Mining the DNA methylome can provide a treasure trove of biological insight, and two new whole-genome DNA methylation sequencing methods have come along to fuel a technological gold rush. NanoMethPhase and scMspJI-seq break new ground in phasing allele-specific methylation (ASM) and generating strand-specific single-cell methylomes. By applying these novel approaches, the talented teams dig deep into complex DNA methylation patterns that have long been particularly challenging to measure.
Megabase Methylation Phasing with NanoMethPhase
Steven Jones’ team (University of British Columbia, Canada) developed NanoMethPhase and SNVoter to help with utilizing nanopore long-read sequencing to map allele-specific methylation (ASM) without the need for sodium bisulfite or enzymatic conversion.
ASM detection requires phasing of sequencing reads, which is achieved by identifying haplotype-specific SNPs. Therefore, both DNA methylation and DNA sequence must be extracted from the same data. Contrary to short reads, long reads are more likely to span multiple SNPs enabling their mapping to a contiguous haplotype. However, a major barrier in nanopore long-read sequencing is the high error rate in identifying single nucleotide variants (SNVs), such as SNPs. Here’s what this dynamic duo of new software offers:
- SNVoter improves SNV calling in low coverage nanopore data (~10x) from a single sample by filtering out low-quality SNPs
- This reduces false-positive calling errors with minimal impact on the number of reads spanning multiple SNPs
- Now that the haplotype phasing hurdle has been surmounted, NanoMethPhase comes into play for DNA methylation phasing by leveraging both DNA methylation and SNVs from one nanopore data sample to map ASM
- NanoMethPhase ASM data is concordant with well-known human imprinting control regions and differentially methylated regions (DMR) on the X chromosome
- DMR mapping showed promoter hypermethylation and gene body hypomethylation of inactive X-linked genes, compared to genes that escape chromosome X inactivation
First author Vahid Akbari shares, “SNPs are used to assign each sequencing read to its paternal or maternal haplotype. On average, each person’s genome has one SNP per kilobase, so you can imagine the challenge of detecting haplotypes genome wide. In the past, whole-genome bisulfite short-read sequencing coupled with genotyping needed to be done on multiple samples to obtain non-overlapping SNPs and fill the missing gaps. With NanoMethPhase and SNVoter, we are able to have near complete coverage of ASM in a single sample, which not only reduces the cost significantly but also enables the study of case-specific haplotype methylomes.”
Strand-Specific Single-Cell Methylomes with scMspJI-seq
Sometimes, there’s more than one way to strike gold. The scMspJI-seq method, developed by Siddharth S. Dey’s (University of California Santa Barbara) and Alexander van Oudenaarden’s (Hubrecht Institute, The Netherlands) labs, generates strand-specific single-cell methylomes on a genome-wide scale. Here’s how scMspJI-seq works:
- Single cells are isolated into 384-well plates and all subsequent sample preparation steps are preformed using a liquid-handling platform, which allows for high-throughput and highly precise low-volume pipetting
- Following cell lysis and chromatin protein removal, T4 phage β-glucosyltransferase (T4-βGT) is added to glucosylate 5-hydroxymethylcytosine (5hmC), which blocks downstream detection of 5hmC thereby allowing exclusive detection of 5mC
- The modification-dependent restriction endonuclease MspJI induces double-stranded DNA breaks 16 bp downstream of the methylated cytosines, leaving a 4-nucleotide 5′ overhang
- DNA adapters designed to target these 4-nucleotide 5′ overhangs are ligated to the fragmented gDNA molecules
- To obtain single-cell and strand-specific methylation data from the same sample, these aptamers contain a cell-specific barcode and a random 3bp unique molecule identifier (UMI) to individually label each 5mC site on each strand separately
- The aptamers also contain a T7 promoter for amplification by in vitro transcription and a 5’ Illumina adapter for library preparation
The scMspJI-seq method provides a significant technological upgrade for studying epigenetic reprogramming in early mammalian development. After fertilization, the embryo must undergo genome-wide demethylation to erase sperm- and oocyte-specific DNA methylation marks through active and passive demethylation processes. The added strand specificity in scMspJI-seq provides additional information about asymmetrical DNA methylation dynamics, which is essential for distinguishing active demethylation (symmetric) versus passive demethylation (asymmetric). Strand-specific DNA methylation analysis with scMspJI-seq could also be highly valuable for studying reprogramming of somatic cells to induced pluripotent stem cells.
The Future of Mining the DNA Methylome
To strike gold in DNA methylation mining, the right method must be meticulously chosen. As we progress towards the next generations of DNA methylation sequencing technologies, specialized methods and software like NanoMethPhase and scMspJI-seq that deliver on the specific needs of an experiment will likely become the new normal and may potentially allow us to dig even deeper into that gold mine.
Single in on all the details with NanoMethPhase in Genome Biology, February 2021 and scMspJI-seq in Nature Communications, February 2021.