Fig 1.
NK cells acquire intrinsic memory features 21 days after in vivo infection.
(A) Experimental scheme. C57BL/6 mice were intranasally injected with either PBS (black symbols) or sub-lethal dose of S. pneumoniae (SPN, red symbols, 5x105 CFU) for two consecutive days. After 21 days, NK cells were highly purified from spleens of D21PBS or D21SPN mice (98% of purity) and stimulated in vitro with cytokines (IL-15 at 2 ng/ml, IL-18 at 1,5 ng/ml, IL-12 at 1,25 ng/ml) and formaldehyde inactivated S. pneumoniae (SPN, MOI 20) for 24 hours. (B) Intensity of IFNγ expression in NK cells (MFI, left panel), percentage of IFNγ+ NK cells (middle panel), representative overlay histogram upon IL-15+IL-18+SPN stimulation (right panel, gray represents isotype control). D21PBS NK cells and D21SPN NK cells are purified from spleen, pooled from n ≥ 4 mice/group and incubated in n ≥ 3 experimental replicates/group. Box plots where each dot represents an experimental replicate (black dots for D21PBS NK cells, red dots for D21SPN NKs cells), lines are median, error bars show min to max. Data are representative of at least three experiments. ns, not significant. **** p < 0.0001. 2way ANOVA test comparing D21PBS NKs and D21SPN NKs values in each condition.
Fig 2.
Sub-lethal infection with S. pneumoniae induces a rapid and transient immune response that returns to basal state before 21 days.
(A) Experimental scheme, showing that C57BL/6 mice were intranasally injected with either PBS (black symbols) or sub-lethal dose of S. pneumoniae (SPN, red symbols, 5x105 CFU) for two consecutive days. Organs were collected at 24h, 72h and 21 days post-infection for CFU counts and cellular infiltrate determination by flow cytometry. (B-D) Organs were collected at 24h (B), 72h (C) and 21 days post-infection (D). CFU counts in the nasal lavage, BALF, lungs, spleen and blood of infected mice (left panel). Box plots where each dot represents an individual mouse, lines are the mean, error bars show min to max and dotted lines represent limit of detection. Percentage of neutrophils (CD11b+ Ly6G+) among CD45+ cells in the lungs and bronchio-alveolar lavage fluid (BALF), percentage of CD69+ neutrophils in the BALF (middle panel). Percentage of NK cells (NK1.1+ CD3-) among CD45+ cells, percentage of IFNγ+, Perforin+, Granzyme B+ NK cells in the lungs and spleen (right panel). Box plots where each dot represents an individual mouse (black dots for uninfected mice, red dots for infected mice), lines are the median, error bar show min to max. Data are pooled from two or three repeats with n ≥ 3 mice/group. ns, not significant. * p < 0.05 and ** p <0.01, Mann-Whitney test for single comparisons and 2way ANOVA test for multiple comparisons.
Fig 3.
Transferred memory NK cells protect mice from lethal S. pneumoniae infection for at least 12 weeks.
(A) Experimental scheme. C57BL/6 mice (donor mice) were intranasally injected with either PBS (black symbols) or sub-lethal dose of S. pneumoniae (SPN, red symbols, 5x105 CFU) for two consecutive days. After 21 days or 12 weeks, NK cells from spleens or lungs were highly purified (98%) and transferred into naïve mice (recipient mice, intravenously, 2x105 cells). One day after, all recipient mice were intranasally infected with a lethal dose of S. pneumoniae (5x106 CFU for survival study, 1x107 CFU for bacterial counts comparison) or L. monocytogenes (1x106 CFU, intravenously). Organs were collected at 24h and 40h post-infection for CFU counts and flow cytometry analysis. (B) Bacterial counts at 24h and 40h post-infection in the nasal lavage, bronchio-alveolar lavage fluid (BALF), lungs, blood and spleen of mice having received either D21PBS NKs (black symbols) or D21SPN NKs (red symbols). Box plots where each dot represents an individual mouse, lines are the mean, error bars show min to max and dotted lines represent limit of detection. Data are pooled from at least two repeats with n ≥ 3 mice/group. (C) Survival curve. Dots represent the percentage of survival of total mice. Data are representative of three repeats with n = 4 mice/group. (D) Clinical score. Dots represent the mean, error bars show the standard error of the mean (SEM). Data are representative of two repeats with n = 4 mice/group. (E) Weight represented as percentage of initial body weight loss. Dots represent the mean, error bars show the standard error of the mean (SEM). After death, mice have the value of 80%. Data are representative of two repeats with n = 4 mice/group. (F) Bacterial counts at 40h post-infection in the lungs and spleen of mice having received either D21PBS NKs (black symbols) or D21SPN NKs (red symbols) isolated from the lungs of donor mice. Box plots where each dot represents an individual mouse, lines are the mean, error bars show min to max and dotted lines represent limit of detection. Data are pooled from three repeats with n ≥ 3 mice/group. (G) Bacterial counts at 40h post-infection in the lungs and spleen of mice having received either W12PBS NKs (black symbols) or W12SPN NKs (red symbols). Box plots where each dot represents an individual mouse, lines are the mean, error bars show min to max and dotted lines represent limit of detection. Data are pooled from three repeats with n ≥ 4 mice/group. (H) Bacterial counts at 40h post-infection in the spleen and liver of mice infected with Listeria monocytogenes (intravenously) and having previously received either D21PBS NKs (black symbols) or D21SPN NKs (red symbols) from the spleens of donor mice. Box plots where each dot represents an individual mouse, lines are the mean, error bars show min to max and dotted lines represent limit of detection. Data are pooled from two repeats with n = 3 mice/group. ns, not significant. * p < 0.05, ** p < 0.01, *** p < 0.001 **** p < 0.0001. 2way ANOVA (B), Log-rank (Mantel-Cox) (C) and Mann-Whitney (F,G,H) tests for statistical significance.
Fig 4.
Protection of mice mediated by memory NK cells is IFNγ independent.
(A-B) Mouse infections are carried out as in the scheme in Fig 3A. Organs were collected at 24h and 40h post-infection for flow cytometry analysis and ELISA assays. (A) Flow cytometry analysis of IFNγ expression in CD45.1+ endogenous NK cells and CD45.2+ transferred NK cells (hatched) into both mice having received D21PBS NK cells (black) and mice having received D21SPN NK cells (red). Violin plot where each dot represents an individual mouse, lines are the median. Data are pooled from two repeats with n ≥ 2 mice/group. (B) IFNγ ELISA assays of lung supernatants from infected mice having received either D21PBS NKs (black) or D21SPN NKs (red) at 24h and 40h post-infection. Bars are the mean of at least four experiments with n ≥ 4 mice/group, each dot represents an individual mouse and error bars are the standard error of the mean (SEM). (C) Mouse infections are carried out as in the scheme in Fig 3A, with the exception that WT mice were used as donor mice and Ifngr KO mice as recipient mice. Bacterial counts at 40h post-infection in the lungs and spleen of Ifngr KO mice having received D21PBS WT NK cells (black symbols) or D21SPN WT NK cells (red symbols). Box plots where each dot represents an individual mouse, lines are the mean, error bars show min to max and dotted lines represent limit of detection. Data are pooled from two repeats with n ≥ 4 mice/group. (D) WT and Ifngr KO mice were intranasally infected with a lethal dose of S. pneumoniae (1x107 CFU). Lungs were collected at 40h post-infection for CFU counts and flow cytometry analysis. Percentages of neutrophils (Ly6G+ CD11b+), NK cells (NK1.1+ CD3-) and CD69+ NK cells. Box plots where each dot represents an individual mouse, lines are the mean, error bars show min to max and dotted lines represent limit of detection. Data are pooled from two repeats with n ≥ 3 mice/group ns, not significant. * p < 0.05, ** p < 0.01. 2way ANOVA (A,B) and Mann-Whitney (C,D) tests for statistical significance.
Fig 5.
Protection of mice appears to be an intrinsic NK cell property.
Mouse infections are carried out as in the scheme in Fig 3A. Organs were collected at 24h and 40h post-infection for flow cytometry analysis, ELISA assays and CFU counts. (A) Cellular infiltrate analysis in the lungs at 24h post-infection. Percentage of respective cell types among CD45+ cells (left panel) and absolute number of cells for each cell type (right panel): neutrophils (Ly6G+ CD11b+), alveolar macrophages (AM) (Ly6G- SiglecF+ CD64+ CD11b-), interstitial macrophages (IM) (Ly6G- SiglecF- CD11bhigh MHCII+ CD64+ CD24-), monocytes (Ly6G- SiglecF- CD11bhigh MHCII-), eosinophils (Ly6G- SiglecF+ CD11b+), dendritic cells (DCs) (CD103+ DCs: Ly6G- SiglecF- CD11blow CD103+ CD24+ + CD11b+ DCs: Ly6G- SiglecF- CD11bhigh MHCII+ CD64- CD24+), NK cells (NK1.1+ CD3-), T cells (NK1.1- CD3+). Radar plots with each value representing the geometric mean of three repeats with n = 4 mice. The blue line for uninfected mice, black line for mice having received D21PBS NKs and red line for mice having received D21SPN NKs. (B) Cellular infiltrate analysis in the bronchio-alveolar lavage fluid (BALF) at 24h post-infection. Percentage of respective cell types among CD45+ cells (left panel) and absolute number of cells for each cell type (right panel): neutrophils (Ly6G+ CD11b+), alveolar macrophages (SiglecF+ CD64+ CD11b-). Box plots where each dot represents an individual mouse (blue dots for uninfected mice, black dots for mice having received D21PBS NKs, red dots for mice having received D21SPN NKs), lines are the median, error bar show min to max. Date are pooled from at least two repeats with n ≥ 2 mice/group. (C) Cellular activation analysis in the lung at 40h post-infection. Percentages of CD69+ neutrophils, CD69+ NK cells and CD86+ interstitial macrophages. Box plots where each dot represents an individual mouse (the blue dots are for uninfected mice, black dots for mice having received D21PBS NKs, red dots for mice having received D21SPN NKs), lines are the median, error bar show min to max. Data are pooled from two repeats with n ≥ 1 mice/group. (D) ELISA assays of lung supernatants from infected mice at 40h post-infection. Bars are the mean of at least 3 experiments with n ≥ 3 mice/group, error bars are the standard error of the mean (SEM). (E-F) At 40h post-infection, lung supernatants of mice having received either D21PBS NKs (black) or D21SPN NKs (red) (C)—either We12PBS NKs (black) or We12SPN NKs (red) (D) were collected to perform Granzyme B ELISA assays. Bars are the mean of at least 3 experiments with n ≥ 3 mice/group, each dot represents an individual mouse and error bars are the standard error of the mean (SEM). ns, not significant. * p < 0.05, ** p < 0.01, *** p < 0.001. Kruskal-Wallis (A,B,C), 2way ANOVA (D) and Mann-Whitney (E,F) tests for statistical significance.
Fig 6.
Cytotoxic proteins are upregulated by memory NK cells and important for protection of mice.
(A-B-C) C57BL/6 mice were intranasally injected with either PBS (black symbols) or sub-lethal dose of S. pneumoniae (SPN, red symbols, 5x105 CFU) for two consecutive days. After 21 days, NK cells were highly purified from spleens of D21PBS or D21SPN mice (98% of purity) and stimulated in vitro with cytokines (IL-15 at 2 ng/ml, IL-18 at 1,5 ng/ml, IL-12 at 1,25 ng/ml) and either formaldehyde inactivated bacteria (MOI 20), Lipopolysaccharide (LPS) or the synthetic lipopeptide Pam3CSK4 (P3C) for 24 hours. SPN: S. pneumoniae, GBS: S. agalactiae, Listeria: L. monocytogenes. (A) Intensity of Granzyme B expression in NK cells (MFI, left panel), percentage of Granzyme B+ NK cells (middle panel), representative overlay histogram upon IL-15+IL-18+SPN stimulation (right panel, gray represents isotype control). D21PBS NK cells and D21SPN NK cells are purified and pooled from n ≥ 3 mice/group and incubated in n ≥ 2 experimental replicates/group. Box plots where each dot represents an experimental replicate (black dots for D21PBS NK cells, red dots for D21SPN NKs cells), lines are median, error bar show min to max. Data are representative of at least two experiments. (B) Intensity of Perforin expression in NK cells (MFI, left panel), percentage of Perforin+ NK cells (middle panel), representative overlay histogram upon IL-15+IL-18+SPN stimulation (right panel, gray represents isotype control). D21PBS NK cells and D21SPN NK cells are purified and pooled from n ≥ 3 mice/group and incubated in n ≥ 2 experimental replicates/group. Box plots where each dot represents an experimental replicate (black dots for D21PBS NK cells, red dots for D21SPN NKs cells), lines are median, error bar show min to max. Data are representative of at least two experiments. (C) NK cells were stimulated in vitro with cytokines, SPN, Lipopolysaccharide (LPS) and the synthetic lipopeptide Pam3CSK4 (P3C). Intensity of Granzyme B expression in NK cells (MFI, left panel), percentage of Granzyme B+ NK cells (right panel). D21PBS NK cells and D21SPN NK cells are purified and pooled from n ≥ 4 mice/group and incubated in n ≥ 3 experimental replicates/group. Box plots where each dot represents an experimental replicate (black dots for D21PBS NK cells, red dots for D21SPN NKs cells), lines are median, error bar show min to max. Data are representative of three experiments. (D) Mouse infections are carried out as in the scheme in Fig 3A, with the exception that Prf1 KO mice were used as both donor and recipient mice. Bacterial counts at 40h post-infection in the nasal lavage, bronchio-alveolar lavage fluid (BALF), lungs and spleen of Prf1 KO mice having received D21PBS Prf1 KO NK cells (black symbols) or D21SPN Prf1 KO NK cells (red symbols). Box plots where each dot represents an individual mouse, lines are the mean, error bars show min to max and dotted lines represent limit of detection. Data are pooled from two repeats with n ≥ 3 mice/group. ns, not significant. * p < 0.05, ** p < 0.01, *** p < 0.001. 2way ANOVA (A,B,C) and Mann-Whitney (D) tests for statistical significance.