green algae [122]. Focusing on these two motifs, 6mA-RE-Seq employs digestion of
DNA by CviAII or DpnII, sonication to fragment the DNA, end repair, adenylation,
and ligation to adapters [122]. The methylated sequences will resist enzymatic
digestion, allowing indirect detection of N6MedA based on the location of the
motif in the sequencing read. The motif at the end of a read indicates a lack
of methylation, while a read with the motif in the middle indicates the presence of
N6MedA. The amount of N6MedA can be calculated from the relative number of
sequencing reads where the specific adenosine is located at the end of a read
(unmethylated) or at an internal position (methylated). Using this technique,
N6MedA was identified at 3–4% of all available motifs in the C. reinhardtii genome
[122]. When N6MedA-containing sites identified by 6mA-RE-Seq were compared
to those from immunoprecipitation (IP), 28% of sites identified by IP were not
identified by 6mA-RE-Seq and did not contain the motifs used in 6mA-RE-Seq.
Using 6mA-RE-Seq and IP sequencing, Fu et al. found that N6MedA was enriched
in the regions of DNA linking nucleosomes possibly used for positioning these
nucleosomes on DNA [122].
4 Direct Detection of Epigenetic Modifications
The requirement for bisulfite conversion and other chemical reactions to locate DNA
epigenetic marks as described above in Sect. 3 introduces an inherent uncertainty
because the conversion rate is never 100%. In addition, these methods usually
involve several rounds of sequencing under different conditions, which increases
the cost, labor, and DNA sample input. Artifactual deamination of cytosine to uracil
during sample preparation and the required read alignment pose additional challenges with this approach. Third-generation sequencing methods allow for reading
the nucleotide sequences at the single-molecule level, making it possible to directly
distinguish between C, 5mC, and its oxidized forms during the sequencing step.
These methodologies are currently under active development.
4.1 Single-Molecule Real-Time (SMRT) Sequencing
As DNA sequencing methodologies entered the third generation, single-molecule
real-time (SMRT) technology has become a new favorite since it has longer read
lengths (10,000 bp up to 60,000 bp), produces higher consensus accuracy, and is
characterized by a lower degree of bias as compared to second-generation sequencing [123]. SMRT-Seq can directly read different base modifications via the kinetics
of nucleotide incorporation during the normal course of sequencing of an intact
DNA [124]. In SMRT-Seq, kinetic signatures of 5mC and 5hmC are very subtle,
presenting an accuracy challenge for detection, whereas 5fC and 5caC show very
distinct kinetics signatures [125]. More recently, it was shown that glycosylation of
Experimental Methodologies for Detection and Mapping of Epigenetic DNA Marks
511
DNA by CviAII or DpnII, sonication to fragment the DNA, end repair, adenylation,
and ligation to adapters [122]. The methylated sequences will resist enzymatic
digestion, allowing indirect detection of N6MedA based on the location of the
motif in the sequencing read. The motif at the end of a read indicates a lack
of methylation, while a read with the motif in the middle indicates the presence of
N6MedA. The amount of N6MedA can be calculated from the relative number of
sequencing reads where the specific adenosine is located at the end of a read
(unmethylated) or at an internal position (methylated). Using this technique,
N6MedA was identified at 3–4% of all available motifs in the C. reinhardtii genome
[122]. When N6MedA-containing sites identified by 6mA-RE-Seq were compared
to those from immunoprecipitation (IP), 28% of sites identified by IP were not
identified by 6mA-RE-Seq and did not contain the motifs used in 6mA-RE-Seq.
Using 6mA-RE-Seq and IP sequencing, Fu et al. found that N6MedA was enriched
in the regions of DNA linking nucleosomes possibly used for positioning these
nucleosomes on DNA [122].
4 Direct Detection of Epigenetic Modifications
The requirement for bisulfite conversion and other chemical reactions to locate DNA
epigenetic marks as described above in Sect. 3 introduces an inherent uncertainty
because the conversion rate is never 100%. In addition, these methods usually
involve several rounds of sequencing under different conditions, which increases
the cost, labor, and DNA sample input. Artifactual deamination of cytosine to uracil
during sample preparation and the required read alignment pose additional challenges with this approach. Third-generation sequencing methods allow for reading
the nucleotide sequences at the single-molecule level, making it possible to directly
distinguish between C, 5mC, and its oxidized forms during the sequencing step.
These methodologies are currently under active development.
4.1 Single-Molecule Real-Time (SMRT) Sequencing
As DNA sequencing methodologies entered the third generation, single-molecule
real-time (SMRT) technology has become a new favorite since it has longer read
lengths (10,000 bp up to 60,000 bp), produces higher consensus accuracy, and is
characterized by a lower degree of bias as compared to second-generation sequencing [123]. SMRT-Seq can directly read different base modifications via the kinetics
of nucleotide incorporation during the normal course of sequencing of an intact
DNA [124]. In SMRT-Seq, kinetic signatures of 5mC and 5hmC are very subtle,
presenting an accuracy challenge for detection, whereas 5fC and 5caC show very
distinct kinetics signatures [125]. More recently, it was shown that glycosylation of
Experimental Methodologies for Detection and Mapping of Epigenetic DNA Marks
511