Hydrogen peroxide sensor HPCA1 is an LRR receptor kinase in Arabidopsis


Hydrogen peroxide (H2O2) is a major reactive oxygen species in unicellular and multicellular organisms, and is produced extracellularly in response to external stresses and internal cues1,2,3,4. H2O2 enters cells through aquaporin membrane proteins and covalently modifies cytoplasmic proteins to regulate signalling and cellular processes. However, whether sensors for H2O2 also exist on the cell surface remains unknown. In plant cells, H2O2 triggers an influx of Ca2+ ions, which is thought to be involved in H2O2 sensing and signalling. Here, by using forward genetic screens based on Ca2+ imaging, we isolated hydrogen-peroxide-induced Ca2+ increases (hpca) mutants in Arabidopsis, and identified HPCA1 as a leucine-rich-repeat receptor kinase belonging to a previously uncharacterized subfamily that features two extra pairs of cysteine residues in the extracellular domain. HPCA1 is localized to the plasma membrane and is activated by H2O2 via covalent modification of extracellular cysteine residues, which leads to autophosphorylation of HPCA1. HPCA1 mediates H2O2-induced activation of Ca2+ channels in guard cells and is required for stomatal closure. Our findings help to identify how the perception of extracellular H2O2 is integrated with responses to various external stresses and internal cues in plants, and have implications for the design of crops with enhanced fitness.

Access options

Rent or Buy article

Get time limited or full article access on ReadCube.


All prices are NET prices.

Fig. 1: Isolation of Arabidopsis mutants defective in H2O2-induced Ca2+ increases.
Fig. 2: Impaired guard-cell eH2O2 signalling in hpca1.
Fig. 3: HPCA1 encodes a LRR-RK that is activated by eH2O2.
Fig. 4: The Cys residues in the hydrogen peroxide domain modified covalently by eH2O2.

Data availability

All datasets generated and/or analysed during the current study are available in the Article, the Source Data files that accompany Figs. 14 and Extended Data Figs. 13, 8, 9, or the Supplementary Information. Additional data, such a raw image files, that support the findings of this study, are available from the corresponding author upon request.


  1. 1.

    Mittler, R. ROS are good. Trends Plant Sci. 22, 11–19 (2017).

    CAS  PubMed  Google Scholar 

  2. 2.

    Sies, H., Berndt, C. & Jones, D. P. Oxidative stress. Annu. Rev. Biochem. 86, 715–748 (2017).

    CAS  PubMed  Google Scholar 

  3. 3.

    Imlay, J. A. Transcription factors that defend bacteria against reactive oxygen species. Annu. Rev. Microbiol. 69, 93–108 (2015).

    CAS  PubMed  PubMed Central  Google Scholar 

  4. 4.

    Waszczak, C., Carmody, M. & Kangasjärvi, J. Reactive oxygen species in plant signaling. Annu. Rev. Plant Biol. 69, 209–236 (2018).

    CAS  PubMed  Google Scholar 

  5. 5.

    Kadota, Y., Shirasu, K. & Zipfel, C. Regulation of the NADPH oxidase RBOHD during plant immunity. Plant Cell Physiol. 56, 1472–1480 (2015).

    CAS  PubMed  Google Scholar 

  6. 6.

    Zipfel, C. & Oldroyd, G. E. D. Plant signalling in symbiosis and immunity. Nature 543, 328–336 (2017).

    ADS  CAS  PubMed  Google Scholar 

  7. 7.

    Oldroyd, G. E. D. Speak, friend, and enter: signalling systems that promote beneficial symbiotic associations in plants. Nat. Rev. Microbiol. 11, 252–263 (2013).

    CAS  PubMed  Google Scholar 

  8. 8.

    Kelliher, T. & Walbot, V. Hypoxia triggers meiotic fate acquisition in maize. Science 337, 345–348 (2012).

    ADS  CAS  PubMed  PubMed Central  Google Scholar 

  9. 9.

    Boisson-Dernier, A. et al. ANXUR receptor-like kinases coordinate cell wall integrity with growth at the pollen tube tip via NADPH oxidases. PLoS Biol. 11, e1001719 (2013).

    PubMed  PubMed Central  Google Scholar 

  10. 10.

    Foreman, J. et al. Reactive oxygen species produced by NADPH oxidase regulate plant cell growth. Nature 422, 442–446 (2003).

    ADS  CAS  Google Scholar 

  11. 11.

    Takeda, S. et al. Local positive feedback regulation determines cell shape in root hair cells. Science 319, 1241–1244 (2008).

    ADS  CAS  PubMed  Google Scholar 

  12. 12.

    Tsukagoshi, H., Busch, W. & Benfey, P. N. Transcriptional regulation of ROS controls transition from proliferation to differentiation in the root. Cell 143, 606–616 (2010).

    CAS  PubMed  Google Scholar 

  13. 13.

    Lee, Y. et al. A lignin molecular brace controls precision processing of cell walls critical for surface integrity in Arabidopsis. Cell 173, 1468–1480.e9 (2018).

    CAS  PubMed  Google Scholar 

  14. 14.

    Pei, Z.-M. et al. Calcium channels activated by hydrogen peroxide mediate abscisic acid signalling in guard cells. Nature 406, 731–734 (2000).

    ADS  CAS  PubMed  Google Scholar 

  15. 15.

    Yi, M. C. & Khosla, C. Thiol-disulfide exchange reactions in the mammalian extracellular environment. Annu. Rev. Chem. Biomol. Eng. 7, 197–222 (2016).

    CAS  PubMed  PubMed Central  Google Scholar 

  16. 16.

    Price, A. H. A possible role for calcium in oxidative plant stress. Free Radic. Res. Commun. 10, 345–349 (1990).

    CAS  PubMed  Google Scholar 

  17. 17.

    Price, A. H. et al. Oxidative signals in tobacco increase cytosolic calcium. Plant Cell 6, 1301–1310 (1994).

    ADS  CAS  PubMed  PubMed Central  Google Scholar 

  18. 18.

    McAinsh, M. R., Clayton, H., Mansfield, T. A. & Hetherington, A. M. Changes in stomatal behavior and guard cell cytosolic free calcium in response to oxidative stress. Plant Physiol. 111, 1031–1042 (1996).

    CAS  PubMed  PubMed Central  Google Scholar 

  19. 19.

    Demidchik, V. & Shabala, S. Mechanisms of cytosolic calcium elevation in plants: the role of ion channels, calcium extrusion systems and NADPH oxidase-mediated ‘ROS-Ca2+ Hub’. Funct. Plant Biol. 45, 9–27 (2018).

    CAS  Google Scholar 

  20. 20.

    Kudla, J. et al. Advances and current challenges in calcium signaling. New Phytol. 218, 414–431 (2018).

    PubMed  Google Scholar 

  21. 21.

    Choi, W. G., Hilleary, R., Swanson, S. J., Kim, S. H. & Gilroy, S. Rapid, long-distance electrical and calcium signaling in plants. Annu. Rev. Plant Biol. 67, 287–307 (2016).

    CAS  PubMed  Google Scholar 

  22. 22.

    Hedrich, R. Ion channels in plants. Physiol. Rev. 92, 1777–1811 (2012).

    ADS  CAS  PubMed  Google Scholar 

  23. 23.

    Edel, K. H., Marchadier, E., Brownlee, C., Kudla, J. & Hetherington, A. M. The evolution of calcium-based signalling in plants. Curr. Biol. 27, R667–R679 (2017).

    CAS  PubMed  Google Scholar 

  24. 24.

    Yuan, F. et al. OSCA1 mediates osmotic-stress-evoked Ca2+ increases vital for osmosensing in Arabidopsis. Nature 514, 367–371 (2014).

    ADS  CAS  PubMed  Google Scholar 

  25. 25.

    Jiang, Z. et al. Plant cell-surface GIPC sphingolipids sense salt to trigger Ca2+ influx. Nature 572, 341–346 (2019).

    ADS  CAS  PubMed  Google Scholar 

  26. 26.

    Boller, T. & Felix, G. A renaissance of elicitors: perception of microbe-associated molecular patterns and danger signals by pattern-recognition receptors. Annu. Rev. Plant Biol. 60, 379–406 (2009).

    CAS  PubMed  Google Scholar 

  27. 27.

    Ranf, S. Sensing of molecular patterns through cell surface immune receptors. Curr. Opin. Plant Biol. 38, 68–77 (2017).

    CAS  PubMed  Google Scholar 

  28. 28.

    Hua, D. et al. A plasma membrane receptor kinase, GHR1, mediates abscisic acid- and hydrogen peroxide-regulated stomatal movement in Arabidopsis. Plant Cell 24, 2546–2561 (2012).

    CAS  PubMed  PubMed Central  Google Scholar 

  29. 29.

    Sierla, M. et al. The receptor-like pseudokinase GHR1 is required for stomatal closure. Plant Cell 30, 2813–2837 (2018).

    CAS  PubMed  PubMed Central  Google Scholar 

  30. 30.

    Macpherson, L. J. et al. Noxious compounds activate TRPA1 ion channels through covalent modification of cysteines. Nature 445, 541–545 (2007).

    ADS  CAS  PubMed  Google Scholar 

  31. 31.

    Bourdais, G. et al. Large-scale phenomics identifies primary and fine-tuning roles for CRKs in responses related to oxidative stress. PLoS Genet. 11, e1005373 (2015).

    PubMed  PubMed Central  Google Scholar 

  32. 32.

    Ortiz-Ramírez, C. et al. GLUTAMATE RECEPTOR-LIKE channels are essential for chemotaxis and reproduction in mosses. Nature 549, 91–95 (2017).

    ADS  PubMed  Google Scholar 

  33. 33.

    Choi, J. et al. Identification of a plant receptor for extracellular ATP. Science 343, 290–294 (2014).

    ADS  CAS  PubMed  Google Scholar 

  34. 34.

    Kutschera, A. et al. Bacterial medium-chain 3-hydroxy fatty acid metabolites trigger immunity in Arabidopsis plants. Science 364, 178–181 (2019).

    ADS  CAS  PubMed  Google Scholar 

  35. 35.

    Knight, M. R., Campbell, A. K., Smith, S. M. & Trewavas, A. J. Transgenic plant aequorin reports the effects of touch and cold-shock and elicitors on cytoplasmic calcium. Nature 352, 524–526 (1991).

    ADS  CAS  PubMed  Google Scholar 

  36. 36.

    Knight, H., Trewavas, A. J. & Knight, M. R. Calcium signalling in Arabidopsis thaliana responding to drought and salinity. Plant J. 12, 1067–1078 (1997).

    CAS  PubMed  Google Scholar 

  37. 37.

    Monshausen, G. B., Messerli, M. A. & Gilroy, S. Imaging of the Yellow Cameleon 3.6 indicator reveals that elevations in cytosolic Ca2+ follow oscillating increases in growth in root hairs of Arabidopsis. Plant Physiol. 147, 1690–1698 (2008).

    CAS  PubMed  PubMed Central  Google Scholar 

  38. 38.

    Ma, Y., Zhao, Y., Walker, R. K. & Berkowitz, G. A. Molecular steps in the immune signaling pathway evoked by plant elicitor peptides: Ca2+-dependent protein kinases, nitric oxide, and reactive oxygen species are downstream from the early Ca2+ signal. Plant Physiol. 163, 1459–1471 (2013).

    CAS  PubMed  PubMed Central  Google Scholar 

  39. 39.

    Caterina, M. J. et al. The capsaicin receptor: a heat-activated ion channel in the pain pathway. Nature 389, 816–824 (1997).

    ADS  CAS  PubMed  Google Scholar 

  40. 40.

    Schroeder, B. C., Cheng, T., Jan, Y. N. & Jan, L. Y. Expression cloning of TMEM16A as a calcium-activated chloride channel subunit. Cell 134, 1019–1029 (2008).

    CAS  PubMed  PubMed Central  Google Scholar 

  41. 41.

    Chalfie, M. Neurosensory mechanotransduction. Nat. Rev. Mol. Cell Biol. 10, 44–52 (2009).

    CAS  PubMed  Google Scholar 

  42. 42.

    Ranf, S. et al. A lectin S-domain receptor kinase mediates lipopolysaccharide sensing in Arabidopsis thaliana. Nat. Immunol. 16, 426–433 (2015).

    CAS  PubMed  Google Scholar 

  43. 43.

    Knight, H., Trewavas, A. J. & Knight, M. R. Cold calcium signaling in Arabidopsis involves two cellular pools and a change in calcium signature after acclimation. Plant Cell 8, 489–503 (1996).

    CAS  PubMed  PubMed Central  Google Scholar 

  44. 44.

    Murata, Y., Pei, Z.-M., Mori, I. C. & Schroeder, J. Abscisic acid activation of plasma membrane Ca2+ channels in guard cells requires cytosolic NAD(P)H and is differentially disrupted upstream and downstream of reactive oxygen species production in abi1-1 and abi2-1 protein phosphatase 2C mutants. Plant Cell 13, 2513–2523 (2001).

    CAS  PubMed  PubMed Central  Google Scholar 

  45. 45.

    Kwak, J. M. et al. NADPH oxidase AtrbohD and AtrbohF genes function in ROS-dependent ABA signaling in Arabidopsis. EMBO J. 22, 2623–2633 (2003).

    CAS  PubMed  PubMed Central  Google Scholar 

  46. 46.

    Pei, Z.-M., Kuchitsu, K., Ward, J. M., Schwarz, M. & Schroeder, J. I. Differential abscisic acid regulation of guard cell slow anion channels in Arabidopsis wild-type and abi1 and abi2 mutants. Plant Cell 9, 409–423 (1997).

    CAS  PubMed  PubMed Central  Google Scholar 

  47. 47.

    Kim, T. H., Böhmer, M., Hu, H., Nishimura, N. & Schroeder, J. I. Guard cell signal transduction network: advances in understanding abscisic acid, CO2, and Ca2+ signaling. Annu. Rev. Plant Biol. 61, 561–591 (2010).

    CAS  PubMed  PubMed Central  Google Scholar 

  48. 48.

    Yoshida, T. et al. Nitric oxide activates TRP channels by cysteine S-nitrosylation. Nat. Chem. Biol. 2, 596–607 (2006).

    CAS  PubMed  Google Scholar 

  49. 49.

    Abe, A. et al. Genome sequencing reveals agronomically important loci in rice using MutMap. Nat. Biotechnol. 30, 174–178 (2012).

    CAS  PubMed  Google Scholar 

  50. 50.

    Katoh, K. & Standley, D. M. MAFFT multiple sequence alignment software version 7: improvements in performance and usability. Mol. Biol. Evol. 30, 772–780 (2013).

    CAS  PubMed  PubMed Central  Google Scholar 

  51. 51.

    Tamura, K., Stecher, G., Peterson, D., Filipski, A. & Kumar, S. MEGA6: molecular evolutionary genetics analysis version 6.0. Mol. Biol. Evol. 30, 2725–2729 (2013).

    CAS  PubMed  PubMed Central  Google Scholar 

  52. 52.

    Waterhouse, A. M., Procter, J. B., Martin, D. M., Clamp, M. & Barton, G. J. Jalview Version 2—a multiple sequence alignment editor and analysis workbench. Bioinformatics 25, 1189–1191 (2009).

    CAS  PubMed  PubMed Central  Google Scholar 

  53. 53.

    Smakowska-Luzan, E. et al. An extracellular network of Arabidopsis leucine-rich repeat receptor kinases. Nature 553, 342–346 (2018).

    ADS  CAS  PubMed  PubMed Central  Google Scholar 

  54. 54.

    Goodstein, D. M. et al. Phytozome: a comparative platform for green plant genomics. Nucleic Acids Res. 40, D1178–D1186 (2012).

    CAS  PubMed  Google Scholar 

  55. 55.

    Karimi, M., Inzé, D. & Depicker, A. GATEWAY vectors for Agrobacterium-mediated plant transformation. Trends Plant Sci. 7, 193–195 (2002).

    CAS  Google Scholar 

  56. 56.

    Clough, S. J. & Bent, A. F. Floral dip: a simplified method for Agrobacterium-mediated transformation of Arabidopsis thaliana. Plant J. 16, 735–743 (1998).

    CAS  PubMed  Google Scholar 

  57. 57.

    Jefferson, R. A., Kavanagh, T. A. & Bevan, M. W. GUS fusions: β-glucuronidase as a sensitive and versatile gene fusion marker in higher plants. EMBO J. 6, 3901–3907 (1987).

    CAS  PubMed  PubMed Central  Google Scholar 

  58. 58.

    Zentella, R. et al. The Arabidopsis O-fucosyltransferase SPINDLY activates nuclear growth repressor DELLA. Nat. Chem. Biol. 13, 479–485 (2017).

    CAS  PubMed  PubMed Central  Google Scholar 

  59. 59.

    Li, J. et al. BAK1, an Arabidopsis LRR receptor-like protein kinase, interacts with BRI1 and modulates brassinosteroid signaling. Cell 110, 213–222 (2002).

    CAS  PubMed  Google Scholar 

  60. 60.

    Asai, T. et al. MAP kinase signalling cascade in Arabidopsis innate immunity. Nature 415, 977–983 (2002).

    ADS  CAS  PubMed  PubMed Central  Google Scholar 

  61. 61.

    Kovtun, Y., Chiu, W. L., Tena, G. & Sheen, J. Functional analysis of oxidative stress-activated mitogen-activated protein kinase cascade in plants. Proc. Natl Acad. Sci. USA 97, 2940–2945 (2000).

    ADS  CAS  PubMed  Google Scholar 

  62. 62.

    Kadota, Y., Macho, A. P. & Zipfel, C. Immunoprecipitation of Plasma Membrane Receptor-Like Kinases for identification of phosphorylation sites and associated proteins. Methods Mol. Biol. 1363, 133–144 (2016).

    CAS  PubMed  Google Scholar 

  63. 63.

    Ankney, J. A., Muneer, A. & Chen, X. Relative and absolute quantitation in mass spectrometry-based proteomics. Annu. Rev. Anal. Chem. (Palo Alto, Calif.) 11, 49–77 (2018).

    ADS  CAS  Google Scholar 

  64. 64.

    Wang, L. et al. Novel RNA-Affinity proteogenomics dissects tumor heterogeneity for revealing personalized markers in precision prognosis of cancer. Cell Chem. Biol. 25, 619–633.e5 (2018).

    CAS  PubMed  Google Scholar 

  65. 65.

    Brewer, T. F., Garcia, F. J., Onak, C. S., Carroll, K. S. & Chang, C. J. Chemical approaches to discovery and study of sources and targets of hydrogen peroxide redox signaling through NADPH oxidase proteins. Annu. Rev. Biochem. 84, 765–790 (2015).

    CAS  PubMed  PubMed Central  Google Scholar 

  66. 66.

    Paulsen, C. E. & Carroll, K. S. Cysteine-mediated redox signaling: chemistry, biology, and tools for discovery. Chem. Rev. 113, 4633–4679 (2013).

    CAS  PubMed  PubMed Central  Google Scholar 

  67. 67.

    Bogeski, I. et al. Differential redox regulation of ORAI ion channels: a mechanism to tune cellular calcium signaling. Sci. Signal. 3, ra24 (2010).

    PubMed  Google Scholar 

  68. 68.

    Ogawa, N., Kurokawa, T. & Mori, Y. Sensing of redox status by TRP channels. Cell Calcium 60, 115–122 (2016).

    CAS  PubMed  Google Scholar 

  69. 69.

    McConnell, E. W. et al. Proteome-wide analysis of cysteine reactivity during effector-triggered immunity. Plant Physiol. 179, 1248–1264 (2019).

    CAS  PubMed  Google Scholar 

  70. 70.

    Ogawa, N. et al. Functional and structural divergence in human TRPV1 channel subunits by oxidative cysteine modification. J. Biol. Chem. 291, 4197–4210 (2016).

    CAS  PubMed  Google Scholar 

  71. 71.

    Guo, Z., Kozlov, S., Lavin, M. F., Person, M. D. & Paull, T. T. ATM activation by oxidative stress. Science 330, 517–521 (2010).

    ADS  CAS  PubMed  Google Scholar 

  72. 72.

    Weerapana, E. et al. Quantitative reactivity profiling predicts functional cysteines in proteomes. Nature 468, 790–795 (2010).

    ADS  CAS  PubMed  PubMed Central  Google Scholar 

  73. 73.

    Backus, K. M. et al. Proteome-wide covalent ligand discovery in native biological systems. Nature 534, 570–574 (2016).

    ADS  CAS  PubMed  PubMed Central  Google Scholar 

  74. 74.

    Hyslop, P. A. et al. Intracellular calcium homeostasis during hydrogen peroxide injury to cultured P388D1 cells. J. Cell. Physiol. 129, 356–366 (1986).

    Google Scholar 

  75. 75.

    Grynkiewicz, G., Poenie, M. & Tsien, R. Y. A new generation of Ca2+ indicators with greatly improved fluorescence properties. J. Biol. Chem. 260, 3440–3450 (1985).

    CAS  PubMed  Google Scholar 

  76. 76.

    Ranf, S., Eschen-Lippold, L., Pecher, P., Lee, J. & Scheel, D. Interplay between calcium signalling and early signalling elements during defence responses to microbe- or damage-associated molecular patterns. Plant J. 68, 100–113 (2011).

    CAS  PubMed  Google Scholar 

  77. 77.

    Macho, A. P. et al. A bacterial tyrosine phosphatase inhibits plant pattern recognition receptor activation. Science 343, 1509–1512 (2014).

    ADS  CAS  PubMed  Google Scholar 

  78. 78.

    Wang, X. et al. Identification and functional analysis of in vivo phosphorylation sites of the Arabidopsis BRASSINOSTEROID-INSENSITIVE1 receptor kinase. Plant Cell 17, 1685–1703 (2005).

    CAS  PubMed  PubMed Central  Google Scholar 

  79. 79.

    Wang, X. et al. Sequential transphosphorylation of the BRI1/BAK1 receptor kinase complex impacts early events in brassinosteroid signaling. Dev. Cell 15, 220–235 (2008).

    CAS  PubMed  Google Scholar 

  80. 80.

    Yun, H. S. et al. Analysis of phosphorylation of the BRI1/BAK1 complex in arabidopsis reveals amino acid residues critical for receptor formation and activation of BR signaling. Mol. Cells 27, 183–190 (2009).

    CAS  PubMed  Google Scholar 

  81. 81.

    Hohmann, U., Lau, K. & Hothorn, M. The structural basis of ligand perception and signal activation by receptor kinases. Annu. Rev. Plant Biol. 68, 109–137 (2017).

    CAS  PubMed  Google Scholar 

  82. 82.

    Perraki, A. et al. Phosphocode-dependent functional dichotomy of a common co-receptor in plant signalling. Nature 561, 248–252 (2018).

    ADS  CAS  PubMed  PubMed Central  Google Scholar 

Download references


We thank M. R. Knight for aequorin-expressing Arabidopsis seeds; S. Gilroy for YC3.6-expressing Arabidopsis lines; G. A. Berkowitz for bak1 and fls2 aequorin expression Arabidopsis mutants; R. Zentella, J. Hu and T. Wang for advice on protein analysis; and T. Sun, D. R. McClay and P. N. Benfey for discussions and critical reading of the manuscript. F.W., Z.J. and Z.H. were supported in part by Shenzhen Peacock Innovative Research Team Program; F.W., Z.J., Y.X., F.Y., J.N., Y.Z. and X.W. were supported by Pandeng Project of Hangzhou Normal University and Zhejiang NSF (Z16C020004); and F.W., W.W., S.Z. and Z.J. were supported by fellowships from China Scholarship Council. This work was supported by grants from the NSF (IOS-1457257, IOS-0848263), USDA (CSREES-2006-35100-17304) and DOE (DE-SC0014077) to Z.-M.P., NIH (1U19AI109965) to X. Chen, and Chinese NSF (31571461) to F.Y.

Author information




Z.-M.P. designed and supervised the overall research with input from F.W., Z.H. and J.N.S. R.Y., F.Y. and B.B. generated EMS-mutagenized aequorin-expression Arabidopsis seed populations and performed initial genetic screens. F.W., W.W., J.N., Y.Z., Z.J., S.Z., Y.C., Y.X., D.W. and S.C. conducted genetic screens. F.W., Y.C., Y.X., D.W., S.C., X.W. and Z.J. performed whole-genome resequencing-based mapping. F.W., F.H., Y.C., C.P. and Z.J. carried out YC3.6 FRET imaging and the stomatal assay. Z.J., F.W. and Y.C. conducted patch-clamp experiments. F.W., F.H., L.X., L.W., Y.C. and X. Chen analysed protein modifications. L.X., L.W. and X. Chen were the major contributors for the mass spectrometry data for HPCA1. F.W., F.H., M.Z. and W.Z. analysed HPCA1 kinase activities. M.S., J.P., D.M.J., S.Z., X. Cao, C.P., Z.Z., Y.L., T.L., G.B.S., H.Z., M.Y. and Z.H. participated in Ca2+-imaging analyses and phenotype characterization. Z.-M.P., F.W. and J.N.S. wrote the paper. All authors discussed the results and commented on the manuscript.

Corresponding author

Correspondence to Zhen-Ming Pei.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

Peer review information Nature thanks Christine Foyer 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 Fig. 1 Similar total amounts of aequorin in the wild-type and hpca1 seedlings.

a, The Arabidopsis wild-type and hpca1 seedlings expressing aequorin (used in Fig. 1a) were treated with a solution containing 0.9 M CaCl2 and 10% (v/v) ethanol (aequorin discharge), and the total amount of remaining aequorin luminescence was imaged. Images are representative of more than 100 separate experiments. The relative [Ca2+]i is scaled by a pseudo-colour bar. b, Quantification of total amount of aequorin luminescence from as in a. Data are mean ± s.d.; n = 12 pools (20 seedlings per pool). P = 0.806, two-sided t-test. About 10 years before Price et al. made the initial observation of eH2O2-induced [Ca2+]i increases in plant cells in 199417, Hyslop et al. investigated the effects of eH2O2 on [Ca2+]i in mouse macrophage cells using the intracellular fluorescent Ca2+ indicator Quin2, and detected [Ca2+]i increases in response to eH2O2 treatment74. Note that, Quin2 was the predecessor of Fura-2, the most important Ca2+ indicator75. Source data

Extended Data Fig. 2 Specific Ca2+ phenotypes of H2O2-sensing hpca1, osmosensing osca1, and salt-sensing moca1 mutants, and the wilt phenotype of hpca1.

a−d, The hpca1 mutant is selective for H2O2 over osmotic stress. a, c, Representative aequorin bioluminescence images from wild-type, hpca1 and osca1 seedlings, which were grown side-by-side and treated with 4 mM H2O2 (a) or 600 mM sorbitol (c). b, d, Quantification of [Ca2+]i in leaves from experiments in a and c with H2O2 (b) and sorbitol (d) treatment. Data are mean ± s.d. and representative of more than eight experiments; n = 16 seedlings. e−h, The hpca1 mutant is selective for H2O2 over salt stress. e, g, Representative aequorin bioluminescence images from wild-type, hpca1 and moca1 seedlings, which were grown side-by-side and treated with 4 mM H2O2 (e) and 200 mM NaCl (g). f, h, Quantification of [Ca2+]i in leaves from experiments in e and g is shown for H2O2 (f) and NaCl (h) treatment. Data are mean ± s.d. and representative of more than eight experiments; n = 16 seedlings. With respect to eH2O2-induced [Ca2+]i increases, osca1, moca1 and wild-type seedlings were similar; whereas with respect to sorbitol- or NaCl-induced [Ca2+]i increases, hpca1 and wild-type seedlings were similar, showing that hpca1 differs from osca1 and moca1, and has eH2O2-specific Ca2+ phenotypes. i, The hpca1 mutant is hypersensitive to desiccation treatment. Rosette leaves from wild-type and hpca1 seedlings were detached and water loss owing to transpiration was analysed at the indicated time points after leaf detachment. Water loss was calculated as a percentage of the initial fresh weight. Data are mean ± s.e.m., n = 4 pools (20 to 30 leaves per pool). P < 0.05, two-way ANOVA. Source data

Extended Data Fig. 3 Mapping-by-sequencing identification of mutations in HPCA1 in three allelic hpca1 mutants and their stomata responses to H2O2, ABA and external Ca2+.

a−c, SNP concurrences across the five chromosomes are analysed and shown from a pooled mapping population of hpca1-1 (a), hpca1-2 (b) and hpca1-3 (c). All three mutants show an enrichment of SNP concurrences in the same area in the long arm of chromosome 5. Genes with SNPs in these mutants were compared, and three allelic mutations in the HPCA1 (At5g49760) gene were found as shown in Fig. 3a and Extended Data Fig. 6a. d, Images of wild-type and three allelic hpca1 mutants were taken after 30 or 60 days of growth under growth chamber conditions (16 h/8 h light/dark photoperiod). e−g, Stomatal apertures were measured in the absence or presence of 1 mM H2O2 (e), 5 μM ABA (f) or 2.5 mM CaCl2 (g). Data are mean ± s.e.m. from three independent experiments, n = 60 stomata. The stomatal apertures without the treatment in e were 3.38 μm in WT, 3.12 μm in hpca1-1, 3.41 μm in hpca1-2, and 3.26 μm in hpca1-3. The stomatal apertures without the treatment in f were 3.21 μm in WT, 3.37 μm in hpca1-1, 3.27 μm in hpca1-2, and 3.26 μm in hpca1-3. The stomatal apertures without the treatment in g were 4.36 μm in WT, 4.54 μm in hpca1-1, 4.40 μm in hpca1-2, and 4.55 μm in hpca1-3. Source data

Extended Data Fig. 4 The phylogeny of HPCA1 subfamily in the LRR-RK family in Arabidopsis.

a, Phylogeny of full-length sequences of the Arabidopsis LRR-RK proteins. Multiple sequence alignments were conducted using MAFFT (v.7.05), and a neighbour-joining tree was created using MEGA (v.6) (Methods). Branch lengths are scaled with the number of amino acids substitution per site. The LRR-RK subfamilies (from I to XV) were indicated along the outer edge of the tree. HPCA1, GHR1 (GUARD CELL HYDROGEN PEROXIDE-RESISTANT1), FLS2 (FLAGELLIN-SENSITIVE 2), EFR (EF-TU RECEPTOR), BRI1 and BAK1 (BRASSINOSTEROID INSENSITIVE 1-ASSOCIATED RECEPTOR KINASE 1) are illustrated. b, Maximum-likelihood phylogeny of the HPCA1 and VIII-1 subfamily that is composed of eight LRR-RK proteins. Statistical supports corresponding to maximum-likelihood and neighbour-joining bootstrap values are labelled beside the nodes, respectively. Branch lengths are scaled with the number of amino acid residues substitution per site. The VIII-1 subfamily is named as the HPCA subfamily, and the eight VIII-1 members are named sequentially as HPCA1 and seven HPCA-LIKE (HPCAL) proteins.

Extended Data Fig. 5 Conserved hydrogen peroxide domains between the last LRR and the transmembrane domain in the HPCA1 LRR-RK subfamily across species.

a, Schematic representation of conserved Cys residues in the region between the last LRR and the kinase domain. Analysis showed that the LRR domain is also flanked by domains with characteristic double-cysteine residue motifs (Cys pair) in these HPCAs. HPCA1 (or VIII-1 subfamily) proteins have six Cys residues that are conserved in this group. VIII-2 proteins have two conserved Cys residues. The others groups (except for V, IX, XIV and XV) have two conserved Cys residues that have been illustrated in a previous evolutionary study of LRR-RKs53 (Methods). These conserved Cys residues in VIII-1, VIII-2 and the other groups appear as pale-blue lines, and their positions were referenced as HPCA1 (or AT5G49760) in VIII-1, AT1G29750 in VIII-2, and AT1G12460 in VIIa, and are shown on a relative scale. The HPCA subfamily contains a unique hydrogen peroxide (HP) domain. The last LRR, transmembrane domain (TM), and protein kinase domains are labelled in VIII-1, VIII-2 and the other LRR-RK groups. The transmembrane and protein kinase domains are conserved among all LRR-RK proteins. VIII-2 also contains a malectin domain. b, Multiple sequence alignment of the region containing the VIII-1-specific Cys residues (hydrogen peroxide domain) in LRR-RK proteins from seven land plant species. The classification of LRR-RKs was retrieved from the previous study53 (Methods). Each protein is depicted as a protein ID, followed by the start and end position of the region. The seven species are the early land plant Physcomitrella patens (Potri), the early vascular plant Selaginella moellendorffii (Smo), two monocot genomes including Sorghum bicolor (Sb) as well as rice Oryza sativa Japonica (Os), and three dicot genomes including Arabidopsis thaliana (At), tomato Solanum lycopersicum (Soly) as well as black cottonwood Populus trichocarpa (Pp). Considering that these Cys residues are conserved in the hydrogen peroxide domain, we speculate that these HPCA1 homologues in Arabidopsis and other species could also function as H2O2 sensors. We also speculate that higher-order Arabidopsis mutants with HPCA1 homologues will show stronger impairment in eH2O2-induced responses. Given that all eight HPCA1 subfamily members in Arabidopsis contain these two Cys pairs—except HPCAL7, which misses Cys412 and Cys424—they could potentially act as eH2O2 sensors with similar kinetics and affinities, consistent with the reduction in amplitudes but not the Kd values of eH2O2-induced Ca2+ increases in hpca1 mutants (Fig. 1f). That is, the similar Kd values between wild-type and hpca1 seedlings (Fig. 1f) may indicate that the eH2O2-sensing systems that are both affected and unaffected by the hpca1 mutation share similar affinities to eH2O2.

Extended Data Fig. 6 HPCA1 genomic structure, mutations in three hpca1 mutants, HPCA1 subcellular localization and expression patterns.

a, Schematic illustration of the exon–intron structure of the HPCA1 gene, with boxes representing exons. The mutations in hpca1-1, hpca1-2 and hpca1-3 are illustrated. Primers for genotyping a splicing error in hpca1-2 are shown. FP, HPCA1 forward primer in the thirteenth exon; RP, HPCA1 reverse primer in the fourteenth exon. Genomic DNA sequence analysis confirmed that the hpca1-1 mutant contains a mutation in the kinase domain (Q856*, stop codon) that leads to a protein with a truncated kinase domain; hpca1-2 contains a mutation that leads to intron mis-splicing and a premature stop codon in the extracellular domain and a truncated HPCA1 protein without the transmembrane and kinase domains (a knockout line); and hpca1-3 contains a mutation in the LRR domain (G159D). The possible loss-of-function mutations are illustrated in the schematic in the kinase domain (C4425T) and LRR domain (C1663T) in hpca1-1 and hpca1-3, respectively, consistent with the importance of these two domains in LRR-RKs. b, The splicing mutation in hpca1-2. The hpca1-2 mutant has a single base mutation (G-to-A in the DNA sense strand) that corresponds to a change from 5′-AG2627 to 5′-AA2627 in the splice acceptor site of the thirteenth intron. This makes the site unrecognizable by the splicing enzymes, and leads to the predicted 76-bp addition in HPCA1 mRNA and a premature stop codon. PCR analyses of cDNA showed that hpca1-2 cDNA was about 76 bp longer than wild-type cDNA, which was confirmed by sequencing the HPCA1-2 cDNA clone of hpca1-2. Experiments were repeated independently three times. c, Expression patterns of HPCA1–YFP in Arabidopsis seedlings stably expressing the pHPCA1::HPCA1-YFP construct, which were also used in Fig. 3f. YFP fluorescence was analysed using a Zeiss stereo microscope, and images were merged to generate the whole-seedling image. Enlargements of upper epidermis and a root tip are shown at the top and bottom right, respectively, illustrating the plasma membrane localization of HPCA1–YFP and the high expression in epidermal cells and guard cells but the low expression in root tips. More than 15 homozygous single-insertion transgenic lines were generated, and similar results were observed from these lines. Scale bars, 3 mm (seedling) or 100 μm (leaf section and root tip). d, YFP fluorescence of Arabidopsis seedlings stably overexpressing YFP (CaMV 35S promoter-driven YFP construct, p35S::YFP) was analysed using confocal microscopy as a control, as in Fig. 3f. YFP fluorescence was observed in the cytosol and nucleus of the turgid cells (left) and plasmolysed cells (right). Scale bar, 20 μm. Data in c and d are representative of more than ten independent lines examined.

Extended Data Fig. 7 Subcellular localization of YFP fusion proteins of HPCA1 and HPCA1 mutants.

a, The fusion proteins HPCA1–YFP, mHPCA1-1–YFP (Q856*; hpca1-1 mutation in the kinase domain), and mHPCA1-3–YFP (G159D; hpca1-3 mutation in the LRR domain) were transiently expressed in Nicotiana benthamiana epidermal cells (CaMV 35S promoter-driven YFP constructs, p35S::HPCA1-YFP, p35S::mHPCA1-1-YFP and p35S::mHPCA1-3-YFP). The ectodomain mutant mHPCA1-3–YFP (G159D) showed lower fluorescence compared to wild-type HPCA1 and truncated mHPCA1-1-YFP (Q856*). These data suggest that the mutation in hpca1-1 may affect only the kinase activity but not the targeting of HPCA1, whereas the mutation in hpca1-3 may affect both the targeting and eH2O2 sensing. The experiments were repeated at least three times for each construct. Scale bars, 30 μm. b, Attenuated activation of MAPK by H2O2 in hpca1 and complementation by expressing wild-type HPCA1. Seedlings were treated with 4 mM H2O2, and phosphorylation of MAPK3 (p-MPK3) and MAPK6 (p-MPK6) was detected using pTEpY antibodies at indicated time points. α-tubulin was used as a loading control. Experiments were repeated independently three times. For gel source data, see Supplementary Fig. 1.

Extended Data Fig. 8 Ca2+ signalling by flg22 and elf26 is not affected in hpca1 mutants, and eH2O2 Ca2+ signalling is not affected in fls2, bak1 or ghr1 receptor kinase mutants.

a, b, Increases in [Ca2+]i in wild-type and hpca1 mutant plotted as a function of increasing concentrations of flg22 (a) or elf26 (b). Data are mean ± s.d.; n = 32 seedlings. P = 0.172 (a), P = 0.062 (b), two way ANOVA. Experiments were performed as in Fig. 1f–h. c, e, Representative images of aequorin bioluminescence from the wild-type, hpca1 and fls2 seedlings, which were grown side-by-side and treated with 4 mM H2O2 (c) or 1 μM flg22 (e). d, f, Quantification of [Ca2+]i in leaves from experiments in c or e is shown for treatment with H2O2 (d) or flg22 (f), treatment, respectively. Data are mean ± s.d. and representative of eight experiments; n = 16 seedlings. g, i, Representative aequorin bioluminescence images from wild-type, hpca1 and bak1 seedlings, which were grown side-by-side and treated with 4 mM H2O2 (g) or 1 μM flg22 (i). h, j, Quantification of [Ca2+]i in leaves from experiments in g or i is shown for H2O2 (h) or flg22 treatment (j), respectively. Data are mean ± s.d. and representative of eight experiments; n = 16 seedlings. k, l, Representative aequorin bioluminescence images (k) and quantification of [Ca2+]i increases (l) from wild-type, hpca1 and ghr1 seedlings, which were grown side-by-side and treated with 4 mM H2O2. Data in l are mean ± s.d. and representative of eight experiments; n = 16 seedlings. The receptors for MAMPs in innate immunity are commonly receptor kinases6,26. The two best-characterized LRR-RKs, FLS2 and EFR, are receptors for flg22 and elf18–elf26, respectively26. Both flg22 and elf18–elf26 induce [Ca2+]i increases, which are abolished in their corresponding receptor mutants, fls2 and efr, respectively26. We analysed whether HPCA1 could distinguish H2O2 produced in response to treatment with MAMPs. [Ca2+]i increases triggered by flg22 or elf26 were not affected in hpca1 mutants. The mutants in FLS2 and its co-receptor BAK1 showed the defects in flg22-triggered [Ca2+]i increases, but displayed wild-type level of eH2O2-induced [Ca2+]i increases. The observation is consistent with the previous finding that the rbohD (NADPH oxidase D) mutant and treatment with the NADPH oxidase inhibitor diphenyleneiodonium do not affect the initial [Ca2+]i increases triggered by 5-min treatment of flg2276. GHR1 is an LRR-RK with an unknown ligand in the III subfamily (Extended Data Fig. 4a), and the ghr1 mutant displays several strong phenotypes in the eH2O2 and ABA signalling28. Source data

Extended Data Fig. 9 HPCA1 kinase activity is required for eH2O2 signalling.

a, Inhibition of H2O2-induced [Ca2+]i increases in wild-type seedlings using the protein kinase inhibitor K252a, which has been shown to attenuate LRR-RK signalling pathways77, such as those for flg22, pep1 and elf26. Seedlings were treated with 2.5 mM H2O2 in the absence (−K252a, blue) or presence (+K252a, brown) of 2 µM K252a, and [Ca2+]i was analysed by aequorin luminometry spectroscopy. Data are mean ± s.e.m.; n = 12 seedlings. P < 0.001, two-way ANOVA. The quantified maximum [Ca2+]i changes from the same experiments are shown in Fig. 3h. b, K252a inhibition of H2O2-induced Ca2+ currents in wild-type guard-cell protoplasts. Representative whole-cell currents in protoplasts treated with 5 mM H2O2 in the absence or presence of 2 µM K252a for 1.5 h are shown. The currents at −180 mV from similar experiments are shown in Fig. 3i. c, Inhibition of H2O2-induced stomatal closure by K252a. Detached leaves were treated with 2.5 mM H2O2 in the absence or the presence of 2 µM K252a, and stomatal apertures were analysed. Data are mean ± s.e.m.; n = 60 stomata. Stomatal apertures without treatment were arbitrarily set to 1 r.u. d, e, Phosphorylation of HPCA1–YFP induced by H2O2 and flg22 in planta. The hpca1-2 seedlings stably expressing the HPCA1::HPCA1-YFP construct were treated with 4 mM H2O2 (d) or 1 µM flg22 (e), and HPCA1–YFP proteins were prepared by immunoprecipitation. Phosphorylation levels were detected using pThr antibodies. YFP was used as a loading control. For gel source data, see Supplementary Fig. 1. f, HPCA1CD was autophosphorylated. GST-tagged HPCA1CD proteins—including wild-type HPCA1CD and two kinase-inactive mutants (HPCA1(K859E)CD and HPCA1(D773L)CD) and kinase domain-truncated HPCA1 found in hpca1-1 (mHPCA1-1; HPCA1(Q856*)CD)—were expressed in Escherichia coli (as in Fig. 3k). Purified proteins were incubated with or without λ-protein phosphatase (Lambda PPase; 30 °C) for 2 h, and separated by SDS–PAGE. Red and black arrows denote phosphorylated and unphosphorylated bands, respectively. Only the HPCA1CD band was shifted owing to autophosphorylation, and the band shift was abolished by λ-protein phosphatase treatment. Experiments in df were repeated independently three times. g, The tandem mass spectrometry (MS/MS) spectra of a HPCA1 peptide (786-THVTTQVK-793; inset). Nine-day-old hpca1-2 plants stably expressing HPCA1–YFP were treated with water or 4 mM H2O2, and HPCA1–YFP was immunoprecipitated. The proteins were purified by PAGE, and in-gel digested with trypsin and GluC, and resulting peptides were extracted and analysed by liquid chromatography–tandem mass spectrometry (LC–MS/MS). Labelled peaks correspond to masses of y and b ions of the modified peptide. Analysis of y and b ion fragmentation patterns with MaxQuant showed that Thr789 and Thr790 were phosphorylated in vivo. These amino acid residues are illustrated in Fig. 3g. h, H2O2 treatment resulted in increased phosphorylation of Thr and Ser residues of HPCA1–YFP in vivo from experiments as in g. Data are mean ± s.d.; n = 3 independent experiments. P < 0.05, two-sided t-test. The functional relevance of these phosphosites could be determined further through complementation experiments of the hpca1-2-null mutant with the respective phosphosite mutants of HPCA1. These amino acid residues are illustrated in Fig. 3g. i, Highly conserved phosphorylated Thr residues in the catalytic core of the kinase domains of HPCA1, BAK1 and BRI1. Alignment of these three protein sequences in subdomain VIb to the invariant E in subdomain VIII. Bold residues are highly conserved in active kinases. Amino acid residues (orange) are confirmed phosphorylation sites in our study and previous reports78,79,80. For the activation of LRR-RKs, it is thought that a conformational change mediated by ligand binding to the extracellular LRR domain activates the kinase by allowing trans-phosphorylation of the kinase domain within its homodimer81,82. This, in turn, could allow further autophosphorylation of the kinase domain, providing docking sites for interacting proteins and funnelling the ligand signal to downstream events27,81. The phosphorylation of these Thr residues has been shown to be required for kinase activation78. Source data

Extended Data Fig. 10 The shift of extracellular Cys residues in the H2O2 domain from the reduced to the oxidative form after eH2O2 treatment.

a, Schematics of the covalent modification of HPCA1 Cys residues by the membrane impermeable alkyne reagents IA-biotin (top) and MTSEA-biotin (bottom) (as in Fig. 4a–d). IA-biotin-mediated irreversible modification of extracellular Cys residues prevents disulfide bond formation between Cys residues. The Cys residues are specifically alkanethiolated by MTSEA-biotin (a cysteine-disulfde-bond-forming reagent) under mild conditions, and then form a covalent disulfide bond. This disulfide bond is readily reduced by neighbour Cys residues to form a disulfide bond between Cys residues (asterisk). b, hpca1 complementation using HPCA1 with Cys mutations could not complement the hpca1-2 phenotype. Wild-type and hpca1-2 mutant seedlings, and hpca1-2 seedlings expressing HPCA1 or HPCA1 with Cys mutations (used in Fig. 4e) were treated with 0.9 M CaCl2 and 10% (v/v) ethanol (aequorin discharge), and the remaining aequorin luminescence was imaged. Similar luminescence intensities were detected for all genotypes. The relative [Ca2+]i is scaled by a pseudo-colour bar. The experiment was repeated independently 20 time with similar results. c, Plasma membrane localization of HPCA1 and Cys-pair-mutated HPCA1 used in Fig. 4e. The hpca1-2 seedlings stably expressing the pHPCA1::HPCA1-YFP construct or mutated Cys-pair HPCA1 were used, and YFP fluorescence was analysed by confocal microscopy. YFP fluorescence of HPCA1 and mutated Cys-pair HPCA1 was observed in the periphery of turgid and plasmolysed cells, which suggests that these Cys mutations did not alter HPCA1 subcellular localization. Data are representative of more than ten independent lines examined. Scale bar, 20 μm. d, Schematics of the covalent modification of HPCA1 Cys residues by iodoacetamide and NEM, for experiments in Fig. 4h–j. HPCA1–YFP proteins from transgenic plants were labelled first with iodoacetamide to form Cys-CAM and represent the reduced form, and then treated with dithiothreitol (DTT) to reduce disulfide bonds, which were then labelled with NEM to represent the oxidized form. Modifications of proteins were analysed by LC–MS/MS. The CAM- and NEM-labelled peptides were analysed and their relative amounts were calculated. e−g, Representative MS/MS spectra of the STLPTNC421SPC424EPGME peptide in the hydrogen peroxide domain from HPCA1–YFP proteins prepared using hpca1-2 plants stably expressing the pHPCA1::HPCA1-YFP construct with or without 4 mM H2O2 treatment, as in Fig. 4h–j. HPCA1–YFP proteins were labelled with iodoacetamide and NEM as in d, and digested using GluC and chymotrypsin for mass spectrometry. Cys421 and Cys424 labelled by CAM (e), Cys421 by CAM and Cys424 by NEM (f), and Cys421 by NEM and Cys424 by CAM (g) were all detected from the HPCA1–YFP proteins prepared with or without H2O2 treatment. Note that H2O2 treatment increased the relative amount of NEM-labelled Cys residues (Fig. 4j). The amounts of Cysred and Cysox residues were represented by CAM- and NEM-labelled peptides, respectively, and normalized to those of CAM-labelled peptides (Fig. 4j). The relative percentages of Cysred and Cysox from the total amounts of Cys residues were determined (Fig. 4j). Note that the relative amount of Cysred + the relative amount of Cysox = 100% Cys residues. Data are from four independent experiments. We have demonstrated that both Cysred and Cysox residues in the hydrogen peroxide domain exist, and that these Cysred residues could be oxidized by eH2O2 to Cysox residues in planta, which transduces the eH2O2 signal to downstream events. To fulfil the receptor function, the HPCA1 receptor should be regenerated. Given that plants grow almost constantly to form new cells, these new cells should carry the HPCA1 receptors with Cysred residues. For the eH2O2-oxidized HPCA1 receptors with Cysox residues, it is possible that these oxidized receptors are either recycled via the endocytic process, or reduced by unknown mechanisms, which need to be determined in the future.

Supplementary information

Supplementary Information

This file contains Supplementary Text (Unbiased screening approaches used to dissect eH2O2 signaling in Arabidopsis) and associated references, and Supplementary Figure 1 (raw gel images).

Reporting Summary

Source data

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Wu, F., Chi, Y., Jiang, Z. et al. Hydrogen peroxide sensor HPCA1 is an LRR receptor kinase in Arabidopsis. Nature 578, 577–581 (2020). https://doi.org/10.1038/s41586-020-2032-3

Download citation

Further reading


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.


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