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.

  • Letter
  • Published:

Radically truncated MeCP2 rescues Rett syndrome-like neurological defects

Abstract

Heterozygous mutations in the X-linked MECP2 gene cause the neurological disorder Rett syndrome1. The methyl-CpG-binding protein 2 (MeCP2) protein is an epigenetic reader whose binding to chromatin primarily depends on 5-methylcytosine2,3. Functionally, MeCP2 has been implicated in several cellular processes on the basis of its reported interaction with more than 40 binding partners4, including transcriptional co-repressors (for example, the NCoR/SMRT complex5), transcriptional activators6, RNA7, chromatin remodellers8,9, microRNA-processing proteins10 and splicing factors11. Accordingly, MeCP2 has been cast as a multi-functional hub that integrates diverse processes that are essential in mature neurons12. At odds with the concept of broad functionality, missense mutations that cause Rett syndrome are concentrated in two discrete clusters coinciding with interaction sites for partner macromolecules: the methyl-CpG binding domain13 and the NCoR/SMRT interaction domain5. Here we test the hypothesis that the single dominant function of MeCP2 is to physically connect DNA with the NCoR/SMRT complex, by removing almost all amino-acid sequences except the methyl-CpG binding and NCoR/SMRT interaction domains. We find that mice expressing truncated MeCP2 lacking both the N- and C-terminal regions (approximately half of the native protein) are phenotypically near-normal; and those expressing a minimal MeCP2 additionally lacking a central domain survive for over one year with only mild symptoms. This minimal protein is able to prevent or reverse neurological symptoms when introduced into MeCP2-deficient mice by genetic activation or virus-mediated delivery to the brain. Thus, despite evolutionary conservation of the entire MeCP2 protein sequence, the DNA and co-repressor binding domains alone are sufficient to avoid Rett syndrome-like defects and may therefore have therapeutic utility.

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

Access options

Rent or buy this article

Prices vary by article type

from$1.95

to$39.95

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

Figure 1: Stepwise truncation of MeCP2 protein to retain only the MBD and NID.
Figure 2: Deletion of the MeCP2 N- and C-terminal regions has minimal phenotypic consequence.
Figure 3: Additional deletion of the intervening region leads to protein instability and mild RTT-like symptoms.
Figure 4: Activation or viral transduction of ΔNIC ameliorates neurological phenotypes in MeCP2-deficient mice.

Similar content being viewed by others

References

  1. 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)

    Article  CAS  Google Scholar 

  2. Lewis, J. D. et al. Purification, sequence, and cellular localization of a novel chromosomal protein that binds to methylated DNA. Cell 69, 905–914 (1992)

    Article  CAS  Google Scholar 

  3. Skene, P. J. et al. Neuronal MeCP2 is expressed at near histone-octamer levels and globally alters the chromatin state. Mol. Cell 37, 457–468 (2010)

    Article  CAS  Google Scholar 

  4. Lyst, M. J. & Bird, A. Rett syndrome: a complex disorder with simple roots. Nat. Rev. Genet. 16, 261–275 (2015)

    Article  CAS  Google Scholar 

  5. Lyst, M. J. et al. Rett syndrome mutations abolish the interaction of MeCP2 with the NCoR/SMRT co-repressor. Nat. Neurosci. 16, 898–902 (2013)

    Article  CAS  Google Scholar 

  6. Chahrour, M. et al. MeCP2, a key contributor to neurological disease, activates and represses transcription. Science 320, 1224–1229 (2008)

    Article  ADS  CAS  Google Scholar 

  7. Jeffery, L. & Nakielny, S. Components of the DNA methylation system of chromatin control are RNA-binding proteins. J. Biol. Chem. 279, 49479–49487 (2004)

    Article  CAS  Google Scholar 

  8. Nan, X. et al. Interaction between chromatin proteins MECP2 and ATRX is disrupted by mutations that cause inherited mental retardation. Proc. Natl Acad. Sci. USA 104, 2709–2714 (2007)

    Article  ADS  CAS  Google Scholar 

  9. Agarwal, N. et al. MeCP2 interacts with HP1 and modulates its heterochromatin association during myogenic differentiation. Nucleic Acids Res. 35, 5402–5408 (2007)

    Article  CAS  Google Scholar 

  10. Cheng, T.-L. et al. MeCP2 suppresses nuclear microRNA processing and dendritic growth by regulating the DGCR8/Drosha complex. Dev. Cell 28, 547–560 (2014)

    Article  CAS  Google Scholar 

  11. Young, J. I. et al. Regulation of RNA splicing by the methylation-dependent transcriptional repressor methyl-CpG binding protein 2. Proc. Natl Acad. Sci. USA 102, 17551–17558 (2005)

    Article  ADS  CAS  Google Scholar 

  12. Della Ragione, F., Vacca, M., Fioriniello, S., Pepe, G. & D’Esposito, M. MECP2, a multi-talented modulator of chromatin architecture. Brief. Funct. Genomics 15, 420–431 (2016)

    CAS  PubMed  Google Scholar 

  13. Nan, X., Meehan, R. R. & Bird, A. Dissection of the methyl-CpG binding domain from the chromosomal protein MeCP2. Nucleic Acids Res. 21, 4886–4892 (1993)

    Article  CAS  Google Scholar 

  14. Kriaucionis, S. & Bird, A. The major form of MeCP2 has a novel N-terminus generated by alternative splicing. Nucleic Acids Res. 32, 1818–1823 (2004)

    Article  CAS  Google Scholar 

  15. Nan, X., Tate, P., Li, E. & Bird, A. DNA methylation specifies chromosomal localization of MeCP2. Mol. Cell. Biol. 16, 414–421 (1996)

    Article  CAS  Google Scholar 

  16. Kudo, S. et al. Heterogeneity in residual function of MeCP2 carrying missense mutations in the methyl CpG binding domain. J. Med. Genet. 40, 487–493 (2003)

    Article  CAS  Google Scholar 

  17. Kruusvee, V. et al. Structure of the MeCP2–TBLR1 complex reveals a molecular basis for Rett syndrome and related disorders. Proc. Natl Acad. Sci. USA 114, E3243–E3250 (2017)

    Article  CAS  Google Scholar 

  18. Guy, J., Gan, J., Selfridge, J., Cobb, S. & Bird, A. Reversal of neurological defects in a mouse model of Rett syndrome. Science 315, 1143–1147 (2007)

    Article  ADS  CAS  Google Scholar 

  19. Cheval, H. et al. Postnatal inactivation reveals enhanced requirement for MeCP2 at distinct age windows. Hum. Mol. Genet. 21, 3806–3814 (2012)

    Article  CAS  Google Scholar 

  20. Brown, K. et al. The molecular basis of variable phenotypic severity among common missense mutations causing Rett syndrome. Hum. Mol. Genet. 25, 558–570 (2016)

    Article  CAS  Google Scholar 

  21. Guy, J., Hendrich, B., Holmes, M., Martin, J. E. & Bird, A. A mouse Mecp2-null mutation causes neurological symptoms that mimic Rett syndrome. Nat. Genet. 27, 322–326 (2001)

    Article  CAS  Google Scholar 

  22. Goffin, D. et al. Rett syndrome mutation MeCP2 T158A disrupts DNA binding, protein stability and ERP responses. Nat. Neurosci. 15, 274–283 (2011)

    Article  Google Scholar 

  23. Shahbazian, M. et al. Mice with truncated MeCP2 recapitulate many Rett syndrome features and display hyperacetylation of histone H3. Neuron 35, 243–254 (2002)

    Article  CAS  Google Scholar 

  24. Samaco, R. C. et al. A partial loss of function allele of methyl-CpG-binding protein 2 predicts a human neurodevelopmental syndrome. Hum. Mol. Genet. 17, 1718–1727 (2008)

    Article  CAS  Google Scholar 

  25. Gadalla, K. K. E. et al. Development of a novel AAV gene therapy cassette with improved safety features and efficacy in a mouse model of Rett syndrome. Mol. Ther. Methods Clin. Dev. 5, 180–190 (2017)

    Article  CAS  Google Scholar 

  26. Lagger, S. et al. MeCP2 recognizes cytosine methylated tri-nucleotide and di-nucleotide sequences to tune transcription in the mammalian brain. PLoS Genet. 13, e1006793 (2017)

    Article  Google Scholar 

  27. Kinde, B., Wu, D. Y., Greenberg, M. E. & Gabel, H. W. DNA methylation in the gene body influences MeCP2-mediated gene repression. Proc. Natl Acad. Sci. USA 113, 15114–15119 (2016)

    Article  CAS  Google Scholar 

  28. Baker, S. A. et al. An AT-hook domain in MeCP2 determines the clinical course of Rett syndrome and related disorders. Cell 152, 984–996 (2013)

    Article  CAS  Google Scholar 

  29. Zhou, Z. et al. Brain-specific phosphorylation of MeCP2 regulates activity-dependent Bdnf transcription, dendritic growth, and spine maturation. Neuron 52, 255–269 (2006)

    Article  CAS  Google Scholar 

  30. Li, H., Zhong, X., Chau, K. F., Williams, E. C. & Chang, Q. Loss of activity-induced phosphorylation of MeCP2 enhances synaptogenesis, LTP and spatial memory. Nat. Neurosci. 14, 1001–1008 (2011)

    Article  CAS  Google Scholar 

  31. Cong, L. et al. Multiplex genome engineering using CRISPR/VCas systems. Science 339, 819–823 (2013)

    Article  ADS  CAS  Google Scholar 

  32. Clément, N. & Grieger, J. C. Manufacturing of recombinant adeno-associated viral vectors for clinical trials. Mol. Ther. Methods Clin. Dev. 3, 16002 (2016)

    Article  Google Scholar 

  33. Gadalla, K. K. E. et al. Improved survival and reduced phenotypic severity following AAV9/MECP2 gene transfer to neonatal and juvenile male Mecp2 knockout mice. Mol. Ther. 21, 18–30 (2013)

    Article  CAS  Google Scholar 

  34. Tao, J. et al. Phosphorylation of MeCP2 at serine 80 regulates its chromatin association and neurological function. Proc. Natl Acad. Sci. USA 106, 4882–4887 (2009)

    Article  ADS  CAS  Google Scholar 

  35. Ebert, D. H. et al. Activity-dependent phosphorylation of MeCP2 threonine 308 regulates interaction with NCoR. Nature 499, 341–345 (2013)

    Article  ADS  CAS  Google Scholar 

  36. Ho, K. L. et al. MeCP2 binding to DNA depends upon hydration at methyl-CpG. Mol. Cell 29, 525–531 (2008)

    Article  CAS  Google Scholar 

  37. Lyst, M. J., Connelly, J., Merusi, C. & Bird, A. Sequence-specific DNA binding by AT-hook motifs in MeCP2. FEBS Lett. 590, 2927–2933 (2016)

    Article  CAS  Google Scholar 

Download references

Acknowledgements

This work was supported by the Rett Syndrome Research Trust, Wellcome, and Sylvia Aitken Charitable Trust. R.T. was funded by a Biotechnology and Biological Sciences Research Council Doctoral Training Partnership studentship. We thank the following people for assistance: A. Cook (advice on designing the truncated proteins), A. McClure (animal husbandry), D. Kelly (microscopy), M. Waterfall (flow cytometry) and A. Kerr (statistics). We also thank members of the Bird, Cobb, M. E. Greenberg and G. Mandel laboratories for discussions. A.B. and S.R.C. are members of the Simons Initiative for the Developing Brain at the University of Edinburgh.

Author information

Authors and Affiliations

Authors

Contributions

R.T., A.B. and S.R.C. designed research. R.T., J.S., M.V.K., K.K.E.G., J.G., D.D.S. and R.D.H. performed the experiments. R.T. and S.R.C. analysed the data. R.T. and A.B. wrote the manuscript. All authors reviewed the manuscript.

Corresponding author

Correspondence to Adrian Bird.

Ethics declarations

Competing interests

A.B. is a member of the Board of ArRETT, a company based in the USA with the goal of developing therapies for Rett syndrome.

Additional information

Reviewer Information Nature thanks B. Davidson and the other anonymous reviewer(s) for their contribution to the peer review of this work.

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

Extended data figures and tables

Extended Data Figure 1 Design of the MeCP2 deletion series.

a, Diagram of the genomic DNA sequences encoding WT and ΔNIC MeCP2, showing the retention of the extreme N-terminal amino acids encoded in exons 1 and 2 and the first 10 bp of exon 3, the deletion of the N- and C-terminal regions, the replacement of the intervening region with a linker and SV40 NLS, and the addition of the C-terminal eGFP tag. Colour key: 5′ untranslated region, white; MBD, blue; NID, pink; other MeCP2 coding regions, grey; SV40 NLS, orange; linkers, dark grey; eGFP, green. b, The N-terminal ends of the sequences of all three truncated proteins (e1 and e2 isoforms) showing the fusion of the extreme N-terminal amino acids to the MBD (starting with P72). c, d, Protein sequence alignment of the MBD (c) and NID (d) regions using ClustalWS, shaded according to BLOSUM62 score. Both alignments are annotated with RTT-causing missense mutations (http://mecp2.chw.edu.au/) (red), activity-dependent phosphorylation sites29,34,35 (orange), sequence conservation, interaction domains and known36/predicted (https://npsa-prabi.ibcp.fr/cgi-bin/npsa_automat.pl?page=/NPSA/npsa_phd.html) structure. Interaction sites: methyl-CpG binding domain (residues 78–162 (ref. 13)), AT hook 1 (residues 183–195 (ref. 37)), AT hook 2 (residues 257–272 (ref. 28)), NCoR/SMRT interaction domain (residues 285–309 (ref. 5)). The bipartite NLS is also shown (residues 253–256 and 266–271). The regions retained in ΔNIC are: MBD resides 72–173 (highlighted by the blue shading in c) and NID resides 272–312 (highlighted by the pink shading in d). Residue numbers correspond to that of mammalian e2 isoforms.

Extended Data Figure 2 Truncated MeCP2 proteins retain the ability to bind methylated DNA and the NCoR/SMRT complex.

a, eGFP-tagged truncated proteins immunoprecipitate components of the NCoR/SMRT co-repressor complex: NCoR, HDAC3 and TBL1XR1. WT and R306C were used as positive and negative controls for binding, respectively. In, input; IP, immunoprecipitate. For gel source data, see Supplementary Information. b, Representative images showing localization of eGFP-tagged truncated MeCP2 proteins to mCpG-rich heterochromatic foci when they are overexpressed in mouse fibroblasts (NIH-3T3 cells). WT and R111G were used as controls to show focal and diffuse localization, respectively. Scale bars, 10 μm. c, Representative images showing recruitment of TBL1X–mCherry to heterochromatin by eGFP-tagged truncated proteins when they are co-overexpressed in NIH-3T3 cells. WT and R306C were used as positive and negative controls for TBL1X–mCherry recruitment, respectively. Scale bars, 10 μm. Quantification (right) shows the percentage of cells with focal TBL1X–mCherry localization, evaluated relative to WT using Fisher’s exact tests: R306C, ****P < 0.0001; ΔN, P = 0.071; ΔNC, P = 0.604; ΔNIC, P = 0.460. Total numbers of cells counted: WT, n = 117; R306C, n = 119; ΔN, n = 113; ΔNC, n = 119; ΔNIC, n = 125; over three independent transfection experiments.

Source data

Extended Data Figure 3 Generation of ΔN and ΔNC mice.

Diagrammatic representation of ΔN (a) and ΔNC (b) knock-in mouse line generation. The endogenous Mecp2 allele was targeted in male ES cells. The site of Cas9 cleavage in the WT sequence is shown by the scissors symbol (used for production of ΔN knock-in ES cells). The selection cassette was removed in vivo by crossing chimaeras with deleter (CMV-cre) transgenic mice. Southern blot analysis shows correct targeting of ES cells and successful cassette deletion in the knock-in mice. The solid black line represents the sequence encoded in the targeting vector and the dotted lines indicate the flanking regions of mouse genomic DNA. For gel source data, see Supplementary Information.

Extended Data Figure 4 ΔN and ΔNC knock-in mice express truncated proteins at approximately WT levels and display minimal phenotypes.

a, Western blot analysis of whole-brain extract showing protein sizes and abundance of MeCP2 in ΔN and ΔNC mice and WT-eGFP controls, detected using a GFP antibody. Histone H3 was used as a loading control. *A non-specific band detected by the GFP antibody. For gel source data, see Supplementary Information. b, Flow cytometry analysis of protein levels in nuclei from whole brain (All) and the high-NeuN subpopulations (Neurons) in WT-eGFP (n = 3), ΔN (n = 3) and ΔNC (n = 3) mice, detected using eGFP fluorescence. Graph shows mean ± s.e.m. and genotypes were compared with WT-eGFP controls by t-test: All ΔN, P = 0.338; ΔNC, **P = 0.003; and Neurons ΔN, P = 0.672; ΔNC, *P = 0.014. au, arbitrary units. c, Flow cytometry analysis of protein levels in WT (n = 3) and WT-eGFP (n = 3) mice, detected using an MeCP2 antibody. Graph shows mean ± s.e.m. and genotypes were compared by t-test: All, P = 0.214; and Neurons, P = 0.085. d, Example histogram (of one WT-eGFP sample) showing how the Neuronal subpopulation was defined according to NeuN-AF647 staining. eg, Growth curves of the backcrossed scoring cohorts (e, f; see Fig. 2a–d) and an outbred (g; 75% C57BL/6J) cohort of ΔNC mice (n = 7) and WT littermates (n = 9). Graphs show mean values ± s.e.m. Genotypes were compared using repeated measures ANOVA: ΔN, P = 0.362; ΔNC, ****P < 0.0001; ΔNC (outbred), P = 0.739. Mecp2-null data (n = 20)20 are shown for comparison to the backcrossed cohorts. h, Behavioural analysis of ΔN (n = 10) and ΔNC mice (n = 10) each compared with their WT littermates (n = 10) at 20 weeks of age (see Fig. 2e–g). Total distance travelled in the open field test was measured during a 20 min trial. Graphs show individual values and medians. Genotypes were compared using t-tests: ΔN, P = 0.691; ΔNC, P = 0.791. NS, not significant.

Source data

Extended Data Figure 5 Generation of ΔNIC and STOP mice.

Diagrammatic representation of ΔNIC and STOP mouse line generation. The endogenous Mecp2 allele was targeted in male ES cells. The site of Cas9 cleavage in the WT sequence is shown by the scissors symbol. The selection cassette was removed in vivo by crossing chimaeras with deleter (CMV-cre) transgenic mice to produce constitutively expressing ΔNIC mice, or retained to produce STOP mice. Southern blot analysis shows correct targeting of ES cells and successful cassette deletion in the ΔNIC knock-in mice. The solid black line represents the sequence encoded in the targeting vector and the dotted lines indicate the flanking regions of mouse genomic DNA. For gel source data, see Supplementary Information.

Extended Data Figure 6 ΔNIC mice have a normal lifespan and no activity phenotype but decreased body weight.

a, Kaplan–Meier plot showing survival of an outbred (75% C57BL/6J) cohort of ΔNIC mice (n = 10) and their WT littermate (n = 1). b, Growth curve of the backcrossed cohort used for phenotypic scoring (see Fig. 3d, e). Graph shows mean ± s.e.m. Genotypes were compared using repeated measures ANOVA ****P < 0.0001. Mecp2-null data (n = 20)20 are shown for comparison. c, Behavioural analysis of ΔNIC mice (n = 10) compared with their WT littermates (n = 10) at 20 weeks of age (see Fig. 3f–h). Total distance travelled the open field test was measured during a 20 min trial. Graph shows individual values and medians. Genotypes were compared using a t-test. P = 0.333. NS, not significant.

Source data

Extended Data Figure 7 ΔNIC mice have a less severe phenotype than the mildest mouse model of RTT, R133C.

ac, Repeat presentation of phenotypic analysis of ΔNIC mice and WT littermates in Fig. 3d, e and Extended Data Fig. 6b, this time including eGFP-tagged R133C mice (n = 10)20 for comparison. a, Phenotypic severity scores (mean ± s.e.m.). b, Growth curve (mean ± s.e.m.). c, Survival (Kaplan–Meier plot).

Extended Data Figure 8 STOP mice with transcriptionally silenced ΔNIC resemble Mecp2-nulls.

a, Western blot analysis of whole-brain extract showing protein sizes and abundance of MeCP2 in STOP mice compared with WT-eGFP and ΔNIC controls, detected using a GFP antibody. Histone H3 was used as a loading control. *A non-specific band detected by the GFP antibody. For gel source data, see Supplementary Information. b, Flow cytometry analysis of protein levels in nuclei from whole brain (All) and the high-NeuN subpopulation (Neurons) in WT-eGFP (n = 3), ΔNIC (n = 3) and STOP (n = 3) mice, detected using eGFP fluorescence. Graph shows mean ± s.e.m. and genotypes were compared using t-tests: ****P < 0.0001. au, arbitrary units. c, Phenotypic scoring of STOP mice (n = 22) compared with published Mecp2-null data (n = 12)20. Graph shows mean scores ± s.e.m. d, Kaplan–Meier plot showing survival of STOP mice (n = 14) compared with Mecp2-null data (n = 24)20.

Source data

Extended Data Figure 9 Successful activation of ΔNIC in tamoxifen-injected STOP creERT mice leads to symptom reversal.

a, Southern blot analysis of genomic DNA to determine the level of recombination mediated by CreERT in tamoxifen-injected (+ Tmx) STOP creERT animals. WT, WT creERT, ΔNIC and STOP samples, with or without tamoxifen injection, were included as controls. (Bsu36I digestion, see restriction map in Extended Data Fig. 5.) b, Protein levels in tamoxifen-injected STOP creERT animals were determined using western blotting (upper, n = 5) and flow cytometry (lower, n = 3). Constitutively expressing ΔNIC mice (n = 3) were used for comparison. Graphs show mean values ± s.e.m. (quantification by western blotting is shown normalized to ΔNIC). Genotypes were compared using t-tests: western blotting, P = 0.434; flow cytometry All nuclei, P = 0.128; and Neuronal nuclei, *P = 0.016. For gel source data, see Supplementary Information. c, Heatmap of the phenotypic scores of the tamoxifen-injected STOP creERT (top; n = 9) and STOP (bottom; n = 9 until 8 weeks of age, see survival plot in Fig. 4c) animals (see Fig. 4b), divided into the six categories. The plot is shaded according to the mean score for each category.

Source data

Extended Data Figure 10 Virus-encoded humanized ΔNIC is expressed in brain and does not have adverse consequences in WT mice.

a, b, Representative confocal images from thalamus and brainstem of Mecp2-null + hΔNIC (a) and WT + hΔNIC (b) mice; stained with an antibody against the Myc epitope (red) and the neuronal marker NeuN (green). Nuclei are stained with DAPI (blue). Scale bars, 20 μm. Graphs show transduction efficiency (mean ± s.e.m.) in different brain regions (n = 3 mice per genotype, 27 fields from each brain region). c, Phenotypic scoring (mean ± s.e.m.) of scAAV-injected and control mice from 5 to 30 weeks: WT + vehicle (n = 15), Mecp2-null + vehicle (n = 20) and WT + hΔNIC (n = 14). d, Kaplan–Meier plot showing survival of the cohort shown in c. One WT + hΔNIC animal was culled owing to injuries at 28 weeks of age (shown by a tick). An arrow indicates the timing of the viral injection.

Source data

Supplementary information

Supplementary Information

This file contains the uncropped gels. (PDF 790 kb)

Reporting Summary (PDF 85 kb)

PowerPoint slides

Source data

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Tillotson, R., Selfridge, J., Koerner, M. et al. Radically truncated MeCP2 rescues Rett syndrome-like neurological defects. Nature 550, 398–401 (2017). https://doi.org/10.1038/nature24058

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1038/nature24058

This article is cited by

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

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