Breast cancer patient tumor samples

After obtaining adequate informed consent, breast cancer tissue and adjacent normal tissue were obtained from 8 patients who underwent curative resection for breast cancer at Nanfang Hospital of Southern Medical University, between November 2019 and May 2020. All patients satisfied the following inclusion criteria: the surgical margins were confirmed to contain no residual carcinoma tissue; clinicopathological information on age, sex, clinical stage, neoadjuvant therapy, ER, PR, HER2, Ki67, histological subtype and histological grade was available and is listed in Supplementary Table 1. All patient-related studies were approved by the institutional review boards of the Seventh Affiliated Hospital of Southern Medical University and Nanfang Hospital of Southern Medical University.

Human cancer cell lines and cell culture

Human breast cancer cell lines (SK-BR-3, MCF-7, T47D, BT474, MDA-MB-468, and MDA-MB-231) were purchased from American Type Culture Collection (ATCC, Manassas, VA, USA). MCF-7 cells were cultured in MEM (Gibco cat:11,095) supplemented with 10% fetal bovine serum (FBS, HyClone, Utah, USA), 1% penicillin‒streptomycin (Gibco), 10 μg/ml insulin (CellCook, cat: CM1007) and 1⨯ nonessential amino acids (CellCook, cat: CM10085/L). SK-BR-3, MCF-7, T47D, BT474, MDA-MB-468, and MDA-MB-231 cells were cultured in RPMI 1640 (Gibco, cat:11,875) supplemented with 10% fetal bovine serum (FBS; HyClone, Utah, USA) and 1% penicillin‒streptomycin (Gibco).

Establishment of transfected cell lines

For CRISPR‒Cas9 cells, 100 μM sgRNA Top/Bottom Oligo solution was configured with ddH2O, and the sgRNA sequences were as follows: sgMETTL3:5’-ATTCTGTGACTATGGAACCA-3’, sgYTHDF1 5’-AGTTTCAAAGCCGACCTCGT-3’, sgYTHDF2 5’-GTCCATTACTAGTAACATCG-3’, and sgYTHDF3: CAACCGAAACTTAAACCCAA-3’. PCR was used to produce linked products that were transformed into DH5α cells. Lentiviral overexpression plasmids for human METTL3 and YTHDF2 were transfected into cells as previously described [15]. Cells were selected using puromycin (2 μg/ml, Gene-Chem) for 3 days for stable transfection. 293 T cells (ATCC Cat# CRL-3216) were used to generate lentiviral particles by cotransfecting the packaging vectors using a transfection reagent (SignaGen, SL100668), following the manufacturer’s instructions.

Western blot analysis

Lysis buffer was used for total protein analysis of extracted cells. Samples were separated on an 8–15% gel by SDS-PAGE. Nitrocellulose membranes were blocked with blocking buffer and incubated with the appropriate primary antibody. The membranes were washed with blocking buffer three times, probed with the appropriate secondary antibody and developed using SuperSignal West Pico or Dura (Thermo Fisher Scientific).

Quantitative PCR analysis

Real-time PCR analysis was performed by using the Biosystems Step-One-Plus Real-Time PCR System with FastStart Universal SYBR Green Master Mix (Roche). The primer sets used were as follows.

GAPDH-F: GAACGGGAAGCTCACTGG,

GAPDH-R: GCCTGCTTCACCACCTTCT,

METTL3-F: AGATGGGGTAGAAAGCCTCCT,

METTL3-R: TGGTCAGCATAGGTTACAAGAGT,

LATS1-F: AATTTGGGACGCATCATAAAGCC,

LATS1-R: TCGTCGAGGATCTTGGTAACTC,

YTHDF2-F: TAGCCAACTGCGACACATTC, and

YTHDF2-R: CACGACCTTGACGTTCCTTT.

The PCR conditions were as follows: 10 min at 95 °C and 40 cycles of 15 s at 95 °C and 1 min at 60 °C. The average Ct value for each gene was determined from triplicate reactions and was normalized as previously reported.

Cell proliferation, apoptosis, migration and invasion assays

For the EdU cell proliferation assay, the cells were trypsinized and seeded onto 6-well plates at a density of 1 × 106 cells per well. After incubation for 24 h at 37 °C, 1 ml medium was added to each well and incubated for 2 h after discarding the old medium. The cells were collected in flow tubes and centrifuged at 1,500 rpm for 5 min. The supernatant was discarded and the cells were resuspended in PBS and centrifuged at 1,500 rpm for 5 min. After discarding the supernatant, the cells were fixed with 4% paraformaldehyde for 15–30 min, neutralized with glycine for 5 min, and resuspended in PBS. The cells were then incubated with 0.5% Triton X-100 penetrant at room temperature for 10 min. One hundred microliters of 1⨯ Apollo staining reaction solution was added to each tube and the cells were resuspended again. After incubation for 10 min at room temperature, the staining reaction solution was discarded by centrifugation at 1,500 rpm for 5 min. Next, 0.5% Triton X-100 penetrant was used to clean the cells 1–3 times at room temperature and the cells were resuspended in PBS. Flow cytometry was performed immediately after staining.

Human cancer cell xenograft model

All animal experiments were approved by the Institutional Animal Care and Use Committee of Southern Medical University. Breast cancer cells (5 × 106) were implanted into the subcutaneous axilla of the forelimb of 3–4-week-old BALB/c nude mice. Seven days after transplantation, the diameter of the tumors was measured, and the tumors were removed after three weeks.

Coimmunoprecipitation assay

Immunoprecipitation assays were performed as previously described [16]. Cells were lysed in radioimmunoprecipitation assay buffer containing protease and phosphatase inhibitors. The cells were then centrifuged at 12,000 × g for 10 min at 4 °C and collected. The supernatants were immunoprecipitated with antibodies and magnetic protein A/G beads (Pierce) were used for incubation for 2 h at 4 °C. The above immune complexes were washed with PBS and resuspended in SDS-PAGE buffer, followed by western blotting analysis.

MeRIP-seq

Total RNA was isolated from cells and tissues using Magzol Reagent (Magen, China), according to the manufacturer’s protocol. The quantity and integrity of the RNA yield were assessed using a K5500 microspectrophotometer (Beijing Kaiao, China) and the Agilent 2200 TapeStation system (Agilent Technologies, USA), respectively. m6A antibody-immunoprecipitated RNA was quality controlled using the Qubit (Thermo Fisher Scientific, USA) and Agilent 2200 TapeStation (Agilent Technologies, USA) systems. Briefly, RNA was fragmented into molecules approximately 200 bp in length. The RNA fragments were then subjected to first-strand and second-strand cDNA synthesis, followed by adaptor ligation and enrichment with a low cycle according to the instructions of the NEBNext® Ultra RNA LibraryPrep Kit for Illumina (NEB, USA). The final library product was assessed using the Agilent 2200 TapeStation and Qubit® system (Life Technologies, USA), and then sequenced on an Illumina platform (Illumina, USA) with a paired-end length of 150 bp at RiboBio Co., Ltd. (Ribobio, China). Adaptor and low-quality bases were trimmed using Trimmomatic tools (version: 0.36), and the clean reads were subjected to rRNA deletion through RNAcentral to obtain effective reads. Genomic alignment was performed using TopHat (version 2.0.13) to obtain uniquely mapped reads. Effective reads from the input sample were used for RNA-seq analysis and the read count value of each transcript was calculated using HTSeq (version 0.6.0). Differentially expressed genes were identified using the DEseq/DESeq2/edgeR/DEGseq R package according to the following criteria: |log2 (fold change)|≥ 1 and P value < 0.05.

Sequence-based RNA adenosine methylation site predictor (SRAMP)

SRAMP is a site prediction tool based on a random forest machine-learning framework that can predict m6A sites based on sequence-derived features. The m6A locus was predicted using the target sequence [17].

Metabolomic analysis

A high-resolution mass spectrometer (QEXactive, Thermo Fisher Scientific, USA) was used to collect positive ion (POS) and negative ion (NEG) data from 18 cell samples for untargeted metabolomics detection using LC‒MS/MS technology to explore the metabolomic composition and biological function of the samples. Compound Discoverer 3.1.0 (Thermo Fisher Scientific, USA) software was used for data processing, including peak extraction, peak alignment, fill gaps, and compound identification. An in-house metabolome information analysis process was used to carry out metabolite annotation, classification (Kyoto Encyclopedia of Genes and Genomes [KEGG], Human Metabolome Database [HMDB]), and enrichment analysis for the identified substances, and to explain the physical and chemical properties and biological functions of the metabolites. The R software package metaX [18] was used for data preprocessing, statistical analysis (univariate and multivariate analyses), and screening of metabolites with significant differences. Discriminant analysis was performed using partial least squares discriminant analysis (PLS-DA). Differentially expressed metabolites were screened by the VIP values of the first two principal components of the model [19] and Student’s t-test was used to analyze the results of the univariate analysis (fold change).

Transcriptome sequencing

Total RNA was extracted using TRIzol reagent (Thermo Fisher, 15,596,018) as previously reported [12]. Total RNA quantity and purity were analyzed using a Bioanalyzer 2100 and RNA 6000 Nano LabChip Kit (Agilent, CA, USA, 5067–1511). Additionally, mRNA was purified from total RNA (5 μg) using Dynabeads Oligo (dT) (Thermo Fisher Scientific, CA, USA) with two rounds of purification and fragmented into short fragments using divalent cations at an elevated temperature (Magnesium RNA Fragmentation Module (NEB, cat. e6150, USA) at 94 °C for 5–7 min). Then, the cleaved RNA fragments were reverse-transcribed to create cDNA with SuperScript™ II Reverse Transcriptase (Invitrogen, cat.1896649, USA). Dual-index adapters were used and size selection was performed by using AMPureXP beads. The U-labeled second-stranded DNAs were treated with the heat-labile UDG enzyme (NEB, cat.m0280, USA). PCR was used to amplify to the ligated products under the following conditions: 95 °C for 3 min; 8 cycles of denaturation at 98 °C for 15 s, 60 °C for 15 s, and 72 °C for 30 s; and 72 °C for 5 min. Finally, 2 × 150 bp paired-end sequencing (PE150) was performed on an Illumina Novaseq 6000 (LC-Bio Technology Co., Ltd., Hangzhou, China).

RNA stability analysis

Cells in the treatment and control groups were collected after 24 h. RNA was extracted using TRIzol reagent, and mRNA expression was determined by qPCR.

Seahorse assay

On the same day that the cells were seeded, 180 μl of hydration solution was added to the lower part of the XF96 Extracellular Flux Assay Kit. Cells were hydrated overnight in an incubator without CO2 at 37 °C. The required drugs and Seahorse XF basic culture medium were prepared simultaneously, and the pH of the medium was adjusted to 7.4. The cells were then placed in a 37 °C water bath for 1 h before use. On the second day, the cells were washed twice with Seahorse XF basic culture medium in a water bath and 175 μl Seahorse XF basic culture medium was added to each well. The cells were then cultured in a CO2-free incubator at 37 °C for 1 h. Each drug was diluted to the required concentration, and 25 μl was added to each well of an XF96 extracellular flux assay kit. Thirty minutes later, the lower part of the XF96 Extracellular Flux Assay Kit was removed and replaced with cell plates that had been cultured in a CO2-free incubator at 37 °C for 1 h and assessed with a computer.

RIP-qPCR

Endogenous RNA was retrieved after trapping in the nucleus or cytoplasm with an antibody or epitope marker. The RNA-binding protein was separated from the bound RNA by immunoprecipitation. The cells were crosslinked with 1% formaldehyde and treated with 150 mM NaCl RIPA buffer containing RNase and protease inhibitors. Cells were lysed in 0.5% sodium deoxycholate, 0.1% SDS, 1% NP40, 1 mM EDTA, and 50 mM Tris (pH 8.0) for 30 min and centrifuged for precipitation. The supernatant was incubated four times with METTL3, YTHDF1, or YTHDF2 primary antibody or the corresponding IgG antibody for 4 h, and protein A/G glycosylated beads were added and incubated for 2 h with shaking. After washing three times with RIPA buffer, RNA was extracted after crosslinking. Quantitative RT-PCR was performed to measure RNA.

RNA pulldown assays

RNA pulldown assays were performed in accordance with the Pierce™ Magnetic RNA‒ Protein Pull-down Kit protocol. RNA was prepared by in vitro transcription and biotin was used. Poly (A) 25 RNA was used as the negative control. Total protein was extracted from each group, and the required protein concentration was greater than 2 mg/ml. RNA pull-down was performed according to a standard protocol, and the expression level was confirmed by western blotting.

Statistical analysis

Data were analyzed using SPSS 20.0. With a two-tailed independent Student’s t-test, p < 0.05 was considered significant. Two patient cohorts were compared using a Kaplan–‒Meier survival plot, and log-rank p value were calculated.