Single-cell genome-wide bisulfite sequencing (scBS-seq): a promising strategy to investigate the DNA methylome
Abstract
Single-cell genome-wide bisulfite sequencing (scBS-seq), a third-generation sequencing technique, was found to assess the DNA methylation profile in rare cells, to explore the 5mC heterogeneity in cell populations and to identify sites in which methylation rate varied more than on the average genome.
DNA methylation is an epigenetic modification, typical of all living organisms, that in vertebrates can only occur at cytosines in the CG dinucleotides by DNA-methyltransferase which converts cytosine residues in 5methylcytosine (5mC). Methylated CG sequences are few and concentrated in genome regions called CpG Island (CGIs), very often located near the regulatory sequences of genes. DNA methylation has numerous biological functions, for example, it controls cell proliferation and differentiation by silencing gene expression and maintaining chromosomal stability in repeated genome sequences. Functional alterations have been associated with human carcinogenesis 1 and for this reason, studying in-depth the epigenome could be essential to understand underlying mechanisms of specific diseases. Elaborating a new technique based on earlier BS-seq2, but applied at single-cell level, Smallwood and collaborators, in 2014, writing in Nature reported a single-cell bisulfite sequencing method (scBS-seq), this new protocol is considered innovative because it allows to determine the heterogeneity of the 5mC on the entire DNA methylome of cell populations3. Cells are first isolated, then lysed and DNA fragments are converted with bisulfite: while 5mC are not susceptible to bisulfite, the unmethylated cells are converted into uracils. Cconverted fragments are then tagged and amplified by PCR and lastly sequenced and mapped (Figure 1). scBS-seq has been applied to two murine cell populations: ovulated metaphase II oocytes (MII), an excellent model for technical assessment and embryonic stem cells (ESC) grown in two different media, 2i and serum, in order to generate a heterogeneous population model.. Indeed it is known that 2i medium leads to the maintenance of the ESC’s pluripotency and therefore this cell population results more homogenous and hypomethylated, unlike the ESC cultivated in serum that express genes for differentiation pathways, resulting in being more heterogeneous and hypermethylated 4. scBS-seq applied to single MII and ESC cells has shown a low mapping efficiency in ESC cells compared to MII oocytes, due to the generation of poly-T contaminants, which has lower impact on a higher amount of chromosome cells (MII). By comparing different sequencing depths, the authors demonstrated that, in general, by increasing sequencing depth it is possible to increase the number of CpGs obtained, up to 48.8% of all CpGs. Furthermore, the analysis of the non-methylated CpGs allowed to determine bisulfite conversion efficiency at 97.7%. scBS-seq has proved to be a reproducible technique, with low technical variability. The correlation rate calculated, for each pair of cells, as the overlap proportion of CpG with identical state of methylation reflects the homogeneity/heterogeneity of the cell methylation pattern. Methylated CGIs analysis of all individual MII compared to the bulk libraries has also showed that ≤8% were incorrectly called methylated, and ≤0.3% were incorrectly called non-methylated, demonstrating a low error rate. The use of this technique enables to obtain methylation profiles information on all genomic context; particularly the enrichment in exons, promoters and CG islands has been observed. Despite scBS-seq is not able to sequences the DNA equally in all cells, the combination of 12 MII cell profiles was found to be representative of a bulk population, demonstrating this method to be an effective instrument to investigate the 5mC landscape in rare materials.
Figure 1. scBS-seq protocol performed on MIIs and ESC cultured either in 2i medium or serum conditions.
Moreover, the authors showed that this technique is able to detect the epigenetic heterogeneity of cells conducting an analysis of three genes loci, Dppa3, Nanog and Slc2a3, involved in maintaining the pluripotency of the ESC cells. The results have shown a very heterogeneous profile of methylation in serum cells, while in 2i medium the genes turned out to be rather hypomethylated. Monitoring the distribution of the methylation sites to the point of the entire genome, the authors observed that regions associated with active enhancers have larger variance, while in the CpGs islands the variance is lower. In fact, previous studies had shown that enhancers are differently methylated depending on the tissue and stage of development of each cell1. scBS-seq is therefore a powerful approach to distinguish different methylation degrees of individual cells and identifies sites whose methylation varies more than the genome average, even in cells that grow under the same conditions.
Taking these evidences together Smallwood et al. have definitely proposed an innovative technique that allows good sensitivity sequencing of the DNA, despite it is presently in a limited quantity compared to mRNA or proteins5. However, it remains an unusable tool for routine analysis because data sets from single cell studies are still small due to the high costs and technical limitations imposed by cell sorting6. The scBS-seq also does not allow the distinction between 5mC and 5hmC, since bisulfite acts only on unmethylated cytosines2 and in case the conversion is incomplete, it could result into false positive results1. Moreover, the authors observed that the efficiency depends on the amount of DNA inside the cell: greater is the number of copies, lower is the contribution of poly-T to the amount of DNA, higher the conversion efficiency. Future developments could therefore lead to overcoming these limitations, increasing information derived from genomic CpGs so that the DNA methylation profile will be considered the starting point for the pluripotent stem cells generation or a useful biomarker to detect the underlying mechanisms of many diseases, to define the transcriptional profile of heterogeneous tissues such as the human brain in order to characterize different types of tumours.
References
- Jones, P. A. Functions of DNA methylation: islands , start sites , gene bodies and beyond. Rev. Genet. 13, 484–492 (2012).
- Frommer, M. et al. A genomic sequencing protocol that yields a positive display of 5-methylcytosine residues in individual DNA strands. 89, 1827–1831 (1992).
- Smallwood, S. A. et al. Single-cell genome-wide bisulfite sequencing for assessing epigenetic heterogeneity. 11, 6–11 (2014).
- Habibi, E. et al. Short Article Whole-Genome Bisulfite Sequencing of Two Distinct Interconvertible DNA Methylomes of Mouse Embryonic Stem Cells. 360–369 (2013). doi:10.1016/j.stem.2013.06.002
- Schübeler, D. Function and information content of DNA methylation. (2015). doi:10.1038/nature14192
- Rahmani, E. et al. Cell-type-specific resolution epigenetics without the need for cell sorting or single-cell biology. Nat. Commun. 1–11 doi:10.1038/s41467-019-11052-9