Abstract
Oxidative modification of 5-methylcytosine (5mC) by ten-eleven translocation (TET) DNA dioxygenases generates 5-hydroxymethylcytosine (5hmC), the most abundant form of oxidized 5mC. Existing single-cell bisulfite sequencing methods cannot resolve 5mC and 5hmC, leaving the cell-type-specific regulatory mechanisms of TET and 5hmC largely unknown. Here, we present joint single-nucleus (hydroxy)methylcytosine sequencing (Joint-snhmC-seq), a scalable and quantitative approach that simultaneously profiles 5hmC and true 5mC in single cells by harnessing differential deaminase activity of APOBEC3A toward 5mC and chemically protected 5hmC. Joint-snhmC-seq profiling of single nuclei from mouse brains reveals an unprecedented level of epigenetic heterogeneity of both 5hmC and true 5mC at single-cell resolution. We show that cell-type-specific profiles of 5hmC or true 5mC improve multimodal single-cell data integration, enable accurate identification of neuronal subtypes and uncover context-specific regulatory effects on cell-type-specific genes by TET enzymes.
This is a preview of subscription content, access via your institution
Access options
Access Nature and 54 other Nature Portfolio journals
Get Nature+, our best-value online-access subscription
$29.99 / 30 days
cancel any time
Subscribe to this journal
Receive 12 print issues and online access
$209.00 per year
only $17.42 per issue
Buy this article
- Purchase on Springer Link
- Instant access to full article PDF
Prices may be subject to local taxes which are calculated during checkout
Similar content being viewed by others
Data availability
Raw and processed data associated with this study were deposited in the Gene Expression Omnibus under accession GSE236798 (ref. 69). The following publicly available datasets were used in this study: GSE97179 (snmC-seq)4. The plasmid used for purification of in-house APOBEC3A is available through Addgene (catalog no. 187822).
Code availability
The source codes for bioinformatic analysis of the paper are available on GitHub repository (https://github.com/wulabupenn/Joint-snhmC-seq)70.
References
Bird, A. DNA methylation patterns and epigenetic memory. Genes Dev. 16, 6–21 (2002).
Luo, C., Hajkova, P. & Ecker, J. R. Dynamic DNA methylation: in the right place at the right time. Science 361, 1336–1340 (2018).
Kelsey, G., Stegle, O. & Reik, W. Single-cell epigenomics: recording the past and predicting the future. Science 358, 69–75 (2017).
Luo, C. et al. Single-cell methylomes identify neuronal subtypes and regulatory elements in mammalian cortex. Science 357, 600–604 (2017).
Mulqueen, R. M. et al. Highly scalable generation of DNA methylation profiles in single cells. Nat. Biotechnol. 36, 428–431 (2018).
Smallwood, S. A. et al. Single-cell genome-wide bisulfite sequencing for assessing epigenetic heterogeneity. Nat. Methods 11, 817–820 (2014).
Plongthongkum, N., Diep, D. H. & Zhang, K. Advances in the profiling of DNA modifications: cytosine methylation and beyond. Nat. Rev. Genet. 15, 647–661 (2014).
Huang, Y. et al. The behaviour of 5-hydroxymethylcytosine in bisulfite sequencing. PLoS ONE 5, e8888 (2010).
Wu, H. & Zhang, Y. Reversing DNA methylation: mechanisms, genomics, and biological functions. Cell 156, 45–68 (2014).
Pastor, W. A., Aravind, L. & Rao, A. TETonic shift: biological roles of TET proteins in DNA demethylation and transcription. Nat. Rev. Mol. Cell Biol. 14, 341–356 (2013).
Wu, H. & Zhang, Y. Charting oxidized methylcytosines at base resolution. Nat. Struct. Mol. Biol. 22, 656–661 (2015).
Colquitt, B. M., Allen, W. E., Barnea, G. & Lomvardas, S. Alteration of genic 5-hydroxymethylcytosine patterning in olfactory neurons correlates with changes in gene expression and cell identity. Proc. Natl Acad. Sci. USA 110, 14682–14687 (2013).
Amir, R. E. et al. Rett syndrome is caused by mutations in X-linked MECP2, encoding methyl-CpG-binding protein 2. Nat. Genet. 23, 185–188 (1999).
Kinde, B., Gabel, H. W., Gilbert, C. S., Griffith, E. C. & Greenberg, M. E. Reading the unique DNA methylation landscape of the brain: Non-CpG methylation, hydroxymethylation, and MeCP2. Proc. Natl Acad. Sci. USA 112, 6800–6806 (2015).
Connolly, D. R. & Zhou, Z. Genomic insights into MeCP2 function: a role for the maintenance of chromatin architecture. Curr. Opin. Neurobiol. 59, 174–179 (2019).
Yu, M. et al. Base-resolution analysis of 5-hydroxymethylcytosine in the mammalian genome. Cell 149, 1368–1380 (2012).
Booth, M. J. et al. Quantitative sequencing of 5-methylcytosine and 5-hydroxymethylcytosine at single-base resolution. Science 336, 934–937 (2012).
Sun, Z. et al. High-resolution enzymatic mapping of genomic 5-hydroxymethylcytosine in mouse embryonic stem cells. Cell Rep. 3, 567–576 (2013).
He, Y. et al. DeepH&M: estimating single-CpG hydroxymethylation and methylation levels from enrichment and restriction enzyme sequencing methods. Sci. Adv. 6, eaba0521 (2020).
Cohen-Karni, D. et al. The MspJI family of modification-dependent restriction endonucleases for epigenetic studies. Proc. Natl Acad. Sci. USA 108, 11040–11045 (2011).
Schutsky, E. K. et al. Nondestructive, base-resolution sequencing of 5-hydroxymethylcytosine using a DNA deaminase. Nat. Biotechnol. 36, 1083–1090 (2018).
Liu, Y. et al. Bisulfite-free direct detection of 5-methylcytosine and 5-hydroxymethylcytosine at base resolution. Nat. Biotechnol. 37, 424–429 (2019).
Zeng, H. et al. Bisulfite-free, nanoscale analysis of 5-hydroxymethylcytosine at single base resolution. J. Am. Chem. Soc. 140, 13190–13194 (2018).
Fullgrabe, J. et al. Simultaneous sequencing of genetic and epigenetic bases in DNA. Nat. Biotechnol. https://doi.org/10.1038/s41587-022-01652-0 (2023).
Wang, T. et al. Direct enzymatic sequencing of 5-methylcytosine at single-base resolution. Nat. Chem. Biol. 19, 1004–1012 (2023).
Mooijman, D., Dey, S. S., Boisset, J. C., Crosetto, N. & van Oudenaarden, A. Single-cell 5hmC sequencing reveals chromosome-wide cell-to-cell variability and enables lineage reconstruction. Nat. Biotechnol. 34, 852–856 (2016).
Sen, M. et al. Strand-specific single-cell methylomics reveals distinct modes of DNA demethylation dynamics during early mammalian development. Nat. Commun. 12, 1286 (2021).
Nabel, C. S. et al. AID/APOBEC deaminases disfavor modified cytosines implicated in DNA demethylation. Nat. Chem. Biol. 8, 751–758 (2012).
Schutsky, E. K., Nabel, C. S., Davis, A. K. F., DeNizio, J. E. & Kohli, R. M. APOBEC3A efficiently deaminates methylated, but not TET-oxidized, cytosine bases in DNA. Nucleic Acids Res. 45, 7655–7665 (2017).
Wu, H., Wu, X., Shen, L. & Zhang, Y. Single-base resolution analysis of active DNA demethylation using methylase-assisted bisulfite sequencing. Nat. Biotechnol. 32, 1231–1240 (2014).
Caldwell, B. A. et al. Functionally distinct roles for TET-oxidized 5-methylcytosine bases in somatic reprogramming to pluripotency. Mol. Cell 81, 859–869 (2021).
Qiu, Q. et al. Massively parallel and time-resolved RNA sequencing in single cells with scNT-seq. Nat. Methods 17, 991–1001 (2020).
Lister, R. et al. Global epigenomic reconfiguration during Mammalian brain development. Science 341, 1237905 (2013).
Roadmap Epigenomics Consortium. et al. Integrative analysis of 111 reference human epigenomes. Nature 518, 317–330 (2015).
van der Velde, A. et al. Annotation of chromatin states in 66 complete mouse epigenomes during development. Commun. Biol. 4, 239 (2021).
Luo, C. et al. Robust single-cell DNA methylome profiling with snmC-seq2. Nat. Commun. 9, 3824 (2018).
Liu, H. et al. DNA methylation atlas of the mouse brain at single-cell resolution. Nature 598, 120–128 (2021).
Hu, P. et al. Dissecting cell-type composition and activity-dependent transcriptional state in mammalian brains by massively parallel single-nucleus RNA-seq. Mol. Cell 68, 1006–1015 (2017).
Kozlenkov, A. et al. A unique role for DNA (hydroxy)methylation in epigenetic regulation of human inhibitory neurons. Sci. Adv. 4, eaau6190 (2018).
Potapov, V. et al. Base modifications affecting RNA polymerase and reverse transcriptase fidelity. Nucleic Acids Res. 46, 5753–5763 (2018).
Bebenek, K., Joyce, C. M., Fitzgerald, M. P. & Kunkel, T. A. The fidelity of DNA synthesis catalyzed by derivatives of Escherichia coli DNA polymerase I. J. Biol. Chem. 265, 13878–13887 (1990).
Johnson, B. S. et al. Biotin tagging of MeCP2 in mice reveals contextual insights into the Rett syndrome transcriptome. Nat. Med. 23, 1203–1214 (2017).
Welch, J. D. et al. Single-cell multi-omic integration compares and contrasts features of brain cell identity. Cell 177, 1873–1887 (2019).
Carrillo-Jimenez, A. et al. TET2 regulates the neuroinflammatory response in microglia. Cell Rep. 29, 697–713 e698 (2019).
Spruijt, C. G. et al. Dynamic readers for 5-(hydroxy)methylcytosine and its oxidized derivatives. Cell 152, 1146–1159 (2013).
Iurlaro, M. et al. A screen for hydroxymethylcytosine and formylcytosine binding proteins suggests functions in transcription and chromatin regulation. Genome Biol. 14, R119 (2013).
Gabel, H. W. et al. Disruption of DNA-methylation-dependent long gene repression in Rett syndrome. Nature 522, 89–93 (2015).
Wu, F. et al. Spurious transcription causing innate immune responses is prevented by 5-hydroxymethylcytosine. Nat. Genet. 55, 100–111 (2023).
Wei, A. & Wu, H. Mammalian DNA methylome dynamics: mechanisms, functions and new frontiers. Development 149, dev182683 (2022).
Shareef, S. J. et al. Extended-representation bisulfite sequencing of gene regulatory elements in multiplexed samples and single cells. Nat. Biotechnol. 39, 1086–1094 (2021).
Nichols, R. V. et al. High-throughput robust single-cell DNA methylation profiling with sciMETv2. Nat. Commun. 13, 7627 (2022).
Angermueller, C. et al. Parallel single-cell sequencing links transcriptional and epigenetic heterogeneity. Nat. Methods 13, 229–232 (2016).
Luo, C. et al. Single nucleus multi-omics identifies human cortical cell regulatory genome diversity. Cell Genom. 2, 100107 (2022).
Pott, S. Simultaneous measurement of chromatin accessibility, DNA methylation, and nucleosome phasing in single cells. eLife 6, e23203 (2017).
Clark, S. J. et al. scNMT-seq enables joint profiling of chromatin accessibility DNA methylation and transcription in single cells. Nat. Commun. 9, 781 (2018).
Guo, F. et al. Single-cell multi-omics sequencing of mouse early embryos and embryonic stem cells. Cell Res. 27, 967–988 (2017).
Lee, D. S. et al. Simultaneous profiling of 3D genome structure and DNA methylation in single human cells. Nat. Methods 16, 999–1006 (2019).
Li, G. et al. Joint profiling of DNA methylation and chromatin architecture in single cells. Nat. Methods 16, 991–993 (2019).
Stoyanova, E., Riad, M., Rao, A. & Heintz, N. 5-Hydroxymethylcytosine-mediated active demethylation is required for mammalian neuronal differentiation and function. eLife 10, e66973 (2021).
Zhao, Y. T. et al. Long genes linked to autism spectrum disorders harbor broad enhancer-like chromatin domains. Genome Res 28, 933–942 (2018).
Hu, P. et al. Single-nucleus transcriptomic survey of cell diversity and functional maturation in postnatal mammalian hearts. Genes Dev. 32, 1344–1357 (2018).
Macosko, E. Z. et al. Highly parallel genome-wide expression profiling of individual cells using nanoliter droplets. Cell 161, 1202–1214 (2015).
Kechin, A., Boyarskikh, U., Kel, A. & Filipenko, M. cutPrimers: a new tool for accurate cutting of primers from reads of targeted next generation sequencing. J. Comput Biol. 24, 1138–1143 (2017).
Krueger, F. & Andrews, S. R. Bismark: a flexible aligner and methylation caller for Bisulfite-Seq applications. Bioinformatics 27, 1571–1572 (2011).
Li, H. et al. The sequence alignment/map format and SAMtools. Bioinformatics 25, 2078–2079 (2009).
Stuart, T. et al. Comprehensive integration of single-cell data. Cell 177, 1888–1902 (2019).
Lun, A. T. L. et al. EmptyDrops: distinguishing cells from empty droplets in droplet-based single-cell RNA sequencing data. Genome Biol. 20, 63 (2019).
Wolock, S. L., Lopez, R. & Klein, A. M. Scrublet: computational identification of cell doublets in single-cell transcriptomic data. Cell Syst. 8, 281–291 (2019).
Fabyanic E. B. et. al. Joint Single-Cell Profiling Resolves 5mC and 5hmC and Reveals their Distinct Gene Regulatory Effects. Datasets (Gene Expression Omnibus, 2023); https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE236798
Fabyanic E. B. et. al. Joint single-cell profiling resolves 5mC and 5hmC and reveals their distinct gene regulatory effects. Source code. GitHub https://github.com/wulabupenn/joint-snhmC-seq (2023).
Acknowledgements
We are grateful to members of the Wu and Kohli laboratories for helpful discussion. We thank W.Z. for the support with data analysis. This work was supported by the Penn Epigenetics Institute pilot grant (to R.M.K. and H.W.), a National Human Genome Research Institute (NHGRI) grant no. R01-HG010646 (to R.M.K. and H.W.), a National Heart Lung and Blood Institute grant no. DP2-HL142044 (to H.W.) and a NHGRI grant no. U01-HG012047 (to H.W.).
Author information
Authors and Affiliations
Contributions
E.B.F. and H.W. conceived and developed snhmC-seq and Joint-snhmC-seq. E.B.F. conducted most of the experiments. P.H. and H.W. performed computational analyses. P.H. developed the computational pipeline for analyzing snhmC-seq and Joint-snhmC-seq datasets with input from H.W. Q.Q., E.B.F., K.N.B. and J.F. performed in-house A3A enzyme production, purification and characterization. Q.Q. and E.B.F. performed sNucDrop-seq experiments. T.W. and R.M.K. contributed reagents and helped with the experimental design. E.B.F., D.R.C. and Z.Z. performed neuronal nuclei isolation and analysis. E.B.F., P.H., Q.Q. and H.W. analyzed the results and wrote the manuscript, with contributions from all the authors.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Peer review
Peer review information
Nature Biotechnology thanks the anonymous reviewers for their contribution to the peer review of this work.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Extended data
Extended Data Fig. 1 Workflow of whole-genome ACE-seq and bACE-seq.
Workflow comparison illustrating stepwise differences in cytosine modification status and 5hmC-protection between ACE-seq (all-enzyme, left) and bACE-seq (hybrid chemical and enzymatic conversion workflow, right), respectively.
Extended Data Fig. 2 Evaluation of WG-bACE-seq in analysis of nuclei from fresh frozen tissue samples.
(a) Workflow of WG-bACE-seq analysis of neuronal nuclei isolated from adult mouse brains (left). Bar plots (right) show the global 5hmCG level in Tet1/2/3 triple KO (Tet TKO) mESCs and mouse cortical neurons analyzed by various sample preparation protocols. As a positive control, ACE-seq (light blue) was used to analyze sheared and denatured gDNA from Tet TKO mESCs or wild-type mouse cortical neurons. WG-bACE-seq was applied to purified Tet TKO mESC gDNA (as a negative control) and sorted NeuN+ nuclei from mouse cortex. The black dot indicates experiments with DNA denaturation, while the grey dot indicates experiments without DNA denaturation. (b) Correlation density plot comparing 5hmCG levels in 1-Mb bins between standard ACE-seq (20ng purified, excitatory neuronal DNA) and WG-bACE-seq (100 neuronal NeuN+ nuclei). (c) Correlation density plots comparing global 5hmC level in 1-Mb bins between 10,000 and 100 NeuN+ nuclei with a pre-A3A denaturation step (left), between100 NeuN+ nuclei with the pre-A3A denaturing step and without denaturing (middle), and between Tet TKO mESC DNA and 100 NeuN+ nuclei (right). (d) Bar plot showing the cytosine modification levels in spike-in phage genomes between standard ACE-seq and WG-bACE-seq. The 5mCG and C deamination efficiency was assessed by the methylated lambda phage, and 5hmC protection was assessed by the mutant T4 (5hmC only) phage genome.
Extended Data Fig. 3 Integrated analysis of excitatory neuronal genomic DNA using paired WG-BS-seq and WG-bACE-seq.
(a) Schematic representation outlining mapping of both 5hmC and true 5mC by paired WG-BS-seq and WG-bACE-seq. (b) Schematic of the sample-splitting workflow for paired WG-BS-seq and WG-bACE-seq on the same pool of mouse excitatory neurons. (c) Scatterplots showing technical reproducibility of WG-BS-seq (top, 5mCG+5hmCG) and WG-bACE-seq (bottom, 5hmCG) between two biological replicates. (d) Genome browser tracks of 5hmCG (blue) and true 5mCG (red) signals at the Neurod6 gene locus. The 5mC level was calculated by directly subtracting WG-bACE-seq signals from WG-BS-seq. Two representative CG sites with different levels of 5mC and 5hmC were highlighted with a grey box. The sequencing coverage of the two methods are shown in grey tracks, respectively. (e) Correlation density plots comparing genome-wide 5hmCG versus 5mCG+5hmCG (left) or 5hmCH versus 5mCH+5hmCH (middle), across 1-Mb bins from WT mouse cortical neurons. The correlation density plot in the right panel compares genome-wide 5hmCH level in 1-Mb bins between WT mouse cortical neurons and Tet TKO mESCs. (f) Dot plots showing the absolute levels of 5hmCH (top) and CH methylation (bottom) in both WT Ex neurons (n = 2) and Tet TKO mESCs (n = 1) across all 12 non-CG sequence contexts. (g) Dot plots showing the absolute levels of 5hmCAC (top) and CAC methylation (bottom) in both WT Ex neurons (n = 2) and Tet TKO mESCs (n = 1) across genomic regions associated with 18 previously annotated chromatin states in mouse genomes (full annotations shown on the right)35.
Extended Data Fig. 4 FANS of excitatory (Ex) and inhibitory (Inh) neuronal subtypes from the mouse cortex.
(a) Schematic representation outlining the workflow for paired analysis of gene expression (using sNucDrop-seq) and DNA (hydoxy)methylomes (using snhmC-seq or Joint-snhmC-seq) in sorted single nuclei from adult mouse cortex. (b) Representative FANS results for isolating excitatory neurons (double positive (DP): NeuN+/Neurod6+), inhibitory neurons (single positive (SP): NeuN+/Neurod6-), and non-neuronal cells (double negative (DN): NeuN-/Neurod6-) from the mouse cortex of NeuroD6/NEX-Cre transgenic reporter mice. (c) UMAP visualization of mouse cortical nuclei (n = 6,000 nuclei) colored by cellular identity defined by single-nucleus RNA sequencing analysis (left panel) or FANS analysis (right panel). NeuN-neg: non-neuronal cells; NeuN-pos: neuronal cells; L2/3: layer 2/3 Ex neurons (n = 2,096 nuclei); L4: layer 4 Ex neurons (n = 754 nuclei); L4/5: layer 4/5 Ex neurons (n = 482 nuclei); L6: layer 6 Ex neurons (n = 900 nuclei); DL: deep-layer Ex neurons (n = 526 nuclei); Pv: parvalbumin-expressing Inh neurons (n = 215 nuclei); Sst: somatostatin-expressing Inh neurons (n = 174 nuclei); Vip/Ndnf: Vip/Ndnf-expressing Inh neurons (n = 125 nuclei); Striatum: striatal cells (n = 48 nuclei); Astro: astrocytes (n = 244 nuclei); MG: microglia (n = 157 nuclei); Oligo: oligodendrocytes (n = 279 nuclei).
Extended Data Fig. 5 Benchmarking snhmC-seq performance.
(a) UMAP visualization of 233 sorted cortical nuclei from wild-type mice (brain #1 and 2) colored by biological replicates (top left) and 127 sorted mouse cortical nuclei from a NEX-Cre transgenic reporter mouse (brain #3) colored by FANS analysis (top right). Also shown are all 360 mouse cortical nuclei colored by FANS sorting (bottom left) or cell types defined by snhmC-seq profiles (bottom right). (b) Boxplots showing mapping rates in snmC-seq (3,377 nuclei4) and snhmC-seq (360 nuclei). The boxes display the median (center line) and interquartile range (from the 25th to 75th percentile), the whiskers represent 1.5 times the interquartile range, and open white circles indicate the mean value. (c) Boxplot showing the mapping rates across 8-plex inline barcodes. The box plot elements are defined as in b. The number of nuclei from each of inline barcodes are shown on top. (d) Scatterplot showing covered CpG sites per nucleus as a function of uniquely mapped reads per nucleus for 8-plex inline barcodes. (e) Violin plots showing the global cytosine modification levels within CG (left) or CH (right) contexts in NeuN+ neuronal nuclei detected via snmC-seq (red: 3,377 nuclei) and snhmC-seq (blue: 256 nuclei). The open white circle indicates mean value in each group. See ‘Data visualization’ in the Methods for definitions of violin plot elements. (f) UMAP visualization of mouse cortical neuronal subtypes in the adult cortex, colored by cell-types defined by snhmC-seq (left: 262 neuronal nuclei) or by snmC-seq (right: 3,377 NeuN+ nuclei4). (g) Scatterplots comparing gene-body 5hmCH from snhmC-seq with genic CH methylation (5mCH+5hmCH) from snmC-seq (top), 5hmCG with CG methylation (5mCG+5hmCG) (middle), and 5hmCG with true 5mCG (bottom) in 7 major neuronal subtypes. True 5mCG was calculated by directly subtracting 5hmCG level (measured by snhmC-seq) from 5mCG+5hmCG level (measured by snmC-seq). The linear regression is indicated by a blue dashed line and Pearson’s correlation coefficient is shown above the plot.
Extended Data Fig. 6 Systematic optimization and benchmarking of A3A deamination reactions in snhmC-seq2.
(a) Boxplot illustrating a comparison of lambda spike-in 5mCG (left) or CH (right) non-conversion rate between snhmC-seq and snhmC-seq2, with mean values below each plot (in red). The boxes display the median (center line) and interquartile range (from the 25th to 75th percentile), the whiskers represent 1.5 times the interquartile range, and open white diamonds indicate the mean value. (b) SDS-PAGE analysis of the concentration and purity of both in-house wild-type A3A and New England Biolabs (NEB) A3A* enzymes. Shown is a representative result from two independent batches of in-house A3A purification using the improved pET-His6-MBP-A3A-Gly6His6 vector (Addgene, catalog no. 187822) and PreScission protease cleavage strategy. Both bathes show similar purity and total yield. The expected molecular weight of MBP-A3A-His (before cleavage) is ~70 kD, while that of untagged A3A is ~24 kD. Bovine serum albumin (BSA, ~66kD) serves as control to quantify the concentration of proteins. (c) Box plots illustrating deamination failure rate (top panel, 5mCG context; bottom panel, CH context) on in vitro CpG-methylated lambda phage gDNA across different concentrations of both NEB A3A* (orange) and in-house A3A (blue), with mean values below each plot. Open white circles indicate mean failure rates (indicated in red). The box plot elements are defined as in a. The number of nuclei (passed sequencing quality control filter) tested in each condition is shown on top.
Extended Data Fig. 7 Benchmarking snhmC-seq2 performance.
(a) Boxplot illustrating mapping rates across various iterations of snhmC-seq2 workflows, with mean values (in red) above each plot. Included are example bioanalyzer traces illustrating QC low (left: short fragment present) and QC high (right: short fragments removed) libraries. Also see Supplementary Note 3. The boxes display the median (center line) and interquartile range (from the 25th to 75th percentile), the whiskers represent 1.5 times the interquartile range, and open white diamonds indicate the mean value. The number of nuclei (passed sequencing quality control filter) tested in each experiment is shown on top. (b) Scatterplot comparing the covered CpG sites per nucleus as a function of sequencing depth (uniquely mapped reads per nucleus) among snhmC-seq2 (non-split, in red) and snhmC-seq (non-split, in grey) workflows. The line depicts linear regression fit of the data and shaded regions depict 95% confidence intervals. (c) Boxplots illustrating mapping rate across 8-plex inline barcodes for snhmC-seq2/split (top) and snmC-seq2/split (bottom). Open white circles indicate the mean value. The boxes are defined as in the b, and open white circles indicate the mean value. The number of nuclei from each of the eight inline barcode groups are shown on the top for both snhmC-seq2/split and snmC-seq2/split modalities. Only QC.high libraries are shown. (d) Violin plots illustrating genomic coverage (averaged single-cell coverage (%) is shown in red) across 100-kb bins in snhmC-seq2/split (blue) and snmC-seq2/split (green) modalities of Joint-snhmC-seq. The boxes are defined as in the b. The number of nuclei from both snhmC-seq2/split and snmC-seq2/split modalities are shown on top.
Extended Data Fig. 8 Cell-type-specific relationship between 5hmC and 5mC or gene expression levels.
(a) Scatterplots comparing gene-body 5hmCG with genic CG methylation (5mCG+5hmCG) (upper panel), and 5hmCG with true 5mCG (bottom panel) in six major neuronal and non-neuronal cell-types identified by Joint-snhmC-seq. True 5mCG was calculated by directly subtracting 5hmCG level measured by snhmC-seq2/split from 5mCG+5hmCG level measured by snmC-seq2/split. The linear regression is indicated by a blue line is superimposed on each plot. Both P-values (upper: P; Pearson correlation test, two-sided) and the Pearson’s correlation coefficients (lower: cor) for each comparison are shown in red. (b) Scatterplots comparing genic CG methylation (5mCG+5hmCG), true 5mCG, 5hmCG, and CH methylation (5mCH+5hmCH) with gene expression (normalized RNA levels) in six major neuronal and non-neuronal cell-types. The linear regression is indicated by a blue line and Pearson’s correlation coefficient is superimposed on each plot.
Extended Data Fig. 9 Joint-snhmC-seq reveals distinct regulatory strategies of neuronal subtype-specific genes.
(a) Scatterplots comparing genic 5hmCG, true 5mCG, unmodified CG and CH methylation (5mCH+5hmCH) with gene expression (log2-scaled, normalized RNA levels) in four representative neuronal subtypes. The linear regression is indicated by a blue line. The log10-scaled gene length (in kb) is color coded (long: red; short: blue). (b) Box plots showing absolute levels of genic 5hmCG, true 5mCG and unmodified CG as well as scaled RNA levels for eight representative cortical Inh neuron-specific marker genes. The canonical marker genes (marked by a genic hypomethylation signature in expressing cell-types) are highlighted by gray boxes (left), whereas non-canonical (marked by a genic hyper-hydroxymethylation signature) are highlighted by blue boxes. In the box plots, the boxes display the median (center line) and interquartile range (from the 25th to 75th percentile), the whiskers represent 1.5 times the interquartile range, and open red diamonds indicate the mean value. L2/3: 87 nuclei, L4: 44 nuclei, L5: 36 nuclei, L6: 41 nuclei, DL: 27 nuclei, Pv: 61 nuclei, Sst: 60 nuclei, and Vip/Ndnf: 35 nuclei.
Extended Data Fig. 10 Joint-snhmC-seq reveals gene length-dependent regulation of neuronal subtype-specific genes by TET enzymes.
(a) Ternary plots showing genic 5mCG, 5hmC and unmodified CG levels (%) for all genes in 12 major cortical neuronal and non-neuronal cell-types. The log2-scaled, normalized RNA levels are color-coded (red: high; blue: low). (b) Scatter plots showing the relationship between 5hmCG and unmodified CG for highly transcribed (log2-scaled, normalized RNA > 2) genes in 9 major cortical neuronal and non-neuronal cell-types. The log10-scaled gene length (in kb) is color coded (long: red; short: blue). The log2-scaled, normalized RNA levels are depicted by circle sizes. The line depicts polynomial regression fit of the data and shaded regions depict 95% confidence intervals.
Supplementary information
Supplementary Information
Supplementary Figs. 1–4, Sequences 1–3, Notes 1–3 and Tables 2 and 3.
Supplementary Tables 1, 4, 5 and 6
Sequencing summary statistics information for snhmC-seq and Joint-snhmC-seq is listed in these Supplementary Tables.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Fabyanic, E.B., Hu, P., Qiu, Q. et al. Joint single-cell profiling resolves 5mC and 5hmC and reveals their distinct gene regulatory effects. Nat Biotechnol (2023). https://doi.org/10.1038/s41587-023-01909-2
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41587-023-01909-2
This article is cited by
-
Joint single-cell profiling resolves 5mC and 5hmC and reveals their distinct gene regulatory effects
Nature Biotechnology (2023)
