CD8+ tissue-resident memory T-cell development depends on infection-matching regulatory T-cell types

Abstract

Immunological memory is critical for immune protection, particularly at epithelial sites, which are under constant risk of pathogen invasions. To counter invading pathogens, CD8⁺ memory T cells develop at the location of infection: tissue-resident memory T cells (T_RM). CD8⁺ T-cell responses are associated with type-1 infections and type-1 regulatory T cells (T_REG) are important for CD8⁺ T-cell development, however, if CD8⁺ T_RM cells develop under other infection types and require immune type-specific T_REG cells is unknown. We used three distinct lung infection models, to show that type-2 helminth infection does not establish CD8⁺ T_RM cells. Intracellular (type-1) and extracellular (type-3) infections do and rely on the recruitment of response type-matching T_REG population contributing transforming growth factor-β. Nevertheless, type-1 T_REG cells remain the most important population for T_RM cell development. Once established, T_RM cells maintain their immune type profile. These results may have implications in the development of vaccines inducing CD8⁺ T_RM cells.

Fig. 1. Type-1 TREG cells are required for TRM cell generation in the lungs upon Influenza infection.

Mice were infected or not intranasally with 1000 plaque-forming units (PFU) of Influenza A X-31 strain (H3N2). 10 days post-infection lungs were collected, and cells were isolated and analysed via flow cytometry for cytokine production. a Numbers of activated (CD44⁺) CD4⁺ T cells producing indicated cytokines (p_{(Non-infected vs. Influenza H3N2)} < 0.000001) and (b) representative flow cytometry plots (n = 6, non-infected, n = 7 infected, N = 3). c Numbers of activated CD8⁺ T cells producing indicated cytokines (p_{(Non-infected vs. Influenza H3N2)} = 0.000013) and d representative flow cytometry plots (n = 6, non-infected, n = 7 infected, N = 3). e, f Foxp3^WT mice (n = 6, non-infected, n = 7 infected, N = 3) were assessed for e percentage (p_{(Non-infected vs. Influenza H3N2)} <0.000001), and f numbers of chemokine receptors CXCR3, ST2 (IL-33R) and CCR6 expression (p_{(Non-infected vs Influenza H3N2)}: CXCR3, p = 001415; ST2, p = 0.000502; CCR6, p = 0.001132), on T_REG cells (n = 6, non-infected, n = 8 infected, N = 3). g, h Mice analysed via flow cytometry for g percentages of total CD8⁺ T cells and h numbers of T_RM cells, defined as CD69⁺KLRG1⁻ CD8⁺ T cells, p(_Non^-_{infected vs Influenza H3N2)} < 0.0001 (n = 6, N = 3). i Foxp3^WT and Foxp3^ΔTbx21 mice received CD8^CD45.1 T cells intravenously, one day prior to infection. 14 days post infection, lung cells were assessed by flow cytometry for the percentage of T_RM cells (CD69⁺KLRG1^-CD103+) in the transferred population (Foxp3^WT n = 17, Foxp3^ΔTbx21 n = 10, N = 4), p_{(Foxp3WT vs Foxp3ΔTbx21)} <0.0001. Two-sided Mann–Whitney analysis was applied to compare groups. Data are presented as bars of mean ± SEM with single data points.

Fig. 2. Type-2 Nippostrongylus brasiliensis infection does not yield CD8 TRM cell formation.

Mice were subcutaneously infected or not with 300 stage L3 larvae of Nippostrongylus brasiliensis. 7 days post-infection lungs were collected, and cells were isolated and analysed via flow cytometry for cytokine production, and T_RM cells. a Numbers of activated (CD44⁺) CD4⁺ T cells producing indicated cytokines, p_{(Non-infected vs Nippostrongylus brasiliensis)}: IFN-γ, p = 0.001749; IL-17, p = 0.002537; IL-4, p = 0.000048; IL-13, p = 0.000729, and b representative flow cytometry plots (non-infected n = 6, infected n = 9, N = 3). c Numbers of activated CD8⁺ T cells producing indicated cytokines and d Representative flow cytometry plots of cytokine production in CD8⁺ T cells (non-infected n = 6, infected n = 9, N = 3). e, f Foxp3^WT mice were assessed for e percentage and f numbers of chemokine receptors CXCR3, ST2 (IL-33R) and CCR6 expression on T_REG cells, p_{(Non-infected vs Nippostrongylus brasiliensis)}: CXCR3, p = 0.014118; ST2, p = 0.023778; CCR6, p = 0.028885, (non-infected n = 6, infected n = 9, N = 3). g, h CD69⁺KLRG1⁻ CD8⁺ T_RM cell g percentage of total CD8⁺ T cells, p_{(Non-infected vs Nippostrongylus brasiliensis)} = 0.0004, and h number, p_{(Non-infected vs Nippostrongylus brasiliensis)}= = 0.0016, in non-infected (control) and infected mice (non-infected n = 6, infected n = 9, N = 3). i–l Foxp3^WT mice received CD8^CD45.1 T cells intravenously, one day prior to infection. 14 days post-infection, lung cells were analysed via flow cytometry for i total number of CD8^CD45.1 T cells, CD69⁺KLRG1⁻ CD8⁺ T_RM cell j percentage and k numbers, and l representative flow plot, within the CD8^CD45.1 T-cell population (n = 7, N = 3). m, n Comparison of transferred CD8^CD45.1 T-cell number in the (m) spleen, p_{(Influenza vs Nippostrongylus brasiliensis)} = 0.0012, or (n) lungs, p_{(Influenza vs Nippostrongylus brasiliensis)} = 0.0002, of mice infected with Influenza and N. brasiliensis (Influenza n = 9, N. brasiliensis n = 7, N = 3). 2-sided Mann–Whitney analysis was applied to compare the differences between groups. Data is presented as bars of mean ± SEM with single data points.

Fig. 3. Type-3 Aspergillus fumigatus infection yields CD8 T-cell recruitment and TRM cell formation.

Mice were intranasally challenged four times every 3 days with 10⁶ spores of Aspergillus fumigatus. 10 days post-infection lungs were collected, and cells were isolated and analysed via flow cytometry for cytokine production, and T_RM cells. a Numbers of activated (CD44⁺) CD4⁺ T cells producing indicated cytokines, p_{(Non-infected vs Aspergillus)}: IFN-γ, p = 0.005452; IL-17, p = 0.000733; IL-4, p = 0.007702; IL-13, p = 0.000005, and b representative flow cytometry plots of cytokine production in CD4⁺ T cells (non-infected n = 10, infected n = 12, N = 4). c Numbers of activated CD8⁺ T cells producing indicated cytokines, p_{(Non-infected vs Aspergillus)}: IFN-γ, p = 0.000951, and d representative flow cytometry plots of cytokine production in CD8⁺ T cells (non-infected n = 10, infected n = 14, N = 4). e, f T_REG cells from Foxp3^WT mice were assessed for expression in e percentage, p_{(Non-infected vs Aspergillus)}: Tbet, p = 0.018008; RORγt, p = 0.001023, and f numbers, p_{(Non-infected vs Aspergillus)}: Tbet, p = 0.022810; RORγt, p = 0.000093, of the transcription factors Tbet, GATA-3 and RORγt in (non-infected n = 6, infected n = 7, N = 3). g, h Analysis of CD69⁺KLRG1^- CD8⁺ T_RM cell (g) percentage, p_{(Non-infected vs Aspergillus)} = 0.000017, within the total CD8⁺ T-cell population, and h number, p_{(Non-infected vs Aspergillus)} = 0.002175 in the lungs of non-infected (control) and infected mice (n = 10, non-infected, n = 14 infected, N = 3). i Comparison of transferred CD8^CD45.1 CD69⁺KLRG1⁻ CD8⁺ T_RM cell numbers in the lungs upon Nippostrongylus brasiliensis (N.b.) or Aspergillus fumigatus (A.f.) infection, p_{(Aspergillus vs Nippostrongylus brasiliensis)}=0.0012, (n = 7, N = 3 for N.b.; n = 6 and N = 3 for A.f). Two-sided Mann-Whitney analysis was applied to compare the differences between groups. Data is presented as bars of mean ± SEM with single data points.

Fig. 4. Type-3 TREG cells are ablated in Foxp3-CreeYFP RORγtfl/fl mice.

Foxp3^WT and Foxp3^ΔRORγt mice were analysed at steady-state for T_REG, CD4⁺, and CD8⁺ T cells in lung, spleen, small intestine lamina propria lymphocytes (LPL), and small intestine intraepithelial lymphocytes (IEL). a Representative flow plot of RORγt expression in LPL T_REG cells in both mouse lines. b Percentage of LPL T_REG cells expressing Tbet, GATA-3 or RORγt, p_{(Foxp3WT vs Foxp3ΔRORγt)} < 0.000001, (Foxp3^WT n = 8, Foxp3^ΔRORγt n = 9, N = 3). c Percentage of RORγt expression in CD4⁺ and CD8⁺ LPL (Foxp3^WT n = 8, Foxp3^ΔRORγt n = 9, N = 3). d–f CD8⁺ T-cell proportions in the spleen were assessed for d naive (T_N CD44⁻CD62L⁺), central memory (T_CM, CD44⁺CD62L⁺) and effector memory/effector (T_EM/T_EFF, CD44⁺CD62L⁻) and e for CXCR3 expression on CD4⁺, CD8 and T_REG cells, with f representative flow plot (n = 8, N = 3). g Percentage of CD69⁺KLRG1^-CD103⁺ CD8⁺ T_RM cell in small intestine LPL and IEL in both mouse lines (LPL Foxp3^WT n = 7, Foxp3^ΔRORγt n = 8, N = 3^; IELs n = 6, N = 2). Two-sided Mann–Whitney analysis was applied to compare the differences between groups. Data is presented as bars of mean ± SEM with single data points.

Fig. 5. Depletion of Type-3 TREG cells reduces TRM cell development in type-3 infections.

a, b In Foxp3^WT and Foxp3^ΔRORγt mice, CD45.1⁺ CD8⁺ T cells were transferred one day prior to infection with 100 mg of segmented filamentous bacteria (SFB)-containing feces, organs were assessed for their T_RM cell phenotype (CD69⁺Eomes^-CD103⁺) 14 days later. Small intestine LPL were collected and assessed for the T-cell phenotype of CD45.1⁺ CD8⁺ T_RM cells. a Representative flow cytometry plots and b percentages of T_RM cells, p_{(Foxp3WT vs Foxp3ΔRORγt)} = 0.0357, (Foxp3^WT n = 3, Foxp3^ΔRORγt n = 5, N = 3). c, d Intranasal challenge with 1000 PFU of Influenza X-31 strain (H3N2). Lungs were collected and transferred T cells were analysed. c Representative flow cytometry plots of T_RM cells within transferred T cells and d percentages of T_RM cells in indicated mouse lines (Foxp3^WT n = 17, Foxp3^ΔRORγt n = 6, N = 4). e–g Mice were infected by four intranasal challenges with 10⁶ spores of Aspergillus fumigatus. Lungs were collected and e total CD4⁺ T cells (Foxp3^WT n = 7, Foxp3^ΔRORγt n = 6, N = 3), f total T_REG cells, p_{(Foxp3WT vs Foxp3ΔRORγt)}=0.0043, and g subsets of T_REG cells based on their transcription factor expression, p_{(Foxp3WT vs Foxp3ΔRORγt)} = 0.000001, were analysed (Foxp3^WT n = 6 Foxp3^ΔRORγt n = 5, N = 3). h, i Mice were transferred with indicated condition receiving control or TGFβ1–deficient T_REG cells prior to the first of four intranasal challenges with 10⁶ spores of Aspergillus fumigatus. Lungs were collected and transferred CD45.1 CD8 T cells were analysed. h Representative flow cytometry plots of T_RM cells within transferred CD45.1 CD8 T cells and i percentages of T_RM cells in all conditions, p_{(Foxp3WT vs Foxp3ΔRORγt)} = 0.0087, p_{(Foxp3ΔRORγt vs Foxp3ΔRORγt+WT Tregs)} = 0.0173, p_{(Foxp3ΔRORγt vs Foxp3ΔRORγt+TregΔTGFβ1)} = 0.0043 (Foxp3^WT; Foxp3^ΔRORγt n = 6, N = 4; Foxp3^ΔRORγt + T_REG^WT n = 5, N = 2; Foxp3^ΔRORγt + T_REG^ΔTGFβ1 n = 5, N = 2). Multiple unpaired t test was used for g, other panels two-sided Mann-Whitney analysis was applied to compare groups. Data is presented as bars of mean ± SEM with single data points.

Fig. 6. Type-1 and type-3 TREG cells enhance TRM cell development, the identity of which are maintained.

Mice were transferred with CD45.1⁺ CD8⁺ T cells one day prior to oral gavage infection. 14 days after transplant, tissues were collected and assessed for their T_RM phenotype. a, b Infection with 5000 oocysts of Eimeria vermiformis, a representative flow cytometry plots T_RM cells (CD69⁺Eomes^-CD103⁺) and b percentages of T_RM cells in both mouse lines, p_{(Foxp3WT vs Foxp3ΔRORγt)} = 0.0020, (Foxp3^WT n = 4, Foxp3^ΔRORγt n = 10, N = 3). c, d Mice were challenged four times intranasally with 10⁶ spores of Aspergillus fumigatus. Lungs were collected and transferred T cells were analysed, c representative flow cytometry plots of T_RM cells and d percentages of T_RM cells in both mouse lines, p_{(Foxp3WT vs Foxp3ΔTbet)} = 0.0238 (Foxp3^WT n = 6, N = 4, Foxp3^ΔTbx21 n = 3, N = 3). e, f Mice were transferred with CD45.1 CD8 T cells prior to the first of four intranasal challenges with 10⁶ spores of Aspergillus fumigatus. Lungs were collected and transferred CD45.1 CD8⁺ T_RM cells were analysed for e IFNγ or f IL-17 production (Foxp3^WT n = 4, Foxp3^ΔRORγt n = 7, Foxp3^ΔTbx21 n = 4, N = 3). Two-sided Mann–Whitney analysis was applied to compare groups. Data is presented as bars of mean ± SEM with single data points.