS and permeabilized with PBS containing 0.1% Triton X-100 for 5 min at room temperature. Thereafter, cells were washed with PBS, blocked with 1% BSA in TBST buffer for 1 h, and incubated with anti-METTL4 antibody (1:500 dilution; HPA) at 4 °C overnight. Next day, cells were washed with PBS and incubated with the secondary antibody (1:500 dilution; goat anti-rabbit Alexa488, Abcam) for 1 h at room temperature followed by DAPI staining (1:10,000 dilution; Invitrogen). For 6mA immunostaining, the procedure was performed with minor modifications [11]. Briefly, the cells were stained with 100–200 nM MitoTracker Deep Red FM (Thermo Fisher) in HEPES buffer at 37°C for 30 min. Permeabilized cells were washed with PBS and then blocked for 1 h in blocking buffer (1% BSA in PBS containing 0.1% Tween, 40 μg/ml RNase A for RNA digestion, and 80 μg/ml DNase for DNA digestion). Cells were then incubated with 6mA antibody (1:500 dilution; Synaptic Systems) overnight at 4 °C. Next day, cells were washed with TBST three times and incubated with a secondary antibody (1:500 dilution; goat anti-rabbit Abberior STAR635). For double staining of mouse tissues, the tissue sections were incubated with anti-METTL4 (1:200 dilution; Abnova), anti-6mA (1:500 dilution; Synaptic Systems), or anti-HIF-1α (1:200 dilution; Abcam) antibodies. Finally, cells and tissues were stained with DAPI (1:10,000 dilution; Invitrogen) in PBS for 1 min at room temperature. The images were acquired using ImageXpress Micro Confocal System (Molecular Devices) and Leica TCS SP8 STED microscope. The images were processed using MetaXpress High-Content Image Acquisition and Analysis Software (MetaXpress version 6.5.4.532). Data analyses are conducted using MetaMorph software version 7.8.0.0 (Universal Imaging, USA) and Leica Application Suite X (LAS X version 3.5.5.19976) software. The cell counts from three independent replicates (n=150 cells) were pooled for further analysis. The hydrochloric acid antigen retrieval procedure was performed as described previously [7]. The antibodies used in immunofluorescence staining are listed (Additional file 2: Table S10).

In vivo tumorigenicity assay

All animal experiment protocols were performed with the approval of the Institutional Animal Care and Use Committee of China Medical University in this study. Five-week-old BALB/c nu/nu mice were purchased from National Science Council Animal Center (Taipei, Taiwan). Briefly, 2 × 106 viable cells were subcutaneously injected into the hind limbs of these BALB/c nu/nu mice, and then, the mice were euthanized after implantation at 30–35 days, and tumor incidence was monitored. Each experimental group contained at least five mice.

In vitro migration/invasion and in vivo tail vein/orthotopic metastatic assays

The in vitro migration/invasion assays and in vivo tail vein/orthotopic metastatic assays were performed as per methods described previously [38, 49, 50]. Cells were injected into 6-week-old male, non-obese diabetic/severe combined immunodeficiency mice (NOD-SCID mice, National Science Council Animal Center, Taipei, Taiwan) through the tail vein (1 × 106 cells) or orthotopic injection (1 × 105 cells). The incidence of lung metastatic nodules was monitored 8–12 weeks after injection. Each experimental group contained at least five mice. All animal experiment protocols were performed with the approval of the Institutional Animal Care and Use Committee of China Medical University in this study.

Fluorescence-activated cell sorting analysis

The procedure was performed as described previously with minor modifications [51, 52]. Briefly, prior to sacrifice and tumor excision, tumor-bearing mice (BALB/c nu/nu) with tumor volumes of ~350–500 mm3 were administered with 100 mg/kg pimonidazole (Hypoxyprobe™ Green Kit, HPI) intraperitoneally for 3 h and 1 mg/mouse Hoechst 33342 (Sigma-Aldrich) intravenously for 30 min. Tumor tissue was harvested and minced in PBS containing collagenase type I (Gibco Invitrogen Corp.) and dispase II (Gibco Invitrogen Corp.) and incubated for 40 min to prepare single-cell suspensions using gentle MACS Dissociator (Miltenyi Biotec). The cell suspension was filtered through a 70-μm sterile nylon mesh filter. Next, red blood cells were lysed with RBC lysis buffer (Invitrogen). After several washes, cells were stained with FITC-conjugated anti-pimonidazole (Hypoxyprobe™ Green Kit, HPI) in sorting buffer (PBS containing 0.5% BSA) following the manufacturer’s instructions. Cells were sorted using a FACSCalibur instrument (BD Bioscience). The hypoxic cell population was defined as Hoechst 33342+ and pimonidazole+ cells, and the normoxic cell population was defined as Hoechst 33342+ and pimonidazole− cells.

Transient transfection, luciferase assays, and cell proliferation assays

The METTL4 and RP11-390F4.3 promoters were cloned; the reporter constructs are shown (Additional file 2: Tables S7 and S8). The ZMIZ1 promoter (from −500 to −1 bp upstream of TSS) was inserted in front of the luciferase gene in pGL3-Basic vector to construct the ZMIZ1 promoter-driven reporter plasmid (ZMIZ1-Luc). To construct an intron-containing luciferase reporter plasmid, the region from 24,990 to 25,469 bp of the ZMIZ1 intron 1 that has the putative 6mA site (GGCACAAGCC) was inserted into the nucleotide 834 position of the firefly luciferase gene of ZMIZ1-Luc construct described above to mimic the artificial luciferase reporter containing an intron as described [53]. Under this construction, the new construct will contain the inserted ZMIZ1 intron 1 sequence that has GT-AG sequence located at the splicing junction of artificial intron inside the luciferase open reading frame (named as ZMIC1-Luc-intron1-WT). Two other plasmids, ZMIZ1-Luc-intron1-Del (deletion of the putative 6mA site) and ZMIZ1-Luc-intron1-mut (mutation of GGCACAAGCC to GGCCCCCGCG), were generated to test the ability of the 6mA site to increase luciferase activity after methylation by METTL4. The reporter constructs are shown in Additional file 2: Tables S7 and S8. The reporter constructs were co-transfected into FADU cells with different expression vectors. A pRL-TK plasmid was used as an internal control. Luciferase activities were measured using the Dual-Luciferase™ reporter assay kit (Promega, Madison, WI) and further normalized with Renilla activity for transfection efficiency [54]. To prepare the in vitro 6mA pre-methylated reporter construct, the RP11-390F4.3 promoter-driven reporter construct was prepared using ECOS™ 2163 competent cells (Escherichia coli K12 strain), and then incubated with METTL4 using in vitro DNA methylation assay as described below. Thereafter, the in vitro 6mA pre-methylated reporter construct was purified and the 6mA modification was confirmed via UPLC-ESI-MS/MS before it was co-transfected with other plasmids into the cells for luciferase assays. Cell proliferation assay were performed in 96-well plate (3.5 × 103 cells/well) as described previously [11]. The proliferation of transfected cells at 24, 48, or 72 h was assessed using CellTiter 96 AQueous One Solution Reagent (Promega) according to the manufacturer’s instructions. The absorbance was measured at 490 nm using a microplate ELISA reader (TECAN Infinite 200).

Lentivirus shRNA experiments

Biotools (Taiwan) designed the oligonucleotides for short hairpin RNA (shRNA) targeting RP11-390F4.3, which was cloned into a pLV2-U6-Puro vector. The sequence and target of RP11-390F4.3 shRNAs (#1 and #2) are listed in Additional file 2: Table S12. shRNA-expressing lentiviral vectors (pLKO.1-puro or pLV2-U6-Puro) were generated in HEK293T cells as previously described [55]. The sequences and clonal names of plasmids against METTL4, RP11-390F4.3, HIF-1α, HIF-1β, HIF-2α, ZMIZ1, KPNA7, N6AMT1, and U2 snRNA are listed in Additional file 2: Table S12. These plasmids and the packaging plasmid pCMVΔR8.91 were provided by the National RNAi Core Facility of Academia Sinica (Taipei, Taiwan). For lentivirus production, HEK293T cells were transfected with 5 μg of lentiviral vectors expressing individual shRNA along with 5 μg of the packaging plasmid pCMVΔR8.91 and 0.5 μg of the envelope plasmid pMD.G. The viruses were collected 48 h after transfection. To prepare METTL4, RP11-390F4.3, HIF-1α, HIF-1β, HIF-2α, ZMIZ1, KPNA7, N6AMT1, and U2 snRNA knockdown cells, cells were infected with the corresponding lentiviruses for 24 h, and the stable clones were then generated by selection with appropriate antibiotics.

Genomic and mitochondria DNA extraction

Briefly, harvested cells (1 × 107) were washed with PBS and suspended in ice-cold pre-lysis buffer (20 mM HEPES [pH 7.4], 10 mM KCl, 2 mM MgCl2, 1 mM EDTA, and 1 mM EGTA). After incubation on ice for 15 min, the cells were ruptured with a 27-gauge needle (BD Precision Glide needle) 15 times and then allowed to stand on ice for 20 min. After centrifugation at 720×g for 5 min at room temperature, pellets were collected and genomic DNA was extracted using the EasyPrep Genomic DNA Extraction Kit (Tools, Taiwan) according to the manufacturer’s protocol. Genomic DNA concentrations were measured using a fluorometer (Qubit® 3.0; Thermo Fisher Scientific). The efficiency of removing mitochondrial DNA from the genomic DNA extraction step was quantified by quantitative real-time PCR analysis using mitochondrial and nuclear-specific primers [11, 56]. The primers used are listed in Additional file 2: Table S11. Mitochondrial DNA extraction was performed as previously described [11, 56]. The quality of the extraction was assessed by comparing the fold change of mitochondrial DNA versus genomic DNA using the ΔCt method.

6mA dot blot assays

The 6mA dot blot assay was performed as described previously, with minor modifications [4, 5]. Briefly, the DNA was diluted using nuclease-free H2O to a final concentration of 150 ng/μl, denatured at 95°C for 10 min, and then loaded onto Hybond N+ membranes (GE Healthcare Life Sciences). The membranes were air-dried, baked at 78°C for 25 min, and then blocked in 5% non-fat dry milk in PBS containing 0.1% Tween 20 (PBST) at 37°C for at least 1 h. Thereafter, the membranes were washed with PBST, probed with anti-6mA antibody (Synaptic Systems), probed with the secondary antibody, and developed. Methylene blue staining was used as the loading control.

Quantification of 6mA in gDNA via ultra-performance liquid chromatography with electrospray ionization tandem mass spectrometry (UPLC-ESI–MS/MS)

The procedure was performed as described with minor modifications [3]. Three hundred nanograms of DNA was denatured in 20 μl of nuclease-free H2O at 95°C for 10 min, rapidly incubated on ice, and digested with 2 U nuclease P1 (Wako USA, 145-08221) in 10 mM ammonium acetate [pH 5.3] at 42°C. After 18 h, 2 μl of phosphodiesterase I (Sigma-Aldrich, P3243-1VL) and 3.4 μl of 1 M NH4HCO3 were added to the digested DNAs, and digestion was continued at 37°C for 4 h. Thereafter, the digested products were treated with 2 U of alkaline phosphatase (Sigma, P5931-500UN) at 37°C for at least 4 h and diluted 1:2 with nuclease-free H2O followed by filtering through a 0.22-μm filter (Millipore, SLGVR04NL). The resulting deoxyribonucleoside samples were analyzed via UPLC-ESI-MS/MS. Deoxyribonucleoside standards (Carbosynth), deoxyadenosine (dA), and N6-deoxyadenosine (6mA) were used for constructing daughter ion scan spectra to identify digested DNA products and also for quantitative calculation by multiple reaction monitoring (MRM). Separation of nucleosides was carried out by reverse phase UPLC on an Acquity UPLC® BEH C18 1.7 μm column, 2.1 mm × 50 mm (Waters, Milford, USA) at 30°C. Mass spectra and chromatograms were acquired in positive ESI mode, using Waters Xevo TQ-XS mass spectrometer. The nucleosides in digested DNAs were determined by LC retention plus MS/MS spectra and quantified based on the ion mass transitions; 6mA (m/z): 266 to 150 and dA (m/z): 252 to 136 in MRM mode. The 6mA/dA ratio of each sample was calculated with quantified 6mA and dA values in accordance with the calibration curves obtained from dA and 6mA standards. The representative chromatogram as quantified by their integrated area in the corresponding chromatogram was shown as peaks for samples tested.

In vitro DNA and RNA methylation assays

Mammalian HEK293T cells were transfected with either pcDNA3-Flag-METTL4 WT or Flag-METTL4 MUT plasmid, and then the overexpressed proteins were purified through immunoprecipitation with anti-FLAG M2 affinity gel (Sigma-Aldrich) according to the manufacturer’s protocol. Enzyme activity of METTL4 was analyzed as described previously with slight modifications [57]. Briefly, in a 50 μl total reaction, 500 ng of DNA substrate was incubated with METTL4 (50–100 μM) which was prepared via immunoprecipitation. The reaction buffer comprised 1.5 mM S-adenosyl-L-methionine (SAM), 80 mM KCl, 1.5 mM MgCl2, 2 mM CaCl2, 10 mM DTT, 4% glycerol, and 15 mM HEPES [pH 7.9]. After incubating the reaction mixture at 16°C for 18 h, the reaction was terminated by heat inactivation at 65°C for 5 min. Thereafter, DNA was cleaned, purified, and the 6mA modification was analyzed via UPLC-ESI-MS/MS, as described above. The 6mA/dA ratio of each sample was normalized with the untreated probe alone. The probe sequences used for the in vitro DNA methylation assay are listed in Additional file 2: Table S13. For the in vitro RNA methylation assay, U2 snRNA and RP11-390F4.3 RNA molecules were generated with an in vitro transcription kit (Thermo Fisher Scientific) using U2 snRNA and RP11-390F4.3 expression constructs, respectively, which were then purified using the RNeasy Mini Kit (Qiagen, Valencia, CA). Each RNA substrate (300 ng) was incubated with METTL4 (50–100 μM), which was purified via immunoprecipitation, in a 50 μl total reaction with reaction buffer containing 1.5 mM SAM, 80 mM KCl, 1.5 mM MgCl2, 0.2 U/μl SUPERaseIn, and 15 mM HEPES [pH 7.9]. After incubation at 30°C for 2 h, the reaction was terminated by heating at 65°C for 5 min. The RNA was then purified with the RNeasy Mini Kit (Qiagen), and the m6Am modification was analyzed via UPLC-ESI-MS/MS as described below.

Quantification of m6Am in U2 snRNA and RP11-390F4.3 via UPLC-ESI–MS/MS

Isolation of U2 snRNA or RP11-390F4.3 from cells for UPLC-ESI-MS/MS analysis was performed as previously described [58]. Briefly, the synthetic biotin-labeled probes against specific RNAs were incubated with total RNA and then purified using streptavidin T1 beads. The biotin-labeled probes used are listed in Additional file 2: Table S14. Five hundred nanograms of U2 snRNA or RP11-390F4.3 were treated with 2 units of nuclease P1 (Sigma-Aldrich) at 37°C in 50 μl of reaction buffer (20 mM sodium acetate [pH 5.3], 5 mM ZnCl2, and 50 mM NaCl). After 4 h, 10 units of antarctic phosphatase (NEB) were added, and the reaction continued for another 4 h. The resulting nucleosides were analyzed by UPLC-ESI-MS/MS with the same settings used to determine 6mA. Nucleoside standards (Toronto Research Chemicals), adenosine, and N6,2′-O-dimethyladenosine (m6Am) were used to construct daughter ion scan spectra to identify digested RNA products and for quantitative calculation by multiple reaction monitoring (MRM). The nucleosides of the digested RNAs were identified using LC retention with MS/MS spectra and quantified based on the ion mass transitions; m6Am (m/z): 296 to 150 and adenosine (m/z): 268 to 136 in MRM mode.

Chromatin immunoprecipitation (ChIP), quantitative chromatin immunoprecipitation (qChIP) and methylated DNA immunoprecipitation (MeDIP) assays

Herein, ChIP and qChIP were performed as described previously [49]. Briefly, the cells were crosslinked, sonicated, and then immunoprecipitated with the antibody. For the re-ChIP assay, after the first ChIP reaction, the DNA–protein complexes were washed and eluted by incubation at 37°C with 25 μl of 15 mM dithiothreitol (DTT) for 30 min. After centrifugation, the supernatant was collected and subsequently diluted with nuclei lysis buffer and subjected to the ChIP procedure again. IgG was used as a negative control. The DNA samples were quantified by quantitative real-time PCR using SYBR® Green PCR Master Mix (Applied Biosystems). The primers and antibodies used in the ChIP assay are listed in Additional file 2: Tables S15 and S10, respectively. MeDIP was performed as previously described [3, 59]. Genomic DNA was isolated from the harvested cells and sonicated using Bioruptor. The fragmented DNA was denatured and then immunoprecipitated with anti-6mA antibodies. The 6mA antibody used is listed (Additional file 2: Table S10), and the primers used in the quantitative MeDIP assay are listed (Additional file 2: Table S15). The qChIP and the qMeDIP values were calculated as previously described [49, 60].

Chromatin isolation by RNA purification (ChIRP)

ChIRP was performed as described with minor modifications [44]. RP11-390F4.3 and lacZ probes were designed using an online program at “www.singlemoleculefish.com” (Additional file 2: Table S16). Anti-sense DNA probes were labeled with a BiotinTEG at the 3′ end. Cells were washed and trypsinized for 5 min at room temperature. The harvested cells were washed with PBS, crosslinked in 20 ml of 1% glutaraldehyde/PBS on an end-to-end rotator for 10 min at room temperature, and then quenched by adding glycine to a final concentration of 125 mM for 5 min at room temperature. The cells were pelleted by centrifuging at 2000g for 5 min at 4°C. This was followed by two PBS washes, and the pellets were resuspended in the lysis buffer (10x mass pellet in grams; 50 mM Tris-HCl [pH 7.0], 10 mM EDTA [pH 8.0], 1% SDS, and protease inhibitor (Sigma)) with freshly added 1 mM of PMSF (Sigma) and Superase IN. The lysate was sonicated by Bioruptor at 4°C for at least 30 min (30 s ON, 30 s OFF) until the sample was no longer turbid. After centrifugation at 16,100g for 10 min at 4°C, the supernatant was subjected to ChIRP. After putting aside samples for DNA and RNA input, the lysates were diluted in 2 volumes of hybridization buffer (50 mM Tris-HCl [pH 7.0], 750 mM NaCl, 1 mM EDTA [pH 8.0], 1% SDS, 15% formamide, and protease inhibitors) with freshly added 1 mM of PMSF and Superase IN. Pools of probes (100 pmol) were added and incubated for 4 h at 37°C. The C-1 magnetic beads (Invitrogen) were washed twice with the lysis buffer and added into the lysate for 30–60 min at 37°C. Next, the beads were washed five times for 5 min at 37°C with mixing in pre-warmed washing buffer (2× saline sodium citrate [SSC], 0.5% SDS, and protease inhibitors) with freshly added 1 mM of PMSF and Superase IN. After the last wash, the beads were resuspended in 1 mL of washing buffer and then separated for RNA (100 μl) and DNA (900 μl) purification as described in the protocols for RNA extraction, ChIP, and qChIP. The primers used in qChIRP assay are listed (Additional file 2: Tables S11 and S15).

RNA sequencing (RNA-seq)

In brief, RNA-Seq libraries were prepared by Kapa HyperPrep Kit with RoboErase and were sequenced using Illumina NextSeq 550 to obtain 150 bp paired-end reads.

Chromatin immunoprecipitation followed by sequencing (ChIP-seq)

After DNAs were obtained from ChIP procedure, libraries were constructed by using KAPA HyperPrep Kit. Libraries were sequenced using Illumina NextSeq 550 to obtain 75 bp single-end reads.

Genomic DNA 6mA-ChIP-exo-seq 5.0

6mA ChIP-exo 5.0 was performed according to the previous reported procedure with some modifications to increase the yield for final library construction [11]. Briefly, 15–20 μg of fragmented genomic DNA was incubated with anti-6mA antibody (Synaptic Systems) at 4°C overnight in 1 × IP buffer (150 mM NaCl, 0.1% IGEPAL CA-630, 10 mM Tris-HCl; pH 7.4). The mixture was irradiated with UV 254 nm with 0.15 mJ/cm2 energy for 6 times. The crosslinked samples were pulled down at 4°C for 2 h, using 80 μl pre-blocked Dynabeads Protein A slurry (Invitrogen). Crosslinked DNAs on beads were washed with FA lysis buffer (50 mM Hepes-KOH, pH 7.5, 150 mM NaCl, 2 mM EDTA, 1% Triton, 0.1% sodium deoxycholate), NaCl lysis buffer (50 mM HEPES-KOH, pH 7.5, 500 mM NaCl, 2 mM EDTA, 1% Triton X-100, 0.1% sodium deoxycholate), LiCl buffer (100 mM Tris-HCl, pH 8.0, 500 mM LiCl, 1% NP-40, 1% sodium deoxycholate), and 10 mM Tris-HCl buffer at 4°C followed by enzyme reaction on beads, including A tailing (15 U Klenow Fragment, -exo (NEB), 1 × NEBuffer 2, and 100 μM dATP) at 37°C for 30 min, the first adapter ligation and kinase reaction (1200 U T4 DNA ligase, 10 U T4 PNK, 1 × NEBNext Quick Ligation Buffer, and 375 nM adapter) at 25°C for 1 h, fill-in reaction (10 U phi29 polymerase, 1 × phi29 reaction buffer, 2 × BSA, and 180 μM dNTPs) at 30°C for 20 min, and lambda exonuclease digestion (10 U λ exonuclease, 1 × λ exonuclease reaction buffer, 0.1% Triton X-100, and 5% DMSO) at 37°C for 30 min. DNAs were then released from beads with proteinase K digestion (30 μg Proteinase K, 25 mM Tris-HCl, pH 7.5, 2 mM EDTA, 200 mM NaCl, and 0.5% SDS) at 56°C overnight and then purified by using Agencourt AMPure XP (Beckman Coulter) with 3× beads clean up. Purified DNAs were subjected to splint ligation (1200 U T4 DNA ligase, 1 × Quick Ligase Buffer, 375 nM splint adapter). The product was purified using AMPure XP with 3× beads clean up and amplified by using PCR primers in TruSeq small RNA preparation kit and Phusion Master Mix with HF buffer (NEB). After the first 5 cycles of pre-amplification, 5 μl of pre-amplified samples was subjected to qPCR reaction to avoid over amplification according to the previous method [61]. Libraries were subjected to high-throughput sequencing by using NextSeq 550 with 75 bp single-end sequence.

DNA whole genome amplification (WGA) protocol for 6mA-ChIP-exo-seq

Whole genome amplified DNA was generated by RPEPLI-g midi kits (Qiagen) with 5 ng input genomic DNA. After amplification, amplified DNA was purified by 1× Ampure XP beads followed by 6mA-ChIP-exo-seq.

High-throughput sequence analysis

Before alignment, we trimmed all the sequenced reads by Trim galore. For RNA-seq, we aligned the reads to hg19 with HISAT2 [62]. After the alignment, gene abundances were done by htseq-count, and edgeR was used for differential gene analysis or Gene Set Enrichment Analysis was used for searching enrichment gene set in our RNA-seq data [63, 64]. For ChIP-seq and ChIP-exo-seq data, we aligned the reads to hg19 with Bowtie [65]. After the alignment, we filtered the aligned reads by mapping quality and removed the duplicated by picard. To get the enrichment region for all types of sequence, we applied MACS2 for peak calling [66]. Normalized signal files and profiles cross the genome were done by deeptools for data visualization [67]. Peak to gene correlation was performed by ChIPseeker with the annotation file of Ensembl version 87 [68]. Functional gene annotation was performed by gene ontology analysis with HOMER (Hypergeometric Optimization of Motif Enrichment) software [69].

6mA-ChIP-exo-seq quality control and genome-wide differential 6mA region analysis

According to the quality control results of each sample, we used 3 out of 5 hypoxia replicates for genome-wide analysis in different comparisons (comparison 1 or 2) (Additional file 1: Fig. S4h-j and Table S6). Due to the low complexity of 6mA ChIP-exo-seq data, we used a different pipeline to obtain enriched 6mA regions. This pipeline was based on previous established pipeline for 6mA signals counting by genome windows with application of quantile-adjusted conditional maximum likelihood method. We defined 3-kb windows across the genome and counted the signals from the de-duplicated aligned reads from each condition (i.e., comparing hypoxia vs. normoxia or METTL4-si vs. scrambled-si under hypoxia with edgeR [63]). For differential analysis, 6mA signal values in each condition were normalized with input sequence control. To define the gain-of-6mA regions regulated by hypoxia and METTL4, we applied the threshold of significant p-value < 0.05 under hypoxia vs. normoxia with the cut-off input-normalized log2 fold change of hypoxia vs. normoxia (> log2(1.1)) and METTL4 knockdown vs. control undergoing hypoxic status (< −log2(1.1)). Through this analysis, we identified 8268 regions in comparison 1 (hypoxia replicate-1,2,3) and 76,695 regions in comparison 2 (hypoxia replicate-1,4,5) (Additional file 1: Fig. S4i). To exclude the potential false discovered gain-of-6mA regions, we calculated the input-normalized log2 fold change of 6mA signals in hypoxia vs. WGA and considered input-normalized log2 fold change > log2(1.1) as hypoxia/WGA positive gain-of-6mA regions. By overlapping gain-of-6mA regions of different comparisons and hypoxia/WGA 6mA-positive regions, we generated 3673 gain-of-6mA regions co-regulated by hypoxia and METTL4 through multiple background subtractions for downstream analysis (Additional file 1: Fig. S4i, j).

Motif calculation

Motifs were calculated by HOMER [69] from comparison of each group of hypoxia 6mA-ChIP-exo-seq datasets as mentioned above (Comparison 1 used hypoxia-1, -2, -3 groups; Comparison 2 used hypoxia-1, -4, -5 groups) (Additional file 2: Table S6). The enriched signals located in hypoxia-induced METTL4 dependent gain-of-6mA regions were used for identifications of de novo motif by using the intersection between comparison 1 and comparison 2 (Additional file 1: Fig. S4i, j). WGA control signals from comparison 1 were used as background to exclude the sequencing noise. The p-value of motifs was calculated accordingly.

Analysis of RNA alternative splicing

RNA alternative splicing analysis was performed by rMAST [70]. After the quantification of spliced events, the results were input to maser, a R package, for downstream analysis. Low counts of spliced events were filtered by the cut-off mean junction counts < 10. Analysis of differential splicing events was performed with each comparison (hypoxia vs normoxia or METTL4-si vs scrambled-si under hypoxia). Significant spliced events were generated with the cut-offs p-value < 0.05, an absolute change in percent spliced in (|ΔPSI|) ≥ 0.05.

Calculation of lncRNA copy number per cell

LncRNA copy number was calculated as described previously [71, 72]. A linear standard amplification curve of Ct values was generated by qPCR using a dilution series of RP11-390F4.3 or NEAT1 plasmid DNA templates with known concentrations, corresponding to the regions amplified during qRT-PCR. The qRT-PCR values of RP11-390F4.3 or NEAT1 transcripts from 500,000 cells were determined under normoxia and hypoxia conditions. The Ct values were then fitted on a standard amplification curve, and the corresponding total transcripts were divided by 500,000 to calculate the number of RP11-390F4.3 or NEAT1 transcripts per cell. The plasmid DNA templates used are listed in Additional file 2: Table S8, and the primers used are listed in Additional file 2: Table S11.

Single-molecule RNA fluorescence in situ hybridization (RNA-FISH)

Single-molecule RNA-FISH was performed as described previously with minor modifications [72, 73]. The RNA-FISH probes targeting RP11-390F4.3 were designed using LGC Biosearch Technologies online software (Stellaris® Probe Designer version 4.2; https://www.biosearchtech.com/support/education/stellaris-rna-fish) and were purchased from LGC Biosearch Technologies. Forty-eight fluorescein-conjugated Stellaris RNA-FISH probes specific to RP11-390F4.3 are listed in Additional file 2: Table S17. Cells were fixed with 4% formaldehyde/5% acetic acid solution for 15 min at room temperature and then washed with PBS. Fixed cells were treated with 1% pepsin and subsequently dehydrated with a gradient of 70–100% ethanol. Cells were incubated with RNA-FISH probes diluted in hybridization buffer (100 mg/ml dextran sulfate, 0.2 mg/ml bovine serum albumin, and 10% formamide in 2× saline sodium citrate buffer (SSC)) at 55 °C. After 4 h, cells were washed three times with 0.1× SSC at 65 °C and then mounted using Prolong Gold An

留言 (0)

沒有登入
gif