Fig 1.
Apoptosis contributes to sepsis-induced systemic loss of NK-cells.
(A) Experimental Design. Sham or CLP mice were sacrificed 12, 24, or 48 hrs after surgery, and the number of NK-cells in the indicated tissues evaluated. (B) Representative flow plots of NK-cell gating. The total number of lymphocytes or NK-cells in spleen (C,D) or liver (E,F) 48 hrs after Sham or CLP surgery. (G) The number of NK-cells in mesenteric (mLN) and inguinal lymph nodes (iLN), peritoneal lavage (PL), blood (PBL/mL), liver, and spleen 24 hrs after sepsis-induction. (H) Representative flow plots of FLICA and PI staining of NK-cells. (I) Frequency of apoptotic (FLICA+PI+) NK-cells in the spleen at 12, 24, and 48 hrs after sepsis induction. (J) NK-cells obtained from Sham and CLP host 48 hrs post-surgery were placed in in vitro culture and the percent of surviving NK-cells was determined at indicated times. Data are representative from 3 independent experiments with 3–5 mice per group. Numbers above bars show fold change between groups. * p<0.05. Error bars represent the standard error of the mean.
Fig 2.
Sepsis induces significant transcriptional changes in NK-cells.
(A) Experimental Design. Mice were sacrificed at day 1 or 2 post-Sham or CLP surgery and NK1.1+/CD3- NK-cells FACS-sorted from the spleen before RNA extraction. (B) Gene expression heatmap of genes with statistically significant change (Fold change greater than or equal to 1.5 and p<0.05) as a result of any combination of comparison. (C) Number of statistically significant gene changes as a result of each comparison. (D) Principal component analysis of significantly changed genes.
Fig 3.
Gene-set enrichment analysis (GSEA) reveals sepsis-induced changes in molecular pathways relevant to NK-cell effector functions.
(A) Enrichment scores of genes in pathways relevant to sepsis (apoptosis) and NK-cells functionality (NK-mediated cytotoxicity and calcium signaling). Red box indicates leading edge of enriched region; negative enrichment—box to left, positive enrichment—box to right. Gene expression heatmap and core enrichment of genes in NK-cell-mediated Cytotoxicity (B) and Calcium Signaling (C). Gene expression heatmap of core enriched genes in NK-cell-mediated Cytotoxicity (D) and Calcium Signaling (E).
Fig 4.
Sepsis impairs NK-cell capacity to produce IFN-γ in response to L. monocytogenes infection and cytokine stimulation.
(A) Experimental Design. 2 days after sham or CLP surgery mice were infected with of virulent Listeria monocytogenes (L.m. - 104 CFU, i.v.). The frequency or number of NK-cells in the spleen (B,C) or liver (D,E) 1 day after L.m. infection. (F) Representative flow plots of IFN-γ producing NK-cells. The frequency or number of IFN-γ+ NK-cells in the spleen (G,H) or liver (I,J). (K) Experimental Design: 2 days after surgery splenocytes from both groups of mice were harvested and stimulated with rIL-12 and rIL-18 or left unstimulated for 8 hrs. BFA was added during the last 4 hrs and intracellular cytokine production was evaluated. (L) Representative flow plots of IFN-γ producing NK-cells. (M) The frequency of IFN-γ+ NK-cells from Sham or CLP hosts in either unstimulated or stimulated wells. Data are representative from 2 independent experiments with 3–5 mice per group. Numbers above bars show fold change between groups. * p<0.05. Error bars represent the standard error of the mean.
Fig 5.
Sepsis results in numerical loss of Ly49H+ NK-cells and impaired in vivo killing.
(A) Experimental Design. Mice were sacrificed 2 days after sham or CLP surgery, and the number and Ly49H expression of Ly49H+ NK-cells in the spleen and liver determined. (B) Representative flow plots. Numbers indicate the frequency of Ly49+ NK-cells. The total number or GMFI of Ly49H by Ly49H+ NK-cells in spleen (C,E) or liver (D,F). (G) Experimental Design. Mice were treated with control IgG or α-NK1.1 depleting antibody prior to sepsis induction. Two days post-surgery all groups of mice received a 1:1 mixture of CFSE-labeled m157 expressing (m157-Tg) target (CFSElo) and m157-deficient littermate (WT) control (CFSEhi) cells. 3 hrs after injection mice were sacrificed and the ratio of m157-Tg to WT cells was determined. NK-depleted mice served as controls. (H) Representative flow plots. (I) m157 specific lysis in the spleen after Sham or CLP. (J) Correlation of m157 specific lysis with the number of Ly49H+ NK-cells in the spleen. Data are representative from 3 independent experiments with 3–5 mice per group. Numbers above bars show fold change between groups. * p<0.05. Error bars represent the standard error of the mean.
Fig 6.
Sepsis-induces NK-cell-intrinsic functional impairments of Ly49H and Ly49D receptors.
(A) Experimental Design. Splenocytes (day 2 post-surgery) were either incubated with: m157-Tg or WT cells for 6 hrs in the presence of monensin to determine ligand-induced degranulation; with m157 and WT cells for 18 hrs to determine in vitro killing; or 8 hrs stimulation with plate-bound antibody and BFA to determine IFN-γ production. (B) Frequency of CD107a+ Ly49H+ NK-cells in response to ligand (Stim[m157 targets]-Unstim[WT targets]) at indicated effector to target ratios. (C) Number of Ly49H+ NK-cells per well. (D) NK-cell-mediated m157 specific lysis after 18 hrs in vitro incubation. Representative flow plots of IFN-γ (E) or GzmB (F) following stimulation. Frequency of IFN-γ+ (G) or fold GzmB GMFI (H) of stimulated receptor+ NK-cells from Sham or CLP mice after 8 hrs stimulation with plate bound control (IgG), αNK1.1, αNKp46, αLy49D, or αLy49H antibody. Data are representative from 2 independent experiments with 3–5 mice per group. Numbers above bars show fold change between groups. * p<0.05, # p<0.05 group comparison by two-way ANOVA. Error bars represent the standard error of the mean.
Fig 7.
Sepsis-induced NK-cell intrinsic functional impairment is associated with reduced DAP12 adaptor protein expression and clustering.
(A) Experimental Design. Splenic NK-cells were purified by negative selection 2 days post-surgery and stimulated as indicated. Immunoblot (B) and ratio quantification (C) for DAP12 and GAPDH in unstimulated NK-cells from Sham and CLP mice, numbers indicate replicates. (D) Ratio of cells obtained from Sham or CLP mice adhering to the glass slide versus non-adherent cells. (E) Representative images of DAP12 staining for TIRF microscopy. (F) Mean DAP12 pixel intensity of adhered cells. Each data set is representative from 2 independent experiments with 5x105 cells per lane in panels B,C; 3 mice per group in panel D; >80 cells were analyzed per group from 3 mice per group in panels E,F. * p<0.05. Error bars represent the standard error of the mean.
Fig 8.
Reduced DAP12 expression is associated with defects in receptor signaling events.
(A) Experimental Design. Splenocytes were obtained 2 days post-surgery and were either labeled with α-Ly49H mAb and crosslinked with beads for 2 hrs and stained with AKT and pAKT antibodies or labeled with a calcium dye before being analyzed for calcium flux (baseline– 20 sec., Ly49H stim– 110 sec., Ionomycin– 50 sec.). Representative profiles or GMFI of AKT (B,D) or pAKT (C,E) in unstimulated and stimulated Ly49H+ NK-cells. Mean calcium dye fluorescence of Sham (F) or CLP (G) Ly49H+ or Ly49H- NK-cells during stimulation time course. (H) Area under the curve (AUC) for Ly49H- and Ly49H+ NK-cells during stimulation with α-Ly49H mAb. Data are representative from 2 independent experiments with 3 mice per group. * p<0.05. Error bars represent the standard error of the mean.
Fig 9.
Sepsis-induced loss of DAP12 is causal in NK-cell-intrinsic impairments.
(A) Experimental Design. BM chimeras containing mCherry+ control (Ctrl) vector or DAP12-overexpression (DAP12-OE) were generated. Following reconstitution mice underwent Sham or CLP surgery. 2 days after surgery IFN-γ production in response to DAP12-dependent receptor (Ly49H and Ly49D) stimulation was assessed. (B) Representative flow plot of mCherry expression on WT, Ctrl, and DAP12-OE Ly49H+ splenic NK-cells, numbers indicate frequency of mCherry+ cells. Frequency of IFN-γ+ of stimulated receptor+ NK-cells from Sham or CLP mice after 8 hrs stimulation with plate bound control (IgG), αLy49H (C), or αLy49D (D) antibody. Data are representative from 2 independent experiments with 2–3 mice per group. Numbers above bars show fold change between groups. * p<0.05. Error bars represent the standard error of the mean.
Fig 10.
Sepsis impairs NK-cell-mediated MCM viral control.
(A) Experimental Design. NK-sufficient or–deficient (IgG and α-NK1.1-treated, respectively) groups of mice were infected with the Smith strain of MCMV (105 PFU, i.p.) 2 days post-Sham or CLP surgery. Viral titers in the spleens (B) or livers (C) of Sham and CLP groups 3 days after MCMV infection. (Data are representative from 3 independent experiments with 3–5 mice per group. * p<0.05. Error bars represent the standard error of the mean.
Fig 11.
IL-2c therapy increases the number of Ly49H+ NK-cells.
(A) Experimental Design. One day after surgery mice were treated with IL-2c or IgG (control). The number of Ly49H+ NK-cells in the PBL (B,C), spleens (D), and livers (E) of Sham and CLP mice at the indicated days post-surgery. Data are representative from 2 independent experiments with 4–5 mice per group. * p<0.05. Error bars represent the standard error of the mean.
Fig 12.
IL-2c therapy does not improve DAP12 expression or intrinsic NK-cell function.
(A) Experimental Design. One day after surgery mice were treated with IL-2c or IgG (control). Cytokine production in response to receptor stimulation and DAP12 expression of NK-cells was determined 6 days later. (B) Frequency of IFN-γ+ of stimulated receptor+ NK-cells from Sham or CLP mice after 8 hrs stimulation with plate bound control (IgG), αLy49H, or αLy49D antibody. Immunoblot (C) and ratio quantification (D) for DAP12 and GAPDH in unstimulated NK-cells from Sham and CLP mice, numbers indicate replicates. Data are representative from 2 independent experiments with 4–5 mice per group in panel B and 1 independent experiment with 5x105 cells per lane in panels C,D. * p<0.05. Error bars represent the standard error of the mean.
Fig 13.
IL-2c therapy numerical recovery of Ly49H+ NK-cells partially rescues MCM viral control in CLP hosts.
(A) Experimental Design. One day after surgery mice were treated with IL-2c or IgG (control) and infected with MCMV 6 days later. Number of Ly49H+ NK-cells in the PBL of Sham (B) and CLP (C) mice at the time of MCMV infection. Viral titers in the spleens (D) or livers (E) of Sham and CLP groups of mice. Data are representative from 3 independent experiments with 3–5 mice per group. * p<0.05. Error bars represent the standard error of the mean.