Patients and samples
Our study was approved by the Ethics Committee of The Affiliated Hospital of Jining Medical University (Approved Number: 2021-12-C013, China), and written consent was provided by each participant before enrolment. From January 2020 to June 2021, we recruited 42 patients who were diagnosed with AD at the memory clinic of the Affiliated Hospital of Jining Medical University. The inclusion criteria for AD patients were: (1) first-time diagnosis of AD according to the criteria of the National Institute on Aging Alzheimer’s Association [31] after clinical evaluation, neuropsychological tests, and magnetic resonance imaging; (2) age between 50 and 80 years; and (3) no history of anti-dementia or mood-stabilizing medications. The exclusion criteria were: (1) history of cerebrovascular disease associated with cognitive impairment; (2) frontotemporal dementia, dementia with Lewy bodies, primary progressive aphasia, and/or Parkinson disease dementia; (3) dementia due to other causes such as infection, poisoning, drugs, and/or metabolic diseases; (4) complications involving functional failure of the heart, lung, liver, kidney, and/or other important organs; or (5) incomplete clinical data.
In addition, we recruited 42 healthy volunteers from the health management center of the same hospital, who were matched to enrolled patients in terms of sex and age. The inclusion criteria for healthy controls were: (1) age older than 50 years; and (2) no history, symptoms, or signs of relevant psychiatric or neurological disease; and (3) no cognitive impairment.
Fasting venous blood samples were collected from all subjects into ethylenediaminetetraacetic acid-containing tubes. PBMCs were isolated using Ficoll-Hypaque density gradient centrifugation as described [32].
RNA isolation and quantitative real-time PCR (qRT-PCR)
Total RNA was isolated from PBMCs and purified using TRIzol reagent (cat# 15596018, Invitrogen, Carlsbad, CA, USA). Sample quality and RNA amount were examined using a NanoDrop ND-1000 (NanoDrop, Wilmington, DE, USA). For inclusion in the experiments, the absorbance ratio A260/A280 of RNA had to be 1.8-2.0, in which case 1 µg was reverse-transcribed into cDNA using the SuperScript III First-Strand Kit (cat# 18080051, Invitrogen). The cDNA was diluted two-fold in RNAase-free ddH2O, then used as template (1 µL) in qRT-PCR with the ChamQ™ Universal SYBR qPCR Master Mix (cat# Q711-02, Vazyme, Nanjing, Jiangsu, China) and primers designed using the online tool https://www.ncbi.nlm.nih.gov/tools/primer-blast/ (Supplementary Table 4). The 2−ΔΔCt method was used to measure mRNA levels relative to GAPDH [33]. Only genes with transcript Ct ≤ 30 were considered to be expressed.
Quantification of m6A methylation
The level of m6A-modified RNA in 1000 ng total RNA from PBMCs was measured using a commercial colorimetric kit (cat# ab185912, Abcam, Cambridge, MA, USA), according to the manufacturer’s instructions. Absorbance was measured at 450 nm, which was converted to m6A levels through a standard curve.
Western blot analysis
Total protein was isolated from PBMCs of AD patients and controls using RIPA lysis buffer (Beyotime Biotechnology, Nanjing, China) containing phenylmethylsulfonyl fluoride (Beyotime Biotechnology). Lysates were left undisturbed for 30 min and then were centrifuged at 12,000 × g for 20 min at 4°C. Total protein concentration in the supernatant was estimated using a bicinchoninic acid assay (Beyotime Biotechnology), and equal amounts of protein (30 µg) were separated by electrophoresis on precast 10% Bis-Tris Gels (Bio-Rad, Laboratories, Hercules, CA, USA) and transferred to polyvinylidene difluoride membranes (Millipore, Billerica, MA, USA). The membranes were incubated for 16 h at 4°C with rabbit primary antibodies against the following proteins: METTL3 (1:1,000, cat# ab195352, Abcam), METTL14 (1:1,000, cat# A8530, ABclonal, Wuhan, Hubei, China), WTAP (1:1,000, cat# 56501, Cell Signaling Technology, Danvers, MA, USA), ALKBH5 (1:1,000, cat# ab195377, Abcam), FTO (1:1,000, cat# ab124892, Abcam), and β-actin (1:50,000, cat# AC026, ABclonal). After washing with TBST, membranes were incubated with horseradish peroxidase-conjugated anti-rabbit IgG (1:5,000, cat# ab6721, Abcam). Antibody binding was revealed using the ECL kit (cat# 32106, Thermo Fisher Scientific, Waltham, MA, USA), images were acquired using a Tanon 5200 imaging analysis system (Tanon Science & Technology, Shanghai, China), and bands were quantified using Image J (version 1.52v, US National Institutes of Health, Bethesda, MD, USA).
Enzyme-linked immunosorbent assay (ELISA)
Plasma was extracted from venous blood by centrifuging at 3,000 × g for 20 min, divided into 1-mL aliquots, and frozen at -80°C until analysis. The plasma was assayed for the following proteins using commercial ELISAs (Meimian, Yancheng, Jiangsu, China): METTL3 (cat# MM-0395H1), METTL14 (cat# MM-0395H2), WTAP (cat# MM-48756R1), ALKBH5 (cat# MM-2545E1) and FTO (cat# MM-51110H2). Absorbance was measured at 450 nm, and values were converted to relative proteins level using a standard curve.
Methylated RNA immunoprecipitation-sequencing (MeRIP-Seq)
Since MeRIP-Seq required at least 100 µg RNA for each sample, the RNAs of ten human PBMCs were pooled as one sample for MeRIP-Seq, and there were three samples for control and AD groups, respectively. RNA was isolated as described. Adequate RNA quality was defined as an RNA integrity number > 7.0 using a Bioanalyzer 2100 (Agilent, Santa Clara, CA, USA) and confirmed by electrophoresis with a denaturing agarose gel.
Poly(A) RNA was purified from 50 µg total RNA using Dynabeads Oligo (dT) 25 (cat# 61005, Thermo Fisher Scientific), fragmented into small pieces at 86°C for 7 min using the Magnesium RNA Fragmentation Module (cat# e6150, New England Biolabs, Ipswich, MA, USA), incubated at 4°C for 2 h with an anti-m6A antibody (cat# 202003, Synaptic Systems, Göttingen, Niedersachsen, Germany) in 50 mM Tris-HCl, 750 mM NaCl, and 0.5% Igepal CA-630. Immunoprecipitated RNA was reverse-transcribed to cDNA by SuperScript™ II Reverse Transcriptase (cat# 1896649, Invitrogen), which was used to synthesize U-labeled second-stranded DNA with Escherichia coli DNA polymerase I (cat# m0209, New England Biolabs), RNase H (cat# m0297, New England Biolabs), and dUTP (cat# R0133, Thermo Fisher Scientific). An A-base was then added to the blunt ends of each strand in order to prepare them for ligation to the indexed adapters. Each adapter contained a T-base end for ligation to A-tailed fragmented DNA. Single- or dual-index adapters were ligated to the fragments, and size selection was performed with AMPureXP beads. After treatment with a heat-labile UDG enzyme (cat# m0280, New England Biolabs), the ligated products were amplified by PCR under the following conditions: initial denaturation at 95°C for 3 min; eight cycles of denaturation at 98°C for 15 sec, annealing at 60°C for 15 sec, and extension at 72°C for 30 sec; then final extension at 72°C for 5 min. The average insert size for the final cDNA library was 300 ± 50 bp. We then performed 2 × 150 bp paired-end sequencing (PE150) on a Novaseq™ 6000 (Illumina, San Diego, CA, USA).
Bioinformatic analyses
RNA sequencing data were cleaned up using Fastp software (https://github.com/OpenGene/fastp) [34] with default parameters in order to remove adapter contamination and low-quality reads, defined as those in which bases with Q ≤ 10 accounted for more than 20% of total reads. The sequence quality of the input and immunoprecipitated samples was also verified using fastp. HISAT2 (http://daehwankimlab.github.io/hisat2) [35] was used to map the reads to the reference genome Homo sapiens (version 101). Mapped reads of immunoprecipitated and input libraries were analyzed using the exomePeak package in R (https://bioconductor.org/packages/exomePeak) [36], which identified m6A peaks in bedGraph or bigWig formats. The output was visualized using IGV software (http://www.igv.org) [37]. MEME (http://meme-suite.org) [38] and HOMER (http://homer.ucsd.edu/homer/motif) were used to identify de novo and known motifs, followed by localization of the motif with respect to the peak summit.
Peaks were annotated based on intersection with gene architecture using the ChIPseeker package in R (https://bioconductor.org/packages/ChIPseeker) [39]. StringTie (https://ccb.jhu.edu/software/stringtie) was used to determine the expression levels of all mRNAs from input libraries, in terms of fragments per kilobase per million (FPKM), calculated as total exon fragments/mapped reads (millions) × exon length (kB).
The mRNAs showing | fold change | ≥ 2 and P < 0.05 between AD patients and controls based on the edgeR package in R (https://bioconductor.org/packages/edgeR) were considered to be differentially expressed [40].
Statistical analysis
All statistical analyses were conducted using GraphPad Prism version 8.0 (Graphpad, San Diego, CA, USA). Data were presented as mean ± standard error of the mean (SEM). Differences between two groups were assessed using Student’s t test for independent samples. Differences in gene expression were expressed in terms of “fold change (FC)” and P values, which were adjusted using Benjamini-Hochberg correction (q-value or false discovery rate). Correlations between continuous variables were analyzed using partial correlation analysis that adjusted for other covariates (age, sex, and disease severity). Statistical results associated with P < 0.05 were considered significant.