The core problem the method solves is heterogeneity. Bulk RRBS, first developed more than a decade and a half ago, averages methylation signals across thousands or millions of cells, producing a consensus methylome that conceals the very variation that matters most in development and disease. In a tumor, or in the blood-forming stem cells of a patient with myelodysplastic syndrome, individual cells can carry strikingly different methylation patterns even when they look identical under the microscope. Single-cell bisulfite sequencing approaches have existed since 2013, but early versions were laborious, low-throughput and prone to losing the tiny amount of DNA contained in one cell — roughly six picograms in a human cell, a quantity easily destroyed by transfers between tubes.

Msc-RRBS attacks that fragility by collapsing the entire early workflow into a single microwell. After fluorescence-activated cell sorting places one cell per well, the cell is lysed, its DNA is cut with the restriction enzyme MspI, the fragment ends are repaired and A-tailed, and short indexed adapters are ligated — all without opening the well or exchanging buffers. Every transfer is a chance to lose DNA, so eliminating them is the single most important design decision in the protocol. The restriction digestion also serves the classic RRBS purpose: MspI cuts at CCGG sites, enriching the library for CpG-dense regions such as promoters and CpG islands, where biologically meaningful methylation changes concentrate.

The multiplexing strategy is where the protocol earns its throughput. Twenty-four wells are pooled into one tube, and each cell’s DNA carries a six-base inline barcode added during adapter ligation. After pooling, adapter dimers — the wasteful self-ligation products that plague low-input library preparations — are removed by magnetic SPRI bead cleanup. Because the Msc-RRBS adapters are deliberately shortened, their dimers measure only about 60 base pairs rather than the roughly 120 base pairs produced by standard Illumina adapters, making them easy to strip away without sacrificing real library fragments. A second layer of PCR-added indices then tags each pool, so every sequencing read carries two barcodes: an internal one identifying the cell and an external one identifying the pool.

Once pooled libraries are sequenced, the bioinformatics begins with demultiplexing on the inline barcodes at the five-prime ends of both read pairs, followed by trimming of barcodes and end-repair motifs. The cleaned reads are aligned to a reference genome, and methylation calls are extracted at individual cytosines. The authors have released a complete analysis pipeline, SingleCellMethPipe, on GitHub, designed to run on standard high-performance computing environments. Basic quality control and primary data processing, they note, can be handled by a researcher with limited bioinformatics training; deeper analyses such as differential methylation testing and lineage inference require more computational experience.

The protocol’s flexibility is one of its quiet strengths. It works not only on sorted single cells but also on pooled cell populations and on sub-nanogram quantities of purified genomic DNA, making it useful for archived clinical material where intact cells may be scarce. The full workflow — from cell sorting through bisulfite conversion, PCR, sequencing and analysis — takes about six days for one or more 96-well plates, and the authors report that the experiments demand only basic molecular biology skills. Cost estimates provided in the supplementary tables suggest the per-cell expense is low enough to support studies of dozens or hundreds of samples, a scale previously out of reach for most single-cell methylation experiments.

The demonstration study puts the method to work on a clinically urgent question. The team profiled hematopoietic stem and progenitor cells from healthy donors and from patients with myelodysplastic syndrome, a bone marrow failure disorder that often progresses to acute myeloid leukemia. Using differentially methylated CpG islands as features, hierarchical clustering with Ward’s linkage and Spearman correlation distance cleanly separated healthy from MDS-derived cells and distinguished stem cells from progenitors. The analysis revealed promoter methylation differences at specific CpG islands, offering a single-cell view of the epigenetic disarray that accompanies the disease — the kind of cell-to-cell variation a bulk assay would average into invisibility.

That capability matters because methylation heterogeneity is increasingly understood to be a driver of disease evolution rather than a byproduct. In chronic lymphocytic leukemia, earlier single-cell methylation studies showed that epigenetic drift can reconstruct the lineage history of a tumor, revealing how subclones diverge over years. Msc-RRBS brings that kind of analysis within reach of laboratories that lack access to droplet-based or combinatorial-indexing platforms, using equipment as ordinary as a fluorescence-activated cell sorter and a multichannel pipette. The protocol’s balance of throughput, cost and data quality positions it as a practical workhorse for epigenomic studies across development, aging and cancer.

The method does have constraints worth noting. Reduced-representation approaches capture only the fraction of the genome adjacent to MspI sites — roughly the CpG-rich promoters and islands where most regulatory methylation lives — rather than the whole genome at base resolution. Coverage per cell is therefore shallower than what whole-genome bisulfite sequencing achieves, and researchers needing complete methylomes must look to alternatives such as single-cell whole-genome bisulfite sequencing or snmC-seq2. Bisulfite conversion itself also degrades DNA, so library complexity depends on starting with intact, high-quality material. Still, for the vast majority of questions about cell identity and disease state, the CpG-island-focused view is precisely the information that matters.

What makes this protocol newsworthy is less any single technical trick than the accumulation of them: one-well enzymatics, shortened adapters that suppress dimers, dual barcoding, magnetic-bead cleanup and a ready-made analysis pipeline. Together they convert a technique once reserved for specialist epigenetics laboratories into something a molecular biology lab can adopt in a week. As single-cell atlases expand from transcriptomes to epigenomes, tools like Msc-RRBS will determine how quickly the methylation dimension of cellular diversity — in healthy tissue, in aging organs and in the evolving chaos of cancer — can be mapped at scale. The sequencing data and analysis code are publicly available, lowering the barrier for laboratories worldwide to begin charting the methylation landscapes hidden inside individual cells.

Subject of Research: Multiplexed single-cell reduced representation bisulfite sequencing for genome-wide DNA methylation profiling

Article Title: Multiplexed single-cell reduced representation bisulfite sequencing for genome-wide DNA methylation profiling

Article References: Liu, F., Huang, J., Zheng, Y., Fu, L., Wu, X., Zhao, N., Wang, Y., Gnirke, A., Meissner, A., Wang, X., Zhang, F., & Gu, H. (2026). Multiplexed single-cell reduced representation bisulfite sequencing for genome-wide DNA methylation profiling. Nature Protocols. https://doi.org/10.1038/s41596-026-01452-9

Image Credits: AI Generated

DOI: 10.1038/s41596-026-01452-9

Keywords: Multiplexed, single-cell, reduced, representation, bisulfite, sequencing, genome-wide, methylation, profiling, scientific research