Single-cell RNA-seq identifies a PD-1hi ILC progenitor and defines its development pathway (original) (raw)

References

  1. Eberl, G., Colonna, M., Di Santo, J. P. & McKenzie, A. N. Innate lymphoid cells. Innate lymphoid cells: a new paradigm in immunology. Science 348, aaa6566 (2015)
    Article Google Scholar
  2. Artis, D. & Spits, H. The biology of innate lymphoid cells. Nature 517, 293–301 (2015)
    Article CAS ADS Google Scholar
  3. Chea, S. et al. Single-cell gene expression analyses reveal heterogeneous responsiveness of fetal innate lymphoid progenitors to Notch signaling. Cell Reports 14, 1500–1516 (2016)
    Article CAS Google Scholar
  4. Constantinides, M. G., McDonald, B. D., Verhoef, P. A. & Bendelac, A. A committed precursor to innate lymphoid cells. Nature 508, 397–401 (2014)
    Article CAS ADS Google Scholar
  5. Klose, C. S. et al. Differentiation of type 1 ILCs from a common progenitor to all helper-like innate lymphoid cell lineages. Cell 157, 340–356 (2014)
    Article CAS Google Scholar
  6. Yang, Q. et al. TCF-1 upregulation identifies early innate lymphoid progenitors in the bone marrow. Nat. Immunol. 16, 1044–1050 (2015)
    Article CAS Google Scholar
  7. Yu, X. et al. The basic leucine zipper transcription factor NFIL3 directs the development of a common innate lymphoid cell precursor. eLife 3, (2014)
  8. Ishizuka, I. E. et al. Single-cell analysis defines the divergence between the innate lymphoid cell lineage and lymphoid tissue-inducer cell lineage. Nat. Immunol. 17, 269–276 (2016)
    Article CAS Google Scholar
  9. Spits, H. et al. Innate lymphoid cells—a proposal for uniform nomenclature. Nat. Rev. Immunol. 13, 145–149 (2013)
    Article CAS Google Scholar
  10. Serafini, N., Vosshenrich, C. A. & Di Santo, J. P. Transcriptional regulation of innate lymphoid cell fate. Nat. Rev. Immunol. 15, 415–428 (2015)
    Article CAS Google Scholar
  11. Brennecke, P. et al. Accounting for technical noise in single-cell RNA-seq experiments. Nat. Methods 10, 1093–1095 (2013)
    Article CAS Google Scholar
  12. Yu, Y. et al. Bcl11a is essential for lymphoid development and negatively regulates p53. J. Exp. Med. 209, 2467–2483 (2012)
    Article CAS Google Scholar
  13. Xu, W. et al. NFIL3 orchestrates the emergence of common helper innate lymphoid cell precursors. Cell Reports 10, 2043–2054 (2015)
    Article CAS Google Scholar
  14. Seillet, C. et al. Nfil3 is required for the development of all innate lymphoid cell subsets. J. Exp. Med. 211, 1733–1740 (2014)
    Article CAS Google Scholar
  15. Seehus, C. R. et al. The development of innate lymphoid cells requires TOX-dependent generation of a common innate lymphoid cell progenitor. Nat. Immunol. 16, 599–608 (2015)
    Article CAS Google Scholar
  16. Zook, E. C. et al. The ETS1 transcription factor is required for the development and cytokine-induced expansion of ILC2. J. Exp. Med. 213, 687–696 (2016)
    Article CAS Google Scholar
  17. Weber, B. N. et al. A critical role for TCF-1 in T-lineage specification and differentiation. Nature 476, 63–68 (2011)
    Article CAS Google Scholar
  18. McConnell, M. J. et al. Growth suppression by acute promyelocytic leukemia-associated protein PLZF is mediated by repression of c-myc expression. Mol. Cell. Biol. 23, 9375–9388 (2003)
    Article CAS Google Scholar
  19. Hoyler, T. et al. The transcription factor GATA-3 controls cell fate and maintenance of type 2 innate lymphoid cells. Immunity 37, 634–648 (2012)
    Article CAS Google Scholar
  20. Pardoll, D. M. The blockade of immune checkpoints in cancer immunotherapy. Nat. Rev. Cancer 12, 252–264 (2012)
    Article CAS Google Scholar
  21. Yagi, R. et al. The transcription factor GATA3 is critical for the development of all IL-7Rα-expressing innate lymphoid cells. Immunity 40, 378–388 (2014)
    Article CAS Google Scholar
  22. Yu, Y. et al. The transcription factor Bcl11b is specifically expressed in group 2 innate lymphoid cells and is essential for their development. J. Exp. Med. 212, 865–874 (2015)
    Article CAS Google Scholar
  23. Walker, J. A. et al. Bcl11b is essential for group 2 innate lymphoid cell development. J. Exp. Med. 212, 875–882 (2015)
    Article CAS Google Scholar
  24. Huang, Y. et al. IL-25-responsive, lineage-negative KLRG1hi cells are multipotential ‘inflammatory’ type 2 innate lymphoid cells. Nat. Immunol. 16, 161–169 (2015)
    Article CAS Google Scholar
  25. Sharma, P. & Allison, J. P. Immune checkpoint targeting in cancer therapy: toward combination strategies with curative potential. Cell 161, 205–214 (2015)
    Article CAS Google Scholar
  26. Agata, Y. et al. Expression of the PD-1 antigen on the surface of stimulated mouse T and B lymphocytes. Int. Immunol. 8, 765–772 (1996)
    Article CAS Google Scholar
  27. Barlow, J. L. et al. Innate IL-13-producing nuocytes arise during allergic lung inflammation and contribute to airways hyperreactivity. J. Allergy Clin. Immunol. 129, 191–198 (2012)
    Article CAS Google Scholar
  28. Kasagi, S. et al. Anti-programmed cell death 1 antibody reduces CD4+PD-1+ T cells and relieves the lupus-like nephritis of NZB/W F1 mice. J. Immunol. 184, 2337–2347 (2010)
    Article CAS Google Scholar
  29. Novey, H. S., Marchioli, L. E., Sokol, W. N. & Wells, I. D. Papain-induced asthma—physiological and immunological features. J. Allergy Clin. Immunol. 63, 98–103 (1979)
    Article CAS Google Scholar
  30. Halim, T. Y., Krauss, R. H., Sun, A. C. & Takei, F. Lung natural helper cells are a critical source of Th2 cell-type cytokines in protease allergen-induced airway inflammation. Immunity 36, 451–463 (2012)
    Article CAS Google Scholar
  31. Jackson, J. T. et al. Id2 expression delineates differential checkpoints in the genetic program of CD8α+ and CD103+ dendritic cell lineages. EMBO J. 30, 2690–2704 (2011)
    Article CAS Google Scholar
  32. Serafini, N. et al. Gata3 drives development of RORγt+ group 3 innate lymphoid cells. J. Exp. Med. 211, 199–208 (2014)
    Article CAS Google Scholar
  33. Picelli, S. et al. Full-length RNA-seq from single cells using Smart-seq2. Nat. Protocols 9, 171–181 (2014)
    Article CAS Google Scholar
  34. Dobin, A. et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics 29, 15–21 (2013)
    Article CAS Google Scholar
  35. Anders, S., Pyl, P. T. & Huber, W. HTSeq—a Python framework to work with high-throughput sequencing data. Bioinformatics 31, 166–169 (2015)
    Article CAS Google Scholar
  36. 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)
    Article Google Scholar
  37. van der Maaten, L. & Hinton, G. Visualizing data using t-SNE. J. Mach. Learn. Res. 4, 1–48 (2008)
    MATH Google Scholar
  38. Robinette, M. L. et al. Transcriptional programs define molecular characteristics of innate lymphoid cell classes and subsets. Nat. Immunol. 16, 306–317 (2015)
    Article CAS Google Scholar
  39. Subramanian, A. et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc. Natl Acad. Sci. USA 102, 15545–15550 (2005)
    Article CAS ADS Google Scholar

Download references

Acknowledgements

We thank F. Colucci and J. Di Santo for providing _Rag2_−/−_Il2rg_−/− mice. We thank the Sanger Institute RSF (J. Bussell, D. Key, A. Kirton, L. Bulman, S. Kemp, P. Green, P. Zielezinski, R. Lacey, C. Rogerson, A. Logan and G. Notley), Flow Cytometry Core Facility (B. L. Ng, J. Graham and C. Hall), Single Cell Genomic Core Facility (S. Loren and I. Bronner) and DNA sequencing pipeline (N. Smerdon) for technical assistances. We thank K. Chen, J. Pramanik and R. Miragaia for technical help. C.W. is supported by the Plan of Youth Growth from Shanghai Municipal Agricultural Committee (Hunongqingzi (2015. No. A-35)). L.L. is funded by National Natural Science Foundation of China (31370904, 81671579). G.T.B. is supported by the Australian Research Council (Future Fellowship FT110100283) and the National Health and Medical Research Council (Fellowship 10402092). A.N.J.M. is supported by the Medical Research Council (U105178805) and Wellcome Trust (100963/Z/13/Z). This work is supported by Wellcome Trust (grant number 098051) (P.L).

Author information

Author notes

  1. Yong Yu, Jason C. H. Tsang and Cui Wang: These authors contributed equally to this work

Authors and Affiliations

  1. Wellcome Trust Sanger Institute, Hinxton, CB10 1HH, Cambridge, UK
    Yong Yu, Jason C. H. Tsang, Cui Wang, Simon Clare, Juexuan Wang, Xi Chen, Cordelia Brandt, Leanne Kane, Lia S. Campos, Sarah A. Teichmann, Gordon Dougan & Pentao Liu
  2. Department of Chemical Pathology, The Chinese University of Hong Kong, Prince of Wales Hospital, Shatin, New Territories, Hong Kong SAR, China
    Jason C. H. Tsang
  3. Li Ka Shing Institute of Health Sciences, The Chinese University of Hong Kong, Shatin, New Territories, Hong Kong SAR, China
    Jason C. H. Tsang
  4. Institute of Animal Husbandry and Veterinary Science, Shanghai Academy of Agricultural Sciences, Shanghai, 201106, China
    Cui Wang
  5. Shanghai Institute of Immunology, Shanghai Jiaotong University School of Medicine, Shanghai, 200025, China
    Liming Lu
  6. The Walter and Eliza Hall Institute of Medical Research, Parkville, Melbourne, 3052, Victoria, Australia
    Gabrielle T. Belz
  7. Department of Medical Biology, University of Melbourne, Melbourne, 3010, Victoria, Australia
    Gabrielle T. Belz
  8. Medical Research Council Laboratory of Molecular Biology, Francis Crick Avenue, Cambridge Biomedical Campus, Cambridge, CB2 0QH, UK
    Andrew N. J. McKenzie
  9. European Molecular Biology Laboratory, European Bioinformatics Institute (EMBL-EBI), Wellcome Trust Genome Campus, Hinxton, CB10 1SD, Cambridge, UK
    Sarah A. Teichmann
  10. Department of Medicine, University of Cambridge, Addenbrooke’s Hospital, Hills Road, Cambridge, CB2 0SP, UK
    Gordon Dougan

Authors

  1. Yong Yu
  2. Jason C. H. Tsang
  3. Cui Wang
  4. Simon Clare
  5. Juexuan Wang
  6. Xi Chen
  7. Cordelia Brandt
  8. Leanne Kane
  9. Lia S. Campos
  10. Liming Lu
  11. Gabrielle T. Belz
  12. Andrew N. J. McKenzie
  13. Sarah A. Teichmann
  14. Gordon Dougan
  15. Pentao Liu

Contributions

Y.Y. designed research, performed experiments and analysed data. J.C.H.T. performed all of the bioinformatics analyses, C.W. performed experiments. S.C. and C.B. performed influenza infection. J.W. generated the Bcl11b tdTomato conditional knockout reporter mice. X.C. performed ChIP–PCR. L.K. did the histologic section staining. L.S.C. analysed the histologic data. L.L. contributed intellectually to the PD-1 experiments. G.T.B. and A.N.J.M. provided _Id2_GFP and Il13+/tdTomato reporter mice, respectively. S.A.T. and G.D. provided intellectual input for the experiments performed in their laboratories. Y.Y., J.C.H.T. and P. L. wrote the paper. Y. Y. and P. L. conceived the PD-1 as an ILC marker concept. P.L. supervised the research.

Corresponding author

Correspondence toPentao Liu.

Ethics declarations

Competing interests

The authors declare no competing financial interests.

Additional information

Reviewer Information

Nature thanks I. Amit and the other anonymous reviewer(s) for their contribution to the peer review of this work.

Extended data figures and tables

Extended Data Figure 1 Quality control of scRNA-seq data of ILC-progenitor-enriched bone marrow cells.

a, FACS sorting strategies of the adult bone marrow cells from wild-type or Vav_-Cre_-Bcl11b fl/fl mice. Flt3lo or IL-7Rαlo cells were included to detect more ILC progenitors. Lin−Flt3lo/−IL-7Rαlo/+α4β7+ cells were further divided into three populations (CD244+CD25−, CD244−CD25− and CD244−CD25+). We sorted two 96-well plates of CD244+CD25−, one plate of CD244−CD25− and one plate of ILC2 progenitors CD244−CD25+ to include most ILC progenitors for scRNA-seq. Two 96-well plates of Lin−Flt3−IL-7Rα+α4β7+ bone marrow cells from _Vav_-Cre-Bcl11b fl/fl mice were purified to investigate early ILC2 development defects. Lin: CD19, CD3, CD4, CD5, CD8, TCRβ, TCRγδ, NK1.1, CD11b, Gr-1, CD11c and Ter119. b, Column charts show the fraction of cells passing specific quality control criteria in each plate: unique count mapped to annotated genes >500,000 (top panel); count mapped to mitochondrial-encoded genes <10% (middle panel); and number of annotated gene detected >2,500 (bottom panel). c, Percentage of cells passing all criteria. d, The percentages of ERCC RNA spike-ins in each plate. The black dots represent the mean of the dataset. e, The total number of unique counts mapped to annotated genes in different plates. The black dots represent the mean of the dataset. f, The fractions of the 92 external ERCC RNA spike-ins in different plates. g, Kolmogorov–Smirnov test of individual ERCC spike-ins between the two plates did not detect any ERCC spike-ins showing significantly different (log2 fold change > 1, and adjusted P value < 0.05) levels. The two vertical lines mark the log2 fold change levels of −1 and 1, the horizontal line marks the adjusted P value threshold of 0.05. h, Identification of highly variable genes. Brown points represent annotated mouse genes. Blue points represent external ERCC RNA spike-ins. The magenta points represent the mouse genes that show significantly higher variability (false discovery rate < 0.1). The solid line represents the fit of the technical noise, the dashed line represents the 50% biological CV (coefficient of variation). i, Biaxial t-SNE clustering of the sequenced wild-type cells. j, Column chart comparing the percentage of cells which show detectable mRNA expression of lineage markers in the wild-type bone marrow cells.

Extended Data Figure 3 Violin plots of genes highly enriched in specific clusters.

The y axis indicates the log2 (normalized count + 1) expression levels. The black point indicates the mean of expression level. The x axis indicates different clusters.

Extended Data Figure 4 Heat map of TCR transcripts in different cell clusters.

Extended Data Figure 5 PD-1hi expression marks ILC progenitors.

a, Correlation of expression levels of Pdcd1 and Zbtb16 in C6 cells. Correlation was calculated by Pearson’s method. The fit represents the line of linear regression. b, Violin plot showing the selected gene expression in PD-1hi cells of C6. The y axis indicates the log2 (normalized count + 1) expression levels. The black point indicates the mean of expression level. c, Expression of ILC markers in PD-1hi cells was analysed by FACS. d, The in vivo developmental potential of PD-1hi cells. CD45.1 _Rag2_−/−_Il2rg_−/− recipients were injected with the equal numbers of CD45.1−CD45.2+ PD-1hi cells and CD45.1+CD45.2+ (F1 of CD45.1 and CD45.2 parents) common lymphoid progenitors (200–800 cells). The progenies of these donor cells were analysed by FACS after 5–7 weeks (n = 3 per donor cell type). e, Clonal analysis of PD-1hi cells in vitro. The PD-1hi, PD-1hiBcl11b+ and PD-1hiBcl11b− bone marrow cells were FACS-purified and cultured on stromal cells and analysed by FACS. ILC1 was defined as CD45+NK1.1+Bcl11b−, ILC2 as CD45+NK1.1−Bcl11b+ and ILC3 as CD45+NK1.1−Bcl11b−RORγt+. Data are representatives of two (c) or three (d) independent experiments.

Extended Data Figure 6 Direct comparison of PD-1hi and PLZFhi ILC progenitors and dissection of the heterogeneity in the ILC progenitor compartments.

a, FACS plots show PLZFhi and PD-1hi cells had the same development potential in vivo. The equal numbers of PD-1hi cells and PLZFhi cells were adoptively transferred into the same recipient and analysed 3–4 weeks later (n = 3 per donor cell type). b, Schematic diagram of the Bcl11b tdTomato conditional knockout reporter allele, where the _lox_P-IRES-tdTomato cassette was inserted to the 3′UTR of the Bcl11b locus. The other _lox_P site was in intron 3. Cre-loxP recombination would delete the exon 4. The selection cassette for initial gene targeting was excised by Flpase-FRT recombination. c, Expression of Bcl11b in CHILPs were analysed in the Id2 GFP;Bcl11b tdTomato duel reporter mice (n = 6). d, Expression of Bcl11b in PD-1hi bone marrow cells was analysed by FACS (n = 6). e, FACS analysis of the in vivo developmental potential of PD-1hiBcl11b− cells (n = 3 per donor cell type). Common lymphoid progenitors were used as the donor cell control. PD-1hiBcl11b− cells predominantly generated ILC1, ILC2 and ILC3. Data are representatives of two (a, e) or three (c, d) independent experiments.

Extended Data Figure 7 scRNA-seq analysis of PD-1hi bone marrow cells.

a, t-SNE clustering analysis of sequenced PD-1hi cells detected two subpopulations. b, Heat map showing the hierarchical clustering result of PD-1hi cells based on selected ILC regulators. The expression levels are log2 transformed and ERCC-size factor normalized.

Extended Data Figure 8 scRNA-seq dissection of early ILC2 development.

a, Analysis of scRNA-seq data identified genes showing expression changes in C6, C7a, C8 and C9 cells. Change and distribution of expression of selected genes are shown. Il17rb and Bcl11b are among the genes showing spike expression from C6 to C9, whereas Il1rl1 (IL-33R) showed steadily increased expression. The bottom t-SNE plots showing expression of representative genes. b, The expression of IL-25R in PD-1hi bone marrow cells in the Bcl11b tdTomato mice (n = 3). c, Clonal differentiation assay of PD-1hiIL-25R+ and PD-1hiIL-25R− cells. Cells were cultured on OP9-DL1 stromal cells in the presence of IL-7 (20.0 ng ml−1) and SCF (50.0 ng ml−1) and were analysed 10 days later. ILC1 was defined as CD45+NK1.1+Bcl11b−, ILC2 as CD45+NK1.1−Bcl11b+ and ILC3 as CD45+NK1.1−Bcl11b−RORγt+. d, FACS analysis of the in vivo developmental potential of PD-1hiIL-25R+ cells. CD45.1 _Rag2_−/−_Il2rg_−/− recipients were injected with equal numbers of CD45.1−CD45.2+ PD-1hiIL-25R+ cells and CD45.1+CD45.2+ common lymphoid progenitors (100–500 of each) and the progenies of these populations were analysed 5–7 weeks later (n = 5 per group). e, Analysis of ILCs in Vav_-Cre_-Bcl11b fl/fl mice. Lin−IL-33R+IL-7Rα+ ILC2s, Lin−KLRG1+IL-7Rαlo ILC2 or Lin–NK1.1+NKp46+ ILCs from the bone marrow, lung or siLP were analysed (n = 4 per genotype), respectively. f, ‘Natural’ ILC2 (nILC2), ‘inflammatory’ ILC2 (iILC2) and BALF IL-5 were analysed in Vav_-Cre_-Bcl11b fl/fl and the control mice after administration of IL-25 (200 ng per mouse per day) for 3 consecutive days (n = 5 per treated group). Error bars denote s.e.m. g, FACS analysis of in vivo developmental potential of PD-1hi cells from _Vav_-Cre-Bcl11b fl/fl mice. CD45.1 _Rag2_−/−_Il2rg_−/− mice were injected with CD45.2 PD-1hi cells sorted from Bcl11b fl/fl or _Vav_-Cre-Bcl11b fl/fl mice. The progenies of these donor cells were analysed 4–7 weeks later by FACS (n = 3 per genotype). Data are representatives of three (b) or two (d–g) independent experiments. *P < 0.05, **P < 0.01 (two-tailed _t_-test).

Extended Data Figure 9 Restoration of development of _Bcl11b_-deficient PD-1hi ILC progenitors to ILC2 by overexpressing IL-25R.

a, FACS plots showing the expression of TCF-1 and Gata3 in mutant PD-1hi bone marrow cells. Protein expression was measured by intracellular antibody staining. b, Expression patterns of Tox, Id2, Tcf7 and Gata3 in the sequenced _Bcl11b_-deficient bone marrow cells. c, Overexpressing Il17rb in _Bcl11b_-deficient PD-1hi bone marrow cells. The rescued cells were analysed by FACS for ILC2 surface markers. PD-1hi cells sorted from Vav_-Cre_-Bcl11b fl/fl mice were transduced with the Il17rb or control retrovirus. The infected cells were cultured on OP9-DL1 stromal cells with the helper CD45.1 ILC2 progenitors in the presence of IL-25 (20.0 ng ml−1), IL-7 (20.0 ng ml−1) and SCF (50.0 ng ml−1). The cells were collected and analysed after two weeks of culture. Data are representatives of two (a, c) independent experiments.

Extended Data Figure 10 PD-1hi marks effector ILCs.

a, FACS analysis of PD-1 expression on peripheral ILCs in steady-state mice (n = 3). b, Gating strategies of lung ILCs. Lung cNK cells were gated as Lin−Id2+IL-7Rα−NK1.1+NKp46+; lung ILC1s as Lin−Id2+IL-7Rα+NK1.1+Bcl11b−; lung ILC2s as Lin−Id2+IL-7Rα+NK1.1−Bcl11b+; and lung ILC3s as Lin−Id2+IL-7Rα+NK1.1−Bcl11b−. The data were from influenza-infected mice at 5 days after infection. cNKs count for at least half of the Lin− leukocytes in these mice (n = 3). c, FACS analysis of PD-1 expression on CD3+ T cells, CD19+ B cells and peripheral ILCs after J43 treatment (n = 3). The tissues were collected at day 14 after J43 treatment. d, FACS plot shows the recognition of different epitopes of PD-1 by PD-1 antibody clones RMP1-30 and J43. The majority of lung PD-1hi ILC2s were stained with both RMP1-30 and J43. e, FACS analysis of lung PD-1hi cNK, ILC1 and ILC3 at 7 days after infection (n = 3). f, BALF cytokines were quantitated as shown (n = 3 per group per time point). The four experimental groups were: _Rag1_−/− mice with mock infection; _Rag1_−/− mice infected with A/X-31 and treated with either an antibody isotype control or J43; _Rag2_−/−_Il2rg_−/− mice infected with A/X-31 as the ILC-deficient control. g, More PD-1hi cells were found after papain challenge (day 6) in _Rag1_−/− mice (n = 3). h, _Rag1_−/− mice were pretreated with PD-1 antibody J43 or the isotype control antibody for 3 days and then administrated with papain (intranasally) for 5 consecutive days. The lung tissue was collected at day 6 for analysis (n = 5 per treatment). Lung ILC2s were reduced in J43 treated mice. PD-1hi and IL-5-producing ILC2 were undetectable after J43 administration. Data are representatives of two (a–h) independent experiments. Error bars (f, h) denote s.e.m. *P < 0.05, **P < 0.01 (two-tailed _t_-test).

Supplementary information

PowerPoint slides

Rights and permissions

About this article

Cite this article

Yu, Y., Tsang, J., Wang, C. et al. Single-cell RNA-seq identifies a PD-1hi ILC progenitor and defines its development pathway.Nature 539, 102–106 (2016). https://doi.org/10.1038/nature20105

Download citation