Abstract

There is a growing appreciation for the importance of the gut microbiota as a therapeutic target in various diseases. However, there are only a handful of known commensal strains that can potentially be used to manipulate host physiological functions. Here we isolate a consortium of 11 bacterial strains from healthy human donor faeces that is capable of robustly inducing interferon-γ-producing CD8 T cells in the intestine. These 11 strains act together to mediate the induction without causing inflammation in a manner that is dependent on CD103+ dendritic cells and major histocompatibility (MHC) class Ia molecules. Colonization of mice with the 11-strain mixture enhances both host resistance against Listeria monocytogenes infection and the therapeutic efficacy of immune checkpoint inhibitors in syngeneic tumour models. The 11 strains primarily represent rare, low-abundance components of the human microbiome, and thus have great potential as broadly effective biotherapeutics.

Access optionsAccess options

Rent or Buy article

Get time limited or full article access on ReadCube.

from$8.99

All prices are NET prices.

Data availability

Genomic sequences of the 21 strains are deposited in the European Nucleotide Archive with accession number PRJEB29940. Microbiome data are deposited in the DNA Data Bank of Japan with accession number DRA007611. Accession numbers for metagenomics datasets: NCBI BioProjects PRJNA275349 (HMP1-2)38, PRJNA48479 (HMP1-2), PRJNA398089 (HMP2)39, PRJNA319574 (500FG)40, PRJDB3601 (healthy Japanese adults)34 and PRJNA399742 (Gajewski)31. EBI study accession EGAS00001001704 (LLDeep)41, PRJEB22894 (Wargo)30, PRJEB22863 (Zitvogel)29. EBI ENA ERP002061 (MetaHIT), ERP003612 (MetaHIT) and ERP004605 (MetaHIT)42,43.

Additional information

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

References

  1. 1.

    Mimee, M., Citorik, R. J. & Lu, T. K. Microbiome therapeutics — advances and challenges. Adv. Drug Deliv. Rev. 105, 44–54 (2016).

  2. 2.

    Kim, S., Covington, A. & Pamer, E. G. The intestinal microbiota: antibiotics, colonization resistance, and enteric pathogens. Immunol. Rev. 279, 90–105 (2017).

  3. 3.

    El Hage, R., Hernandez-Sanabria, E. & Van de Wiele, T. Emerging trends in “smart probiotics”: functional consideration for the development of novel health and industrial applications. Front. Microbiol. 8, 1889 (2017).

  4. 4.

    Blander, J. M., Longman, R. S., Iliev, I. D., Sonnenberg, G. F. & Artis, D. Regulation of inflammation by microbiota interactions with the host. Nat. Immunol. 18, 851–860 (2017).

  5. 5.

    Atarashi, K. et al. Treg induction by a rationally selected mixture of Clostridia strains from the human microbiota. Nature 500, 232–236 (2013).

  6. 6.

    Faith, J. J., Ahern, P. P., Ridaura, V. K., Cheng, J. & Gordon, J. I. Identifying gut microbe–host phenotype relationships using combinatorial communities in gnotobiotic mice. Sci. Transl. Med. 6, 220ra11 (2014).

  7. 7.

    Sefik, E. et al. Individual intestinal symbionts induce a distinct population of RORγ+ regulatory T cells. Science 349, 993–997 (2015).

  8. 8.

    Atarashi, K. et al. Th17 cell induction by adhesion of microbes to intestinal epithelial cells. Cell 163, 367–380 (2015).

  9. 9.

    Tan, T. G. et al. Identifying species of symbiont bacteria from the human gut that, alone, can induce intestinal Th17 cells in mice. Proc. Natl Acad. Sci. USA 113, E8141–E8150 (2016).

  10. 10.

    Ichinohe, T. et al. Microbiota regulates immune defense against respiratory tract influenza A virus infection. Proc. Natl Acad. Sci. USA 108, 5354–5359 (2011).

  11. 11.

    Thackray, L. B. et al. Oral antibiotic treatment of mice exacerbates the disease severity of multiple flavivirus infections. Cell Rep. 22, 3440–3453 (2018).

  12. 12.

    Spranger, S., Sivan, A., Corrales, L. & Gajewski, T. F. Tumor and host factors controlling antitumor immunity and efficacy of cancer immunotherapy. Adv. Immunol. 130, 75–93 (2016).

  13. 13.

    Bhatt, A. P., Redinbo, M. R. & Bultman, S. J. The role of the microbiome in cancer development and therapy. CA Cancer J. Clin. 67, 326–344 (2017).

  14. 14.

    Zitvogel, L., Ma, Y., Raoult, D., Kroemer, G. & Gajewski, T. F. The microbiome in cancer immunotherapy: diagnostic tools and therapeutic strategies. Science 359, 1366–1370 (2018).

  15. 15.

    Park, C. O. & Kupper, T. S. The emerging role of resident memory T cells in protective immunity and inflammatory disease. Nat. Med. 21, 688–697 (2015).

  16. 16.

    Atarashi, K. et al. Ectopic colonization of oral bacteria in the intestine drives TH1 cell induction and inflammation. Science 358, 359–365 (2017).

  17. 17.

    Hildner, K. et al. Batf3 deficiency reveals a critical role for CD8α+ dendritic cells in cytotoxic T cell immunity. Science 322, 1097–1100 (2008).

  18. 18.

    Ohta, T. et al. Crucial roles of XCR1-expressing dendritic cells and the XCR1–XCL1 chemokine axis in intestinal immune homeostasis. Sci. Rep. 6, 23505 (2016).

  19. 19.

    Joeris, T., Müller-Luda, K., Agace, W. W. & Mowat, A. M. Diversity and functions of intestinal mononuclear phagocytes. Mucosal Immunol. 10, 845–864 (2017).

  20. 20.

    Satpathy, A. T. et al. Notch2-dependent classical dendritic cells orchestrate intestinal immunity to attaching-and-effacing bacterial pathogens. Nat. Immunol. 14, 937–948 (2013).

  21. 21.

    Sheridan, B. S. et al. Oral infection drives a distinct population of intestinal resident memory CD8+ T cells with enhanced protective function. Immunity 40, 747–757 (2014).

  22. 22.

    Foulds, K. E. et al. Cutting edge: CD4 and CD8 T cells are intrinsically different in their proliferative responses. J. Immunol. 168, 1528–1532 (2002).

  23. 23.

    Gao, J. et al. Loss of IFN-γ pathway genes in tumor cells as a mechanism of resistance to anti-CTLA-4 therapy. Cell 167, 397–404 (2016).

  24. 24.

    Wei, S. C. et al. Distinct cellular mechanisms underlie anti-CTLA-4 and anti-PD-1 checkpoint blockade. Cell 170, 1120–1133 (2017).

  25. 25.

    Yang, J. C. & Perry-Lalley, D. The envelope protein of an endogenous murine retrovirus is a tumor-associated T-cell antigen for multiple murine tumors. J. Immunother. 23, 177–183 (2000).

  26. 26.

    Cooper, Z. A. et al. Response to BRAF inhibition in melanoma is enhanced when combined with immune checkpoint blockade. Cancer Immunol. Res. 2, 643–654 (2014).

  27. 27.

    Dankort, D. et al. Braf V600E cooperates with Pten loss to induce metastatic melanoma. Nat. Genet. 41, 544–552 (2009).

  28. 28.

    Dubin, K. et al. Intestinal microbiome analyses identify melanoma patients at risk for checkpoint-blockade-induced colitis. Nat. Commun. 7, 10391 (2016).

  29. 29.

    Routy, B. et al. Gut microbiome influences efficacy of PD-1-based immunotherapy against epithelial tumors. Science 359, 91–97 (2018).

  30. 30.

    Gopalakrishnan, V. et al. Gut microbiome modulates response to anti-PD-1 immunotherapy in melanoma patients. Science 359, 97–103 (2018).

  31. 31.

    Matson, V. et al. The commensal microbiome is associated with anti-PD-1 efficacy in metastatic melanoma patients. Science 359, 104–108 (2018).

  32. 32.

    Frankel, A. E. et al. Metagenomic shotgun sequencing and unbiased metabolomic profiling identify specific human gut microbiota and metabolites associated with immune checkpoint therapy efficacy in melanoma patients. Neoplasia 19, 848–855 (2017).

  33. 33.

    Johansson, M. E. et al. The inner of the two Muc2 mucin-dependent mucus layers in colon is devoid of bacteria. Proc. Natl Acad. Sci. USA 105, 15064–15069 (2008).

  34. 34.

    Nishijima, S. et al. The gut microbiome of healthy Japanese and its microbial and functional uniqueness. DNA Res 23, 125–133 (2016).

  35. 35.

    Miyajima, M. et al. Metabolic shift induced by systemic activation of T cells in PD-1-deficient mice perturbs brain monoamines and emotional behavior. Nat. Immunol. 18, 1342–1352 (2017).

  36. 36.

    Hu, S. et al. Targeted metabolomic analysis of head and neck cancer cells using high performance ion chromatography coupled with a Q exactive HF mass spectrometer. Anal. Chem. 87, 6371–6379 (2015).

  37. 37.

    Oka, M. et al. Arl8b is required for lysosomal degradation of maternal proteins in the visceral yolk sac endoderm of mouse embryos. J. Cell Sci. 130, 3568–3577 (2017).

  38. 38.

    Lloyd-Price, J. et al. Strains, functions and dynamics in the expanded Human Microbiome Project. Nature 550, 61–66 (2017).

  39. 39.

    The Integrative HMP (iHMP) Research Network Consortium. The Integrative Human Microbiome Project: dynamic analysis of microbiome-host omics profiles during periods of human health and disease. Cell Host Microbe 16, 276–289 (2014).

  40. 40.

    Schirmer, M. et al. Linking the human gut microbiome to inflammatory cytokine production capacity. Cell 167, 1897 (2016).

  41. 41.

    Zhernakova, A. et al. Population-based metagenomics analysis reveals markers for gut microbiome composition and diversity. Science 352, 565–569 (2016).

  42. 42.

    Qin, J. et al. A human gut microbial gene catalogue established by metagenomic sequencing. Nature 464, 59–65 (2010).

  43. 43.

    Li, J. et al. An integrated catalog of reference genes in the human gut microbiome. Nat. Biotechnol. 32, 834–841 (2014).

Download references

Acknowledgements

K.H. was funded through AMED LEAP under grant number JP18gm0010003, the Takeda Science Foundation, and the Mitsubishi Foundation. T.T. was funded through AMED PRIME under grant number JP18gm6010013 and the Nakajima Foundation. R.J.X. was funded through NIH DK43351, AT009708, and the Center for Microbiome Informatics and Therapeutics, MIT. The LC–MS platform used in this study was provided by the JST ERATO Gas Biology Project, which was led by M.S. until March 2015. We thank N. Palm, J. Faith and J. S. Weber for discussion, J. Baginska and O. Ohara for their technical support, P. Burrows for comments, and RIKEN BRC and the International Mouse Phenotyping Consortium for generating and providing the H2-M3 knockout mice.

Reviewer information

Nature thanks S. Mazmanian and the anonymous reviewer(s) for their contribution to the peer review of this work.

Author information

Author notes

  1. These authors contributed equally: Takeshi Tanoue, Satoru Morita

Affiliations

  1. Department of Microbiology and Immunology, Keio University School of Medicine, Tokyo, Japan

    • Takeshi Tanoue
    • , Satoru Morita
    • , Ashwin N. Skelly
    • , Seiko Narushima
    • , Kayoko Sugita
    • , Atsushi Shiota
    • , Kozue Takeshita
    • , Keiko Yasuma-Mitobe
    • , Koji Atarashi
    •  & Kenya Honda
  2. JSR-Keio University Medical and Chemical Innovation Center, Tokyo, Japan

    • Takeshi Tanoue
    • , Atsushi Shiota
    • , Koji Atarashi
    •  & Kenya Honda
  3. RIKEN Center for Integrative Medical Sciences, Yokohama, Japan

    • Takeshi Tanoue
    • , Wataru Suda
    • , Seiko Narushima
    • , Iori Motoo
    • , Masahira Hattori
    • , Koji Atarashi
    •  & Kenya Honda
  4. Infectious Disease and Microbiome Program, Broad Institute of MIT and Harvard, Cambridge, MA, USA

    • Damian R. Plichta
    • , Hera Vlamakis
    •  & Ramnik J. Xavier
  5. Cooperative Major in Advanced Health Science, Graduate School of Advanced Science and Engineering, Waseda University, Tokyo, Japan

    • Wataru Suda
    •  & Masahira Hattori
  6. Graduate School of Frontier Sciences, The University of Tokyo, Kashiwa, Japan

    • Wataru Suda
    •  & Masahira Hattori
  7. Department of Biochemistry, Keio University School of Medicine, Tokyo, Japan

    • Yuki Sugiura
    •  & Makoto Suematsu
  8. Department of Biomedical Sciences, Nazarbayev University School of Medicine, Astana, Kazakhstan

    • Dieter Riethmacher
  9. Department of Immunology, Institute of Advanced Medicine, Wakayama Medical University, Wakayama, Japan

    • Tsuneyasu Kaisho
  10. Vedanta Biosciences, Cambridge, MA, USA

    • Jason M. Norman
    • , Bernat Olle
    •  & Bruce Roberts
  11. Laboratory of Mucosal Immunology, The Rockefeller University, New York, NY, USA

    • Daniel Mucida
  12. Division of Cellular Signaling, Institute for Advanced Medical Research, Keio University School of Medicine, Tokyo, Japan

    • Tomonori Yaguchi
    •  & Yutaka Kawakami
  13. Department of Bioengineering, Program in Biotechnology and Biomedical Engineering, University of Massachusetts Dartmouth, North Dartmouth, MA, USA

    • Vanni Bucci
  14. Marmoset Research Department, Central Institute for Experimental Animals, Kawasaki, Japan

    • Takashi Inoue
  15. Center for Computational and Integrative Biology, Massachusetts General Hospital and Harvard Medical School, Boston, MA, USA

    • Ramnik J. Xavier

Authors

  1. Search for Takeshi Tanoue in:

  2. Search for Satoru Morita in:

  3. Search for Damian R. Plichta in:

  4. Search for Ashwin N. Skelly in:

  5. Search for Wataru Suda in:

  6. Search for Yuki Sugiura in:

  7. Search for Seiko Narushima in:

  8. Search for Hera Vlamakis in:

  9. Search for Iori Motoo in:

  10. Search for Kayoko Sugita in:

  11. Search for Atsushi Shiota in:

  12. Search for Kozue Takeshita in:

  13. Search for Keiko Yasuma-Mitobe in:

  14. Search for Dieter Riethmacher in:

  15. Search for Tsuneyasu Kaisho in:

  16. Search for Jason M. Norman in:

  17. Search for Daniel Mucida in:

  18. Search for Makoto Suematsu in:

  19. Search for Tomonori Yaguchi in:

  20. Search for Vanni Bucci in:

  21. Search for Takashi Inoue in:

  22. Search for Yutaka Kawakami in:

  23. Search for Bernat Olle in:

  24. Search for Bruce Roberts in:

  25. Search for Masahira Hattori in:

  26. Search for Ramnik J. Xavier in:

  27. Search for Koji Atarashi in:

  28. Search for Kenya Honda in:

Contributions

K.H. and T.T. planned experiments, analysed data, and wrote the paper together with A.N.S., A.S., B.O. and B.R.; T.T., S.M., A.N.S., S.N., I.M., K.S., K.T., K.Y.-M, T.I. and K.A. performed immunological analyses and bacterial cultures; W.S., M.H., J.M.N., V.B., H.V., D.R.P. and R.J.X. performed bacterial sequencing and microbiome analysis; M.S. and Y.S. performed metabolomic analysis; D.M., D.R. and T.K. provided essential materials; T.Y. and Y.K. supervised tumour models.

Competing interests

B.O., B.R. and J.M.N. are employees of Vedanta Biosciences Inc. A.S. is an employee of JSR Corporation. K.H. is co-founder and scientific advisor to Vedanta Biosciences since 11 August 2011.

Corresponding author

Correspondence to Kenya Honda.

Extended data figures and tables

  1. Extended Data Fig. 1 Characterization of the intestinal IFNγ+ CD8 T cell population.

    a, b, Representative flow cytometry histograms and plots showing the expression of CD8β, CD44, T-bet, ICOS, CD103, KLRG1 and GrB by colonic lamina propria IFNγ+ CD8 T cells isolated from SPF mice from two independent experiments. c, Number of IFNγ+ CD8 T cells in the colonic or small intestinal lamina propria of germ-free and SPF mice. d, Percentage of IFNγ+ cells among CD8 T cells in the colon of SPF mice (between 8 and 12 weeks old) treated with or without AVMN (ampicillin, vancomycin, metronidazole and neomycin) via the drinking water for 4 weeks. e, Faeces from SPF mice (SPFfae) were orally inoculated into germ-free mice (7 weeks old), and 4 weeks later the percentage of IFNγ+ cells among CD8 T cells was analysed. f, Left, percentage of IFNγ+ cells among CD8 T cells in the colonic lamina propria of SPF C57BL/6 mice (10 weeks old), reared at the indicated institute or vendor. Right, SPF C57BL/6 mice (8 weeks old) obtained from Charles River were cohoused with C57BL/6 mice from CLEA for 0, 2 or 6 weeks, and the percentage of IFNγ+ cells among colonic lamina propria CD8 T cells was analysed. Each circle represents an individual animal, and the number of mice in each group is shown. Data are mean and s.d. ***P < 0.001; *P < 0.05; one-way ANOVA with Tukey’s test (c, f) or two-tailed unpaired t-test (d, e). See Source Data for exact P values. Source Data

  2. Extended Data Fig. 2 Bacteria from the healthy human gut microbiota associated with IFNγ+ CD8 T cell induction.

    a, Schematic representation of the strategy for isolating IFNγ+ CD8 T cell-inducing bacteria from the healthy human gut microbiota. b, Representative flow cytometry plots showing the expression of IFNγ and CD8α by colonic or small intestinal lamina propria CD3+TCRβ+ T cells from germ-free mice orally inoculated with healthy human faecal samples (from donors A to F). c, Spearman’s correlation coefficient and P values for the relationship between the number of individual bacterial 228 OTUs (among 3,000 reads) detected in mice shown in Fig. 1e and the percentage of colonic IFNγ+ CD8 T cell population shown in Fig. 1d was calculated using GraphPad Prism. OTUs positively and negatively correlated with the frequency of colonic IFNγ+ CD8 T cells with statistical significance (P < 0.05) are highlighted in red and dark blue, respectively. OTUs detected in GF+CHL B5 mice are marked in light and dark blue. OTUs showing no significant correlation (P > 0.05) are marked in grey. The closest species/strain and individual Rho (ρ) value for each OTU are shown. Source Data

  3. Extended Data Fig. 3 Induction of IFNγ+CD8 T cells by the 11-mix via non-inflammatory immunomodulation.

    a, Percentage of IFNγ+ cells among colonic CD8 T cells from mice of the indicated genetic background, colonized with or without the 11-mix. b, c, Representative flow cytometry histograms and plots showing the expression of CD103, PD-1, CD44, T-bet, ICOS, KLRG1 and GrB by colonic IFNγ+ CD8 T cells from germ-free mice colonized with or without the 11-mix for 4 weeks, from two independent experiments. d, Expression of the indicated pro-inflammatory genes in colonic epithelial cells from germ-free, GF+11-mix (4 weeks after colonization), and SPF mice, as determined by qPCR. e, Representative photographs of caecums and colons from germ-free, GF+11-mix (4 weeks after colonization) and SPF mice from several independent experiments. fh, Haematoxylin and eosin-stained colon photomicrographs (f), histology score (g), and percentage of IFNγ+ cells among CD8 T cells in the colonic lamina propria (h) of germ-free and GF+11-mix mice (4 weeks or 6 months after colonization). Scale bar, 50 μm. i, Percentages of IL-17A+ (TH17) and IFNγ+ (TH1) cells among CD3+TCRβ+CD4 T cells and IFNγ+ cells among CD3+TCRβ+CD4 T cells (representing primarily CD8 T cells) in the colons of gnotobiotic mice colonized with the 11 strains, a reported TH1 cell-inducing Klebsiella pneumoniae strain (Kp2H7), or 20 reported TH17 cell-inducing strains (TH17 20-mix). Each circle represents an individual animal. The number of mice in each group is shown. Data are mean and s.d. ***P < 0.001; **P < 0.01; *P < 0.05; two-tailed unpaired t-test (a) or one-way ANOVA with Tukey’s test (d, g, h, i) was used. See Source Data for exact P values. Source Data

  4. Extended Data Fig. 4 Characterization of the 11 strains.

    a, DNA was extracted from each of the 26 strains and 16S rRNA gene sequences were determined by PCR and Sanger sequencing. Genome sequencing was conducted for 21 strains using the Illumina MiSeq or the MinION nanopore sequencer. To identify the closest reference species or strain, 16S rRNA gene and genes encoding 42 ribosomal proteins predicted from the assembled draft genome of each strain were blasted to the RDP and NCBI genome databases (RefSeq genome representative), respectively. Top-hit strains were defined as those with the highest 16S rRNA sequence similarity or those with the highest ribosomal gene sequence similarity for the maximum percentage of the 42 queried genes (strains for which this is less than 37 out of 42 are listed in parentheses). Percentage similarity refers to average sequence similarity between ribosomal genes of the isolated strain and those of the top-hit reference strain. b, A phylogenetic tree was constructed by comparing 42 concatenated ribosomal gene sequences of each isolate using the MEGAv7.0 package and the neighbour-joining method with a bootstrap of 1,000 replicates. c, Relative expression of the indicated genes in colonic epithelial cells (ECs) of germ-free mice colonized with or without the 11-mix for 1 week, as determined by qPCR. d, Frequencies of TCR Vβ gene usage among IFNγ+ CD8 T cells (left) and IFNγ CD8 T cells (right) from the colons of GF (grey) or GF+11-mix (red) mice, as determined by flow cytometry. 2-way ANOVA interaction p-values: 0.0001 (IFNγ+ subset), 0.31 (IFNγ subset). e, Luminal contents from the indicated anatomical positions of the gut as well as faecal samples were collected from three GF+11-mix mice 3-4 weeks post-colonization. The relative abundance of each of the 11 strains’ DNA was determined by qPCR. f, MLNs were collected from GF+11-mix mice 1 week after colonization and bacterial DNA was extracted. The relative abundance of each of the 11 strains’ DNA was determined by qPCR. Each circle represents an individual animal (c, e) or a pool of mice (d), and the height of each bar indicates the mean, and the number of mice in each group is shown. n.d., not detected. Error bars, s.d. ***P < 0.001; **P < 0.01; *P < 0.05; two-tailed unpaired t-test. See Source Data for exact P values. Source Data

  5. Extended Data Fig. 5 Systemic effects of 11 strain colonization.

    a, Representative flow cytometry histograms showing the expression of CD103, PD-1, CD44, ICOS and KLRG1 by inguinal lymph node (iLN) or colonic IFNγ+ CD8 T cells from the indicated mice from two independent experiments. b, Frequencies of TCR Vβ gene usage among IFNγ+ CD8 T cells (left) and IFNγ CD8 T cells (right) from the inguinal lymph nodes of germ-free (grey) or GF+11-mix (red) mice, and those from the colons of GF+11-mix mice (blue). c, Principal component analysis plots of the water-soluble caecal metabolome from gnotobiotic mice colonized with the indicated bacterial mixtures. df, Heat maps depicting differentially elevated metabolites in the caecum (d, e) or serum (f) of gnotobiotic mice colonized with the indicated bacterial mixtures. Raw metabolomic data was screened for metabolites specifically increased in GF+11-mix mice only (d, e) or in GF+11-mix and GF+4-mix mice only (d) with at least fourfold enrichment as compared to GF controls and at least twofold enrichment compared to the other gnotobiotic groups, with a cut-off P value of 0.05. Overlapping metabolites between the two datasets were selected (highlighted in green), and their levels in the serum were queried (f). The most promising candidate effector molecules for the local and systemic effects observed in GF+11-mix mice were identified by considering the metabolomic data in light of the phenotypic immunomodulatory data, and are highlighted in red font (df). For a more detailed discussion of how these metabolites were chosen, see Supplementary Discussion. The area under each individual metabolite peak was normalized by dividing by the area under the peak of the corresponding internal standard. Heat map colours represent the z-score (red and blue indicate high and low abundance, respectively). Caecal contents were isolated 4 weeks after gavage unless otherwise noted (1w, 1 week; 4w, 4 weeks). 2PG, 2-phospho-d-glycerate; 3PG, 3-phospho-d-glycerate; GlcNAc 6P, N-acetyl-d-glucosamine 6-phosphate; GlcNAc 1P, N-acetyl-d-glucosamine 1-phosphate; Rib5P, ribose-5-phosphate; Ru5P, ribulose-5-phosphate; Xu5P, xylulose-5-phosphate. Each circle represents an individual mouse (b, iLN, c), a pool of mice (b, colon). The number of mice in each group is shown. Data are mean and s.d. *P < 0.05; two-tailed unpaired t-test. See Source Data for exact P values. Source Data

  6. Extended Data Fig. 6 11-mix-mediated enhancement of host protection against Listeria monocytogenes infection.

    a, Experimental design for germ-free mouse-based studies in b, c (and Fig. 3a–d), C57BL/6 germ-free mice were colonized with 11- or 10-mix, or were left uncolonized as a control. The mice were then orally infected with Lm-WT. b, Listeria CFU in the colon on day 3 after infection. c, Representative H&E staining on day 3 after infection. Scale bar, 50 μm. d, Experimental design for SPF mouse-based studies in el (and Fig. 3e–g). C57BL/6 SPF mice were treated with AVMN, then reconstituted with SPFfae by oral gavage. In the 11-mix treatment group (+11mix), initial administration of the 11-mix was done simultaneously with SPFfae, followed by repetitive oral gavage of the 11-mix alone at the indicated time points. Mice were then orally (ek) or intraperitoneally (l) treated with Lm-InlAm (e, f), Lm-WT (gj, l) or Lm-OVA (k). e, g, Representative H&E staining on days 5 (e) and 3 (g) after infection. Scale bar, 50 μm. f, l, Listeria CFU in the indicated tissues on day 5 (f) or 3 (l) after infection. h, Colonic histology score at the indicated time points after Lm-WT infection. i, j, Percentage weight change over the course of Lm-WT infection. k, Number of OVA peptide SIINFEKL-specific or general (determined by PMA+ionomycin stimulation) colonic IFNγ+ CD8 T cells induced after Lm-OVA infection, as enumerated by flow cytometry at the indicated time points. The mean of each group is represented by a line. Note that the expansion of general IFNγ+ CD8 T cells occurred as early as 3 days after infection, sooner than that of antigen-specific IFNγ+ CD8 T cells. Therefore, the 11-mix seems to elicit both general and antigen-specific induction. To deplete CD8 T cells, anti-CD8α antibody was administered intraperitoneally 1 day before the administration of 11-mix, and every 3–4 days thereafter. Each circle represents an individual animal. The number of mice in each group is shown. Data are mean and s.d. **P < 0.01; *P < 0.05; Kruskal–Wallis with Dunn’s test (b, f), two-tailed unpaired t-test (h), two-way ANOVA with Bonferroni’s test (j, k), or two-tailed Mann–Whitney test (l). See Source Data for exact P values. Source Data

  7. Extended Data Fig. 7 Characterization of IFNγ+CD8 TILs induced by the 11-mix.

    a, Experimental design for germ-free mouse-based studies in b and c (and Fig. 4a–c). C57BL/6 germ-free mice were colonized with 11- or 10-mix on day −7, or left uncolonized, and were subjected to subcutaneous implantation of MC38 adenocarcinoma cells on day 0. Anti-PD-1 monoclonal antibody was injected intraperitoneally every third day between days 3 and 9. b, Representative flow cytometry histograms showing the expression of CD103, PD-1, ICOS, CD44 and KLRG1 by colonic lamina propria or TIL IFNγ+ CD8 T cells from the indicated mice. c, Left, Frequencies of TCR Vβ gene usage among IFNγ+ CD8 TILs from GF+MC38+anti-PD-1 (grey) or GF+11-mix+MC38+anti-PD-1 (red) mice. Right, Frequencies of TCR Vβ gene usage among IFNγ+ CD8 T cells from the colons of GF+11-mix mice (blue) or the tumours of GF+11-mix+MC38+anti-PD-1 (red). Each circle represents an individual mouse (TIL) or a pool of mice (colon). The number of mice in each group is shown. Data are mean and s.d. **P < 0.01; *P < 0.05; two-tailed unpaired t-test. See Source Data for exact values. Source Data

  8. Extended Data Fig. 8 Efficacy of 11-mix in enhancing treatment of MC38 tumours in the context of a complex microbiota.

    a, Experimental design for SPF mouse-based studies in b and c (and in Fig. 4d–k). SPF mice were subjected to treatment with AVMN (from day −7 to day 2) and subcutaneous implantation of MC38 adenocarcinoma or BrafV600E Pten−/− melanoma cells on day 0. The mice were reconstituted with SPFfae on day 3. For the 11- or 10-mix treatment groups, the initial oral administration was done simultaneously with SPFfae on day 3, followed by repetitive dosing of the 11- or 10-mix alone, two or three times per week until the end of the experiment. An anti-PD-1 or anti-CTLA-4 antibody was injected intraperitoneally every third day between days 3 and 9. b, Representative photograph of excised MC38 tumours on day 23. Scale bar, 10 mm. c, H&E staining, along with histology score, of the colon on day 27 (SPF, SPF+anti-PD-1, SPF+11mix, and SPF+11mix+anti-PD-1 groups) and on day 44 (SPF+anti-CTLA-4 and SPF+11mix+anti-CTLA-4 groups). Scale bar, 50 μm. For comparison, the colonic histology score of SPF Il10−/− (colitis-prone) mice is shown. d, Experimental design for GF+donor C human microbiota-based studies in e and f. C57BL/6 germ-free mice were colonized with donor C human faecal microbiota or left uncolonized on day 0. For the 11- or 10-mix treatment groups, the initial oral administration was done simultaneously with the donor C human faecal sample on day 0, followed by repetitive dosing of the 11- or 10-mix alone two or three times per week until the end of the experiment. e, Faeces were collected at the indicated time points. The relative abundance of each of the 11 strains’ DNA was determined by qPCR. Colonization with all 11 strains was confirmed. f, Percentage of IFNγ+ CD8 T cells in the colonic lamina propria and iLNs at day 21 was enumerated by flow cytometry. g, h, C57BL/6 germ-free mice were inoculated with donor C faecal samples with or without 11- or 10-mix, following the same protocol as in d. Mice were then subjected to subcutaneous implantation of MC38 cells on day 0. Anti-PD-1 was injected intraperitoneally every third day between days 3 and 9. Experimental design is shown in g and tumour growth data of MC38 is shown in h. Each circle represents an individual animal, except in h, where each circle represents the mean. The number of mice in each group is shown. Red, grey and brown asterisks show significance versus the C+11-mix+anti-PD-1, C+anti-PD-1, and C+10-mix+anti-PD-1 groups, respectively. Data are mean and s.e.m. (e) or s.d. (all others). ***P < 0.001; **P < 0.01; *P < 0.05; one-way (f) or two-way (h) ANOVA with Tukey’s test. See Source Data for exact P values. Source Data

  9. Extended Data Fig. 9 Strain- and species-level abundance of the 11 isolates in the human gut microbiome.

    A total of 3,327 gut metagenome samples with at least 1 million quality-controlled reads across various data sets (HMP1-2, LLDeep, MetaHIT, 500FG, HMP2 (only the first time point was used), healthy Japanese adults, and three microbiome in cancer immunotherapy studies) were mapped to 1-kb regions in the 11 strains (filtered by 95% mapping identity). The mapped read counts were normalized to RPKM. a, Detection and abundance of the strain-specific marker regions. Median abundance across all marker regions was plotted against marker gene coverage. Points with high abundance at low coverage, for example, less than 75%, indicate limited marker region resolution and likely represent other strains of a species. b, Among the 11 strains, 4 were detected in 16 out of 3,327 microbiome samples evaluated. A strain was deemed detected if at least 95% of 1-kb regions in the marker region were detected. c, Abundance at the species level was calculated as the median abundance across all 1-kb regions in a genome using 3,327 metagenome samples with at least 1 million quality-controlled reads (mapped reads were filtered at 95% mapping identity). The mapped read counts were normalized to reads per RPKM. d, Abundance of the strain-specific marker regions in the faecal microbiome of the six healthy Japanese volunteers (donors A to F) was examined, as shown in a. A strain was deemed detected if at least 95% of the 1-kb regions in the marker region were detected. At our sequencing depth, three strains were detected in the donor B sample, and one in that of donor A. e, Abundance at the species-level was calculated as in c. The mapped read counts were normalized to RPKM.

  10. Extended Data Fig. 10 Antigen-specificity and TCR Vβ usage of IFNγ+CD8 T cells induced by the 11-mix differ by anatomical location.

    a, Percentage of IFNγ+ cells among CD8 T cells in either colons from GF+11-mix mice or MC38 tumours from SPF+11mix+anti-PD-1 mice, after ex vivo stimulation with the indicated antigens. b, Principal component analysis plots of TCR Vβ usage by the IFNγ+ and IFNγ- CD8 T cell subsets isolated from the indicated tissues of GF+11-mix (right) or germ-free (left) mice. For TIL data, IFNγ+ and IFNγ CD8 T cells were isolated from germ-free ± 11mix+MC38+anti-PD-1 mice. Two-way ANOVA interaction P values comparing two populations at a time are as follows. GF group: 0.006 (CLP IFNγ+ versus iLN IFNγ+), 0.50 (CLP IFNγ+ versus TIL IFNγ+), 0.20 (iLN IFNγ+ versus TIL IFNγ+), <0.0001 (CLP IFNγ versus iLN IFNγ), <0.0001 (CLP IFNγ versus TIL IFNγ), 0.0024 (iLN IFNγ versus TIL IFNγ). GF+11-mix group: <0.0001 (CLP IFNγ+ versus iLN IFNγ+), <0.0001 (CLP IFNγ+ versus TIL IFNγ+), 0.0001 (iLN IFNγ+ versus TIL IFNγ+), <0.0001 (CLP IFNγ versus iLN IFNγ), <0.0001 (CLP IFNγ versus TIL IFNγ), 0.0009 (iLN IFNγ versus TIL IFNγ). Each circle represents an individual mouse (iLN and TIL) or a pool of mice (colon). The number of mice in each group is shown. Data are mean and s.d. ***P < 0.001; **P < 0.01; one-way ANOVA with Tukey’s test. See Source Data for exact P values. Source Data

Supplementary information

  1. Supplementary Information

    This file contains the Supplementary Discussion.

  2. Reporting Summary

  3. Supplementary Tables

    This file contains Supplementary Tables 1-3.

Source data

About this article

Publication history

Received

Accepted

Published

Issue Date

DOI

https://doi.org/10.1038/s41586-019-0878-z

Further reading

Comments

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