Trace-gas metabolic versatility of the facultative methanotroph Methylocella silvestris

This article has been updated (view changelog)

Abstract

The climate-active gas methane is generated both by biological processes and by thermogenic decomposition of fossil organic material, which forms methane and short-chain alkanes, principally ethane, propane and butane1,2. In addition to natural sources, environments are exposed to anthropogenic inputs of all these gases from oil and gas extraction and distribution. The gases provide carbon and/or energy for a diverse range of microorganisms that can metabolize them in both anoxic3 and oxic zones. Aerobic methanotrophs, which can assimilate methane, have been considered to be entirely distinct from utilizers of short-chain alkanes, and studies of environments exposed to mixtures of methane and multi-carbon alkanes have assumed that disparate groups of microorganisms are responsible for the metabolism of these gases. Here we describe the mechanism by which a single bacterial strain, Methylocella silvestris, can use methane or propane as a carbon and energy source, documenting a methanotroph that can utilize a short-chain alkane as an alternative to methane. Furthermore, during growth on a mixture of these gases, efficient consumption of both gases occurred at the same time. Two soluble di-iron centre monooxygenase (SDIMO) gene clusters were identified and were found to be differentially expressed during bacterial growth on these gases, although both were required for efficient propane utilization. This report of a methanotroph expressing an additional SDIMO that seems to be uniquely involved in short-chain alkane metabolism suggests that such metabolic flexibility may be important in many environments where methane and short-chain alkanes co-occur.

Access options

Rent or Buy article

Get time limited or full article access on ReadCube.

from$8.99

All prices are NET prices.

Figure 1: Propane enhanced Methylocella silvestris growth, and methane and propane were consumed by cultures at rates corresponding to their relative concentrations.
Figure 2: Methane induced transcription of sMMO only, whereas propane induced both sMMO and PrMO.
Figure 3: Inactivation of PrMO had a marginal effect during growth on a mixture of methane and propane.

Change history

  • 01 May 2014

    The Supplementary Information file was removed on 1 May 2014; the table it contained is presented in Extended Data Fig. 3.

References

  1. 1

    Etiope, G. & Ciccioli, P. Earth’s degassing: a missing ethane and propane source. Science 323, 478 (2009)

    ADS  CAS  Article  Google Scholar 

  2. 2

    Mango, F. D., Hightower, J. W. & James, A. T. Role of transition-metal catalysis in the formation of natural gas. Nature 368, 536–538 (1994)

    ADS  CAS  Article  Google Scholar 

  3. 3

    Heider, J., Spormann, A. M., Beller, H. R. & Widdel, F. Anaerobic bacterial metabolism of hydrocarbons. FEMS Microbiol. Rev. 22, 459–473 (1998)

    CAS  Article  Google Scholar 

  4. 4

    Etiope, G., Lassey, K. R., Klusman, R. W. & Boschi, E. Reappraisal of the fossil methane budget and related emission from geologic sources. Geophys. Res. Lett. 35, L09307 (2008)

    ADS  Article  Google Scholar 

  5. 5

    Reddy, C. M. et al. Composition and fate of gas and oil released to the water column during the Deepwater Horizon oil spill. Proc. Natl Acad. Sci. USA 109, 20229–20234 (2012)

    ADS  CAS  Article  Google Scholar 

  6. 6

    Osborn, S. G., Vengosh, A., Warner, N. R. & Jackson, R. B. Methane contamination of drinking water accompanying gas-well drilling and hydraulic fracturing. Proc. Natl Acad. Sci. USA 108, 8172–8176 (2011)

    ADS  CAS  Article  Google Scholar 

  7. 7

    Etiope, G., Drobniak, A. & Schimmelmann, A. Natural seepage of shale gas and the origin of ‘eternal flames’ in the Northern Appalachian Basin, USA. Mar. Petrol. Geol. 43, 178–186 (2013)

    CAS  Article  Google Scholar 

  8. 8

    Etiope, G. & Klusman, R. W. Microseepage in drylands: flux and implications in the global atmospheric source/sink budget of methane. Glob. Planet. Change 72, 265–274 (2010)

    ADS  Article  Google Scholar 

  9. 9

    Hinrichs, K. U. et al. Biological formation of ethane and propane in the deep marine subsurface. Proc. Natl Acad. Sci. USA 103, 14684–14689 (2006)

    ADS  CAS  Article  Google Scholar 

  10. 10

    Reeburgh, W. S. in Treatise on Geochemistry (eds Holland, H. D. & Turekian, K. K. ) 1–32 (Elsevier, 2007)

    Google Scholar 

  11. 11

    Hamamura, N. & Arp, D. J. Isolation and characterization of alkane-utilizing Nocardioides sp. strain CF8. FEMS Microbiol. Lett. 186, 21–26 (2000)

    CAS  Article  Google Scholar 

  12. 12

    Ashraf, W., Mihdhir, A. & Murrell, J. C. Bacterial oxidation of propane. FEMS Microbiol. Lett. 122, 1–6 (1994)

    CAS  Article  Google Scholar 

  13. 13

    Kotani, T., Kawashima, Y., Yurimoto, H., Kato, N. & Sakai, Y. Gene structure and regulation of alkane monooxygenases in propane-utilizing Mycobacterium sp. TY-6 and Pseudonocardia sp. TY-7. J. Biosci. Bioeng. 102, 184–192 (2006)

    CAS  Article  Google Scholar 

  14. 14

    Kotani, T., Yamamoto, T., Yurimoto, H., Sakai, Y. & Kato, N. Propane monooxygenase and NAD+-dependent secondary alcohol dehydrogenase in propane metabolism by Gordonia sp. strain TY-5. J. Bacteriol. 185, 7120–7128 (2003)

    CAS  Article  Google Scholar 

  15. 15

    Johnson, E. L. & Hyman, M. R. Propane and n-butane oxidation by Pseudomonas putida GPo1. Appl. Environ. Microbiol. 72, 950–952 (2006)

    CAS  Article  Google Scholar 

  16. 16

    Semrau, J. D., DiSpirito, A. A. & Vuilleumier, S. Facultative methanotrophy: false leads, true results, and suggestions for future research. FEMS Microbiol. Lett. 323, 1–12 (2011)

    CAS  Article  Google Scholar 

  17. 17

    Cooley, R. B., Dubbels, B. L., Sayavedra-Soto, L. A., Bottomley, P. J. & Arp, D. J. Kinetic characterization of the soluble butane monooxygenase from Thauera butanivorans, formerly ‘Pseudomonas butanovora’. Microbiology 155, 2086–2096 (2009)

    CAS  Article  Google Scholar 

  18. 18

    Dedysh, S. N., Knief, C. & Dunfield, P. F. Methylocella species are facultatively methanotrophic. J. Bacteriol. 187, 4665–4670 (2005)

    CAS  Article  Google Scholar 

  19. 19

    Rahman, M. T. et al. Environmental distribution and abundance of the facultative methanotroph Methylocella. ISME J. 5, 1061–1066 (2011)

    CAS  Article  Google Scholar 

  20. 20

    Mason, O. U. et al. Metagenome, metatranscriptome and single-cell sequencing reveal microbial response to Deepwater Horizon oil spill. ISME J. 6, 1715–1727 (2012)

    CAS  Article  Google Scholar 

  21. 21

    Theisen, A. R. et al. Regulation of methane oxidation in the facultative methanotroph Methylocella silvestris BL2. Mol. Microbiol. 58, 682–692 (2005)

    CAS  Article  Google Scholar 

  22. 22

    Mackay, D. & Shiu, W. Y. A critical review of Henry’s law constants for chemicals of environmental interest. J. Phys. Chem. Ref. Data 10, 1175–1199 (1981)

    ADS  CAS  Article  Google Scholar 

  23. 23

    Chen, Y. et al. Complete genome sequence of the aerobic facultative methanotroph Methylocella silvestris BL2. J. Bacteriol. 192, 3840–3841 (2010)

    CAS  Article  Google Scholar 

  24. 24

    Coleman, N. V., Bui, N. B. & Holmes, A. J. Soluble di-iron monooxygenase gene diversity in soils, sediments and ethene enrichments. Environ. Microbiol. 8, 1228–1239 (2006)

    CAS  Article  Google Scholar 

  25. 25

    Colby, J., Stirling, D. I. & Dalton, H. The soluble methane mono-oxygenase of Methylococcus capsulatus (Bath). Its ability to oxygenate n-alkanes, n-alkenes, ethers, and alicyclic, aromatic and heterocyclic compounds. Biochem. J. 165, 395–402 (1977)

    CAS  Article  Google Scholar 

  26. 26

    Patel, N. A. et al. Comparison of one- and two-dimensional liquid chromatography approaches in the label-free quantitative analysis of Methylocella silvestris. J. Proteome Res. 11, 4755–4763 (2012)

    CAS  Article  Google Scholar 

  27. 27

    Furuya, T., Hirose, S., Osanai, H., Semba, H. & Kino, K. Identification of the monooxygenase gene clusters responsible for the regioselective oxidation of phenol to hydroquinone in mycobacteria. Appl. Environ. Microbiol. 77, 1214–1220 (2011)

    CAS  Article  Google Scholar 

  28. 28

    Prior, S. D. & Dalton, H. Acetylene as a suicide substrate and active site probe for methane monooxygenase from Methylococcus capsulatus (Bath). FEMS Microbiol. Lett. 29, 105–109 (1985)

    CAS  Article  Google Scholar 

  29. 29

    McDonald, I. R., Bodrossy, L., Chen, Y. & Murrell, J. C. Molecular ecology techniques for the study of aerobic methanotrophs. Appl. Environ. Microbiol. 74, 1305–1315 (2008)

    CAS  Article  Google Scholar 

  30. 30

    Crombie, A. & Murrell, J. C. Development of a system for genetic manipulation of the facultative methanotroph Methylocella silvestris BL2. Methods Enzymol. 495, 119–133 (2011)

    CAS  Article  Google Scholar 

  31. 31

    Guerlava, P., Izac, V. & Tholozan, J.-L. Comparison of different methods of cell lysis and protein measurements in Clostridium perfringens: application to the cell volume determination. Curr. Microbiol. 36, 131–135 (1998)

    CAS  Article  Google Scholar 

  32. 32

    Tamura, K., Dudley, J., Nei, M. & Kumar, S. MEGA4: molecular evolutionary genetics analysis (MEGA) software version 4.0. Mol. Biol. Evol. 24, 1596–1599 (2007)

    CAS  Article  Google Scholar 

  33. 33

    Sambrook, J., Fritsch, E. F. & Maniatis, T. Molecular Cloning: A Laboratory Manual 3rd edn (Cold Spring Harbor Laboratory Press, 2001)

    Google Scholar 

  34. 34

    Gilbert, B. et al. Molecular analysis of the pmo (particulate methane monooxygenase) operons from two type II methanotrophs. Appl. Environ. Microbiol. 66, 966–975 (2000)

    CAS  Article  Google Scholar 

  35. 35

    Marx, C. J. & Lidstrom, M. E. Broad-host-range cre-lox system for antibiotic marker recycling in Gram-negative bacteria. Biotechniques 33, 1062–1067 (2002)

    CAS  Article  Google Scholar 

  36. 36

    Patel, V. J. et al. A comparison of labeling and label-free mass spectrometry-based proteomics approaches. J. Proteome Res. 8, 3752–3759 (2009)

    ADS  CAS  Article  Google Scholar 

  37. 37

    Green, M. J. & Hill, H. A. O. Chemistry of dioxygen. Methods Enzymol. 105, 3–22 (1984)

    CAS  Article  Google Scholar 

  38. 38

    Anthony, C. & Zatman, L. J. The microbial oxidation of methanol. 2. The methanol-oxidizing enzyme of Pseudomonas sp. M 27. Biochem. J. 92, 614–621 (1964)

    CAS  Article  Google Scholar 

  39. 39

    Leahy, J. G., Batchelor, P. J. & Morcomb, S. M. Evolution of the soluble diiron monooxygenases. FEMS Microbiol. Rev. 27, 449–479 (2003)

    CAS  Article  Google Scholar 

  40. 40

    Coleman, N. V., Bui, N. B. & Holmes, A. J. Soluble di-iron monooxygenase gene diversity in soils, sediments and ethene enrichments. Environ. Microbiol. 8, 1228–1239 (2006)

    CAS  Article  Google Scholar 

  41. 41

    Barrios, H., Valderrama, B. & Morett, E. Compilation and analysis of σ54-dependent promoter sequences. Nucleic Acids Res. 27, 4305–4313 (1999)

    CAS  Article  Google Scholar 

Download references

Acknowledgements

This work was funded by a Natural Environmental Research Council (NERC) grant (NE/E016855/1) and a NERC studentship to A.T.C. Additional funding was provided by the University of East Anglia and the Earth and Life Systems Alliance of the Norwich Research Park. We thank A. Johnston for critical reading of the manuscript.

Author information

Affiliations

Authors

Contributions

A.T.C. designed and performed experiments, analysed data and wrote the paper. J.C.M. designed experiments and wrote the paper.

Corresponding author

Correspondence to J. Colin Murrell.

Ethics declarations

Competing interests

The authors declare no competing financial interests.

Extended data figures and tables

Extended Data Figure 1 Methylocella silvestris grown on propane (20% (v/v) propane in air) in a fermenter (period between 62 and 106 days shown).

The 2-propanol and acetone in the culture medium reached approximately 16 mM and 9 mM, respectively. When the propane supply was shut off (days 92–96), the 2-propanol and acetone concentrations decreased to nearly zero, whereas the culture density continued to increase without interruption. Following the resumption of the propane supply, accumulation of 2-propanol and acetone was observed once again. The cells and medium were removed, and fresh medium was added on day 99. These data demonstrate that 2-propanol and acetone result from propane oxidation and that cells growing on propane in the presence of these intermediates can metabolize them without an appreciable lag phase. The 2-propanol and acetone concentrations show the mean ± s.d. of triplicate measurements.

Extended Data Figure 2 Phylogeny of SDIMOs.

a, Phylogenetic relationships between the two M. silvestris SDIMOs (underlined) and other representative enzymes. The tree, constructed using the maximum likelihood method, is based on an alignment of amino acid sequences of the α-subunit of the hydroxylases. The sequences were aligned using Clustal; positions containing gaps or missing data were eliminated; and the tree was constructed with a final data set of 356 amino acids using MEGA 5 (ref. 32). Bootstrap values (based on 500 replications) greater than 95% are shown as filled circles at nodes, and those between 75% and 95% as open circles. The SDIMO subgroups39,40 are indicated on the right of the figure. GenBank accession numbers (in order from the top): AAC45289.1, ABD46892.1, ZP_06887019.1, ABD46898.1, CAD30366.1, YP_002361593.1, CAD88243.1, BAE86875.1, YP_113659.1, BAA84751.1, BAJ17645.1, AAM19727.1, BAF34294.1, ACZ56324.1, AAO48576.1, YP_919254.1, BAA07114.1, AAS19484.1, CAC10506.1, YP_700435.1, BAF34308.1, BAD03956.2, YP_002361961.1, YP_001834443.1, YP_001020147.1, NP_770317.1, YP_352924.1, AAL50373.1, P19732.1, AAT40431.1, YP_001409304.1 and CAB55825.1. b, The propane monooxygenase gene cluster (shown in red). The structural genes (hydroxylase α-subunit, reductase, hydroxylase β-subunit and coupling protein) are followed by those encoding a putative chaperone prmG (ACK50595.1) and regulatory protein prmR (ACK50594.1), with homology to mmoG and mmoR of sMMO. The putative promoter sequence is shown above in relation to the ATG start codon, together with the consensus sequence of σ54 promoters as described previously41.

Extended Data Figure 3 M. silvestris polypeptide profiles.

a, An SDS–PAGE gel loaded with soluble extract from wild-type cells grown on methane (M), propane (P), succinate (S), 2-propanol (2-P) or acetone (A). Prominent bands (identified with boxes in the right-hand photograph of the same gel) evident following growth on methane or propane, but not on succinate, were excised from the gel and analysed by mass spectrometry. The polypeptide identifications are shown in b. The data show that both sMMO and PrMO were expressed during growth on propane but that PrMO subunits were not expressed at a high level during growth on methane. PrMO subunits, but not sMMO subunits, were expressed during growth on 2-propanol. In addition, gel-free analysis of the complete soluble proteome26 did not result in the detection of PrMO polypeptides in succinate-grown or methane-grown cell extracts. b, Polypeptide identifications of the gel bands shown in a. For each band, the four most abundant polypeptides are shown, except where fewer than four were detected. In addition, all sMMO-related and PrMO-related polypeptides identified (irrespective of the number of peptides used for identification) are included. Otherwise, polypeptides identified with at least three peptides are included (except for the succinate lane; four peptides). Other identified polypeptides are not shown. The number of peptides used for the identification of each polypeptide is shown. The total number of peptides detected from all of the polypeptides identified in each band is shown for comparison. DH, dehydrogenase; MM, theoretical molecular mass; PEP, phosphoenolpyruvate.

Extended Data Figure 4 Growth of M. silvestris strains.

a, Growth of wild-type M. silvestris (solid lines, filled symbols) and the ΔPrmA strain (dashed lines, open symbols) on methane (circles) or propane (triangles) (2.5% (v/v)). The inset shows the substrate carbon conversion efficiency (CCE) (mg of cellular protein formed per mg of substrate carbon consumed) for the wild-type M. silvestris (dark grey) and the ΔPrmA strain (light grey) during growth on methane (M) or propane (P). Data are the mean ± s.d. for three (methane) or two (propane) vials. b, Growth and substrate consumption of wild-type M. silvestris on a mixture of methane (red, crosses) and propane (purple, triangles) (2.5% (v/v) each). The inoculum for the cultures was grown on methane and propane (2.5% (v/v) each). Data points show the mean ± s.d. of triplicate vials. c, d, Growth of and consumption of methane and propane by wild-type M. silvestris (c) and the ΔMmoX strain (d) when supplied with an approximately 1:10 (v/v) ratio of gases, showing that PrMO did not oxidize appreciable amounts of methane. Methane concentrations are shown as % (v/v) × 10 on the secondary (right) y axes. Data are the mean ± s.d. of triplicate vials.

Extended Data Figure 5 Methane and propane oxidation kinetics.

a, Maximum methane-induced and propane-induced specific oxygen consumption rate (nmol min−1 per mg dry weight) of whole M. silvestris cells grown on methane, propane or succinate, as determined by oxygen electrode studies, with approximately 200 μM oxidation substrate. Data are the mean ± s.d. of three measurements. b, c, Kinetics of methane-induced (b) or propane-induced (c) oxygen consumption in whole cells grown on that substrate. The Hanes–Woolf plots show substrate concentration (S) divided by oxygen consumption rate (v) as a function of S, and the trend line cuts the x axis at −Km.

Extended Data Figure 6 Wild-type M. silvestris activity assays.

a, Stoichiometry of whole-cell oxygen consumption in response to addition of 100–250 nmol methanol, 1-propanol or propanal. Cells from all growth conditions consumed approximately equimolar amounts of oxygen when methanol was added; however, when 1-propanol or propanal replaced methanol, the amount of oxygen consumed by propane-grown cells increased to threefold (1-propanol) or fivefold (propanal) that of methane-grown or succinate-grown cells. These data suggest that the oxidation of terminal intermediates proceeds further in propane-grown cells, although enzymes that can oxidize 1-propanol and propanal may be present in all cell types, as shown in b and c. Data are the mean ± s.d. of the number of measurements shown. b, Oxygen uptake rates of M. silvestris cells grown on methane, propane or succinate in response to the addition of the substrates shown. High rates of oxygen consumption were recorded in response to methanol, 1-propanol and propanal in cells grown on either methane, propane or succinate. However, the oxygen consumption rates in response to 2-propanol, acetone and acetol were at least twofold, fourfold or sevenfold higher, respectively, in propane-grown cells than in methane-grown or succinate-grown cells, demonstrating that the ability to oxidize intermediates of the sub-terminal oxidation pathway was induced during growth on propane. Substrates were used at a final concentration of 5 mM. Data are the mean ± s.d. of three measurements (except for methane/methanol, n = 7). c, Quinoprotein alcohol dehydrogenase activity (mean ± s.d. of three measurements) assayed as DCPIP reduction (nmol min−1 per mg protein) in soluble extract from methane-grown or propane-grown cells. The activity in cell extracts from both growth conditions was high when 1-propanol was the substrate (probably as a result of the constitutive expression of methanol dehydrogenase); however, with 2-propanol as the substrate, the activity was sixfold higher in the extract from cells grown on propane than cells grown on methane. By contrast, the rates of NAD+-linked or NADP+-linked 1-propanol or 2-propanol dehydrogenase activity were less than 10 nmol min−1 per mg protein in cell extracts from each growth condition (data not shown).

Extended Data Figure 7 1-Propanol-metabolizing ability was induced in propane-grown cells.

a, When vials were inoculated with methane-grown cells, growth was possible on 2-propanol but not on 1-propanol at any concentration tested. b, In addition, 1-propanol completely inhibited growth on 2-propanol. c, When using succinate-grown inoculum, 1-propanol also greatly inhibited growth on succinate. d, However, when using an inoculum grown on propane, limited growth occurred on 1-propanol, and 1-propanol did not inhibit growth on succinate, suggesting that 1-propanol-metabolizing potential was induced in cells grown on propane and could be maintained during growth on, or in the presence of, 1-propanol. The concentrations used were 0.05% (v/v) 1-propanol, 0.05% (v/v) 2-propanol, 3 mM succinate (except d, 5 mM), or as indicated. Data are the mean ± s.d. of duplicate (a, c d) or triplicate (b) vials.

Extended Data Figure 8 Detection of oxidation products in wild-type M. silvestris cultures.

a, During growth on 4% (v/v) propane, 2-propanol was detected in the culture medium during the mid-exponential and late exponential phase, before declining from a maximum of approximately 0.5 mM at 210 h to below the limit of detection at stationary phase (300 h). The culture density (OD540) is shown in black; propane concentration, in purple; and 2-propanol concentration, in red; with solid lines and filled symbols. Control vials, which contained cells killed by autoclaving, are shown as dotted lines and open symbols. The 2-propanol concentration is shown as ×4 on the secondary (right) y axis. Data are the mean ± s.d. of triplicate vials. b, For the cultures shown in a, the amount of 2-propanol present at each time point (including the amount previously removed during sampling) is expressed as a percentage of the propane consumed. At 94 h, 25% of the propane consumed could be accounted for as 2-propanol in the growth medium. This value therefore represents the minimum percentage of propane that is oxidized to 2-propanol, because no allowance has been made for the consumption of 2-propanol by the cultures. Data are calculated from the mean ± s.d. of triplicate vials.

Extended Data Figure 9 Oxidation products of the ΔPrmA strain and growth rates of the ΔPrmA, wild-type and inhibited wild-type strains.

a, Culture density (black lines, triangles) and 2-propanol (red lines, circles) and acetone (purple lines, diamonds) concentrations during growth of the ΔPrmA strain on 4% (v/v) propane. This strain accumulated 1.3 mM 2-propanol, but neither 2-propanol nor acetone were detected during growth of the ΔMmoX strain on 20% (v/v) propane (data not shown), suggesting that 2-propanol is one (or the major) product of propane oxidation by sMMO. b, Growth rate of wild-type M. silvestris (dark grey) and the ΔPrmA strain (light grey) on the substrates shown. Disruption of PrMO decreased growth on 2-propanol compared with the wild-type, although growth on acetone, acetate, methanol or ethanol was not significantly affected, implicating PrMO in 2-propanol metabolism. Data are the mean ± s.d. of three vials (except 2-propanol, n = 6 (ΔPrmA strain) or n = 5 (wild-type)). Significance was determined by Student’s t-test (**, P < 0.01). c, Growth of wild-type M. silvestris on acetone (squares), 2-propanol (circles) or propane (10% (v/v)) (triangles) either without an inhibitor (filled symbols, solid lines) or with acetylene (2% (v/v)) as an inhibitor (open symbols, dashed lines). Data are the mean ± s.d. of triplicate vials. For 2-propanol, the specific growth rates in the early exponential phase were 0.020 ± 0.0002 h−1 and 0.018 ± 0.0001 h−1 (mean ± s.d.) for uninhibited or inhibited cultures, respectively.

Extended Data Table 1 Substrate utilization by Methylocella silvestris

PowerPoint slides

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Crombie, A., Murrell, J. Trace-gas metabolic versatility of the facultative methanotroph Methylocella silvestris . Nature 510, 148–151 (2014). https://doi.org/10.1038/nature13192

Download citation

Further reading

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

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