Research ArticleT CELLS

PD-1hi CD8+ resident memory T cells balance immunity and fibrotic sequelae

See allHide authors and affiliations

Science Immunology  14 Jun 2019:
Vol. 4, Issue 36, eaaw1217
DOI: 10.1126/sciimmunol.aaw1217
  • Fig. 1 Epitope-specific manifestation of exhaustion gene signature in lung TRM cells.

    WT C57BL/6 mice were infected with influenza PR8. Spleens or lungs were harvested after intravenous (i.v.) administration of CD45 Ab at the indicated d.p.i. (A) Expression of CD69 and CD103 on lung NP366–374 or PA224–233 circulating memory (intravenous Ab+, TM-Circ) cells or TRM cells (intravenous Ab) by flow cytometry at 40 d.p.i. (n = 4). (B and C) Transcriptional profiles of NP366–374 and PA224–233 TRM cells were determined by RNA-seq at 42 d.p.i. (pooled from 16 mice). (B) Differential gene expression between NP366–374 TRM cells and PA224–233 TRM cells. FPKM, fragments per kilobase million. (C) GSEA showing positive enrichment in NP366–374 TRM cells of the genes up-regulated in exhausted CD8+ T cells (top) or showing negative enrichment in NP366–374 TRM cells of the genes down-regulated in exhausted CD8+ T cells (bottom). NES, normalized enrichment score. (D and E) Expression of immune-related genes in NP366–374 (NP) or PA224–233 (PA) lung effector (TE-LUNG) and TRM cells, and spleen effector (TE-SPL) and memory (TM-SPL) cells were determined by NanoString at 8 (effector) or 38 (memory) d.p.i. (pooled from 4 to 12 mice per group). (D) Heat map representing expression levels of 560 immune-associated genes. (E) Heat map representing expression levels of exhaustion-associated genes.

  • Fig. 2 Exhausted-like TRM cells coexhibit exhausted and memory T cell features.

    (A to G) WT C57BL/6 mice were infected with influenza PR8. Spleens and lungs were harvested after intravenous administration of CD45 Ab at the indicated d.p.i. (A) PD-1 expression levels [mean fluorescence intensity (MFI)] on intravenous Ab lung NP366–374 or PA224–233 T cells were assessed by flow cytometry at the indicated d.p.i. (B) Expression of inhibitory receptors on lung NP366–374 or PA224–233 TRM cells was assessed by flow cytometry at 40 d.p.i. (C) Total numbers of inhibitory receptors expressed on lung NP366–374 or PA224–233 TRM cells were assessed by flow cytometry at 40 d.p.i. (D and E) IFN-γ and TNF-α production by TRM cells was assessed by flow cytometry after ex vivo stimulation with the NP366–374 or PA224–233 peptide at 40 d.p.i. (D) Representative plots of tetramer (Tet), IFN-γ, and TNF-α staining in lung-resident (intravenous Ab) CD8+ cells. (E) Frequencies of IFN-γ+ TNF-α+ cells were normalized to the frequencies of tetramer+ TRM cells in resident CD8+ cells. Percentages of IFN-γ+ TNF-α+ cells in tetramer+ TRM cells were assessed. (F) NP366–374 or PA224–233 TRM cell TCF-1 and CD127 expression was assessed by flow cytometry at 42 d.p.i. (G) Expression of CD8+ memory-associated genes in lung effector (TE-LUNG) or TRM cells, and spleen effector (TE-SPL) or memory (TM-SPL) cells was determined by NanoString at 8 or 38 d.p.i. (H and I) Tgfbr2fl/fl or Tgfbr2Δdlck mice were infected with influenza PR8. Spleens and lungs were harvested after intravenous administration of CD45 Ab at 42 d.p.i. (H) Representative flow cytometry plots. (I) Frequencies of NP366–374 or PA224–233 TRM and TM-SPL cells. (J) WT C57BL/6 mice were infected with influenza PR8, treated with FTY720 (39 to 41 d.p.i.), and then rechallenged with influenza X31 at 40 d.p.i. KI-67 expression in NP366–374 or PA224–233 TRM cells was assessed by flow cytometry before and 2 days after rechallenge. Representative of two to three experiments (n = 2 to 7) except (G). Data are mean ± SD; ns, not significant. *P < 0.05, **P < 0.01, ****P < 0.0001, unpaired two-tailed t test.

  • Fig. 3 TCR signaling is required for exhausted-like TRM cell formation and maintenance.

    (A) Nur77-GFP mice were infected with influenza PR8. Spleens and lungs were harvested after intravenous administration of CD45 Ab. Green fluorescent protein (GFP) expression in TRM cells, lung-circulating memory (TM-Circ, intravenous Ab+), and TM-SPL cells was assessed by flow cytometry at 40 d.p.i. (B to D) Nur77-GFP mice were infected with influenza PR8 and received vehicle or FTY720 daily starting at 21 d.p.i. Mice were euthanized after intravenous administration of CD45 Ab at 40 d.p.i. (B) Schematic of experimental design (top) and representative flow cytometry plots of Nur77-GFP expression in NP366–374 or PA224–233 TRM cells. (C) Quantification of percentages of Nur77-GFP+ cells in NP366–374 or PA224–233 TRM cells after vehicle or FTY720 treatment. (D) PD-1, TIM-3, CD69, or CD103 expression on NP366–374 or PA224–233 TRM cells after vehicle or FTY720 treatment was assessed by flow cytometry. (E and F) Thy1.1+ C57BL/6 WT and Thy1.2+ Nr4a1−/− (Nur77) mixed bone marrow (BM) chimeric mice were infected with influenza PR8. Spleens and lungs were harvested after intravenous administration of CD45 Ab at 40 d.p.i. (E) PD-1 expression on NP366–374 or PA224–233 TRM cells was determined by flow cytometry. (F) Representative plots (left) and percentages (right) of NP366–374 or PA224–233 TRM cells in Thy1.1+ WT or Thy1.2+ Nr4a1−/−-resident CD8+ T cells. Representative of two to three experiments (n = 3 to 5). Data are mean ± SD; ns, not significant. *P < 0.05, **P < 0.01, unpaired two-tailed t test.

  • Fig. 4 Persistent TCR–pMHC-I signaling drives the formation and maintenance of exhausted-like TRM cells.

    H2dbfl/fl (Ctl) and H2dbΔUbc-creERT2 (KO) mice were infected with influenza PR8 and then treated with tamoxifen starting at 22 d.p.i. Spleens and lungs were harvested after intravenous administration of CD45 Ab at 42 d.p.i. (A) Schematic of the experimental design. (B) Average PD-1 or TIM-3 expression levels (MFI) on NP366–374 TRM cells evaluated by flow cytometry. (C) Average PD-1 or TIM-3 expression levels (MFI) on PA224–233 TRM cells evaluated by flow cytometry. (D) Representative plots and average frequencies of TRM cells (top) or TM-SPL cells (bottom) in lung-resident or splenic CD8+ T cells, respectively. (E) Cell numbers of NP366–374 or PA224–233 TRM cells. (F) Representative plot of CD103 expression on NP366–374 TRM cells. (G) Average CD103 expression levels (MFI) on NP366–374 or PA224–233 TRM cells were evaluated by flow cytometry. Representative of three experiments (n = 4 mice per group). Data are mean ± SD; ns, not significant. *P < 0.05, **P < 0.01, unpaired two-tailed t test.

  • Fig. 5 B7 blockade diminishes exhausted-like TRM cell maintenance.

    (A) WT mice were infected with influenza PR8. CD28 gene expression in NP366–374 or PA224–233 TRM cells or spleen memory T cells was determined by NanoString at 38 d.p.i. (B) WT mice were infected with influenza PR8. CD28 expression levels (MFI) on TRM cells were assessed by flow cytometry at 42 d.p.i. (C to F) WT mice were infected with influenza PR8 and B7 costimulation was blocked through the administration of α-B7.1 plus α-B7.2 at 21 d.p.i. Spleens or lungs were harvested after intravenous administration of CD45 Ab at 42 d.p.i. (C) Schematic of experimental design and PD-1 expression levels (MFI) on TRM cells. (D) Frequencies of NP366–374 or PA224–233 TRM cells in total lung-resident (intravenous Ab) CD8+ T cells. (E) Cell numbers of NP366–374 or PA224–233 TRM cells. (F) Representative plots (left) and frequencies (right) of active caspase-3/7+ cells in NP366–374 or PA224–233 TRM cells. (G) WT mice were infected with influenza PR8 with or without B7 blockade at 21 d.p.i. and then rechallenged with X31 (1.2 × 104 pfu) at 42 d.p.i. in the presence of FTY720. Percentages of original weight after rechallenge were assessed daily. Representative of two to three experiments except (A) (n = 3 to 4 mice per group). Data are mean ± SD; ns, not significant. *P < 0.05, **P < 0.01, ***P < 0.001, unpaired two-tailed t test.

  • Fig. 6 PD-L1 blockade rejuvenates exhausted-like TRM cells.

    (A to G) WT mice were infected with influenza PR8 and received control IgG (Ctl) or α-PD-L1 from 21 to 37 d.p.i. Spleens and lungs were harvested after intravenous administration of CD45 Ab at the indicated d.p.i. (A) Experimental design and representative plots of NP366–374 or PA224–233 TRM cells at 40 or 60 d.p.i. (B) Frequencies of NP366–374 or PA224–233 TRM cells in total resident (intravenous Ab) CD8+ T cells at 40 or 60 d.p.i. (C) Cell numbers of NP366–374 or PA224–233 TRM cells at 40 or 60 d.p.i. (D) Frequencies of active caspase-3/7+ cells in NP366–374 TRM cells at 35 d.p.i. (E) IFN-γ and TNF-α production by TRM cells was determined after ex vivo NP366–374 peptide stimulation at 40 d.p.i. Left panel, representative plots. Right panel, average frequencies of IFN-γ and TNF-α double-positive cells with or without PD-L1 blockade. (F and G) CD103 expression on NP366–374 lung TRM cells was determined at 40 d.p.i. Representative plots (F) and frequencies of CD103+ cells (G, left) or CD103 expression levels (MFI) (G, right) in NP366–374 TRM cells. (H and I) WT mice were infected with influenza PR8 and received control Ab, α-PD-L1, and/or α-B7 as indicated starting at 21 d.p.i. Lungs were harvested after intravenous administration of CD45 Ab at 40 d.p.i. (H) Representative plots and percentages of TRM cells in lung-resident (intravenous Ab) CD8+ T cells. (I) CD103 expression levels (MFI) on TRM cells were determined by flow cytometry. Representative of two to four experiments (n = 3 to 6). Mean ± SD; ns, not significant. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 unpaired two-tailed t test or one-way analysis of variance (ANOVA) with Tukey multiple comparison test.

  • Fig. 7 TRM cell exhaustion balances protective immunity and fibrotic sequelae.

    (A) WT mice were infected with influenza PR8 and received control IgG (Ctl) or α-PD-L1 from 21 to 37 d.p.i. Mice were rechallenged with influenza X31 (2.4 × 104 pfu) in the presence of FTY720 at 42 d.p.i. Percentages of original weight were determined daily after rechallenge. (B to D) WT mice were infected with influenza PR8 and received control IgG (Ctl) or α-PD-L1 from 21 to 37 d.p.i. Lung pathology and hydroxyproline levels were determined at 60 d.p.i. (B) Hematoxylin and eosin (H&E) and Masson’s trichrome C staining of lung sections. (C) Hydroxyproline levels (micrograms per milligram of lung tissue) of the lungs. (D) Hydroxyproline levels of the lungs from mice received control Ab, α-PD-L1, or α-PD-L1 plus α-CD8 (CD8 depletion). (E and F) CD8, PD-1, and CD103 staining was performed on lung sections from control (n = 10) or patients with IPF (n = 10). (E) Representative of CD8, PD-1, and CD103 staining. Blue, DAPI (4′,6-diamidino-2-phenylindole). (F) Frequencies of CD8+ cells, CD8+ CD103+, or CD8+ PD-1+ cells in DAPI+ cells of control or IPF lungs. (A to D) Representative of two to five experiments (n = 3 to 6). Mean ± SD, *P < 0.05, **P < 0.01, ***P < 0.001, unpaired two-tailed t test.

Supplementary Materials

  • immunology.sciencemag.org/cgi/content/full/4/36/eaaw1217/DC1

    Fig. S1. Epitope-specific expression of CD103 and PD-1 on TRM cells.

    Fig. S2. Epitope-specific expression of inhibitory receptors on TRM cells.

    Fig. S3. Exhausted-like TRM cells are present in the X31 influenza model.

    Fig. S4. Exhausted-like TRM cells respond effectively to recall in situ.

    Fig. S5. Cognate peptide inoculation directly promotes PD-1 expression on PA224–233 TRM cells.

    Fig. S6. Nur77 (Nr4a1) is required for the induction and/or maintenance of exhausted-like TRM cells.

    Fig. S7. Limiting NP antigen dose decreases the exhaustion phenotype.

    Fig. S8. TCR signaling and inhibitor receptor expression on exhausted-like TRM cells are lost 4 months after infection.

    Fig. S9. Persistent MHC-I engagement maintains exhausted-like TRM phenotypes and cells.

    Fig. S10. NP366–374 TRM cells, PD-1 expression, and heterologous protection are lost over time.

    Fig. S11. Continuous CD28 signaling is required for the maintenance of NP366–374 TRM cells.

    Fig. S12. PD-L1 blockade promotes exhausted-like TRM cell survival and CD103 expression.

    Fig. S13. PD-L1 blockade promotes cytokine production of exhausted-like TRM cells to antigenic restimulation.

    Fig. S14. PD-L1 blockade promotes CD103 expression.

    Fig. S15. B7 signaling is required for the effects of PD-L1 blockade.

    Fig. S16. PD-L1 blockade promotes heterologous immunity and causes tissue pathology.

    Fig. S17. CD8 T cells are responsible for the development of tissue pathology after α-PD-L1 blockade.

    Fig. S18. ILD lungs exhibit elevated PD-1, CD103, and TIM-3 expression.

    Table S1. Raw data.

  • Supplementary Materials

    The PDF file includes:

    • Fig. S1. Epitope-specific expression of CD103 and PD-1 on TRM cells.
    • Fig. S2. Epitope-specific expression of inhibitory receptors on TRM cells.
    • Fig. S3. Exhausted-like TRM cells are present in the X31 influenza model.
    • Fig. S4. Exhausted-like TRM cells respond effectively to recall in situ.
    • Fig. S5. Cognate peptide inoculation directly promotes PD-1 expression on PA224–233 TRM cells.
    • Fig. S6. Nur77 (Nr4a1) is required for the induction and/or maintenance of exhausted-like TRM cells.
    • Fig. S7. Limiting NP antigen dose decreases the exhaustion phenotype.
    • Fig. S8. TCR signaling and inhibitor receptor expression on exhausted-like TRM cells are lost 4 months after infection.
    • Fig. S9. Persistent MHC-I engagement maintains exhausted-like TRM phenotypes and cells.
    • Fig. S10. NP366–374 TRM cells, PD-1 expression, and heterologous protection are lost over time.
    • Fig. S11. Continuous CD28 signaling is required for the maintenance of NP366–374 TRM cells.
    • Fig. S12. PD-L1 blockade promotes exhausted-like TRM cell survival and CD103 expression.
    • Fig. S13. PD-L1 blockade promotes cytokine production of exhausted-like TRM cells to antigenic restimulation.
    • Fig. S14. PD-L1 blockade promotes CD103 expression.
    • Fig. S15. B7 signaling is required for the effects of PD-L1 blockade.
    • Fig. S16. PD-L1 blockade promotes heterologous immunity and causes tissue pathology.
    • Fig. S17. CD8 T cells are responsible for the development of tissue pathology after α-PD-L1 blockade.
    • Fig. S18. ILD lungs exhibit elevated PD-1, CD103, and TIM-3 expression.

    Download PDF

    Other Supplementary Material for this manuscript includes the following:

    • Table S1 (Microsoft Excel format). Raw data.

    Files in this Data Supplement:

Stay Connected to Science Immunology

Navigate This Article