Skip to main content

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

  • Article
  • Published:

The lung microbiome regulates brain autoimmunity

Nature volume 603, pages 138–144 (2022)Cite this article

Subjects

Abstract

Lung infections and smoking are risk factors for multiple sclerosis, a T-cell-mediated autoimmune disease of the central nervous system1. In addition, the lung serves as a niche for the disease-inducing T cells for long-term survival and for maturation into migration-competent effector T cells2. Why the lung tissue in particular has such an important role in an autoimmune disease of the brain is not yet known. Here we detected a tight interconnection between the lung microbiota and the immune reactivity of the brain. A dysregulation in the lung microbiome significantly influenced the susceptibility of rats to developing autoimmune disease of the central nervous system. Shifting the microbiota towards lipopolysaccharide-enriched phyla by local treatment with neomycin induced a type-I-interferon-primed state in brain-resident microglial cells. Their responsiveness towards autoimmune-dominated stimulation by type II interferons was impaired, which led to decreased proinflammatory response, immune cell recruitment and clinical signs. Suppressing lipopolysaccharide-producing lung phyla with polymyxin B led to disease aggravation, whereas addition of lipopolysaccharide-enriched phyla or lipopolysaccharide recapitulated the neomycin effect. Our data demonstrate the existence of a lung–brain axis in which the pulmonary microbiome regulates the immune reactivity of the central nervous tissue and thereby influences its susceptibility to autoimmune disease development.

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

Access options

Buy this article

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

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

Fig. 1: Manipulations of the lung microbiota affect CNS autoimmunity.
Fig. 2: Lung dysbiosis does not influence T cell activation and migration.
Fig. 3: Lung dysbiosis affects microglia immune reactivity.
Fig. 4: Lung dysbiosis shifts microglia to a type I IFN signature.
Fig. 5: Pulmonary LPS controls CNS autoimmunity.

Similar content being viewed by others

Data availability

RNA-seq datasets have been deposited online in the Gene Expression Omnibus (GEO) and BioProject with accession codes GSE191287, GSE192411 and PRJNA789820. Source data are provided with this paper.

References

  1. Olsson, T., Barcellos, L. F. & Alfredsson, L. Interactions between genetic, lifestyle and environmental risk factors for multiple sclerosis. Nat. Rev. Neurol. 13, 25–36 (2017).

    Article  CAS  PubMed  Google Scholar 

  2. Odoardi, F. et al. T cells become licensed in the lung to enter the central nervous system. Nature 488, 675–679 (2012).

    Article  ADS  CAS  PubMed  Google Scholar 

  3. O’Dwyer, D. N., Dickson, R. P. & Moore, B. B. The lung microbiome, immunity, and the pathogenesis of chronic lung disease. J. Immunol. 196, 4839–4847 (2016).

    Article  PubMed  CAS  Google Scholar 

  4. Jin, C. et al. Commensal microbiota promote lung cancer development via γδ T cells. Cell 176, 998–1013 (2019).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  5. Yokote, H. et al. NKT cell-dependent amelioration of a mouse model of multiple sclerosis by altering gut flora. Am. J. Pathol. 173, 1714–1723 (2008).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  6. Ochoa-Repáraz, J. et al. Role of gut commensal microflora in the development of experimental autoimmune encephalomyelitis. J. Immunol. 183, 6041–6050 (2009).

    Article  PubMed  CAS  Google Scholar 

  7. Berer, K. et al. Commensal microbiota and myelin autoantigen cooperate to trigger autoimmune demyelination. Nature 479, 538–541 (2011).

    Article  ADS  CAS  PubMed  Google Scholar 

  8. Rothhammer, V. et al. Type I interferons and microbial metabolites of tryptophan modulate astrocyte activity and central nervous system inflammation via the aryl hydrocarbon receptor. Nat. Med. 22, 586–597 (2016).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  9. Miyauchi, E. et al. Gut microorganisms act together to exacerbate inflammation in spinal cords. Nature 585, 102–106 (2020).

    Article  ADS  CAS  PubMed  Google Scholar 

  10. Flügel, A., Willem, M., Berkowicz, T. & Wekerle, H. Gene transfer into CD4+ T lymphocytes: green fluorescent protein-engineered, encephalitogenic T cells illuminate brain autoimmune responses. Nat. Med. 5, 843–847 (1999).

    Article  PubMed  Google Scholar 

  11. Lodygin, D. et al. β-Synuclein-reactive T cells induce autoimmune CNS grey matter degeneration. Nature 566, 503–508 (2019).

    Article  ADS  CAS  PubMed  Google Scholar 

  12. Bartholomäus, I. et al. Effector T cell interactions with meningeal vascular structures in nascent autoimmune CNS lesions. Nature 462, 94–98 (2009).

    Article  ADS  PubMed  CAS  Google Scholar 

  13. Kivisäkk, P. et al. Localizing central nervous system immune surveillance: meningeal antigen-presenting cells activate T cells during experimental autoimmune encephalomyelitis. Ann. Neurol. 65, 457–469 (2009).

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  14. Lodygin, D. et al. A combination of fluorescent NFAT and H2B sensors uncovers dynamics of T cell activation in real time during CNS autoimmunity. Nat. Med. 19, 784–790 (2013).

    Article  CAS  PubMed  Google Scholar 

  15. Starossom, S. C. et al. Galectin-1 deactivates classically activated microglia and protects from inflammation-induced neurodegeneration. Immunity 37, 249–263 (2012).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  16. Kawakami, N. et al. The activation status of neuroantigen-specific T cells in the target organ determines the clinical outcome of autoimmune encephalomyelitis. J. Exp. Med. 199, 185–197 (2004).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  17. Odoardi, F. et al. Instant effect of soluble antigen on effector T cells in peripheral immune organs during immunotherapy of autoimmune encephalomyelitis. Proc. Natl Acad. Sci. USA 104, 920–925 (2007).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  18. Heppner, F. L. et al. Experimental autoimmune encephalomyelitis repressed by microglial paralysis. Nat. Med. 11, 146–152 (2005).

    Article  CAS  PubMed  Google Scholar 

  19. Hanisch, U. K. & Kettenmann, H. Microglia: active sensor and versatile effector cells in the normal and pathologic brain. Nat. Neurosci. 10, 1387–1394 (2007).

    Article  CAS  PubMed  Google Scholar 

  20. Rock, R. B. et al. Transcriptional response of human microglial cells to interferon-γ. Genes Immun. 6, 712–719 (2005).

    Article  CAS  PubMed  Google Scholar 

  21. Popovic, N. et al. Inhibition of autoimmune encephalomyelitis by a tetracycline. Ann. Neurol. 51, 215–223 (2002).

    Article  CAS  PubMed  Google Scholar 

  22. Elmore, M. R. et al. Colony-stimulating factor 1 receptor signaling is necessary for microglia viability, unmasking a microglia progenitor cell in the adult brain. Neuron 82, 380–397 (2014).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  23. Prinz, M. et al. Distinct and nonredundant in vivo functions of IFNAR on myeloid cells limit autoimmunity in the central nervous system. Immunity 28, 675–686 (2008).

    Article  CAS  PubMed  Google Scholar 

  24. Khorooshi, R. et al. Induction of endogenous type I interferon within the central nervous system plays a protective role in experimental autoimmune encephalomyelitis. Acta Neuropathol. 130, 107–118 (2015).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  25. McNab, F., Mayer-Barber, K., Sher, A., Wack, A. & O’Garra, A. Type I interferons in infectious disease. Nat. Rev. Immunol. 15, 87–103 (2015).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  26. Bradley, K. C. et al. Microbiota-driven tonic interferon signals in lung stromal cells protect from influenza virus infection. Cell Rep. 28, 245–256 (2019).

    Article  CAS  PubMed  Google Scholar 

  27. d’Hennezel, E., Abubucker, S., Murphy, L. O. & Cullen, T. W. Total lipopolysaccharide from the human gut microbiome silences toll-like receptor signaling. mSystems 2, e00046-17 (2017).

    Article  PubMed  PubMed Central  Google Scholar 

  28. Yang, D. et al. Dysregulated lung commensal bacteria drive interleukin-17B production to promote pulmonary fibrosis through their outer membrane vesicles. Immunity 50, 692–706 (2019).

    Article  CAS  PubMed  Google Scholar 

  29. Bhor, V. M., Thomas, C. J., Surolia, N. & Surolia, A. Polymyxin B: an ode to an old antidote for endotoxic shock. Mol. Biosyst. 1, 213–222 (2005).

    Article  CAS  PubMed  Google Scholar 

  30. Vargas-Caraveo, A. et al. Lipopolysaccharide enters the rat brain by a lipoprotein-mediated transport mechanism in physiological conditions. Sci. Rep. 7, 13113 (2017).

    Article  ADS  PubMed  PubMed Central  CAS  Google Scholar 

  31. Sandiego, C. M. et al. Imaging robust microglial activation after lipopolysaccharide administration in humans with PET. Proc. Natl Acad. Sci. USA 112, 12468–12473 (2015).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  32. Dickson, R. P., Erb-Downward, J. R., Martinez, F. J. & Huffnagle, G. B. The microbiome and the respiratory tract. Annu. Rev. Physiol 78, 481–504 (2016).

    Article  CAS  PubMed  Google Scholar 

  33. Belkaid, Y. & Hand, T. W. Role of the microbiota in immunity and inflammation. Cell 157, 121–141 (2014).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  34. Erny, D. et al. Host microbiota constantly control maturation and function of microglia in the CNS. Nat. Neurosci. 18, 965–977 (2015).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  35. Braniste, V. et al. The gut microbiota influences blood–brain barrier permeability in mice. Sci. Transl. Med. 6, 263ra158 (2014).

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  36. Wang, Y. et al. An intestinal commensal symbiosis factor controls neuroinflammation via TLR2-mediated CD39 signalling. Nat. Commun. 5, 4432 (2014).

    Article  ADS  CAS  PubMed  Google Scholar 

  37. Luu, M. et al. The short-chain fatty acid pentanoate suppresses autoimmunity by modulating the metabolic-epigenetic crosstalk in lymphocytes. Nat. Commun. 10, 760 (2019).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  38. Sonner, J. K. et al. Dietary tryptophan links encephalogenicity of autoreactive T cells with gut microbial ecology. Nat. Commun. 10, 4877 (2019).

    Article  ADS  PubMed  PubMed Central  CAS  Google Scholar 

  39. Jakimovski, D., Kolb, C., Ramanathan, M., Zivadinov, R. & Weinstock-Guttman, B. Interferon β for multiple sclerosis. Cold Spring Harb. Perspect. Med. 8, a032003 (2018).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  40. Guo, B., Chang, E. Y. & Cheng, G. The type I IFN induction pathway constrains Th17-mediated autoimmune inflammation in mice. J. Clin. Invest. 118, 1680–1690 (2008).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  41. Bauer, H., Horowitz, R. E., Levenson, S. M. & Popper, H. The response of the lymphatic tissue to the microbial flora. Studies on germfree mice. Am. J. Pathol. 42, 471–483 (1963).

    CAS  PubMed  PubMed Central  Google Scholar 

  42. Smith, K., McCoy, K. D. & Macpherson, A. J. Use of axenic animals in studying the adaptation of mammals to their commensal intestinal microbiota. Semin. Immunol. 19, 59–69 (2007).

    Article  CAS  PubMed  Google Scholar 

  43. Kennedy, E. A., King, K. Y. & Baldridge, M. T. Mouse microbiota models: comparing germ-free mice and antibiotics treatment as tools for modifying gut bacteria. Front. Physiol. 9, 1534 (2018).

    Article  PubMed  PubMed Central  Google Scholar 

  44. Wypych, T. P., Wickramasinghe, L. C. & Marsland, B. J. The influence of the microbiome on respiratory health. Nat. Immunol. 20, 1279–1290 (2019).

    Article  CAS  PubMed  Google Scholar 

  45. Balmer, M. L. et al. The liver may act as a firewall mediating mutualism between the host and its gut commensal microbiota. Sci. Transl. Med. 6, 237ra266 (2014).

    Article  CAS  Google Scholar 

  46. Määttä, J. A., Coffey, E. T., Hermonen, J. A., Salmi, A. A. & Hinkkanen, A. E. Detection of myelin basic protein isoforms by organic concentration. Biochem. Biophys. Res. Commun. 238, 498–502 (1997).

    Article  PubMed  Google Scholar 

  47. Murray, C. et al. Interdependent and independent roles of type I interferons and IL-6 in innate immune, neuroinflammatory and sickness behaviour responses to systemic poly I:C. Brain Behav. Immun. 48, 274–286 (2015).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  48. Rittirsch, D. et al. Acute lung injury induced by lipopolysaccharide is independent of complement activation. J. Immunol. 180, 7664–7672 (2008).

    Article  CAS  PubMed  Google Scholar 

  49. Klindworth, A. et al. Evaluation of general 16S ribosomal RNA gene PCR primers for classical and next-generation sequencing-based diversity studies. Nucleic Acids Res. 41, e1 (2013).

    Article  CAS  PubMed  Google Scholar 

  50. von Hoyningen-Huene, A. J. E. et al. Bacterial succession along a sediment porewater gradient at Lake Neusiedl in Austria. Sci. Data 6, 163 (2019).

    Article  CAS  Google Scholar 

  51. Rognes, T., Flouri, T., Nichols, B., Quince, C. & Mahé, F. VSEARCH: a versatile open source tool for metagenomics. PeerJ 4, e2584 (2016).

    Article  PubMed  PubMed Central  Google Scholar 

  52. Yilmaz, P. et al. The SILVA and “All-species Living Tree Project (LTP)” taxonomic frameworks. Nucleic Acids Res. 42, D643–D648 (2014).

    Article  CAS  PubMed  Google Scholar 

  53. Quast, C. et al. The SILVA ribosomal RNA gene database project: improved data processing and web-based tools. Nucleic Acids Res. 41, D590–D596 (2013).

    Article  CAS  PubMed  Google Scholar 

  54. Chen, L. et al. GMPR: a robust normalization method for zero-inflated count data with application to microbiome sequencing data. PeerJ 6, e4600 (2018).

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  55. Andersen, K. S., Kirkegaard, R. H., Karst, S. M. & Albertsen, M. ampvis2: an R package to analyse and visualise 16S rRNA amplicon data. Preprint at https://doi.org/10.1101/299537 (2018).

  56. Wickham, H. ggplot2: Elegant Graphics for Data Analysis (Springer, 2016).

  57. Schläger, C. et al. Effector T-cell trafficking between the leptomeninges and the cerebrospinal fluid. Nature 530, 349–353 (2016).

    Article  ADS  PubMed  CAS  Google Scholar 

  58. Cabeza, R. et al. An RNA sequencing transcriptome analysis reveals novel insights into molecular aspects of the nitrate impact on the nodule activity of Medicago truncatula. Plant Physiol. 164, 400–411 (2014).

    Article  CAS  PubMed  Google Scholar 

  59. Love, M. I., Huber, W. & Anders, S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 15, 550 (2014).

    PubMed  PubMed Central  Google Scholar 

  60. Huang, D. W., Sherman, B. T. & Lempicki, R. A. Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat. Protoc. 4, 44–57 (2009).

    Article  CAS  Google Scholar 

  61. Doorn, K. J. et al. Brain region-specific gene expression profiles in freshly isolated rat microglia. Front. Cell. Neurosci. 9, 84 (2015).

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  62. Klinkert, W. E. et al. TNF-α receptor fusion protein prevents experimental auto-immune encephalomyelitis and demyelination in Lewis rats: an overview. J. Neuroimmunol. 72, 163–168 (1997).

    Article  CAS  PubMed  Google Scholar 

Download references

Acknowledgements

We thank S. Schwarz for helping in the quantification of the tuf gene expression; M. Ulisse for contributing to characterizing the lung immune milieu; G. Salinas for performing the transcriptome analyses; O. Shomroni for the analysis of the transcriptome data; S. Hamann, M. Weig and M. Heinemann for technical assistance; A. Poehlein for performing the 16S rRNA sequencing; D. Schneider for providing the bioinformatic amplicon processing pipeline; D. Miljković for helping with animal experiments and reading the manuscript; and C. Ludwig for text editing. This work was supported by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) RK-Grant FL 377/3-1; FL 377/2-2; SFB 1328/1 project A01, project no. 335447717; OD 87/1-1, OD 87/3-1; SFB TRR 274/1 2020 projects A03 and A04, project no. 408885537; and by the European Commission under the European Union’s Horizon 2020 research and innovation programme, grant agreement no. 101021345 (T-Neuron). L.H. is supported by the Klaus Faber Stiftung.

Author information

Author notes

  1. These authors contributed equally: Alexander Flügel, Francesca Odoardi

Authors and Affiliations

  1. Institute for Neuroimmunology and Multiple Sclerosis Research, University Medical Center Göttingen, Göttingen, Germany

    Leon Hosang, Roger Cugota Canals, Felicia Joy van der Flier, Alexander Flügel & Francesca Odoardi

  2. Department of Genomic and Applied Microbiology, University of Göttingen, Göttingen, Germany

    Jacqueline Hollensteiner & Rolf Daniel

  3. Center for Biostructural Imaging of Neurodegeneration, University Medical Center Göttingen, Göttingen, Germany

    Francesca Odoardi

Authors
  1. Leon Hosang
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Roger Cugota Canals
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Felicia Joy van der Flier
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Jacqueline Hollensteiner
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Rolf Daniel
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Alexander Flügel
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Francesca Odoardi
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

L.H. performed most experimental work and together with A.F. and F.O. wrote the paper. R.C.C. contributed to immune cell characterizations by quantitative PCR analyses and flow cytometry and by performing EAE experiments. F.J.v.d.F. performed the intravital TPLSM and supported L.H. with inducing and analysing autoimmune models. J.H. contributed with the microbiome analyses, and R.D. contributed with his expertise in microbiome biology and  interpretation of the microbiota sequencing data. A.F. together with F.O. designed the study, coordinated the experimental work and wrote the manuscript with inputs from co-authors.

Corresponding authors

Correspondence to Alexander Flügel or Francesca Odoardi.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature thanks Sarkis Mazmanian, Vijay Kuchroo and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.

Additional information

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 Establishment of a lung EAE model and targeted manipulation of the lung microbiome.

a, Lung EAE. Rats were intravenously (i.v.) transferred with resting TMBP cells and 6 h later were intratracheally (i.tr.) immunized with CFA ± MBP. Clinical parameters: Body weight change (lines) and clinical scores (bars) over the EAE course, incidence (%), onset (days after immunization), peak score, cumulative score. Mean ± s.e.m. Representative data of 3 independent experiments. n = 4 (−MBP); n = 5 (+MBP). ND, not determinable. b–e, I.tr. neomycin (Neo) treatment does not change the lung immune cell composition. Lung samples were isolated from rats i.tr. treated for 7 days with PBS or neomycin . b, Number of endogenous αβTCR+ CD4+ and CD8+ T cells, CD45RA+ B cells, CD11b+ ED9− interstitial macrophages, CD11b+ ED9+ MΦ (alveolar macrophages and infiltrating monocytes) and RP3+ neutrophils. Flow cytometry. Mean ± s.e.m. c, Percentage expression of FoxP3, IFNγ and IL17 in endogenous αβTCR+ CD4+ and CD8+ T cells. Flow cytometry. IFNγ and IL17 were analysed in steady state condition and upon stimulation with PMA and Ionomycin (PMA/I). Mean ± s.e.m. d, e, Corresponding surface expression of the activation markers CD134 (OX40) and CD25 (IL2R) by flow cytometry (d) and expression of the indicated cytokines measured by quantitative PCR (e) Mean ± s.e.m. b-e, Cumulative data of 2 independent experiments. n = 5 (all groups). f, I.tr. treatment for 7 days with PBS or neomycin  does not change MHC expression in lung immune cells. Expression of the indicated genes in total lung, pulmonary stromal cells (CD45−), as well as lymphoid (CD45+ CD11b−) or myeloid (CD45+ CD11b+) hematopoietic pulmonary cell populations. Quantitative PCR. Housekeeping gene: β-actin. Mean ± s.e.m. Cumulative data of 3 independent experiments. n = 9–14 (PBS); n = 8–12 (Neo) per condition. a–f, Statistical significance determined by unpaired two-tailed t-test (Gaussian distribution) and Mann–Whitney test (non-Gaussian distribution). *P < 0.05, **P < 0.01, ***P < 0.001.

Source data

Extended Data Fig. 2 Neomycin treatment outside the lung does not affect CNS autoimmunity.

a, b, Oral treatment with neomycin  does not ameliorate EAE. a, Left, PCA of the microbiota composition of faecal samples from rats that were orally treated with neomycin (1 or 10 mg) or PBS for 7 days. Middle, corresponding Shannon and phylogenetic diversity indices. Right, quantification of bacterial abundance based on tuf gene expression via 16S rRNA-based quantitative PCR. Mean ± s.e.m. Cumulative data of 2 independent experiments. n = 6 (all groups). b, Clinical parameters observed in lung EAE of rats pre-treated orally with PBS or neomycin (1 or 10 mg) for 7 days. Mean ± s.e.m. Cumulative data of 2 independent experiments. n = 8 (PBS, 1 mg Neo); n = 7 (10 mg Neo). c, d, I.tr. neomycin treatment does not affect TMBP cell proliferation and effector function. c, Quantification of TMBP cells cultured in presence of neomycin at the indicated concentrations. Flow cytometry on D2, D3 and D4 after antigen challenge. Representative data of 2 independent experiments. n = 3 (all groups). d, Clinical outcome of EAE induced by transfer of TMBP cells previously stimulated in vitro in presence of neomycin at the indicated concentrations. Clinical parameters. Mean ± s.e.m. Representative data of 3 independent experiments. n = 3 (both groups). e, f, Subcutaneous (s.c.) treatment with neomycin does not affect EAE. e, Lung EAE was induced in rats s.c. pre-treated for 7 days with PBS or neomycin. Clinical parameters. Mean ± s.e.m. Representative data of 3 independent experiments. n = 5 (both groups). f, Rats s.c. pre-treated for 7 days with PBS or neomycin were i.v. transferred with resting TMBP cells. 6 h later, they were s.c. immunized with MBP. Clinical parameters. Mean ± s.e.m. Representative data of 3 independent experiments. n = 3 (both groups). a–d, Statistical significance determined by one-way ANOVA with Tukey’s multiple comparisons test (Gaussian distribution) and Kruskal–Wallis test with Dunn’s multiple comparisons test (non-Gaussian distribution). e, f, Statistical significance determined by unpaired two-tailed t-test (Gaussian distribution) and Mann–Whitney test (non-Gaussian distribution). *P < 0.05, **P < 0.01, ***P < 0.001.

Source data

Extended Data Fig. 3 Lung dysbiosis does not modify T cell activation and expression profile in the lung.

a–e, Lung EAE was induced in rats that were pre-treated i.tr. with neomycin or PBS for 7 days. a, Lung immunization induces reprogramming in the gene expression profile of TMBP cells. Volcano plots depicting the differential gene expression profiles of lung-derived TMBP cells from PBS- (left) or neomycin- (right) pre-treated rats between D1 and D0 after immunization. Red and blue dots represent significantly up- or downregulated genes (P < 0.05), respectively. Indicated are representative genes involved in cell division and cell cycle. b, c, Genes differentially expressed after immunization are mainly involved in cell cycle. b, Significantly regulated KEGG pathways for genes upregulated between D1 versus D0 after immunization in TMBP cells isolated from lung of PBS- (left) or neomycin-treated (right) rats. Bold, pathways significantly enriched in both treatments. c, Heat map of the 50 most upregulated genes in D1 versus D0 after immunization in PBS- and neomycin-treated rats. d, Lung immunization does not change effector T cell differentiation. Total reads of transcription factors, cytokines and chemokine receptors in TMBP cells isolated from lung of PBS- or neomycin-treated rats on D0 and D1 after immunization Mean ± s.e.m. n = 3 (all groups). e, I.tr. neomycin treatment does not impair TMBP cell activation and migratory program. Relative expression of chemokine receptors and genes involved in cell cycle and cell egress in TMBP cells isolated from the lung of PBS- or neomycin-pre-treated rats on D1 after immunization. Quantitative PCR. Mean ± s.e.m. Representative data of two independent experiments. n = 3 (PBS); n = 4 (Neo) per condition. d, e, Statistical significance determined by unpaired two-tailed t-test. *P < 0.05.

Source data

Extended Data Fig. 4 Lung dysbiosis does not impair the lung immune response after immunization but impairs grey matter autoimmunity.

a–f, Lung EAE was induced in rats pre-treated i.tr. with PBS or neomycin. Characterization of the lung immune milieu was performed 24 h after i.tr. immunization. a, Lung EAE was induced in rats that were pre-treated i.tr. with neomycin or PBS for 7 days. b–d, Lung microbiome dysbiosis does not impair local T cell responses. b, Absolute numbers of TMBP cells and endogenous αβTCR+ CD4+ and CD8+ T cells. Flow cytometry. c, Corresponding percentage expression of FoxP3 and proinflammatory cytokines in stimulated (PMA/I) or non-stimulated T cell subsets. Flow cytometry. Representative data from 2 independent experiments. n = 5 (all groups). d, Relative expression of the indicated T cell lineage-signature cytokines in each T cell subset. Quantitative PCR. Housekeeping gene: β-actin. Mean ± s.e.m. b-d, Representative data from 2 independent experiments. n = 5 (all groups). e, f, Lung microbiome dysbiosis does not impair local myeloid cell responses. e, Absolute number of CD11b+ ED9− interstitial macrophages, CD11b+ ED9+ MΦ and RP3+ neutrophils. Flow cytometry. f, Corresponding expression of chemokines, iNOS (Nos2), MHC-II (Rt1ba) and M2 macrophage markers. Quantitative PCR. Housekeeping gene: β-actin. e–f, Mean ± s.e.m. Representative data of 2 independent experiments. n = 5 (all groups). g, h, Lung dysbiosis prevents TMBP cell entry into the CNS and ameliorates peripheral EAE induced by s.c. immunization. Rats were i.tr. treated with neomycin or PBS for 7 days. Subsequently, they were i.v. transferred with resting TMBP cells and s.c. immunized with MBP. g, Clinical parameters. Mean ± s.e.m. Representative data of 3 independent experiments. n = 4 (both groups). h, Number of TMBP cells detected in the indicated organs by flow cytometry on D11 after immunization. Mean ± s.e.m. Representative data of 3 independent experiments. n = 3 (both groups). i, j, I.tr. neomycin treatment reduces TbSYN cell entry in the brain and ameliorates autoimmune grey matter disease. Grey matter autoimmunity was induced in rats pre-treated with neomycin or PBS by transfer of TbSYN cells. i, Clinical parameters. Mean ± s.e.m. Cumulative data of 2 independent experiments. n = 6 (both groups). j, Quantification of the indicated immune cell subsets in blood or brain on D5 after transfer. Flow cytometry. Mean ± s.e.m. Representative data of 2 independent experiments. n = 4 (both groups). b–j, Statistical significance determined by unpaired two-tailed t-test (Gaussian distribution) and Mann–Whitney test (non-Gaussian distribution). *P < 0.05, **P < 0.01, ***P < 0.001.

Source data

Extended Data Fig. 5 Lung dysbiosis affects neither T cell and endothelial cell interactions at the CNS borders nor the barrier integrity.

a, b, I.tr. neomycin treatment does not influence the expression of either chemokine receptors or adhesion molecules in TMBP cells. Lung EAE or transfer EAE were induced in PBS- or neomycin-pre-treated rats. a, Chemokine receptor and integrin expression in TMBP cells isolated from blood on D5 after immunization (lung EAE, n = 5 per condition) or D4 after transfer (transfer EAE, n = 4 per condition). Quantitative PCR. Housekeeping gene: β-actin. Mean ± s.e.m. Representative data of 2 independent experiments. b, Corresponding protein expression of LFA-1 and VLA-4. Flow cytometry. Representative data of 2 independent experiments. c, d, I.tr. neomycin treatment does not affect T cell motility at the CNS borders. TMBP (c) or TOVA (d) cells were i.v. transferred into rats pre-treated with PBS or neomycin. c, Intravascular TMBP cell motility was recorded in the leptomeninges by TPLSM on D2.5 post transfer. Depicted are representative time-projection images over a period of 30 min, percentage of crawling versus rolling cells (n = 8 videos per group) and quantification of the indicated motility parameters. Number of analysed T cells is indicated. Mean ± s.e.m. Representative data of 2 independent experiments. n = 2-4 (both groups). Turquoise, TMBP cells; Red, 70 kDa Texas Red Dextran labelled vessels; Blue, Collagen. d, Intravascular (upper panel) and extravascular (lower panel) TOVA cell motility was recorded in the leptomeninges by TPLSM on D3 and D4 post transfer, respectively. Time projection images over a period of 30 min, percentage of crawling versus rolling cells (n = 9 videos per group) and intravascular and extravascular motility parameters derived from the indicated number of T cells. Mean ± s.e.m. Turquoise, TOVA cells; Red, 70 kDa Texas Red Dextran labelled vessels; Blue, Collagen. e, I.tr. neomycin treatment does not affect TOVA cell diapedesis. Intravital TPLSM overviews and corresponding magnified pictures depicting the distribution of TOVA cells (turquoise) in the leptomeningeal milieu at the indicated time points post transfer in PBS- or neomycin- pre-treated rats. Red, 70 kDa dextran Texas-Red labelled vessels; Blue, Collagen. Arrows, Representative TOVA cells. Graph, Corresponding quantification of TOVA cells in the extravascular environment. Each dot represents a single 30 min video. Mean ± SEM. Representative data of two independent experiments. n = 8 (both groups). f, I.tr. neomycin treatment does not change the expression of tight junction molecules and integrin ligands. Expression of the indicated genes in endothelial cells isolated from spinal cord leptomeninges and parenchyma of rats pre-treated for 7 days with PBS or neomycin. Quantitative PCR. Housekeeping gene: β-actin. Mean ± s.e.m. Cumulative data of 4 independent experiments. n = 17-18 (all groups). g, I.tr. neomycin treatment does not alter the permeability of leptomeningeal vessels. Intravital TPLSM overviews and corresponding magnified pictures of the thoracic spinal cord recorded 7 days after i.tr. PBS or neomycin treatment. Images were acquired 0, 30, 60 and 90 min after i.v. injection of 3 kDa Texas Red Dextran. No leakage of the dye was observed at any time point. Representative images from two independent experiments. a, e, f, Statistical significance determined by unpaired two-tailed t-test (Gaussian distribution) and Mann–Whitney test (non-Gaussian distribution). c, d, Statistical significance of percentage crawling versus rolling was determined with a two-way ANOVA; for the other motility parameters unpaired two-tailed t-test (Gaussian distribution) and Mann–Whitney test (non-Gaussian distribution) were used.

Source data

Extended Data Fig. 6 Lung dysbiosis induces quantitative but not qualitative changes in the immune infiltrates in the CNS.

a, I.tr. neomycin treatment reduces TMBP cell-mediated CNS inflammation. Expression of iNOS (Nos2), MHC-II (Rt1ba), chemokines, proinflammatory cytokines and regulatory genes in total spinal cord at the initiation stage of the EAE (i.e. 24 h after the onset of the clinical symptoms). Quantitative PCR. Housekeeping gene: β-actin. Mean ± s.e.m. n = 11 (PBS); n = 9 (Neo). b, c, T cell responses in the CNS are not affected by lung dysbiosis in transfer EAE. EAE was induced in rats pre-treated i.tr. with neomycin or PBS by transfer of TMBP cells. Immune cell characterization was performed at the initiation stage of the disease. b, Percentage of TMBP cells, endogenous αβTCR+ CD4+ and CD8+ T cells expressing FoxP3 and proinflammatory cytokines (steady state and PMA/I-stimulated conditions) in the CNS and in the indicated peripheral compartments. Flow cytometry. Mean ± s.e.m. Representative data of 2 independent experiments. n = 5 (all groups). c, Corresponding expression of the specified T cell lineage signature cytokines in the indicated T cell subsets isolated from the spinal cord. Quantitative PCR. Housekeeping gene: β-actin. Mean ± s.e.m. n = 2–4 (all groups). d, Myeloid cells recruited to the CNS are not impaired by lung dysbiosis. Expression of iNOS (Nos2), MHC-II (Rt1ba), chemokines, M2 macrophage marker and IFNβ in spleen-derived CD45high CD11b+ MΦ and in CNS-derived CD45high CD11b+ MΦ at the initiation stage of EAE. Quantitative PCR. Housekeeping gene: β-actin. Mean ± s.e.m. Representative data of 2 independent experiments. n = 3-4 (PBS); n = 4 (Neo) per condition. e, T cell response in the CNS is not affected by lung dysbiosis also in lung active EAE. Lung EAE was induced in rats pre-treated i.tr. with neomycin or PBS. Immune cell analysis was performed at the initiation stage of the disease. e,  Percentage of TMBP cells, endogenous αβTCR+ CD4+ and CD8+ T cells expressing FoxP3 and proinflammatory cytokines (steady state and PMA/I-stimulated conditions) in the CNS and in the indicated peripheral compartments. Flow cytometry. Mean ± s.e.m. Representative data of 2 independent experiments. n = 5 (all groups). f, Corresponding expression of the specified T cell lineage-signature cytokines measured as in c. Quantitative PCR. Mean ± s.e.m. n = 2–5 (all groups). g, Myeloid cells recruited to the CNS are not functionally impaired. Expression of the indicated genes measured as in d in spleen-derived CD45high CD11b+ MΦ and in CNS-derived CD45high CD11b+ MΦ. Quantitative PCR. Mean ± s.e.m. Representative data of 2 independent experiments. n = 2–5 (all groups). a–g, Statistical significance was determined by unpaired two-tailed t-test (Gaussian distribution) and Mann–Whitney test (non-Gaussian distribution). *P < 0.05, ***P < 0.001; ND, not detected.

Source data

Extended Data Fig. 7 Lung dysbiosis induces changes in microglial morphology and expression profile.

a, I.tr. neomycin treatment dampens microglial response in EAE. EAE was induced in rats pre-treated i.tr. with neomycin or PBS by transfer of TMBP cells. Confocal images acquired in spinal cord white matter and grey matter depicting the morphology of Iba1+ microglia (red) at the peak of EAE. b, c, PLX3397 ameliorates EAE but does not add to the disease-ameliorating effects of i.tr. neomycin treatment. b, Quantification of CD45low CD11b+ microglia in the brain after 7 days of oral treatment with PLX3397 or vehicle. Flow cytometry. Mean ± s.e.m. Cumulative data of 3 independent experiments. n = 9 (both groups). c, Rats were treated orally with PLX3397 or vehicle and i.tr. with neomycin or PBS. After 7 days EAE was induced by transfer of TMBP cells. The treatments were continued throughout the entire disease course. Clinical parameters. Mean ± s.e.m. Cumulative data of 2 independent experiments. n = 10 (vehicle and PBS, vehicle and Neo); n = 9 (PLX and PBS); n = 11 (PLX and Neo). d, I.tr. neomycin treatment does not induce quantitative microglia changes. Rats were treated i.tr. with PBS or neomycin for 7 days. Histological quantification of Iba1+ microglia in the grey matter of the spinal cord (n = 5 (PBS); n = 6 (Neo)), and cytofluorometric quantification of CD45low CD11b+ microglia in the spinal cord (n = 10 (PBS); n = 9 (Neo)). Mean ± s.e.m. e, f, I.tr. neomycin treatment changes microglia morphology in spinal cord and brain cortex without EAE induction. Rats were treated i.tr. with PBS or neomycin for 7 days. e, Quantification of the indicated morphological parameters extracted from confocal images of microglia in the spinal cord of PBS- or neomycin-pre-treated rats. Mean ± s.e.m. 16 cells from 3 different rats per group were analysed. f, Iba1+ microglia in cortical grey matter after 7 days of i.tr. treatment with PBS or neomycin. Representative confocal 3D-reconstructions and corresponding morphological parameters derived from 13 cells from 3 different rats per group. Mean ± s.e.m. g, I.tr. neomycin treatment induces a type I IFN signature in spinal cord microglia. Significantly enriched (P < 0.05) GO terms belonging to biological processes (BP) in genes upregulated in microglia of rats treated with neomycin compared to PBS. h, I.tr. neomycin treatment induces type I IFN-stimulated gene expression in brain-derived microglia. Differential expression of the indicated genes in microglial cells sorted from the brain of rats pre-treated with PBS or neomycin. Representative data of 2 independent experiments. Quantitative PCR. Housekeeping gene: β-actin. Mean ± s.e.m. n = 5 (both groups). i, I.tr. neomycin treatment induces upregulation of type I IFN-stimulated genes in the total spinal cord. Comparison of differential gene expression between neomycin- and PBS-treated rats. Light red dots, genes significantly upregulated (P < 0.05) but below the 0.5-fold change cut-off. Type I IFN-regulated genes are indicated. Bold, genes upregulated in both spinal cord and sorted microglia (Fig. 4b). j, Lung dysbiosis does not induce a shift to a type I IFN profile in astrocytes. Expression of type I IFN-regulated genes, β2MG (B2m), MHC-II (Rt1ba), and TNF. Note that no signal was detectable in most of the samples. Cumulative data of 2 independent experiments. n = 6 (PBS); n = 7 (Neo) per condition. b, d, e, h, j, Statistical significance determined by unpaired two-tailed t-test (Gaussian distribution) and Mann–Whitney test (non-Gaussian distribution). c, Statistical significance determined by one-way ANOVA with Tukey’s multiple comparisons test. *P < 0.05, **P < 0.01, ***P < 0.001; ND, not detected.

Source data

Extended Data Fig. 8 Effects of lung dysbiosis on microglia and the lung milieu.

a, I.tr. neomycin treatment reduces microglial reactivity towards inflammatory cytokines in the CNS. Rats were i.tr. treated daily with neomycin or PBS. After 7 days, PBS or TNF & IFNγ were administered intrathecally (i.th.). Percentage of MHC-II+ CD45low CD11b+ microglia and number of CD45high CD11b+ MΦ in spinal cord and brain 4 h and 18 h after i.th. injection. Flow cytometry. Mean ± s.e.m. Representative data from 3 independent experiments. For each CNS compartment, n = 4 (PBS and PBS); n = 5 (PBS and TNF & IFNγ); n = 3 (Neo and PBS); n = 5 (Neo and TNF & IFNγ). b, Lung dysbiosis impairs the capacity of microglia to respond to proinflammatory stimuli in vitro. Microglial cells, isolated from rats treated for 7 days with PBS or neomycin, were stimulated in vitro with increasing doses of IFNγ. Expression of chemokines, cytokines β2MG (B2m), MHC-II (Rt1ba) and iNOS (Nos2) 4h after stimulation. Quantitative PCR. Representative data of 2 independent experiments. Each value represents the pooled microglia of at least 6 rats per group. c, Oral administration of inactivated P. melaninogenica does not affect EAE. EAE was induced by transfer of TMBP cells. Clinical parameters after daily oral treatment started 7 days before TMBP cell transfer and continued throughout the entire disease course. Mean ± s.e.m. Cumulative data of 2 independent experiments. n = 5 (PBS); n = 4 (P. melaninogenica). d, Neomycin treatment increases pulmonary LPS. Concentration of LPS in BALF of rats treated for 7 days with PBS, neomycin or vancomycin (Vanco). ELISA. Mean ± s.e.m. Cumulative data of 2 independent experiments. n = 9 (PBS); n = 6 (Vanco); n = 8 (Neo). e, Neomycin treatment induces a shift to a type I IFN phenotype in pulmonary immune cells. Expression of type I IFN-stimulated genes in pulmonary stromal cells (CD45−) and immune cell subsets in PBS- or neomycin-pre-treated rats. Quantitative PCR. Housekeeping gene: β-actin. Mean ± s.e.m. Cumulative data of 3 independent experiments. n = 4–13 (PBS); n = 4–11 (Neo) per condition. a, c, e, Statistical significance determined by unpaired two-tailed t-test (Gaussian distribution) and Mann–Whitney test (non-Gaussian distribution). d, Statistical significance determined by one-way ANOVA with Tukey’s multiple comparisons test. *P < 0.05, **P < 0.01, ***P < 0.001.

Source data

Extended Data Fig. 9 Vancomycin does not induce a shift towards LPS and does not affect CNS autoimmunity.

a, b, Vancomycin does not increase LPS producing phyla in the lung microbiota. a, Average relative abundance of bacterial phyla of lung microbiota in rats treated with PBS or vancomycin for 7 days. b, Corresponding heat map depicting the most regulated inhabitants of lung microbiota at family level. c, d, I.tr. vancomycin treatment does not affect the microglial gene expression profile. c, Volcano plots depicting the differential expression profile between vancomycin- and PBS-pre-treated rats in spinal cord derived CD45low CD11b+ microglia or in total spinal cord. d, Expression of type I IFN-regulated genes in CD45low CD11b+ microglial cells isolated from the spinal cord of rats pre-treated i.tr. for 7 days with PBS or vancomycin. Quantitative PCR. Please note that the experiment was performed in parallel with the one depicted in Fig. 4c and therefore the values in the PBS group are the same. Housekeeping gene: β-actin. Mean ± s.e.m. Cumulative data of 3 independent experiments. n = 5–12 (PBS); n = 5–11 (Vanco) per condition. e, f, I.tr. vancomycin treatment does not affect transfer EAE or EAE induced via the lung. e, Transfer EAE was induced in rats pre-treated i.tr. with PBS or vancomycin for 7 days. Clinical parameters. Mean ± s.e.m. Representative data of 3 independent experiments. n = 6 (both groups). f, Lung EAE was induced in rats pre-treated i.tr. with PBS or vancomycin for 7 days. Clinical parameters. Mean ± s.e.m. Representative data of 3 independent experiments. n = 3 (both groups). g, Oral treatment with vancomycin does not affect EAE. Transfer EAE was induced in rats pre-treated orally with PBS or vancomycin for 7 days. Clinical parameters. Mean ± s.e.m. Cumulative data of 2 independent experiments. n = 8 (PBS); n = 8 (1 mg Vanco); n = 7 (10 mg Vanco). d–f, Statistical significance determined by unpaired two-tailed t-test (Gaussian distribution) and Mann–Whitney test (non-Gaussian distribution). g, Statistical significance determined by one-way ANOVA with Tukey’s multiple comparisons test. *P < 0.05, **P < 0.01, ***P < 0.001.

Source data

Extended Data Fig. 10 LPS regulates EAE severity.

a, I.tr. treatment with polymyxin B aggravates EAE. Transfer EAE was induced in rats pre-treated i.tr. for 7 days with polymyxin B or PBS. Clinical parameters. Mean ± s.e.m. Cumulative data of 3 independent experiments. n = 11 (PBS); n = 14 (polymyxin B). b, I.tr. treatment with E. coli LPS ameliorates EAE. Transfer EAE was induced in rats pre-treated daily i.tr. for 7 days with LPS or PBS. The treatment was continued throughout the entire disease course. Clinical parameters. Mean ± s.e.m. Cumulative data of 2 independent experiments. n = 8 (both groups). c, I.th.  E. coli LPS administration ameliorates EAE. EAE was induced by transfer of TMBP cells. LPS was administered on D0, D2 and D4 after transfer. Clinical parameters. Mean ± s.e.m. Representative data of 2 independent experiments. n = 5 (both groups). d, Graphical abstract: Lung microbiota controls the immune reactivity of the CNS in steady state condition and in the case of autoimmunity. Created with BioRender.com. a–c, Statistical significance determined by unpaired two-tailed t-test (Gaussian distribution) and Mann–Whitney test (non-Gaussian distribution). *P < 0.05, **P < 0.01, ***P < 0.001.

Source data

Supplementary information

Supplementary Fig. 1

Flow cytometry gating strategies for the characterization of immune and resident cells. a, Stromal and myeloid cells in lung EAE on D1 after immunization. b, IFNγ, IL-17 and FOXP3 expression in TMBP cells, endogenous αβTCR+ CD4+ and CD8+ T cells in the spinal cord on D5 after transfer of TMBP cells. Proinflammatory cytokines are depicted in steady state condition (−PMA/I) and after stimulation with PMA and ionomycin (+PMA/I). c, Microglia and MΦ (comprising both recruited blood derived monocytes and resident macrophages) in the spinal cord of rats pre-treated with neomycin (D0) and on D5 after TMBP cell transfer. d, Myeloid cells and astrocytes in the spinal cord of rats pre-treated with neomycin (D0). e, Endothelial cells isolated from spinal cord meninges of rats pre-treated with neomycin. Numbers indicate the percentage fraction of each cell population.

Reporting Summary

Supplementary Table 1

Genes differentially regulated in microglial cells isolated from PBS- or neomycin-treated rats. RNA-seq data. P values were extracted from Wald test statistics generated for differential gene expression analysis by the DESeq2 package for R. Multiple test correction was done with the FDR/Benjamini–Hochberg method.

Supplementary Video 1

Lung dysbiosis does not affect TMBP cell motility at the CNS borders. Locomotion behaviour of TMBP cells in the leptomeningeal vessels of spinal cord of rats pre-treated with PBS or neomycin. D2.5 post transfer. Intravital TPLSM. 30 min time-lapse recordings and corresponding time-projections. Scale bar, 50 µm. Time interval, 15 sec. Turquoise, TMBP cells; Red, 70 kDa Texas Red Dextran labelled vessels.

Supplementary Video 2

Lung dysbiosis does not affect CNS-ignorant TOVA cell motility at the CNS borders. Intravascular and extravascular locomotion behaviour of TOVA cells in the leptomeninges recorded 3 or 4 days post transfer, respectively in rats pre-treated with PBS or neomycin. Intravital TPLSM recordings. 30 min time-lapse recordings and corresponding time-projections. Scale bar, 50 µm. Time interval, 15 sec. Turquoise, TOVA cells; Red, 70 kDa Texas Red Dextran labelled vessels.

Supplementary Video 3

Lung dysbiosis leads to morphological changes in brain and spinal cord microglia. 3D reconstructions of Iba1+ microglial cells from cortical grey matter and lumbar spinal cord of rats pre-treated with PBS or neomycin. Confocal microscopy. Scheme of the experimental design created with BioRender.com.

Source data

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Hosang, L., Canals, R.C., van der Flier, F.J. et al. The lung microbiome regulates brain autoimmunity. Nature 603, 138–144 (2022). https://doi.org/10.1038/s41586-022-04427-4

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1038/s41586-022-04427-4

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

Advanced search

Quick links

Nature Briefing

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

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