The methylated RNA immunoprecipitation assay is an antibody-based method used to enrich for methylated RNA fragments. Coupled with deep sequencing, it leads to the identification of transcripts carrying the m5C modification.
Secondary base modifications on RNA, such as m5C, affect the structure and function of the modified RNA molecules. Methylated RNA Immunoprecipitation and sequencing (MeRIP-seq) is a method that aims to enrich for methylated RNA and ultimately identify modified transcripts. Briefly, sonicated RNA is incubated with an antibody for 5-methylated cytosines and precipitated with the assistance of protein G beads. The enriched fragments are then sequenced and the potential methylation sites are mapped based on the distribution of the reads and peak detection. MeRIP can be applied to any organism, as it does not require any prior sequence or modifying enzyme knowledge. In addition, besides fragmentation, RNA is not subjected to any other chemical or temperature treatment. However, MeRIP-seq does not provide single-nucleotide prediction of the methylation site as other methods do, although the methylated area can be narrowed down to a few nucleotides. The use of different modification-specific antibodies allows MeRIP to be adjusted for the different base modifications present on RNA, expanding the possible applications of this method.
In all three kingdoms of life, RNA species undergo post-transcriptional modifications and research on these functionally relevant biochemical modifications is called “epitranscriptomics”. Epitranscriptomics is a growing field and various methods are being developed to study and map the modifications on RNA molecules (reviewed in1,2). More than a hundred RNA modifications have been found, detected in rRNAs, tRNAs, other ncRNAs, as well as mRNAs3,4. Although the presence and function of chemically diverse post-transcriptional modifications in tRNAs and rRNAs are extensively studied5,6,7,8, only recently have mRNA modifications been characterized. In plants, many mRNA modifications have been identified to date, including m7G at the cap structure9, m1A10, hm5C11,12, and uridylation13. However, only m6A10,14,15, m5C11,16,17, and pseudouridine18 have been mapped transcriptome-wide in Arabidopsis. Post-transcriptional mRNA base modifications are involved in several developmental processes19,20.
One of the most commonly used approaches in epitranscriptomics is the methylated RNA immunoprecipitation coupled with deep sequencing (MeRIP-seq). MeRIP-seq was developed in 2012 to study m6A in mammalian cells21,22. It requires the use of an antibody for the desired modification and aims to enrich for RNA fragments carrying the modified nucleotide(s). It is usually followed by deep sequencing to identify and map the enriched fragments or quantitative PCR to verify specific RNA targets. The accuracy of MeRIP is based on the specificity of the antibody to recognise the modified nucleotide over similar modifications (e.g., m5C and hm5C11,23). Besides m6A, MeRIP-seq has been also applied to study m1A and m5C RNA methylation in several organisms11,17,23,24,25.
Methylation of the cytosine at the fifth carbon position (m5C) is the most prevalent DNA modification26,27 and one of the most common RNA modifications too3,4. While m5C was detected in eukaryotic mRNAs in 197528, only recently have studies focused on mapping the modification transcriptome-wide, in coding and non-coding RNAs11,16,17,23,29,30,31,32,33,34.
Alternative methods used in m5C RNA research include chemical conversion of non-methylated cytosines into uracils (bisulfite sequencing) and immunoprecipitation assays based on an irreversible binding of a known RNA cytosine methyltransferase to its RNA targets (miCLIP, aza-IP). In brief, bisulfite sequencing exploits the feature of 5-methylated cytosine to be resistant to sodium bisulfite treatment that deaminates unmodified cytosines to uracil. The method was first developed for DNA but adapted for RNA too and many studies have chosen this approach to detect m5C sites in RNA16,23,29,32,34,35. Both miCLIP and aza-IP require previous knowledge of the RNA cytosine methyltransferase and use of the respective antibody. In the case of miCLIP (methylation individual-nucleotide-resolution crosslinking and immunoprecipitation), the methyltransferase carries a single amino acid mutation so that it binds to the RNA substrate but cannot be released30. In aza-IP (5-azacytidine–mediated RNA immunoprecipitation), the irreversible binding is formed between the 5-azaC nucleoside and the RNA cytosine methyltransferase when exogenously provided 5-azaC is incorporated by RNA polymerases into a target RNA molecule31.
The main advantage of these three methods is that they allow single nucleotide resolution mapping of m5C. In addition, miCLIP and aza-IP provide information about the specific targets of a selected RNA cytosine methyltransferase, deciphering deeper the mechanism and role of post-transcriptional RNA modifications. However, the MeRIP-seq approach can identify transcriptome-wide m5C regions without any previous knowledge required and avoids harsh chemical and temperature conditions, such as bisulfite treatment or incubation with 5-azaC. Both MeRIP and bisulfite sequencing can be inhibited by secondary RNA structures36. The fragmentation step that is included in the MeRIP assay prior to immunoprecipitation aims to facilitate antibody binding and increase the resolution of m5C identification.
Another method worth mentioning is mass spectrometry (MS) of RNA nucleosides. MS can detect and distinguish any type of modification both on DNA and on RNA. Briefly, RNA is extracted and DNase digested, then desalted and digested to single nucleosides. The RNA nucleosides are analyzed by a mass spectrometer. This method can be used to quantify the levels of each modification and it does not rely on an antibody or a chemical conversion. However, a major drawback is that it provides bulk information about the presence of RNA modifications. In order to map the modifications, MS needs to be combined with RNase digestion and sequencing information about specific RNA molecules, as in the case of human tRNALeu(CAA)37.
Here, we describe and discuss the MeRIP assay as used to study m5C RNA methylation in Arabidopsis17.
1. Preparing the RNA
2. Methylated RNA Immunoprecipitation (MeRIP)
3. Downstream analysis
A schematic of the method is provided in Figure 1. The first critical steps of the protocol are to obtain RNA of good quality (RIN ≥ 7) and sonicate it to approximately 100 nt fragments. The efficiency of both steps is examined by a chip-based capillary electrophoresis machine. In Figure 2A, a representative run of a good RNA sample is shown. The sample is diluted 1:10 before loading on the chip in order to have a concentration that is in the range of detection of the kit used (5-500 ng/µL). The same sample is also run after sonication and is shown in Figure 2B. Notice the presence of one uniform peak shifted to the left of the diagram, at a size of around 100 nucleotides. The lower concentration is caused both by loss of RNA during fragmentation but also because of the increased volume of the samples (60-100 µL, step 1.7.1).
The quality of the IP and Mock samples can be evaluated by qRT-PCR. To this end, the spiked-in IVTs serve as positive and negative controls: the methylated IVT, where in all cytosine positions there is m5C, is expected to be highly enriched in the IP sample; on the contrary, the non-methylated IVT should not have a difference between IP and Mock. The primers used in the qRT-PCR assay for the two control IVTs (EGFP and Renilla luciferase) are listed in Table 1. Indeed, as shown in Figure 3, around 80% of the methylated IVT was recovered in the IP sample, and only approximately 2% in the Mock. For the non-methylated control, the recovery was below 1% in both IP and Mock samples. This verifies the efficiency of MeRIP that methylated RNA fragments were precipitated and enriched, and is a good indicator that the samples can be used for downstream analysis. In addition, the fold enrichment of the non-methylated IVT (IP to Mock ratio) can be applied as a threshold to estimate significance of enrichment in the qRT-PCR assays.
After aligning the reads to the genome (Figure 4), the peak calling algorithms described in steps 3.4.1 and 3.4.2 are applied to identify the statistically significant windows, enriched in the IP samples compared to the Input. The sequences that correspond to these windows can be used further, for example to search for conserved methylation-related motifs11,17.
Figure 1: Schematic representation of MeRIP-seq protocol.
RNA samples are incubated with an antibody for 5-methylated cytosines and the complexes are pulled down with protein G magnetic beads that capture the antibodies along with the bound RNA. The eluted RNA samples are analyzed by deep sequencing and qRT-PCR. Please click here to view a larger version of this figure.
Figure 2: Representative results from quality analysis of RNA samples.
(A) Representative profile of a qualified total RNA plant sample. (B) Representative profile of an RNA sample after sonication to 100 nt fragments. Output files from capillary electrophoresis software. Please click here to view a larger version of this figure.
Figure 3: qRT-PCR analysis of control IVTs.
Methylated and non-methylated in vitro transcripts were used as positive and negative controls of the MeRIP assay, respectively. After immunoprecipitation, the methylated IVT is highly enriched in the IP sample (green) but not in the Mock sample (without anti-m5C antibody; purple). The non-methylated IVT showed no enrichment and no difference between IP and Mock. Please click here to view a larger version of this figure.
Figure 4: Read alignment before and after MeRIP around a representative transcript.
Reads aligned to a specific transcript in the Input sample (top row) and in two IP replicates (middle and bottom row). The black box shows an identified enriched 50 nt long window based on MACS243 and MeRIP-seq22 peak-calling analyses. Please click here to view a larger version of this figure.
Target | Primer pair | Product sequence | Product length | ||
EGFP | For: 5'- GGCAACTACAAGACCCGCGCC -3' | GGCAACTACAAGACCCGCGCCGAG GTGAAGTTCGAGGGCGACACCCTG GTGAACCGCATCGAGCTGAAGGGC | 72 bp | ||
Rev: 5'- GCCCTTCAGCTCGATGCGGTT -3' | |||||
Renilla luciferase | For: 5'- GGAGAATAACTTCTTCGTGGAAAC -3' | GGAGAATAACTTCTTCGTGGAAACC ATGTTGCCATCAAAAATCATGAGAA AGTTAGAACCAGAAGAATTTGCAGC | 75 bp | ||
Rev: 5'- GCTGCAAATTCTTCTGGTTCTAA -3' |
Table 1: Information about the primers and generated products for the qRT-PCR analysis.
Buffer Recipes. Please click here to download this file.
RNA carries more than one hundred distinct base modifications4 that form the epitranscriptome44. These modifications add an additional layer of regulation of translation and signalling (reviewed in5,6,8,20,45,46). Early studies were able to detect the presence of post-transcriptional modifications on RNA28,47 but the specific modified RNAs need to be identified in order to understand the role of epitranscriptomics. MeRIP was designed as a method to map RNA methylation sites transcriptome-wide21,22. It can be adapted to any modification, if a specific antibody is available.
The main strength of this protocol is that is relatively simple, safe for the RNA and the user (e.g., 5-azaC is highly toxic for plants and humans) and does not require sequence or modifying enzyme information. Moreover, the enrichment of methylated RNAs by the IP increases the chances of low abundant mRNAs to be detected, unlike bisulfite sequencing that does not contain an enrichment step. When two serial rounds of MeRIP are performed, enrichment in RNA fragments containing methylation sites increases further22. One of the limitations of MeRIP, especially when applied to mRNA methylation studies, is the high quantity of RNA required as input for the assay. The ribodepletion – or poly(A) enrichment – step will reduce the background caused by the heavily modified ribosomal RNA but it removes more than 90% of total RNA. DNA must also be completely removed as it is rich in 5-methylated cytosines. Another drawback is the lower resolution of the exact position of methylation. Sonication of the RNA prior to incubation with the antibody helps towards this direction by narrowing down the region containing the modification to 100-200 nucleotides. When MeRIP is combined with deep sequencing, the resolution of m5C site prediction increases as the sequencing reads form a Gaussian distribution around the potential methylation site. Additionally, the specificity of the antibody needs to be confirmed prior to the assay (e.g., with RNA dot blot assays, performed with oligos synthesized with modified nucleotides), however, to what extent an antibody can actually distinguish between closely related modifications (e.g., m6A and m6Am) is a point of argument in the field48,49. Moreover, highly structured RNAs might interfere with the antibody–antigen interaction, another restriction that is mostly addressed with fragmentation and denaturation of RNA prior to IP. On the contrary, bisulfite sequencing that is also affected by secondary structures, does not include a fragmentation step and this might be one reason that causes discrepancy between the m5C sites and mRNAs predicted by bisulfite sequencing16 and MeRIP-seq11,17. Other cytosine modifications (e.g., hm5C) are also resistant to bisulfite-mediated deamination35.
Modifications of MeRIP-seq include a crosslinking step, either with the introduction of a photoactivatable ribonucleoside (photo-crosslinking-assisted m6A-seq, PA-m6A-seq 50) or using UV light to create antibody-RNA crosslinks after the IP (miCLIP49, different method than the miCLIP described in introduction30, but also with individual-nucleotide resolution). In the future, and as knowledge on RNA methylation is accumulating, more targeted approaches might be preferable, based on the modifying enzymes and/or the consensus sequences where methylation is appearing. The identification of reader proteins is essential to the understanding of the molecular and signalling function of post-transcriptional modifications. Nanopore sequencing technology already allows the direct identification of modified nucleotides without prior treatment of the RNA17 but there is still room for improvement on this field regarding sequence depth and bioinformatic analysis. Overall, MeRIP-seq is currently an established, reliable, and unbiased approach to identify methylated RNA transcripts.
This work was supported by an IMPRS PhD stipend to E.S, an EMBO Long-Term Fellowship to V.P., and MPI-MPP internal funds to F.K. Part of this work was also funded through PLAMORF, a project that has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (Grant agreement No. 810131). The authors would like to thank Federico Apelt for the bioinformatic analysis and comments on the manuscript, and Mathieu Bahin and Amira Kramdi for the bioinformatic analysis. Work in the Colot lab is supported by Investissements d’Avenir ANR-10-LABX-54 MEMO LIFE, 506 ANR-11-IDEX-0001-02 PSL* Research University.
Name | Company | Catalog Number | Comments |
α‐m5C antibody | ZymoResearch | A3001 (discontinued), A3002 available | Clone 10G4 (discontinued), 7D21 available |
Chip and reagents for capillary electrophoresis | Agilent | 5067-1511 | RNA 6000 Nano kit |
Dnase | Ambion | AM1907 | TURBO DNA-free kit |
Co-precipitant of nucleic acids | Ambion | AM9515 | Glycoblue, 15 mg/mL |
In vitro transcription kit | Promega | P1300 | T7 RiboMAX Large Scale RNA Production System |
Low-binding reaction tubes | Kisker Biotech | G017 | 1.7 ml microcentrifuge tubes |
Protein G magnetic beads | Invitrogen | 10003D | Dynabeads Protein G |
Proteinase K | Ambion | AM2546 | 20mg/mL stock solution, RNA grade |
RNase inhibitor | Promega | N2615 | RNasin Plus 40 U/μl |
rRNA removal kit | Epicentre | RZPL11016 | Ribo-Zero rRNA removal kit (Plant leaf) |
Sonication tubes | Covaris | 520045 | microtube AFA Fiber Pre-Slit Snap-Cap 6x16mm |
RNA extraction reagent | Ambion | 15596018 | TRIzol reagent |
Equipment | |||
Capillary electrophoresis machine | Agilent | G2939BA | 2100 Bioanalyzer System |
Magnetic rack | Invitrogen | 12321D | DynaMag-2 |
Microentrifuge | Eppendorf | discontinued | 5417R model |
Rotator | Benchmark | R4040 | RotoBot Mini |
Sonicator | Covaris | 500217 | S220 Focused-ultrasonicator |
Spectrophotometer | ThermoFisher Scientific | ND-ONEC-W | NanoDrop OneC |
Waterbath | GFL | 1012 | 1012 Incubation bath |
This article has been published
Video Coming Soon
ABOUT JoVE
Copyright © 2024 MyJoVE Corporation. All rights reserved