Competition between DNA methylation and transcription factors determines binding of NRF1 (original) (raw)

Accession codes

Primary accessions

Gene Expression Omnibus

Data deposits

Genome-wide datasets generated for this study are deposited at GEO under the accession number GSE67867.

References

  1. Slattery, M. et al. Absence of a simple code: how transcription factors read the genome. Trends Biochem. Sci. 39, 381–399 (2014)
    Article CAS PubMed PubMed Central Google Scholar
  2. Iwafuchi-Doi, M. & Zaret, K. S. Pioneer transcription factors in cell reprogramming. Genes Dev. 28, 2679–2692 (2014)
    Article PubMed PubMed Central Google Scholar
  3. Tate, P. H. & Bird, A. P. Effects of DNA methylation on DNA-binding proteins and gene expression. Curr. Opin. Genet. Dev. 3, 226–231 (1993)
    Article CAS PubMed Google Scholar
  4. Becker, P. B., Ruppert, S. & Schütz, G. Genomic footprinting reveals cell type-specific DNA binding of ubiquitous factors. Cell 51, 435–443 (1987)
    Article CAS PubMed Google Scholar
  5. Weih, F., Nitsch, D., Reik, A., Schütz, G. & Becker, P. B. Analysis of CpG methylation and genomic footprinting at the tyrosine aminotransferase gene: DNA methylation alone is not sufficient to prevent protein binding in vivo . EMBO J. 10, 2559–2567 (1991)
    Article CAS PubMed PubMed Central Google Scholar
  6. Bell, A. C. & Felsenfeld, G. Methylation of a CTCF-dependent boundary controls imprinted expression of the Igf2 gene. Nature 405, 482–485 (2000)
    Article CAS ADS PubMed Google Scholar
  7. Hark, A. T. et al. CTCF mediates methylation-sensitive enhancer-blocking activity at the H19/Igf2 locus. Nature 405, 486–489 (2000)
    Article CAS ADS PubMed Google Scholar
  8. Stadler, M. B. et al. DNA-binding factors shape the mouse methylome at distal regulatory regions. Nature 480, 490–495 (2011)
    Article CAS ADS PubMed Google Scholar
  9. Maurano, M. T. et al. Role of DNA methylation in modulating transcription factor occupancy. Cell Rep . 12, 1184–1195 (2015)
    Article CAS PubMed Google Scholar
  10. Feldmann, A. et al. Transcription factor occupancy can mediate active turnover of DNA methylation at regulatory regions. PLoS Genet. 9, e1003994 (2013)
    Article PubMed PubMed Central Google Scholar
  11. Wu, H. & Zhang, Y. Reversing DNA methylation: mechanisms, genomics, and biological functions. Cell 156, 45–68 (2014)
    Article CAS PubMed PubMed Central Google Scholar
  12. Ziller, M. J. et al. Charting a dynamic DNA methylation landscape of the human genome. Nature 500, 477–481 (2013)
    Article CAS ADS PubMed PubMed Central Google Scholar
  13. Jones, P. A. Functions of DNA methylation: islands, start sites, gene bodies and beyond. Nature Rev. Genet. 13, 484–492 (2012)
    Article CAS PubMed Google Scholar
  14. Schübeler, D. Function and information content of DNA methylation. Nature 517, 321–326 (2015)
    Article ADS PubMed Google Scholar
  15. Tsumura, A. et al. Maintenance of self-renewal ability of mouse embryonic stem cells in the absence of DNA methyltransferases Dnmt1, Dnmt3a and Dnmt3b. Genes Cells 11, 805–814 (2006)
    Article CAS PubMed Google Scholar
  16. Karimi, M. M. et al. DNA methylation and SETDB1/H3K9me3 regulate predominantly distinct sets of genes, retroelements, and chimeric transcripts in mESCs. Cell Stem Cell 8, 676–687 (2011)
    Article CAS PubMed PubMed Central Google Scholar
  17. Neph, S. et al. An expansive human regulatory lexicon encoded in transcription factor footprints. Nature 489, 83–90 (2012)
    Article CAS ADS PubMed PubMed Central Google Scholar
  18. Virbasius, C. A., Virbasius, J. V. & Scarpulla, R. C. NRF-1, an activator involved in nuclear–mitochondrial interactions, utilizes a new DNA-binding domain conserved in a family of developmental regulators. Genes Dev. 7, 2431–2445 (1993)
    Article CAS PubMed Google Scholar
  19. Kumari, D. & Usdin, K. Interaction of the transcription factors USF1, USF2, and α-Pal/Nrf-1 with the FMR1 promoter. Implications for Fragile X mental retardation syndrome. J. Biol. Chem. 276, 4357–4364 (2001)
    Article CAS PubMed Google Scholar
  20. Spruijt, C. G. et al. Dynamic readers for 5-(hydroxy)methylcytosine and its oxidized derivatives. Cell 152, 1146–1159 (2013)
    Article CAS PubMed Google Scholar
  21. Hu, S. et al. DNA methylation presents distinct binding sites for human transcription factors. eLife 2, e00726 (2013)
    Article PubMed PubMed Central Google Scholar
  22. Baubec, T., Ivanek, R., Lienert, F. & Schübeler, D. Methylation-dependent and -independent genomic targeting principles of the MBD protein Family. Cell 153, 480–492 (2013)
    Article CAS PubMed Google Scholar
  23. Borgel, J. et al. Targets and dynamics of promoter DNA methylation during early mouse development. Nature Genet. 42, 1093–1100 (2010)
    Article CAS PubMed Google Scholar
  24. Ficz, G. et al. FGF signaling inhibition in ESCs drives rapid genome-wide demethylation to the epigenetic ground state of pluripotency. Cell Stem Cell 13, 351–359 (2013)
    Article CAS PubMed PubMed Central Google Scholar
  25. Habibi, E. et al. Whole-genome bisulfite sequencing of two distinct interconvertible DNA methylomes of mouse embryonic stem cells. Cell Stem Cell 13, 360–369 (2013)
    Article CAS PubMed Google Scholar
  26. Hon, G. C. et al. Global DNA hypomethylation coupled to repressive chromatin domain formation and gene silencing in breast cancer. Genome Res. 22, 246–258 (2012)
    Article CAS PubMed PubMed Central Google Scholar
  27. ENCODE Project Consortium An integrated encyclopedia of DNA elements in the human genome. Nature 489, 57–74 (2012)
  28. Lienert, F. et al. Identification of genetic elements that autonomously determine DNA methylation states. Nature Genet. 43, 1091–1097 (2011)
    Article CAS PubMed Google Scholar
  29. Krebs, A. R., Dessus-Babus, S., Burger, L. & Schübeler, D. High-throughput engineering of a mammalian genome reveals building principles of methylation states at CG rich regions. eLife 3, e04094 (2014)
    Article PubMed PubMed Central Google Scholar
  30. Sherwood, R. I. et al. Discovery of directional and nondirectional pioneer transcription factors by modeling DNase profile magnitude and shape. Nature Biotechnol. 32, 171–178 (2014)
    Article CAS Google Scholar
  31. Jørgensen, H. F., Chen, Z. -F., Merkenschlager, M. & Fisher A. G. Is REST is required for ESC pluripotency? Nature 457, E4–E5, E7 (2009)
    Article ADS PubMed Google Scholar
  32. Chen, Z. F., Paquette, A. J. & Anderson, D. J. NRSF/REST is required in vivo for repression of multiple neuronal target genes during embryogenesis. Nature Genet. 20, 136–142 (1998)
    Article CAS PubMed Google Scholar
  33. Bibel, M., Richter, J., Lacroix, E. & Barde, Y.-A. Generation of a defined and uniform population of CNS progenitors and neurons from mouse embryonic stem cells. Nature Protocols 2, 1034–1043 (2007)
    Article CAS PubMed Google Scholar
  34. John, S. et al. Genome-scale mapping of DNase I hypersensitivity. Curr. Protoc. Mol. Biol. Chapter 27, Unit 21.27–21.27.20 (2013)
    PubMed Google Scholar
  35. Jermann, P., Hoerner, L., Burger, L. & Schübeler, D. Short sequences can efficiently recruit histone H3 lysine 27 trimethylation in the absence of enhancer activity and DNA methylation. Proc. Natl Acad. Sci. USA 111, E3415–E3421 (2014)
    Article CAS ADS PubMed PubMed Central Google Scholar
  36. Schübeler, D. et al. Genomic targeting of methylated DNA: influence of methylation on transcription, replication, chromatin structure, and histone acetylation. Mol. Cell. Biol. 20, 9103–9112 (2000)
    Article PubMed PubMed Central Google Scholar
  37. Trapnell, C., Pachter, L. & Salzberg, S. L. TopHat: discovering splice junctions with RNA-seq. Bioinformatics 25, 1105–1111 (2009)
    Article CAS PubMed PubMed Central Google Scholar
  38. Langmead, B., Trapnell, C., Pop, M. & Salzberg, S. L. Ultrafast and memory-efficient alignment of short DNA sequences to the human genome. Genome Biol. 10, R25 (2009)
    Article PubMed PubMed Central Google Scholar
  39. Gaidatzis, D., Lerch, A., Hahne, F. & Stadler, M. B. QuasR: quantification and annotation of short reads in R. Bioinformatics 31, 1130–1132 (2015)
    Article CAS PubMed Google Scholar
  40. Kent, W. J. et al. The Human Genome Browser at UCSC. Genome Res. 12, 996–1006 (2002)
    Article CAS PubMed PubMed Central Google Scholar
  41. Bardet, A. F. et al. Identification of transcription factor binding sites from ChIP-seq data at high resolution. Bioinformatics 29, 2705–2713 (2013)
    Article CAS MathSciNet PubMed PubMed Central Google Scholar
  42. Anders, S. & Huber, W. Differential expression analysis for sequence count data. Genome Biol. 11, R106 (2010)
    Article CAS PubMed PubMed Central Google Scholar
  43. Xie, W. et al. Base-resolution analyses of sequence and parent-of-origin dependent DNA methylation in the mouse genome. Cell 148, 816–831 (2012)
    Article CAS PubMed PubMed Central Google Scholar
  44. Siepel, A. Evolutionarily conserved elements in vertebrate, insect, worm, and yeast genomes. Genome Res. 15, 1034–1050 (2005)
    Article CAS PubMed PubMed Central Google Scholar
  45. Sandelin, A. JASPAR: an open-access database for eukaryotic transcription factor binding profiles. Nucleic Acids Res. 32, D91–D94 (2004)
    Article CAS PubMed PubMed Central Google Scholar
  46. Jolma, A. et al. DNA-binding specificities of human transcription factors. Cell 152, 327–339 (2013)
    Article CAS PubMed Google Scholar
  47. Newburger, D. E. & Bulyk, M. L. UniPROBE: an online database of protein binding microarray data on protein–DNA interactions. Nucleic Acids Res. 37, D77–D82 (2009)
    Article CAS PubMed Google Scholar
  48. Bailey, T. L. & Gribskov, M. Combining evidence using p-values: application to sequence homology searches. Bioinformatics 14, 48–54 (1998)
    Article CAS PubMed Google Scholar
  49. Marks, H. et al. The transcriptional and epigenomic foundations of ground state pluripotency. Cell 149, 590–604 (2012)
    Article CAS PubMed PubMed Central Google Scholar
  50. Tippmann, S. C. et al. Chromatin measurements reveal contributions of synthesis and decay to steady-state mRNA levels. Mol. Syst. Biol. 8, 593 (2012)
    Article PubMed PubMed Central Google Scholar
  51. Yue, F. et al. A comparative encyclopedia of DNA elements in the mouse genome. Nature 515, 355–364 (2014)
    Article CAS PubMed PubMed Central Google Scholar
  52. Arnold, P. et al. Modeling of epigenome dynamics identifies transcription factors that mediate Polycomb targeting. Genome Res. 23, 60–73 (2013)
    Article CAS PubMed PubMed Central Google Scholar
  53. Chen, X. et al. Integration of external signaling pathways with the core transcriptional network in embryonic stem cells. Cell 133, 1106–1117 (2008)
    Article CAS PubMed Google Scholar
  54. Marson, A. et al. Connecting microRNA genes to the core transcriptional regulatory circuitry of embryonic stem cells. Cell 134, 521–533 (2008)
    Article CAS PubMed PubMed Central Google Scholar

Download references

Acknowledgements

We are grateful to S. Dessus-Babus, K. Jacobeit and T. Roloff (FMI) for processing deep-sequencing samples, to C. Wirbelauer for technical assistance and to A. Arnold for technical advice. We thank M. Stadler and D. Gaidatzis for bioinformatic advice and members of our laboratory, N. Thomae (FMI) and M. Lorincz (UBC Vancouver) for comments on the manuscript. We apologize to colleagues whose work we could not cite owing to space limitations. Research in the laboratory of D.S. is supported by the Novartis Research Foundation, the European Union (NoE ‘EpiGeneSys’ FP7-HEALTH-2010-257082 and the ‘Blueprint’ consortium FP7-282510), the European Research Council (EpiGePlas) and the Swiss initiative in Systems Biology (RTD Cell Plasticity). A.F.B. and P.A.G. are supported by EMBO postdoctoral long-term fellowships and S.D. and D.H. by predoctoral fellowships from the Boehringer Ingelheim Fonds.

Author information

Author notes

  1. Silvia Domcke and Anaïs Flore Bardet: These authors contributed equally to this work.

Authors and Affiliations

  1. Friedrich Miescher Institute for Biomedical Research, Maulbeerstrasse 66, CH 4058, Basel, Switzerland
    Silvia Domcke, Anaïs Flore Bardet, Paul Adrian Ginno, Dominik Hartl, Lukas Burger & Dirk Schübeler
  2. University of Basel, Faculty of Sciences, Petersplatz 1, CH 4003, Basel, Switzerland
    Silvia Domcke, Dominik Hartl & Dirk Schübeler
  3. Swiss Institute of Bioinformatics, Maulbeerstrasse 66, CH 4058, Basel, Switzerland
    Lukas Burger

Authors

  1. Silvia Domcke
    You can also search for this author inPubMed Google Scholar
  2. Anaïs Flore Bardet
    You can also search for this author inPubMed Google Scholar
  3. Paul Adrian Ginno
    You can also search for this author inPubMed Google Scholar
  4. Dominik Hartl
    You can also search for this author inPubMed Google Scholar
  5. Lukas Burger
    You can also search for this author inPubMed Google Scholar
  6. Dirk Schübeler
    You can also search for this author inPubMed Google Scholar

Contributions

A.F.B., L.B., S.D. and D.S. initiated and designed the study; S.D. performed the experiments; A.F.B. performed the data analysis; S.D. contributed to data analysis; S.D. and P.A.G. generated the TKO cell line; D.H. generated the overexpression construct; L.B. advised on data analysis; D.S. supervised all aspects of the project; the manuscript was prepared by S.D., A.F.B. and D.S. All authors discussed results and commented on the manuscript.

Corresponding author

Correspondence toDirk Schübeler.

Ethics declarations

Competing interests

The authors declare no competing financial interests.

Extended data figures and tables

Extended Data Figure 1 Characterization of an isogenic DNMT TKO cell line created with CRISPR/Cas9.

a, Frameshift deletions (brown) introduced at the active PCQ/N loops of the three DNA methyltransferases by CRISPR/Cas9 genome editing. b, Levels of 5-methyl-C and 5-hydroxy-methyl-C in the wild-type, isogenic (mouse ES cell line 159) and traditional (J1) TKO cell lines as determined by mass spectrometry. c, Average CpG methylation in wild-type and TKO cell lines determined by whole-genome bisulfite sequencing. Methylation in the TKO cell line is comparable to background levels represented by the methylation in chromosome M. d, Gene expression levels (RPKM) in isogenic wild type and TKO (159). Black dots represent significantly differentially expressed genes in wild type or TKO, with expected unpregulation of germline genes16. The Dnmt genes are among the most downregulated genes (purple), while the majority of genes that reside within imprinted domains are upregulated roughly twofold (orange). Prominent marker genes of ES cells (Oct4, Sox2 and Nanog, blue) remain unaltered. e, Hierarchical clustering of gene expression correlations for three independent 159 ES cell line wild-type and TKO replicates, and published J1 wild-type and TKO RNA-seq samples16. Overall, gene expression clusters by strain rather than presence of DNA methylation. This reflects the strong influence of genetic background on the global gene expression program and supports our approach of focusing further analysis on the isogenic TKO.

Extended Data Figure 2 Characteristics of DNase-hypersensitive sites.

a, DNase-seq signal in our 159 ES cell line (wild-type) and an ENCODE WW6 ES cell (wild-type) DNase-seq sample27 using a tiling window (500 bp) over the whole genome in mappable regions not blacklisted by ENCODE, illustrating that our protocol for genome-wide detection of DHSs matches available data sets in mouse ES cells. PCC was calculated on all DHSs. b, c, DNase-seq signal and PCC at all DHSs for independent biological replicates of wild type (b) and TKO (c). d, Wild-type methylation and replicates for DNase-seq signal in the 159 ES cell line (wild-type and TKO) and ENCODE WW6 (wild-type) at the genomic region from Fig. 1a (chr17: 25,920,000–25,972,499), illustrating that most DHSs remain unchanged upon removal of DNA methylation, in agreement with the overall similarity in gene expression. e, Change in DNase-seq signal and PCC between wild type and TKO using different replicate samples, illustrating a high reproducibility of quantitative DHS changes between wild type and TKO. f, Distance of all wild-type, wild-type-specific or TKO-specific DHSs from closest gene transcriptional start site (TSS). Proximal and distal separation is at 2 kb. g, Change in DNase-seq signal between TKO and wild-type as a function of CpG content for all wild-type and TKO DHSs, illustrating that most changes occur in CpG-poor regions. h, Change in DNase-seq signal between TKO and wild-type versus average CpG methylation of all wild-type and TKO DHSs matching Fig. 1c, showing that TKO-specific DHSs (right) lie in regions with high methylation in wild type. Black dots represent significantly enriched DHSs (see Methods) in wild type (n = 2,837) or TKO (n = 1,543) from Fig. 1b.

Extended Data Figure 3 Motif enrichment in cell-line-specific DNase-hypersensitive sites.

a, Occurrence of all possible hexamers in TKO-specific DHSs compared to all wild-type DHSs. Blue colouring illustrates hexamer CpG content. Hexamers representing the NRF1 motif are highlighted by a circle. Most strongly enriched hexamers are labelled (only one of two reverse complements). b, Gene expression levels (RPKM) of candidate methylation-sensitive TFs in wild type and TKO indicating that differential abundance does not account for DHS formation upon loss of DNA methylation. Error bars are standard deviation from three biological replicates. c, Footprints of candidate TF motifs enriched in TKO-specific (NRF1, MYCN, GABPA) or wild-type-specific (SOX2, TEAD1) DHSs shown as metaplot of wild-type (brown) or TKO (red) DNase-seq signal for all motifs in all wild-type and TKO (left), TKO-specific (middle) and wild-type-specific (right) DHSs. Number of regions is indicated above each metaplot. A DNase footprint is apparent at the NRF1 motif and, to a lesser extent, at MYCN and GABPA motifs specifically in TKO-specific sites in the TKO sample, whereas footprints at SOX2 and TEAD1 motifs in wild-type-specific sites are less unique to that cell state. d, Motif occurrences in wild-type-specific DHSs compared to all wild-type DHSs. Blue colouring illustrates motif CpG content.

Extended Data Figure 4 Characteristics of NRF1 binding sites.

a, Wild-type methylation, and wild-type and TKO DNase-seq, NRF1 ChIP-seq, H3K27ac ChIP-seq and RNA-seq signal also upon Nrf1 and mock knockdown in TKO at TKO-specific distal (left, chr4: 99,235,170–99,237,170; from Fig. 2a) and proximal (middle, chr5: 31,409,700–31,411,700; right, chrX: 70,341,500–70,343,500) genomic regions. The transcripts initiated directly at the NRF1 binding sites in TKO cells are specifically reduced upon knockdown of Nrf1, implying that they are indeed NRF1-dependent. b, c, NRF1 ChIP-seq signal at all NRF1 peak regions for independent biological replicates of wild type (b) and TKO (c). d, Change in NRF1 ChIP-seq signal and PCC between wild type and TKO using different replicate samples, illustrating a high reproducibility of quantitative NRF1 changes between wild type and TKO. e, Change in NRF1 ChIP-seq signal between TKO and wild type versus CpG content of all wild-type and TKO NRF1 peak regions, illustrating that most changes occur in CpG-poor regions. f, RNA expression levels (RPKM) in wild type and TKO at all wild-type and TKO NRF1 peak regions, illustrating the appearance of a few aberrant TKO-specific transcripts directly at NRF1 binding sites. g, H3K27ac ChIP-seq signal in wild type and TKO at all wild-type and TKO NRF1 peak regions, illustrating appearance of TKO-specific acetylation at a few NRF1 binding sites. h, Knockdown efficiency for the pool of three siRNAs and most efficient single siRNA targeting Nrf1 in TKO cells. Mean of three independent biological replicates normalized to GAPDH; error bars reflect standard deviation. Genetic deletion of Nrf1 with CRISPR/Cas9 was lethal (data not shown). i, Reduction in nuclear NRF1 levels upon siRNA knockdown with pool of three siRNAs and most efficient single siRNA targeting Nrf1 as measured by western blot. Blot was cropped for clarity, all samples were loaded on the same gel (for uncropped gels see Supplementary Fig. 1). j, Expression change (in RPKM) of genes closest to shared and TKO-specific NRF1 peaks between TKO cells treated either with negative control siRNA or the most efficient single siRNA targeting Nrf1, showing highly significant loss in expression after knockdown. P values from Wilcoxon tests. k, Number of CpGs in NRF1 motifs closest to peak summit in all wild-type (top) or TKO-specific (bottom) NRF1 peaks, illustrating that motifs in TKO-specific NRF1 peaks contain at least one CpG. l, Change in NRF1 ChIP-seq signal between TKO and wild type versus average methylation in wild type at all NRF1 sites corresponding to Fig. 2g, illustrating that increased NRF1 binding in TKO occurs at regions that were methylated in wild type. m–o, Average wild-type MeCP2 ChIP-seq signal22 (m), wild-type methylation in NRF1 peak regions or in NRF1 motifs closest to peak summits (n) and change of NRF1 signal between wild type and TKO (o) within 500 bp regions around TKO-specific NRF1 peak summits grouped according to CpG density (0–5 CpGs, n = 3,680; 5–10 CpGs, n = 2,477; >10 CpGs, n = 680). If indirect repression could contribute to differential NRF1 binding, we would expect a more pronounced increase of NRF1 binding at sites with higher CpG density upon demethylation of the genome, as methyl-CpG binding domain proteins (MBDs) such as MeCP2 bind preferentially to regions with a high density of methylated CpGs rather than fully methylated regions with low CpG density. TKO-specific binding of NRF1 is independent of CpG density and MeCP2 enrichment in the methylated genome, strongly arguing against an involvement of indirect repression in NRF1 binding site restriction.

Extended Data Figure 5 NRF1 binding in different culture conditions.

a, Nrf1 gene expression levels (RPKM) in 2i and serum culture conditions49. b, NRF1 ChIP-seq signal in wild-type cells adapted to 2i culture conditions (after culture with serum) for two biological replicates. c, NRF1 ChIP-seq signal in wild-type cells adapted to 2i (after culture with serum) and TKO. d, Methylation in wild-type cells cultured in serum and 2i (after culture with serum) at all NRF1 motifs. e, Methylation in serum and 2i (after culture with serum) measured by amplicon Bis-seq for fully methylated (FMR), low methylated (LMR), unmethylated (UMR) controls, 6 unbound NRF1 sites and 56 TKO-specific NRF1 sites. f, Comparison and PCC of DNA methylation levels by amplicon Bis-seq and whole-genome Bis-seq upon culture in 2i (after culture with serum). g, Average 2i (after culture with serum) methylation in NRF1 peak regions or NRF1 motifs within peaks versus change in NRF1 signal between TKO and 2i (after culture with serum) at all NRF1 peaks, illustrating that reduced NRF1 binding in 2i compared to TKO can be explained by residual methylation. h, Methylation in wild-type cells cultured in serum, cultured in 2i (after culture with serum) and cultured in serum (after culture in 2i) and NRF1 ChIP-seq signal in wild type, TKO, cultured in 2i (after culture with serum) and cultured in serum (after culture with 2i) at TKO-specific regions with higher 2i methylation in NRF1 motifs (grey lines) than surrounding region (left, chr10: 66,251,100–66,251,700; middle, chr4: 15,976,050–15,976,650; right, chr19: 55,833,420–55,834,020). NRF1 is unable to bind if CpGs in the motif remain methylated in 2i, even if the surrounding region is unmethylated. i, NRF1 ChIP-seq signal in wild-type cells adapted back to serum (after culture with 2i) for two biological replicates. j, Methylation in wild-type cells cultured in serum and adapted back to serum (after culture with 2i) at all NRF1 motifs. k, Methylation in wild-type cells cultured in serum and adapted back to serum (after culture with 2i) measured by amplicon Bis-seq for FMR, LMR and UMR controls, 6 unbound NRF1 sites and 56 TKO-specific NRF1 sites. l, NRF1 ChIP-seq signal in wild-type cells adapted back to serum (after culture with 2i) and original serum conditions. m, NRF1 ChIP-seq signal in wild-type cells adapted back to serum (after culture with 2i) and adapted to 2i (after culture with serum).

Extended Data Figure 6 Overexpression of NRF1 is unable to induce binding to TKO-specific sites.

a, Transient overexpression of NRF1 under control of the CMV (middle) or CAG promoter (right, used for ChIP experiments) leads to strong increase in nuclear NRF1 protein levels compared to endogenous levels (left) as measured by western blot (for uncropped gel data see Supplementary Fig. 1). The overexpressed protein contains a protein tag accounting for the higher molecular weight. b, NRF1 ChIP-seq signal upon transient NRF1 overexpression for two biological replicates. c, NRF1 ChIP-seq signal in wild type and upon overexpression. d, NRF1 ChIP-seq signal in TKO and overexpression conditions only at TKO- and overexpression-specific NRF1 peak regions, illustrating that TKO-specific NRF1 sites are distinct from overexpression-specific sites. e, Change in NRF1 ChIP-seq signal between overexpression and wild type versus the score (MAST position P value) of NRF1 motifs closest to the summit, illustrating that sites gaining most NRF1 upon overexpression do not contain high-confidence motifs.

Extended Data Figure 7 Cell-type-specific binding of NRF1 correlates with methylation and expression changes.

a–e, Comparison of NRF1 binding in ES and neuronal progenitor cells. Methylation in ES and neural progenitors8 at all NRF1 motifs (a), NRF1 ChIP-seq signal in ES and neuronal progenitors at all NRF1 peaks (b), neuronal progenitor minus ES methylation of peak regions or NRF1 motifs in ES-specific (n = 4,934) and shared (n = 4,951) NRF1 peaks (negligible number of neuronal-progenitor-specific peaks) (c), expression of the genes50 closest to ES-specific and shared NRF1 peaks (d), selection of gene ontology (GO) biological functions enriched in genes closest to ES-specific and shared NRF1 peaks (e). P values from Wilcoxon tests. f–i, Comparison of NRF1 binding in HMEC and HCC1954 cells. Methylation in HMEC and HCC195426 at all NRF1 motifs (f), NRF1 ChIP-seq signal in HMEC and HCC1954 at all NRF1 peaks (g), HCC1954 minus HMEC methylation of peak regions or NRF1 motifs in HMEC-specific (n = 2,726), HCC1954-specific (n = 2,685) and shared (n = 12,180) NRF1 peaks (h), expression of the genes26 closest to HMEC-specific, HCC1954-specific and shared NRF1 peaks (i). j–m, Comparison of NRF1 binding in H1-hESC and GM12878 cells. Methylation in H1-hESC and GM1287827 at all NRF1 motifs (j), NRF1 ChIP-seq signal in H1-hESC and GM1287827 at all NRF1 peaks (k), GM12878 minus H1-hESC methylation of peak regions or NRF1 motifs in H1-hESC- (n = 618), GM12878-specific (n = 561) and shared (n = 3,198) NRF1 peaks (l), expression of the genes27 closest to H1-hESC-specific, GM12878-specific and shared NRF1 peaks (m).

Extended Data Figure 8 NRF1 binding to the unmethylated motif can be recapitulated at an ectopic site.

a, Wild-type and TKO DNase-seq and NRF1 ChIP-seq signal for two biological replicates at the endogenous counterparts of the inserted regions profiled in Extended Data Fig. 8b (left, chr8: 123,019,920–123,021,030) and Extended Data Fig. 8c (right, chr8: 113,271,460–113,272,690). b, Methylation (amplicon Bis-seq, left, coloured lines indicate position and methylation status of CpGs) and NRF1 binding (ChIP-qPCR, right) for an endogenous methylation-dependent NRF1 site (chr8: 123,020,293–123,020,670) and upon insertion of this region into a defined ectopic genomic locus. The position of the two NRF1 motifs containing two CpGs each is indicated in blue. The reporter construct was inserted either unmethylated or in vitro premethylated with M.SssI. In the untreated construct one motif becomes completely methylated upon insertion, whereas the other only gains roughly 50% methylation, and NRF1 binding is detected. The pre-methylated construct maintains at least one CpG with almost complete methylation in both core motifs present and shows strongly reduced NRF1 binding by comparison. Thus, the methylation sensitivity of NRF1 can be recapitulated in an ectopic site even in the absence of global changes in DNA methylation. As expected, forcing complete demethylation of both core motifs in the premethylated insert by treatment of the cells with 5-aza-2′-deoxycytidine leads to further increased NRF1 binding compared to the untreated template. ChIP–qPCR enrichments are the mean of three independent biological replicates; error bars reflect standard deviation. See Supplementary Table 3 for methylation source data. c, Methylation (amplicon Bis-seq, left, coloured lines indicate position and methylation status of CpGs) and NRF1 binding (ChIP–qPCR, right) for an endogenous methylation-dependent NRF1 site (chr8: 113,271,870–113,272,282) and upon insertion of this region into a defined ectopic genomic locus. The untreated template gains full methylation in the core motif (blue) and does not show detectable NRF1 binding. Forcing complete demethylation by treatment with 5-aza-2′-deoxycytidine enables NRF1 to bind the site in the ectopic locus. ChIP–qPCR enrichments are mean of three independent biological replicates; error bars reflect standard deviation. See Supplementary Table 3 for methylation source data.

Extended Data Figure 9 Constitutive NRF1 sites are co-bound by other TFs.

a, Change in NRF1 ChIP-seq signal between TKO and wild type versus size of DHSs overlapping NRF1 peak regions, illustrating that wild-type NRF1 sites tend to overlap with larger DHSs. b, Overlap of wild-type and TKO-specific NRF1 peak regions with published ChIP-seq peak regions from other TFs expressed in ES cells8,53,54, illustrating that wild-type NRF1 sites coincide with other TF binding events. P values from hypergeometric tests. c, Wild-type methylation, wild-type and TKO DNase-seq, and NRF1 and CTCF8 ChIP-seq signal for two biological replicates at the endogenous Gtf2a1l promoter (chr17: 89,067,600–89,068,350). The region used for the insertion experiments in Fig. 4b is indicated below. d, Wild-type methylation, wild-type and TKO DNase-seq for two biological replicates and NRF1 and REST52 ChIP-seq signal at adjacent NRF1 and REST binding sites (left, chr15: 100,703,260–100,704,500; middle, chr2: 180,152,200–180,153,150; right, chr2: 118,604,800–118,605,900). Regions profiled with amplicon Bis-seq in REST wild-type and REST KO cells in Fig. 4c and the position of the TF motifs are indicated below.

Extended Data Table 1 Number of raw and mapped reads and enriched regions for all high-throughput sequencing samples

Full size table

Supplementary information

Supplementary Figure 1

This file contains gel source data for Extended Data Figures 4h and 6a. (PDF 2662 kb)

Supplementary Table 1

This table contains occurrences of known transcription factor motifs in DHS, ranked by P-value. Logos for top motifs corresponding to expressed TFs have been manually assigned to categories NRF1-, MYC-, GABPA-like and others. (XLSX 364 kb)

Supplementary Table 2

This table contains occurrences of all possible NRF1 motif variants in NRF1 binding sites (151 bp around peak summit), ranked according to occurrences in TKO-specific NRF1 binding sites. (XLSX 237 kb)

Supplementary Table 3

This table contains Amplicon Bis-seq data for endogenous NRF1 sites and ectopic insertions (Extended Data Fig. 8b = Ectopic_insert1; Extended Data Fig. 8c = Ectopic_insert2; Fig. 4a = Ectopic_Mrap), showing primers, genomic location, methylation ratio and coverage in different conditions for entire amplicons as well as for individual CpGs. (XLSX 121 kb)

PowerPoint slides

Rights and permissions

About this article

Cite this article

Domcke, S., Bardet, A., Adrian Ginno, P. et al. Competition between DNA methylation and transcription factors determines binding of NRF1.Nature 528, 575–579 (2015). https://doi.org/10.1038/nature16462

Download citation