Skip to main content

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

  • Article
  • Published:

A unique bipartite Polycomb signature regulates stimulus-response transcription during development

Nature Genetics volume 53, pages 379–391 (2021)Cite this article

Subjects

Abstract

Rapid cellular responses to environmental stimuli are fundamental for development and maturation. Immediate early genes can be transcriptionally induced within minutes in response to a variety of signals. How their induction levels are regulated and their untimely activation by spurious signals prevented during development is poorly understood. We found that in developing sensory neurons, before perinatal sensory-activity-dependent induction, immediate early genes are embedded into a unique bipartite Polycomb chromatin signature, carrying active H3K27ac on promoters but repressive Ezh2-dependent H3K27me3 on gene bodies. This bipartite signature is widely present in developing cell types, including embryonic stem cells. Polycomb marking of gene bodies inhibits mRNA elongation, dampening productive transcription, while still allowing for fast stimulus-dependent mark removal and bipartite gene induction. We reveal a developmental epigenetic mechanism regulating the rapidity and amplitude of the transcriptional response to relevant stimuli, while preventing inappropriate activation of stimulus-response genes.

This is a preview of subscription content, access via your institution

Access options

Buy this article

  • Purchase on Springer Link
  • Instant access to full article PDF
Buy now

Prices may be subject to local taxes which are calculated during checkout

Fig. 1: Immediate early- and late-response genes carry distinct Polycomb signatures during sensory neuron development.
Fig. 2: Developmental cellular representation and genome-wide distribution of bipartite chromatin, and promoter H3K27ac and gene body H3K27me3 coexistence at bipartite genes.
Fig. 3: Bipartite chromatin dynamics during barrelette neuron development.
Fig. 4: RNA polymerase II transcripts of bipartite genes are not efficiently processed to productive mRNA.
Fig. 5: Polycomb marking of gene bodies inhibits productive mRNA processing and regulates rapidity and amplitude of transcriptional response to stimuli.
Fig. 6: Polycomb marking of bipartite genes bodies hampers productive mRNA elongation by inhibition of stimulus-dependent NELF release and chromatin compaction.
Fig. 7: Mechanism of stimulus-dependent transition from bipartite to active chromatin.
Fig. 8: Polycomb-dependent regulation of stimulus-response genes during development.

Similar content being viewed by others

Data availability

All raw sequencing data and processed data used for this study are available through ArrayExpress and will be released to the public without restrictions: mRNA-seq (Smart-seq2), E-MTAB-8314; total RNA-seq (Solo RNA-seq), E-MTAB-8311; ChIP–seq (ChIPmentation), E-MTAB-8317; ATAC–seq, E-MTAB-8313; 4C–seq, E-MTAB-8295; and single-cell RNA-seq (10x Genomics), E-MTAB-8312. FACS gating strategies/source data are presented in Supplementary Figs. 1–4. Public sequencing datasets were obtained from the Gene Expression Omnibus and ENCODE as follows: mouse cortical culture (GSE21161 and GSE60192), mouse embryonic forebrain (GSE93011 and GSE52386), mouse adult cortical excitatory neuron (GSE63137), mouse ESCs (GSE36114 and GSE94250), mouse ESCs for Ezh2-KO experiments (GSE116603), mouse E14.5 heart tissues (GSE82764, GSE82637, GSE82640 and GSE78441; ENCSR068YGC), mouse E14.5 liver tissues (GSE78422, GSE82407, GSE82615 and GSE82620; ENCSR032HKE) and E10.5 mouse NCCs isolated from the frontal nasal process (GSE89437).

Code availability

Computational analyses were performed in R using the mentioned publicly available packages (Methods, Reporting Summary and Supplementary Methods). The custom tool monaLisa (v0.1.28), used for motif enrichment, can be found on GitHub at https://github.com/fmicompbio/monaLisa/. The custom tool swissknife (v0.10) is available on https://github.com/fmicompbio/swissknife/.

References

  1. Fowler, T., Sen, R. & Roy, A. L. Regulation of primary response genes. Mol. Cell 44, 348–360 (2011).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  2. West, A. E. & Greenberg, M. E. Neuronal activity-regulated gene transcription in synapse development and cognitive function. Cold Spring Harb. Perspect. Biol. https://doi.org/10.1101/cshperspect.a005744 (2011).

  3. Greenberg, M. E. & Ziff, E. B. Stimulation of 3T3 cells induces transcription of the c-fos proto-oncogene. Nature 311, 433–438 (1984).

    Article  CAS  PubMed  Google Scholar 

  4. Malik, A. N. et al. Genome-wide identification and characterization of functional neuronal activity-dependent enhancers. Nat. Neurosci. 17, 1330–1339 (2014).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  5. Vierbuchen, T. et al. AP-1 transcription factors and the BAF complex mediate signal-dependent enhancer selection. Mol. Cell 68, 1067–1082 (2017).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  6. Stroud, H. et al. An activity-mediated transition in transcription in early postnatal neurons. Neuron 107, 874–890 (2020).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  7. Mayer, A., Landry, H. M. & Churchman, L. S. Pause & go: from the discovery of RNA polymerase pausing to its functional implications. Curr. Opin. Cell Biol. 46, 72–80 (2017).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  8. Yap, E. L. & Greenberg, M. E. Activity-regulated transcription: bridging the gap between neural activity and behavior. Neuron 100, 330–348 (2018).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  9. Lonze, B. E. & Ginty, D. D. Function and regulation of CREB family transcription factors in the nervous system. Neuron 35, 605–623 (2002).

    Article  CAS  PubMed  Google Scholar 

  10. Toth, A. B., Shum, A. K. & Prakriya, M. Regulation of neurogenesis by calcium signaling. Cell Calcium 59, 124–134 (2016).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  11. Ginty, D. D., Glowacka, D., Bader, D. S., Hidaka, H. & Wagner, J. A. Induction of immediate early genes by Ca2+ influx requires cAMP-dependent protein kinase in PC12 cells. J. Biol. Chem. 266, 17454–17458 (1991).

    Article  CAS  PubMed  Google Scholar 

  12. Greenberg, M. E., Greene, L. A. & Ziff, E. B. Nerve growth factor and epidermal growth factor induce rapid transient changes in proto-oncogene transcription in PC12 cells. J. Biol. Chem. 260, 14101–14110 (1985).

    Article  CAS  PubMed  Google Scholar 

  13. Erzurumlu, R. S., Murakami, Y. & Rijli, F. M. Mapping the face in the somatosensory brainstem. Nat. Rev. Neurosci. 11, 252–263 (2010).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  14. Kitazawa, T. & Rijli, F. M. Barrelette map formation in the prenatal mouse brainstem. Curr. Opin. Neurobiol. 53, 210–219 (2018).

    Article  CAS  PubMed  Google Scholar 

  15. Erzurumlu, R. S. & Gaspar, P. Development and critical period plasticity of the barrel cortex. Eur. J. Neurosci. 35, 1540–1553 (2012).

    Article  PubMed  PubMed Central  Google Scholar 

  16. Hrvatin, S. et al. Single-cell analysis of experience-dependent transcriptomic states in the mouse visual cortex. Nat. Neurosci. 21, 120–129 (2018).

    Article  CAS  PubMed  Google Scholar 

  17. Tyssowski, K. M. et al. Different neuronal activity patterns induce different gene expression programs. Neuron 98, 530–546 (2018).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  18. Valles, A. et al. Genomewide analysis of rat barrel cortex reveals time- and layer-specific mRNA expression changes related to experience-dependent plasticity. J. Neurosci. 31, 6140–6158 (2011).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  19. Mohn, F. et al. Lineage-specific polycomb targets and de novo DNA methylation define restriction and potential of neuronal progenitors. Mol. Cell 30, 755–766 (2008).

    Article  CAS  PubMed  Google Scholar 

  20. Ferrai, C. et al. RNA polymerase II primes Polycomb-repressed developmental genes throughout terminal neuronal differentiation. Mol. Syst. Biol. 13, 946 (2017).

    Article  PubMed  PubMed Central  Google Scholar 

  21. Hirabayashi, Y. et al. Polycomb limits the neurogenic competence of neural precursor cells to promote astrogenic fate transition. Neuron 63, 600–613 (2009).

    Article  CAS  PubMed  Google Scholar 

  22. Aranda, S., Mas, G. & Di Croce, L. Regulation of gene transcription by Polycomb proteins. Sci. Adv. 1, e1500737 (2015).

    Article  PubMed  PubMed Central  Google Scholar 

  23. Schuettengruber, B., Bourbon, H. M., Di Croce, L. & Cavalli, G. Genome regulation by Polycomb and Trithorax: 70 years and counting. Cell 171, 34–57 (2017).

    Article  CAS  PubMed  Google Scholar 

  24. Bernstein, B. E. et al. A bivalent chromatin structure marks key developmental genes in embryonic stem cells. Cell 125, 315–326 (2006).

    Article  CAS  PubMed  Google Scholar 

  25. Minoux, M. et al. Gene bivalency at Polycomb domains regulates cranial neural crest positional identity. Science https://doi.org/10.1126/science.aal2913 (2017).

  26. Piunti, A. & Shilatifard, A. Epigenetic balance of gene expression by Polycomb and COMPASS families. Science 352, aad9780 (2016).

    Article  PubMed  Google Scholar 

  27. Bonnefont, J. et al. Cortical neurogenesis requires Bcl6-mediated transcriptional repression of multiple self-renewal-promoting extrinsic pathways. Neuron 103, 1096–1108 (2019).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  28. Chen, F. X., Smith, E. R. & Shilatifard, A. Born to run: control of transcription elongation by RNA polymerase II. Nat. Rev. Mol. Cell Biol. 19, 464–478 (2018).

    Article  CAS  PubMed  Google Scholar 

  29. Brookes, E. & Pombo, A. Modifications of RNA polymerase II are pivotal in regulating gene expression states. EMBO Rep. 10, 1213–1219 (2009).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  30. Zaborowska, J., Egloff, S. & Murphy, S. The pol II CTD: new twists in the tail. Nat. Struct. Mol. Biol. 23, 771–777 (2016).

    Article  CAS  PubMed  Google Scholar 

  31. Brookes, E. et al. Polycomb associates genome-wide with a specific RNA polymerase II variant, and regulates metabolic genes in ESCs. Cell Stem Cell 10, 157–170 (2012).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  32. Stock, J. K. et al. Ring1-mediated ubiquitination of H2A restrains poised RNA polymerase II at bivalent genes in mouse ES cells. Nat. Cell Biol. 9, 1428–1435 (2007).

    Article  CAS  PubMed  Google Scholar 

  33. Simon, J. A. & Kingston, R. E. Mechanisms of polycomb gene silencing: knowns and unknowns. Nat. Rev. Mol. Cell Biol. 10, 697–708 (2009).

    Article  CAS  PubMed  Google Scholar 

  34. Blackledge, N. P., Rose, N. R. & Klose, R. J. Targeting Polycomb systems to regulate gene expression: modifications to a complex story. Nat. Rev. Mol. Cell Biol. 16, 643–649 (2015).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  35. Shen, X. et al. EZH1 mediates methylation on histone H3 lysine 27 and complements EZH2 in maintaining stem cell identity and executing pluripotency. Mol. Cell 32, 491–502 (2008).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  36. Schoeftner, S. et al. Recruitment of PRC1 function at the initiation of X inactivation independent of PRC2 and silencing. EMBO J. 25, 3110–3122 (2006).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  37. Lavarone, E., Barbieri, C. M. & Pasini, D. Dissecting the role of H3K27 acetylation and methylation in PRC2-mediated control of cellular identity. Nat. Commun. 10, 1679 (2019).

    Article  PubMed  PubMed Central  Google Scholar 

  38. Kim, T. K. et al. Widespread transcription at neuronal activity-regulated enhancers. Nature 465, 182–187 (2010).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  39. Schaukowitch, K. et al. Enhancer RNA facilitates NELF release from immediate early genes. Mol. Cell 56, 29–42 (2014).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  40. Voiculescu, O. et al. Hindbrain patterning: Krox20 couples segmentation and specification of regional identity. Development 128, 4967–4978 (2001).

    Article  CAS  PubMed  Google Scholar 

  41. Madisen, L. et al. A robust and high-throughput Cre reporting and characterization system for the whole mouse brain. Nat. Neurosci. 13, 133–140 (2010).

    Article  CAS  PubMed  Google Scholar 

  42. Bechara, A. et al. Hoxa2 selects barrelette neuron identity and connectivity in the mouse somatosensory brainstem. Cell Rep. 13, 783–797 (2015).

    Article  CAS  PubMed  Google Scholar 

  43. Moreno-Juan, V. et al. Prenatal thalamic waves regulate cortical area size prior to sensory processing. Nat. Commun. 8, 14172 (2017).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  44. Luche, H., Weber, O., Nageswara Rao, T., Blum, C. & Fehling, H. J. Faithful activation of an extra-bright red fluorescent protein in ‘knock-in’ Cre-reporter mice ideally suited for lineage tracing studies. Eur. J. Immunol. 37, 43–53 (2007).

    Article  CAS  PubMed  Google Scholar 

  45. Di Meglio, T. et al. Ezh2 orchestrates topographic migration and connectivity of mouse precerebellar neurons. Science 339, 204–207 (2013).

    Article  PubMed  PubMed Central  Google Scholar 

  46. Maheshwari, U. et al. Postmitotic Hoxa5 expression specifies pontine neuron positional identity and input connectivity of cortical afferent subsets. Cell Rep. 31, 107767 (2020).

    Article  CAS  PubMed  Google Scholar 

  47. Seidler, B. et al. A Cre-loxP-based mouse model for conditional somatic gene expression and knockdown in vivo by using avian retroviral vectors. Proc. Natl Acad. Sci. USA 105, 10137–10142 (2008).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  48. Picelli, S. et al. Full-length RNA-seq from single cells using Smart-seq2. Nat. Protoc. 9, 171–181 (2014).

    Article  CAS  PubMed  Google Scholar 

  49. Schmidl, C., Rendeiro, A. F., Sheffield, N. C. & Bock, C. ChIPmentation: fast, robust, low-input ChIP–seq for histones and transcription factors. Nat. Methods 12, 963–965 (2015).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  50. Buenrostro, J. D., Giresi, P. G., Zaba, L. C., Chang, H. Y. & Greenleaf, W. J. Transposition of native chromatin for fast and sensitive epigenomic profiling of open chromatin, DNA-binding proteins and nucleosome position. Nat. Methods 10, 1213–1218 (2013).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  51. Ritchie, M. E. et al. limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 43, e47 (2015).

    Article  PubMed  PubMed Central  Google Scholar 

  52. O’Connell, J. & Hojsgaard, S. Hidden semi-Markov models for multiple observation sequences: the mhsmm package for R. J. Stat. Softw. 39, 4 (2011).

    Google Scholar 

  53. Zheng, G. X. et al. Massively parallel digital transcriptional profiling of single cells. Nat. Commun. 8, 14049 (2017).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  54. Heinz, S. et al. Simple combinations of lineage-determining transcription factors prime cis-regulatory elements required for macrophage and B cell identities. Mol. Cell 38, 576–589 (2010).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  55. Maaten, L. J. P. V. D. & Hinton, G. E. Visualizing high-dimensional data using t-SNE. J. Mach. Learn. Res. 9, 2579–2605 (2008).

    Google Scholar 

  56. Venables, W. N. & Ripley, B. D. Modern Applied Statistics with S 4th edn (Springer, 2002).

  57. Durinck, S., Spellman, P. T., Birney, E. & Huber, W. Mapping identifiers for the integration of genomic datasets with the R/Bioconductor package biomaRt. Nat. Protoc. 4, 1184–1191 (2009).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

Download references

Acknowledgements

F.M.R. wishes to dedicate this paper to the memory of P. Sassone-Corsi (1956–2020), dear friend and eminent scientist who made seminal contributions to immediate early gene and activator protein-1 transcriptional regulation and with whom F.M.R. discussed this project at an early stage. We thank A. Pombo for very useful discussion. We thank N. Vilain, S. Smallwood, D. Gaidatzis and the members of the Rijli group and FMI facilities for excellent technical support and discussion. We thank S. H. Orkin (Harvard Medical School) and A. Wutz (ETH Zurich) for the kind gifts of the Ezh2flox mouse line and EedKO ESCs, respectively. The LAL-R26TVA-LacZ line was a kind gift from D. Saur (Technische Universitat Munchen). T.K. was supported by a Japan Society for the Promotion of Science fellowship, and O.J. was supported by an EMBO Long-Term fellowship. F.M.R. was supported by the Swiss National Science Foundation (31003A_149573 and 31003A_175776). This project has also received funding from the European Research Council under the European Union’s Horizon 2020 research and innovation programme (grant no. 810111-EpiCrest2Reg). F.M.R. and M.B.S. were also supported by the Novartis Research Foundation.

Author information

Author notes

  1. Henrik Gezelius

    Present address: Science for Life Laboratory, Solna, Sweden

  2. These authors contributed equally: Taro Kitazawa, Dania Machlab.

Authors and Affiliations

  1. Friedrich Miescher Institute for Biomedical Research, Basel, Switzerland

    Taro Kitazawa, Dania Machlab, Onkar Joshi, Nicola Maiorano, Hubertus Kohler, Sebastien Ducret, Sandra Kessler, Charlotte Soneson, Panagiotis Papasaikas, Michael B. Stadler & Filippo M. Rijli

  2. Swiss Institute of Bioinformatics, Basel, Switzerland

    Dania Machlab, Charlotte Soneson, Panagiotis Papasaikas & Michael B. Stadler

  3. University of Basel, Basel, Switzerland

    Dania Machlab & Filippo M. Rijli

  4. Instituto de Neurociencias de Alicante, Universidad Miguel Hernández-Consejo Superior de Investigaciones Científicas (UMH-CSIC), Sant Joan d’Alacant, Spain

    Henrik Gezelius & Guillermina López-Bendito

Authors
  1. Taro Kitazawa
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Dania Machlab
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Onkar Joshi
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Nicola Maiorano
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Hubertus Kohler
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Sebastien Ducret
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Sandra Kessler
    View author publications

    You can also search for this author in PubMed Google Scholar

  8. Henrik Gezelius
    View author publications

    You can also search for this author in PubMed Google Scholar

  9. Charlotte Soneson
    View author publications

    You can also search for this author in PubMed Google Scholar

  10. Panagiotis Papasaikas
    View author publications

    You can also search for this author in PubMed Google Scholar

  11. Guillermina López-Bendito
    View author publications

    You can also search for this author in PubMed Google Scholar

  12. Michael B. Stadler
    View author publications

    You can also search for this author in PubMed Google Scholar

  13. Filippo M. Rijli
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

T.K. and F.M.R. conceived the study, designed experiments and analyzed experimental data. T.K. performed most of the experiments. H.K. carried out cell sorting. D.M., T.K. and M.B.S. performed computational analysis. O.J. performed 4C–seq. S.K. carried out some ChIP–seq assays. S.D., H.G. and G.L.-B. contributed to Kir-OE mouse generation and characterization. N.M. analyzed the phenotype of Kir-OE mice. C.S. performed analysis of scRNA-seq. C.S. and P.P. contributed to the Solo RNA-seq analysis; T.K., D.M. and M.B.S. wrote the first draft. F.M.R. revised and wrote the final manuscript.

Corresponding author

Correspondence to Filippo M. Rijli.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

Peer review information Nature Genetics thanks the anonymous reviewers for their contribution to the peer review of this work. Peer reviewer reports are available.

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Extended data

Extended Data Fig. 1 Genetic strategy of barrelette neuron isolation and identification of activity response genes.

a and b, Representative FACS gating for barrelette neurons (Supplementary Figs. 1–4). c and d, Intersectional strategies to FACS isolate E14.5, E18.5 or P4 postmitotic barrelette neurons (Drg11vPrV-ZsGreen/+ (c), Drg11vPrV-tdTomato/+ (d)) from ventral principal trigeminal nucleus (vPrV). (Supplementary Table 1, Supplementary Note). e, Intersectional strategy to FACS isolate Kir2.1(Kir)-mCherry overexpressing, neuronal activity-deprived, vPrV barrelette neurons (Drg11vPrV-Kir/+; Supplementary Table 1, Supplementary Note). f, Volumes of PrV in control (K20tdTomato/+;r2EGFP/+) and Kir overexpressing (K20Kir/+;r2EGFP/+) mice (n = 3, biologically independent animals). g and h, P8 cytochrome oxidase (CO) staining in wild-type (WT; g) and K20Kir/+ (h) mice. Representative images of n = 3 biologically independent animals. Scale bars: 200 μm. i–p, P10 barrelette neuron dendrite orientation by GFP expression after pseudotyped rabies virus (EnvA-SADΔG-GFP) injection in P3 thalamus in control Krox20::Cre;LSL-R26TVA-LacZ (K20TVA/+) (j) and Kir overexpressing Krox20::Cre;R26Kir-mCherry;LSL-R26TVA-LacZ (K20TVA/Kir) (k) mice. Scale bars: 10 μm. Symmetry index (l) and surface ratio (p) are compared (Supplementary Methods). n = 8 (K20TVA/+) and n = 6 (K20TVA/Kir) biologically independent animals were used, and j, k, n, o are representative images. q and r, MA-plots comparing E18.5 and E14.5 mRNA levels in control Drg11vPrV-ZsGreen/+ barrelette neurons (q), and E18.5 Drg11vPrV-Kir/+ (Kir-OE) and E18.5 Drg11vPrV-ZsGreen/+ wild-type (WT) barrelette neurons (r). 702 genes (green dots, q) increase their expression at E18.5 as compared to E14.5 (log2(fold change) > 1.5), while 102 genes (red dots, r) decrease their expression in E18.5 Kir-OE barrelette neurons (log2(fold change) < –1) (Methods). s, Identification of 56 bsARGs. t, Fractions of the 56 bsARGs (left) and 83 nbARGs (right) with H3K27me3- and H3K27me3+ profiles in E14.5 Drg11vPrV-ZsGreen/+ barrelette neurons (see Fig. 1c). u, Scatterplots showing ATAC–seq (x axis) and H3K4me2 (y axis) signals on promoters (1 kb around TSS) in E14.5 Drg11vPrV-ZsGreen/+ barrelette neurons. Dashed lines indicate thresholds corresponding to a 5% false discovery rate (FDR) (Methods) (see Fig. 1c). f, l, p, Bars indicate median and P values are from Welch’s two-sample two-sided t-tests. NS: not significant (P > 0.05).

Extended Data Fig. 2 Chromatin profiles of IEGs carrying the bipartite signature.

a, Fos, Egr1, Fosb and Nr4a3 genome browser views. ATAC (violet), H3K4me2 (yellow, E14.5 barrelette neurons), H3K4me3 (yellow, adult cortical neurons and cultured embryonic neurons), H3K27ac (red) and H3K27me3 (blue) are shown in E14.5 Drg11vPrV-ZsGreen/+ barrelette neurons, E15.5 embryonic cortical neural progenitors (NPC) and postmitotic neurons (PMID: 28793256), adult cortical excitatory neurons (PMID: 26087164), and embryonic cortical 7 day ex vivo cultured neurons (PMID: 20393465). Shaded boxes highlight promoters (pink) and gene bodies (blue). IEGs display a bipartite chromatin signature characterized by promoter-H3K27ac and gene body-H3K27me3 in E14.5 barrelette neurons and E15.5 cortical progenitors and postmitotic neurons. In contrast, H3K27me3 is not present on IEG gene bodies in postnatal cortical excitatory neurons, similar to postnatal barrelette neurons (see Fig. 3e and Extended Data Fig. 6b). Also, culturing embryonic cortical neurons for one week results in depletion of the H3K27me3 mark, similar to embryonic hindbrain neuron culture (see Extended Data Fig. 9a). In long genes (for example Nr4a3), H3K27me3 deposition on gene body does not stretch throughout the gene body, but is restricted only to the proximal region downstream of the promoter (also see Extended Data Fig. 3c). b, IEG (Fos, Egr1 and Fosb) genome browser views at E14.5, pre-fixed prior to the dissociation procedure. Shaded boxes highlight promoters (pink) and gene bodies (blue). The presence of the bipartite pattern (H3K27ac + promoter/H3K27me3+ gene body) in the pre-fixed tissue indicates that it is not induced by the dissociation procedure.

Extended Data Fig. 3 Genome-wide characterization of the bipartite signature.

a and b, Bipartite (a) and bivalent (b) gene rank (x axis) and corresponding rate of correct bipartite and bivalent classification obtained through genome browser visual inspection of individual loci (y axis, fraction of bipartite true-positive, Methods) in E10.5 K20tdTomato/+ progenitors and E14.5, E18.5 and P4 Drg11vPrV-ZsGreen/+ barrelette neurons. c, Aggregate plot showing profile of H3K27me3 around the transcription start site (TSS) in E14.5Bip genes (top 100 genes ranked by bipartiteness scores in E14.5 Drg11vPrV-ZsGreen/+ barrelette neurons) with long gene length (>10 kb). Promoters (defined as 1 kb upstream to 500 bp downstream of TSS) and gene bodies (from 1 kb to 3 kb downstream of TSS) are highlighted. Note that H3K27me3 on gene bodies does not stretch further than 2-3 kb downstream of the TSS, even when genes are long. d, Genome browser profiles of representative bipartite IEGs (Fos, Egr1) in mouse ESCs. Note that the H3K27me3 mark is deposited not only on downstream (gene body) but also upstream regions of these H3K27ac promoters. e, Scatterplots showing promoter H3K27ac (x axis) and gene body H3K27me3 (y axis) signals in E14.5 rhombomere 3 (r3)-derived K20tdTomato/+ hindbrain cells. E14.5Bip genes identified in Drg11vPrV-ZsGreen/+ barrelette neurons are mapped (red dots). Dashed lines indicate thresholds corresponding to a 5% false discovery rate (FDR) based on a gaussian mixture model with two components (for foreground and background, see Methods). Barrelette neuron E14.5Bip genes show high levels of promoter H3K27ac and gene body H3K27me3 indicating that they are bipartite also in r3-derived K20tdTomato/+ hindbrain cells. f, CpG average observed/expected (o/e) ratios in a 100 bp window in E14.5Bip and E14.5Biv gene loci. Bins overlapping with the promoter, TSS, and gene body are indicated. g, Transcription factor binding motifs specifically enriched in E14.5Bip as compared to E14.5Biv promoters (Methods).

Extended Data Fig. 4 Coexistence of H3K27ac and H3K27me3 at promoter and gene body of bipartite genes.

a, Scheme of sequential ChIP-seq protocol (Supplementary Methods). b and c, Scatterplots comparing H3K27ac (x axis) and H3K27me3 (y axis) signals detected in regions from −1 kb to +3 kb around the transcription start site (TSS) by single ChIP-seq performed with large (2-3 kb) chromatin fragments in E14.5 hindbrain tissue. The colors indicate the corresponding H3K27me3/H3K27ac sequential-ChIP-seq signals, either for all autosomal genes (b), or only for E14.5Bip genes (c). Dashed lines indicate thresholds corresponding to a 5% FDR based on a gaussian mixture model with two components (for foreground and background, see Methods). Stratified by one single ChIP signal (for example H3K27ac), the sequential ChIP signal still correlates with the other (for example H3K27me3), which indicates that single chromatin fragments have been double-marked and thus have been enriched at both steps of the sequential ChIP experiments. d, Genome browser view of bipartite Egr1 gene locus displaying chromatin accessibility (ATAC-seq), 2-3kb-fragment H3K27ac, H3K27me3 and H3K27me3/H3K27ac sequential-ChIP-seq in E14.5 hindbrain tissue. H3K27me3 and H3K27ac coexist on gene body and promoter, respectively. e, Violin plots displaying promoter H3K27ac (left), bulk mRNA-seq (middle, Smart-seq2), and single cell fraction with detected mRNA-seq (right, 10X genomics) of E10.5 K20tdTomato/+ progenitors. E10.5 bipartite (E10.5Bip) genes (green, n = 99 genes, see Methods) and E10.5 non-bipartite genes with Bip-matching promoter H3K27ac levels (blue, n = 99 genes) are compared. E10.5Bip gene transcripts are only detected in as little as 6% of single cells on average. Plots extend from the data minima to the maxima with the white dot indicating median, the box showing the interquartile range and whiskers extending to the most extreme data point within 1.5X the interquartile range. P values are from two-sided Wilcoxon’s tests. NS: not significant (P > 0.05).

Extended Data Fig. 5 t-SNE visualization of mouse genes according to chromatin pattern.

a, Aggregate plot of chromatin profiles (ATAC-seq, ChIP-seq) around transcript start sites (TSSs) of all autosomal genes in E14.5 Drg11vPrV-ZsGreen/+ barrelette neurons. Promoter and gene body regions are highlighted (Methods). b-d, Two-dimensional (2D) projection on a E14.5, E18.5 and P4-combined t-SNE map of autosomal genes according to chromatin accessibility, H3K27me3, H3K4me2, and H3K27ac levels at promoters and gene bodies in Drg11vPrV-ZsGreen/+ barrelette neurons (Methods). b, E14.5 (red), E18.5 (blue) and P4 (green) genes are indicated (Supplementary Note). c, Color-coded t-SNE gene maps according to promoter (top row) and gene body (bottom row) chromatin profiles (columns). d, Color-coded t-SNE gene maps indicating mRNA levels. Numbers 1–3: example genes in E14.5 barrelette neurons. e, Chromatin profiles of the example genes in d, namely H3K4me2 + /H3K27ac + /H3K27me3-/ATAC + (active, Actb), Polycomb-dependent H3K4me2 + /H3K27ac-/H3K27me3 + /ATAC + (permissive, Dlx5) and H3K4me2-/H3K27ac-/H3K27me3-/ATAC- (repressed, Olfr67). f, color-coded E14.5, E18.5 and P4-combined t-SNE gene maps displaying bipartiteness (left) or bivalency (right) scores in E14.5, E18.5 and P4 barrelette neurons (Methods). g, E14.5, E18.5 and P4-combined t-SNE map with contour lines indicating regions enriched with bipartite (green) and bivalent (red) genes (Methods). h, E14.5, E18.5 and P4-combined t-SNE map in which all E14.5 genes are colored according to their developmental change of chromatin state from E14.5 to P4 (Supplementary Note and Methods). i, E14.5Bip genes are subdivided into two subgroups according to their localization on the t-SNE plot (n = 57, E14.5Bip-a genes, orange dots; n = 43, E14.5Bip-b genes, blue dots). j, Violin plots showing promoter H3K27ac (left), gene body H3K27me3 (middle) and mRNA (right) levels in E14.5Bip-a (orange) and E14.5Bip-b genes (blue) in E14.5 barrelette neurons (Supplementary Note and Methods). Plots extend from the data minima to the maxima with the white dot indicating median, the box showing the interquartile range and whiskers extending to the most extreme data point within 1.5X the interquartile range. P values are from two-sided Wilcoxon’s tests. k and l, E10.5 K20tdTomato/+ progenitor t-SNE maps with bipartiteness or bivalency scores (k) and contour lines (l) at E10.5 (Methods).

Extended Data Fig. 6 Developmental dynamics of the bipartite chromatin signature.

a, Developmental dynamics of chromatin profiles (ATAC-seq and ChIP-seq signals in promoter and gene body regions) and mRNA levels of E14.5Bip genes in E14.5, E18.5 and P4 Drg11vPrV-ZsGreen/+ barrelette neurons. Log2 fold changes are calculated with reference to E14.5. At P4, 20% of E14.5Bip genes become expressed (Exp, that is RPKM > = 3 at P4), 25% become bivalent (Biv, red dots in Fig. 3d), and the rest (55%) remain bipartite (Bip), (blue, red and green lines, respectively). Bottom: summary diagram. b, Genome browser view of the Egr1 locus at the E10.5, E14.5, E18.5 and P4 stages. c, 3D interaction map (4C-seq) using the Fos promoter (top left), enhancer 2 (e2, bottom left) and enhancer 5 (e5, bottom right) as viewpoints in E10.5, E14.5, E18.5 and P4 hindbrain tissue. Normalized read per 4 C fragment is visualized. d, Genome browser views of Fos and Egr1 at the E14.5, E18.5 and P4 stages. e, Violin plots showing transcription end site (TES, Methods) RNAPII-S7P (left), RNAPII-S2P (middle), and H3K36me3 (right) levels of E14.5Bip genes at E14.5 and P4 (see a). E14.5Bip genes that become expressed (Exp, blue, n = 25) at P4 displayed significantly higher levels of RNAPII-S7P, -S2P and H3K36me3 marks as compared to E14.5Bip genes that become bivalent (Biv, red, n = 20) or remain bipartite (Bip, green, n = 55) at P4: violin deviations between groups are to be compared within the same time point since the time points reflect different batches. Plots extend from the data minima to the maxima with the white dot indicating median, the box showing the interquartile range and whiskers extending to the most extreme data point within 1.5X the interquartile range. P values are from two-sided Wilcoxon’s tests. NS: not significant (P > 0.05).

Extended Data Fig. 7 Polycomb marking of gene bodies inhibits productive mRNA elongation.

a, Spliced transcript expression (transcripts per million, TPM, Methods) of E14.5Bip genes (red) and control genes (blue) with matching distributions of spliced transcripts (Methods) (n = 3 biologically independent replicates). b, Unspliced over total transcript fractions for each gene set in a (Methods). Note the larger fraction of unspliced transcripts of E14.5Bip genes compared to control genes. c, Fractions of spliced over total transcripts of E14.5Bip genes (red) in E14.5 control K20tdTomato/+ (WT) and Ezh2cKOr3-RFP hindbrain cells (n = 3, biologically independent littermates). d, Violin plots comparing the TSS/whole gene ratios of total RNA-seq reads between Ezh2cKOr3-RFP and WT hindbrain cells (see Fig. 5c). e, Scatter plot (left) comparing H3K36me3 levels on the coding region between E14.5 wild-type Hoxa2tdTomato/+ (WT) and Ezh2cKOHB-RFP hindbrain cells, indicating H3K36me3-positive and negative E14.5Bip genes in red and gray, respectively (Methods). Increased H3K36me3 levels of H3K36me3-positive E14.5Bip genes in Ezh2-depleted cells are further illustrated (right panel). Bars indicate the median. f, (left, middle) Genes with non-zero expression in EedKO and wild-type (WT), and carrying H3K27me3 on gene bodies in mouse ESCs (n = 3457) were subdivided into genes that show up-regulated (red, n = 1067), down-regulated (blue, n = 93) and unchanged (gray, n = 2297) levels of mRNA in EedKO compared with WT ESCs (Supplementary Note, Methods). (right) Violin plots showing log2 fold changes of transcription end site (TES) RNAPII-S2P levels in EedKO mutant compared with WT ESCs. g, Chromatin profiles of Fos in WT and EedKO ESCs. While stalled RNAPII-S5P showed decrease in the promoter (pink highlight), elongating RNAPII-S2P was increased (green highlight) in EedKO ESCs. h, MA-plots comparing mRNA levels of bipartite (Bip) genes (n = 100, red) between full Ezh1KO;Ezh2KO (left) or Ezh2 catalytically inactive Ezh1KO;Ezh2Y726D (right) with WT ESCs. d and f, Plots extend from the data minima to the maxima with the white dot or middle bar indicating median, the box showing the interquartile range and whiskers extending to the most extreme data point within 1.5X the interquartile range. P value is from a two-sided Wilcoxon’s test between the two groups.

Extended Data Fig. 8 IEG bipartite chromatin is necessary to prevent precocious activity-dependent neuronal maturation.

a, Short time (two days) E12.5 ex vivo-cultured Drg11tdTomato/+ hindbrains. After over-night (o/n) treatment with a cocktail of neuronal activity blockers (TDN cocktail = TTX + D-AP5 + NBQX, inhibitors of sodium channel, NMDAR and AMPAR), cultured neurons were treated by 55 mM KCl for 1 hour. Drg11-positive immature trigeminal neurons were FACS-isolated for ATAC-seq analysis. Violin plots visualize log2 fold changes of enhancer chromatin accessibilities in 1 hour KCl-treated neurons as compared to non-treated control neurons. Increased accessibility is selectively detected in KCl-treated neurons at activity-dependent Fos-binding enhancers that normally become open only at P4 (green, n = 85 enhancers) (purple, all non-Fos-binding enhancers that gain accessibilities only at P4, n = 3882 enhancers). Plots extend from the data minima to the maxima with the white dot indicating median, the box showing the interquartile range and whiskers extending to the most extreme data point within 1.5X the interquartile range from the box. P value is from a two-sided Wilcoxon’s test. b, Scatterplots comparing enhancer accessibilities (ATAC) in E14.5 Ezh2 heterozygous control (ctrl) and homozygous mutant (Ezh2cKOHB-RFP) hindbrain cells. All the barrelette enhancers in Drg11vPrV-ZsGreen/+ barrelette neurons (left), non-Fos-binding enhancers that gain accessibilities at P4 as compared with E14.5 in Drg11vPrV-ZsGreen/+ barrelette neurons (n = 3882 enhancers, middle, purple), neuronal activity-dependent Fos-binding enhancers that gain accessibilities at P4 as compared with E14.5 Drg11vPrV-ZsGreen/+ barrelette neurons (n = 85 enhancers, left, green) are shown (Methods). 85 activity-dependent Fos-binding enhancers show precocious opening upon H3K27me3 removal at E14.5. Also see Fig. 5d.

Extended Data Fig. 9 dCas9-UTX overexpression in E12.5 short-term ex vivo cultured neurons.

a, H3K27me3 profiles of the Fos locus in E14.5 Drg11vPrV-ZsGreen/+ barrelette neurons and E12.5 cultured hindbrain neurons at day 1 and day 7 of culture. One week hindbrain neuron culture results in the loss of the H3K27me3 mark from the Fos gene body, similarly to one-week embryonic cortical neuron culture (Extended Data Fig. 2a); in contrast, short-term (day 1) cultured hindbrain neurons still maintain H3K27me3 levels comparable to E14.5 barrelette neurons. b, H3K27me3 levels at the Fos locus in short-term cultured E12.5 hindbrain neurons overexpressing control dCas9 (green) or dCas9-UTX (red) targeted to Fos gene body: three biological replicates overlaid. Genomic regions targeted by guide-RNAs (gRNAs) are indicated. c, Averaged H3K27me3 profiles of Fos (left) and the rest of the E14.5Bip genes (right). Overexpression of control dCas9 (green) or dCas9-UTX (red) are compared. d, (left) MA-plot comparing H3K27me3 levels of dCas9-UTX against dCas9 targeting to Fos locus. Fos (red dot) shows a loss of gene body H3K27me3 compared to control genes carrying similar levels of H3K27me3 (green dots). (right) Density plot (green line) shows the distribution of the logFC values of the selected control genes (greed dots), highlighting the Fos logFC (red line) in the 1.48 % lower tail of the density plot and logFC of E14.5Bip genes (yellow line), indicating slight but significant decrease in Fos H3K27me3. n = 3 biologically independent neuron cultures. e, mRNA levels of Actb, Gapdh and Fos determined by RT-qPCR in short-term cultured E12.5 hindbrain neurons overexpressing control dCas9 (green) or dCas9-UTX (purple) targeted to Actb (left) or Gapdh (right) loci (n = 6 biologically independent neuron cultures). The median expression is indicated by bars. P values are from Welch’s two-sample two-sided t-tests. NS: not significant (P > 0.05).

Extended Data Fig. 10 Polycomb marking of bipartite gene bodies regulates the rapidity and amplitude of transcriptional response to relevant stimuli.

a, Genome browser view of bivalent Junb in E14.5 Drg11vPrV-ZsGreen/+ barrelette neurons. Chromatin accessibility (ATAC), H3K4me2, H3K27ac and H3K27me3 are shown. b, mRNA levels of bipartite (Fos, Egr1) and bivalent (Junb) ARGs, determined by RT-qPCR in E14.5 hindbrain cells treated by KCl for 8 (left) or 30 min (right) (n = 4, biologically independent embryos). The median expression is indicated by bars. P values are from Welch’s two-sample two-sided t-tests. NS: not significant (p > 0.05). c, mRNA levels of Fos (left) and Egr1 (right), determined by RT-qPCR in serum-starved WT (green) and EedKO (purple) mouse ESCs (n = 4, biologically independent cultured cells) treated with a low (1%) or high (10%) concentration of Fetal Calf Serum (FCS) for 16 minutes. WT ESCs Fos and Egr1 could be induced only after prolonged exposure (that is 16 minutes) to 1% FCS: also see the effects of shorter (that is 8 minutes) time exposure to 1% FCS in Fig. 5f and g. The median expression is indicated by bars. P values are from Welch’s two-sample two-sided t-tests.

Supplementary information

Supplementary Information

Supplementary Note, Discussion, Methods, Figs. 1–4 and References

Reporting Summary

Peer Review Information

Supplementary Tables 1–5

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Kitazawa, T., Machlab, D., Joshi, O. et al. A unique bipartite Polycomb signature regulates stimulus-response transcription during development. Nat Genet 53, 379–391 (2021). https://doi.org/10.1038/s41588-021-00789-z

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1038/s41588-021-00789-z

This article is cited by

Search

Advanced search

Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing