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.

  • Protocol
  • Published:

Selective ribosome profiling as a tool for studying the interaction of chaperones and targeting factors with nascent polypeptide chains and ribosomes

Abstract

A plethora of factors is involved in the maturation of newly synthesized proteins, including chaperones, membrane targeting factors and enzymes. Many factors act co-translationally through association with ribosome-nascent chain complexes (RNCs), but their target specificities and modes of action remain poorly understood. We developed selective ribosome profiling (SeRP) to identify substrate pools and points of RNC engagement of these factors. SeRP is based on sequencing mRNA fragments covered by translating ribosomes (general ribosome profiling (RP)), combined with a procedure to selectively isolate RNCs whose nascent polypeptides are associated with the factor of interest. Factor-RNC interactions are stabilized by cross-linking; the resulting factor-RNC adducts are nuclease-treated to generate monosomes, and then they are affinity purified. The ribosome-extracted mRNA footprints are converted to DNA libraries for deep sequencing. The protocol is specified for general RP and SeRP in bacteria. It was first applied to the chaperone trigger factor (TF) and is readily adaptable to other co-translationally acting factors, including eukaryotic factors. Factor-RNC purification and sequencing library preparation takes 7–8 d, and sequencing and data analysis can be completed in 5–6 d.

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

Access options

Buy this article

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

Figure 1: Outline of the procedure for general ribosome profiling (RP) in bacteria (in black) and selective ribosome profiling (SeRP) of factor-RNCs (in red).
Figure 2: Translatome analyses of cells harvested according to the conventional or the rapid harvesting method.
Figure 3: The distribution of ribosomes along an average message (Meta-gene analysis) from cells harvested according to the conventional or the rapid harvesting method.
Figure 4: The effect of cross-linking on the purification of TF-RNCs.
Figure 5: Comparison of samples cross-linked in vivo and ex vivo in translatome and interactome analyses.
Figure 6: Comparison of translatome and interactome samples after in vivo and ex vivo cross-linking.
Figure 7: Footprint fragment lengths and rRNA contamination variation according to digestion conditions.
Figure 8: Effect of salt concentrations on the stability of ribosomes during ribosome purification.
Figure 9: Evidence that translating ribosomes can be specifically pulled down with TF.
Figure 10: Preparation of a deep-sequencing library.

Similar content being viewed by others

References

  1. Pechmann, S., Willmund, F. & Frydman, J. The ribosome as a hub for protein quality control. Mol. Cell 49, 411–421 (2013).

    Article  CAS  Google Scholar 

  2. Kramer, G., Boehringer, D., Ban, N. & Bukau, B. The ribosome as a platform for co-translational processing, folding and targeting of newly synthesized proteins. Nat. Struct. Mol. Biol. 16, 589–597 (2009).

    Article  CAS  Google Scholar 

  3. Giglione, C., Boularot, A. & Meinnel, T. Protein N-terminal methionine excision. Cell Mol. Life Sci. 61, 1455–1474 (2004).

    Article  CAS  Google Scholar 

  4. Starheim, K.K., Gevaert, K. & Arnesen, T. Protein N-terminal acetyltransferases: when the start matters. Trends Biochem. Sci. 37, 152–161 (2012).

    Article  CAS  Google Scholar 

  5. Jones, J.D. & O'Connor, C.D. Protein acetylation in prokaryotes. Proteomics 11, 3012–3022 (2011).

    Article  CAS  Google Scholar 

  6. Hartl, F.U. & Hayer-Hartl, M. Converging concepts of protein folding in vitro and in vivo. Nat. Struct. Mol. Biol. 16, 574–581 (2009).

    Article  CAS  Google Scholar 

  7. Hartl, F.U., Bracher, A. & Hayer-Hartl, M. Molecular chaperones in protein folding and proteostasis. Nature 475, 324–332 (2011).

    Article  CAS  Google Scholar 

  8. Preissler, S. & Deuerling, E. Ribosome-associated chaperones as key players in proteostasis. Trends Biochem. Sci. 37, 274–283 (2012).

    Article  CAS  Google Scholar 

  9. Egea, P.F., Stroud, R.M. & Walter, P. Targeting proteins to membranes: structure of the signal recognition particle. Curr. Opin. Struct. Biol. 15, 213–220 (2005).

    Article  CAS  Google Scholar 

  10. Ingolia, N.T., Brar, G.A., Rouskin, S., McGeachy, A.M. & Weissman, J.S. The ribosome profiling strategy for monitoring translation in vivo by deep sequencing of ribosome-protected mRNA fragments. Nat. Protoc. 7, 1534–1550 (2012).

    Article  CAS  Google Scholar 

  11. Oh, E. et al. Selective ribosome profiling reveals the cotranslational chaperone action of trigger factor in vivo. Cell 147, 1295–1308 (2011).

    Article  CAS  Google Scholar 

  12. Ingolia, N.T., Ghaemmaghami, S., Newman, J.R. & Weissman, J.S. Genome-wide analysis in vivo of translation with nucleotide resolution using ribosome profiling. Science 324, 218–223 (2009).

    Article  CAS  Google Scholar 

  13. Brar, G.A. et al. High-resolution view of the yeast meiotic program revealed by ribosome profiling. Science 335, 552–557 (2012).

    Article  CAS  Google Scholar 

  14. Ingolia, N.T., Lareau, L.F. & Weissman, J.S. Ribosome profiling of mouse embryonic stem cells reveals the complexity and dynamics of mammalian proteomes. Cell 147, 789–802 (2011).

    Article  CAS  Google Scholar 

  15. Li, G.W., Oh, E. & Weissman, J.S. The anti-Shine-Dalgarno sequence drives translational pausing and codon choice in bacteria. Nature 484, 538–541 (2012).

    Article  CAS  Google Scholar 

  16. Han, Y. et al. Monitoring cotranslational protein folding in mammalian cells at codon resolution. Proc. Natl. Acad. Sci. USA 109, 12467–12472 (2012).

    Article  CAS  Google Scholar 

  17. Patzelt, H. et al. Three-state equilibrium of Escherichia coli trigger factor. Biol. Chem. 383, 1611–1619 (2002).

    Article  Google Scholar 

  18. Kramer, G. et al. L23 protein functions as a chaperone docking site on the ribosome. Nature 419, 171–174 (2002).

    Article  CAS  Google Scholar 

  19. Mitkevich, V.A. et al. Thermodynamic characterization of ppGpp binding to EF-G or IF2 and of initiator tRNA binding to free IF2 in the presence of GDP, GTP, or ppGpp. J. Mol. Biol. 402, 838–846 (2010).

    Article  CAS  Google Scholar 

  20. Nakatogawa, H. & Ito, K. The ribosomal exit tunnel functions as a discriminating gate. Cell 108, 629–636 (2002).

    Article  CAS  Google Scholar 

  21. Gong, F. & Yanofsky, C. Instruction of translating ribosome by nascent peptide. Science 297, 1864–1867 (2002).

    Article  CAS  Google Scholar 

  22. Wilson, D.N. On the specificity of antibiotics targeting the large ribosomal subunit. Ann. NY Acad. Sci. 1241, 1–16 (2011).

    Article  CAS  Google Scholar 

  23. Kerner, M.J. et al. Proteome-wide analysis of chaperonin-dependent protein folding in Escherichia coli. Cell 122, 209–220 (2005).

    Article  CAS  Google Scholar 

  24. Calloni, G. et al. DnaK functions as a central hub in the E. coli chaperone network. Cell Rep. 1, 251–264 (2012).

    Article  CAS  Google Scholar 

  25. Datta, A.K. & Burma, D.P. Association of ribonuclease I with ribosomes and their subunits. J. Biol. Chem. 247, 6795–6801 (1972).

    CAS  PubMed  Google Scholar 

  26. Dingwall, C., Lomonossoff, G.P. & Laskey, R.A. High sequence specificity of micrococcal nuclease. Nucleic Acids Res. 9, 2659–2673 (1981).

    Article  CAS  Google Scholar 

  27. Martin, M. Cutadapt removes adapter sequences from high-throughput sequencing reads. EMBnet. J. 17, 10–12 (2011).

    Article  Google Scholar 

  28. 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  Google Scholar 

  29. Langmead, B. & Salzberg, S.L. Fast gapped-read alignment with Bowtie 2. Nat. Methods 9, 357–359 (2012).

    Article  CAS  Google Scholar 

  30. Merz, F. et al. Molecular mechanism and structure of Trigger Factor bound to the translating ribosome. EMBO J. 27, 1622–1632 (2008).

    Article  CAS  Google Scholar 

  31. Buskiewicz, I. et al. Trigger factor binds to ribosome-signal-recognition particle (SRP) complexes and is excluded by binding of the SRP receptor. Proc. Natl. Acad. Sci. USA 101, 7902–7906 (2004).

    Article  CAS  Google Scholar 

  32. Raine, A., Lovmar, M., Wikberg, J. & Ehrenberg, M. Trigger factor binding to ribosomes with nascent peptide chains of varying lengths and sequences. J. Biol. Chem. 281, 28033–28038 (2006).

    Article  CAS  Google Scholar 

  33. Bornemann, T., Jockel, J., Rodnina, M.V. & Wintermeyer, W. Signal sequence-independent membrane targeting of ribosomes containing short nascent peptides within the exit tunnel. Nat. Struct. Mol. Biol. 15, 494–499 (2008).

    Article  CAS  Google Scholar 

  34. Vorderwulbecke, S. et al. Low temperature or GroEL/ES overproduction permits growth of Escherichia coli cells lacking trigger factor and DnaK. FEBS Lett. 559, 181–187 (2004).

    Article  CAS  Google Scholar 

  35. Martinez-Hackert, E. & Hendrickson, W.A. Promiscuous substrate recognition in folding and assembly activities of the trigger factor chaperone. Cell 138, 923–934 (2009).

    Article  CAS  Google Scholar 

  36. Deuerling, E., Schulze-Specking, A., Tomoyasu, T., Mogk, A. & Bukau, B. Trigger factor and DnaK cooperate in folding of newly synthesized proteins. Nature 400, 693–696 (1999).

    Article  CAS  Google Scholar 

  37. Deuerling, E. et al. Trigger Factor and DnaK possess overlapping substrate pools and binding specificities. Mol. Microbiol. 47, 1317–1328 (2003).

    Article  CAS  Google Scholar 

  38. del Alamo, M. et al. Defining the specificity of cotranslationally acting chaperones by systematic analysis of mRNAs associated with ribosome-nascent chain complexes. PLoS Biol. 9, e1001100 (2011).

    Article  CAS  Google Scholar 

  39. Willmund, F. et al. The cotranslational function of ribosome-associated Hsp70 in eukaryotic protein homeostasis. Cell 152, 196–209 (2013).

    Article  CAS  Google Scholar 

  40. Guo, H., Ingolia, N.T., Weissman, J.S. & Bartel, D.P. Mammalian microRNAs predominantly act to decrease target mRNA levels. Nature 466, 835–840 (2010).

    Article  CAS  Google Scholar 

  41. Homann, O.R. & Johnson, A.D. MochiView: versatile software for genome browsing and DNA motif analysis. BMC Biol. 8, 49 (2010).

    Article  Google Scholar 

  42. van den Berg, S., Lofdahl, P.A., Hard, T. & Berglund, H. Improved solubility of TEV protease by directed evolution. J. Biotechnol. 121, 291–298 (2006).

    Article  CAS  Google Scholar 

Download references

Acknowledgements

We thank B. Zachmann-Brand and N. Reifenberger for cloning and initial experiments; the DKFZ sequencing facility for experimental support; A. Jonasson and B. Haldemann for help in the data analysis; and members of the Bukau and Weissman labs for discussions on troubleshooting and comments on the manuscript. This work was supported by the US National Institutes of Health (P01 AG10770) and the Howard Hughes Medical Institute (J.S.W.); the Peter und Traudl Engelhorn-Stiftung and the European Molecular Biology Organization (EMBO) (A.H.B.); and SFB638 and FOR1805 of the Deutsche Forschungsgemeinschaft (B.B. and G.K.).

Author information

Authors and Affiliations

Authors

Contributions

G.K., J.S.W. and B.B. designed the study. A.H.B. and E.O. performed experiments. E.O. and J.S.W. set up the protocol for general ribosome profiling in bacteria. A.H.B., G.K. and B.B. established the protocol for selective ribosome profiling. A.H.B. and E.O. analyzed the data. A.H.B., G.K. and B.B. wrote the manuscript.

Corresponding authors

Correspondence to Günter Kramer or Bernd Bukau.

Ethics declarations

Competing interests

The authors declare no competing financial interests.

Integrated supplementary information

Supplementary Figure 1 Influence of chloramphenicol treatment on translation.

Depicted is an autoradiograph of samples after ultracentrifugation to isolate ribosomes. E. coli MC4100 cells were grown in 40 ml of M9 minimal media to an OD600 of 0.45 at 37 °C. Cultures were transferred to 50-ml conical tubes in a 37 °C water bath. For radioactive pulse labeling, 6 μl of 10 μCi/μl (10 μM) 35S-methionine (Hartmann, SRM-01) were added to the cultures. After a 50-s pulse at 37 °C, translation was arrested with 1 mM chloramphenicol, and the crosslinker DSP was added to 0.25 mg/ml (0.62 mM) end concentration (sample 1, lane 1). In case of sample 2 (lane 2), 35S-methionine and chloramphenicol were added at the same time and incubated for 50 s, then DSP was added for crosslinking. For samples 3 (lane 3) and 4 (lane 4) chloramphenicol was added first, followed by a 50 s (3) or 3 min (4) incubation at 37 °C before 35S-methionine pulse (50 s) and crosslinking. All samples were crosslinked at 37 °C for 30 min. The crosslinker was quenched with 20 mM Tris pH 7.5, cultures were incubated for 5 more min at 37 °C and transferred to ice. Cells were harvested for 10 min at 4 °C and 4000g (4306 rpm), resuspended in 500 μl of buffer B (50 mM Tris pH 7.5, 1 M potassium acetate, 10 mM MgAc2, 1 mg/ml lysozyme, 1 mM PMSF, 1 mM chloramphenicol, 0.4% Triton X-100, 0.1% NP-40), snap-frozen in liquid nitrogen and stored at −80 °C. Lysis was performed with four times repetition of freezing (in liquid nitrogen) and thawing (in a 30 °C water bath). Then 5 mM CaCl2, 75 U of MNase and 2.5 U of RNase-free DNase I were added and the lysate was incubated for 1 h at RT. After addition of EGTA to 6 mM to inactivate the MNase, the lysate (250-300 μl) was loaded onto 750 μl of sucrose cushion (50 mM Tris pH 7.5, 1 M potassium acetate, 10 mM MgAc2, 1 mM chloramphenicol, 1 mM PMSF, 25% (wt/vol) sucrose) and ultracentrifuged at 254,000 g (75,000 rpm) for 90 min at 4 °C in an S120 AT2 rotor. The pellet was washed once with buffer C (50 mM Tris pH 7.5, 100 mM potassium acetate, 10 mM MgAc2, 1 mM chloramphenicol, 1 mM PMSF, 0.4% Triton X100, 0.1% NP-40) and resuspended in 50 μl of buffer C for 1 h on ice. 5 μl of resuspended ribosomes were mixed with 2x reducing sample buffer and separated on a 10% tricine gel. The gel was coomassie-stained and dried for autoradiography.

Supplementary Figure 2 Purification and activity of recombinant MNase.

(a) Design of the MNase overexpression construct pET24a-ompA-nucB(MNase). Nt 178–684 (corresponding to amino acids (aa) 60–228) of the nuc gene from Staphylococcus aureus (687 nt) were codon-optimized for recombinant expression in E. coli and cloned into an expression vector to allow overexpression by T7-polymerase under IPTG-control. The protein was produced with the signal sequence of the precursor of outer membrane protein A (OmpA; aa 1–20) at its N-terminus to facilitate the export of the active, overproduced nuclease to the periplasm to avoid degradation of cytosolic nucleic acids. The signal sequence was cleaved off during translocation by endogenous signal peptidase. Furthermore, the protein was produced with a C-terminal His6-tag for purification. (b) The new MNase has a slightly higher molecular weight compared to purchased MNase. 1.4 μg of each protein were loaded on 14% SDS-PAGE and the gel was stained with coomassie. While the new, mature, His6-tagged protein has a molecular weight of 20.1 kDa (18.9 kDa without His6-tag), the purchased protein has a slightly lower molecular weight and presumably consists of aa 80–228 from the nuc gene product43. (c–e) Comparison of translatomes derived from lysates treated with new or purchased MNase. E. coli MC4100 cells grown in LB medium were harvested according to step 1, option C. The thawed lysate (according to step 7, option A) was digested with 15 U/A260 of new or purchased MNase and loaded onto sucrose gradients (step 18, option B). Sequencing libraries (without rRNA depletion) were cloned (according to Supplementary Methods) and data were analyzed as described in steps 35–59. (c) Read lengths of footprint fragments generated with different MNases according to analysis step 41. (d,e) Gene expression levels (d) and read densities in protein coding regions (e) were calculated as described in the legend of Fig. 2a and c, respectively.

Supplementary Figure 3 Effect of different salts and salt concentrations on polysome profiles after sucrose cushion and sucrose gradient centrifugations.

(a–d) Polysome profiles of samples that were harvested and digested as described in the legend of Fig. 8. Digested lysates were either loaded directly onto sucrose gradients ('gradient') containing different salts and salt concentrations in the buffer: 100 mM NH4Cl ('low NH4Cl'), 1 M NH4Cl ('high NH4Cl'), 100 mM NaCl ('low NaCl'), and 1 M NaCl ('high NaCl'). Then, the lysates were centrifuged according to step 18, option B. Alternatively, lysates were first loaded onto sucrose cushions (step 18, option A) ('cushion') containing the same salts and concentrations as the gradients. After sucrose cushion centrifugation, ribosomes were then resuspended and 100 μg of ribosomes were subsequently loaded onto sucrose gradients containing the standard concentration of 100 mM NH4Cl and centrifuged. '30S' and '50S' depict the peaks of the small and large ribosomal subunits, respectively. The monosome peak is labeled with '70S'. Polysome profiles were normalized to the area under the curves as explained in the legend of Fig. 4a,b.

Supplementary Figure 4 Comparison of different protein purification procedures for TEV protease.

Plasmid pTH24TEVsh42 contains nt 6255–6981 from the tobacco etch virus (TEV) RNA genome44 corresponding to aa 2038–2279 of the TEV polyprotein encoding nuclear inclusion protein A (49 kDa proteinase; NP_734212.1). The gene is flanked by attachment sides used for Gateway cloning. Nt sequences of the V5 epitope and a His6-tag are fused to the 3' end. The plasmid was transformed into BL21(DE3) Star-Rosetta for TEV protease production. (a) Samples of TEV protease after standard ('old') and nucleic acid free ('new'; according to box 3) protein purification. 1 μg was loaded on 10% SDS-PAGE and proteins were stained with coomassie. Pictures from the two purifications were derived from the same gel, but samples in between were cut out for this illustration. (b) The same samples as in (a) were loaded on a 15% TBE-Urea polyacrylamide gel and nucleic acids were stained with SYBR gold.

Supplementary Figure 5 Gel purification steps and Bioanalyzer results during the deep sequencing library preparation after size selection.

(a) Quantification of dephosphorylated RNA on a Bioanalyzer Small RNA chip (step 38 of the Supplementary Methods). Depicted are samples of translatome, interactome and control oligonucleotide. In all three samples a marker peak with a length of 4 nt is included. (b–d) RNA or DNA samples were loaded on polyacrylamide gels. Gels were stained with SYBR gold and shown before (pre-cut) and after (post-cut) the region of interest (marked with the red box) was excised. (b) Separation of 3'-linked mRNA footprint fragments as well as 3'-linked RNA control oligonucleotide from unligated linker on a 10% TBE-Urea polyacrylamide gel (step 45 and following of the Supplementary Methods). The phosphorylated RNA control oligonucleotide was loaded for size comparison. (c) Purification of the reverse transcribed ssDNA of footprint fragments and control oligonucleotide after the second ligation step on a 10% TBE-Urea polyacrylamide gel (step 60 and following of the Supplementary Methods). (d) Purification of dsDNA after PCR on an 8% TB polyacrylamide gel (step 90 and following of the Supplementary Methods). The PCR was run for 6, 8, 10 and 12 cycles, but only products after 6 and 8 cycles were excised due to the occurrence of higher molecular weight products in later cycles indicating non-linear amplification. (e) Quantification of the dsDNA library for deep sequencing on a Bioanalyzer High Sensitivity DNA chip (step 95 of the Supplementary Methods). Markers with a size of 35 and 10380 nt were included. A peak with the size of around 160 nt (containing footprint fragments with an average of 31 nt in length) shows the expected PCR product. The occurrence of an additional smaller peak with around 130 nt (right panel only) indicates circularization and PCR amplification of free linker L1'L2' lacking a footprint fragment and should, therefore, be avoided.

Supplementary Figure 6 Comparison of translatome data including and excluding an rRNA depletion step during the library preparation.

E. coli MC4100 Δtig::Kan + pTrc-tig-TEV-Avi cells were grown in LB medium, harvested according to the rapid protocol (step 1, option C), and ex vivo crosslinked with EDC (step 7, option B). After polysome digest ribosomes were isolated in sucrose gradient ultracentrifugation (step 18, option B). Isolated footprint fragments were used to prepare a sequencing library including ('with rRNA depl.') or excluding ('no rRNA depl.') an rRNA depletion step. Footprint fragments were sequenced (step 96 of the Supplementary Methods) and data were analyzed in the basic and specific analysis (steps 35–59). (a) Gene expression analysis and (b) read densities along protein coding regions were determined as described in the legend to Fig. 2a and c, respectively.

Supplementary information

Supplementary Figure 1

Influence of chloramphenicol treatment on translation. (PDF 268 kb)

Supplementary Figure 2

Purification and activity of recombinant MNase. (PDF 327 kb)

Supplementary Figure 3

Effect of different salts and salt concentrations on polysome profiles after sucrose cushion and sucrose gradient centrifugations. (PDF 254 kb)

Supplementary Figure 4

Comparison of different protein purification procedures for TEV protease. (PDF 268 kb)

Supplementary Figure 5

Gel purification steps and Bioanalyzer results during the deep sequencing library preparation after size selection. (PDF 1466 kb)

Supplementary Figure 6

Comparison of translatome data including and excluding an rRNA depletion step during the library preparation. (PDF 162 kb)

Supplementary Methods

(PDF 186 kb)

Supplementary Table 1

Biotinylated oligonucleotides for the depletion of rRNA contaminations derived from E. coli ribosomes. (PDF 45 kb)

Supplementary Table 2

PCR primers for the amplification of cloned footprint fragments and introduction of a barcode. (PDF 49 kb)

Supplementary Table 3

List of primers to generate Illumina compatible adaptors. (PDF 68 kb)

Supplementary Note 1

Supplementary Note 1 (TXT 1 kb)

Supplementary Note 2

Supplementary Note 2 (TXT 4 kb)

Supplementary Note 3

Supplementary Note 3 (TXT 1 kb)

Supplementary Note 4

Supplementary Note 4 (TXT 5 kb)

Supplementary Note 5

Supplementary Note 5 (TXT 4 kb)

Supplementary Note 6

Supplementary Note 6 (TXT 2 kb)

Supplementary Note 7

Supplementary Note 7 (TXT 2 kb)

Supplementary Note 8

Supplementary Note 8 (TXT 3 kb)

Supplementary Note 9

Supplementary Note 9 (TXT 1 kb)

Supplementary Note 10

Supplementary Note 10 (TXT 1 kb)

Supplementary Note 11

Supplementary Note 11 (TXT 1 kb)

Supplementary Note 12

Supplementary Note 12 (TXT 5 kb)

Supplementary Note 13

Supplementary Note 13 (TXT 5 kb)

Supplementary Note 14

Supplementary Note 14 (TXT 3 kb)

Rights and permissions

Reprints and permissions

About this article

Cite this article

Becker, A., Oh, E., Weissman, J. et al. Selective ribosome profiling as a tool for studying the interaction of chaperones and targeting factors with nascent polypeptide chains and ribosomes. Nat Protoc 8, 2212–2239 (2013). https://doi.org/10.1038/nprot.2013.133

Download citation

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1038/nprot.2013.133

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