Transcriptome-wide m6A detection with DART-Seq


 m6A is the most abundant internal mRNA modification and plays diverse roles in gene expression regulation. Much of our current knowledge about m6A has been driven by recent advances in the ability to detect this mark transcriptome-wide. Antibody-based approaches have been the method of choice for global m6A mapping studies. These methods rely on m6A antibodies to immunoprecipitate methylated RNAs, followed by next-generation sequencing to identify m6A-containing transcripts1,2. While these methods enabled the first identification of m6A sites transcriptome-wide and have dramatically improved our ability to study m6A, they suffer from several limitations. These include requirements for high amounts of input RNA, costly and time-consuming library preparation, high variability across studies, and m6A antibody cross-reactivity with other modifications. Here, we describe DART-Seq (deamination adjacent to RNA modification targets), an antibody-free method for global m6A detection. In DART-Seq, the C to U deaminating enzyme, APOBEC1, is fused to the m6A-binding YTH domain. This fusion protein is then introduced to cellular RNA either through overexpression in cells or with in vitro assays, and subsequent deamination of m6A-adjacent cytidines is then detected by RNA sequencing to identify m6A sites. DART-Seq can successfully map m6A sites throughout the transcriptome using as little as 10 nanograms of total cellular RNA, and it is compatible with any standard RNA-seq library preparation method.


Introduction
General notes: The DART-Seq approach assumes that APOBEC1-YTH and APOBEC1-YTH mut fusion proteins have been cloned into an expression vector appropriate for the application. We have tested mammalian expression under the CMV promoter, but others will likely work as well.
Prior to starting, pilot studies are recommended to ensure that expression of fusion proteins are not toxic to cells and that adequate expression is achieved.
Practice good RNA handling technique during library preparation.
For standard RNA-seq, we use 1 microgram of total cellular RNA for library preparation with the NEBNext Ultra II Directional RNA Library Prep Kit for Illumina (New England Biolabs). We have also prepared libraries from as little as 10 nanograms of total RNA with good results. As with any RNA-seq library preparation, lower amounts of starting RNA may require more sensitive library preparation methods. We have used the Single Cell/Low Input RNA Library Prep Kit (New England Biolabs) for our low-input libraries with good results. Other library preparation methods are also compatible with DART-Seq and can be tested by the user.

Procedure
A. Transfect and grow cells The protocol described here has been used for adherent HEK293T cells. Adjustments may be needed for other cell types. Sequencing can be carried out using paired-end or single-read sequencing. We typically use SR 50bp sequencing on an Illumina HiSeq4000 or similar machine. We achieve 30-50 million reads per sample using this platform.
To further ensure high-confidence identification of m 6 A sites, we filter C to U editing sites to include only those with a minimum threshold of mutations per read compared to sites obtained by expression of APOBEC1-YTH mut : A. First, use bedtools intersect to find C to T transitions that are present only in APOBEC1-YTH samples. Ensure that the YTH and mut .bed files you use have mutations/read (m/k) in column 5.
These will be merged back in later.