Type I interferon signaling in hematopoietic cells is required for survival in mouse polymicrobial sepsis by regulating CXCL10

Abstract

Type I interferon (IFN) alpha/beta is critical for host defense. During endotoxicosis or highly lethal bacterial infections where systemic inflammation predominates, mice deficient in IFN-alpha/beta receptor (IFNAR) display decreased systemic inflammation and improved outcome. However, human sepsis mortality often occurs during a prolonged period of immunosuppression and not from exaggerated inflammation. We used a low lethality cecal ligation and puncture (CLP) model of sepsis to determine the role of type I IFNs in host defense during sepsis. Despite increased endotoxin resistance, IFNAR(-/-) and chimeric mice lacking IFNAR in hematopoietic cells display increased mortality to CLP. This was not associated with an altered early systemic inflammatory response, except for decreased CXCL10 production. IFNAR(-/-) mice display persistently elevated peritoneal bacterial counts compared with wild-type mice, reduced peritoneal neutrophil recruitment, and recruitment of neutrophils with poor phagocytic function despite normal to enhanced adaptive immune function during sepsis. Importantly, CXCL10 treatment of IFNAR(-/-) mice improves survival and decreases peritoneal bacterial loads, and CXCL10 increases mouse and human neutrophil phagocytosis. Using a low lethality sepsis model, we identify a critical role of type I IFN-dependent CXCL10 in host defense during polymicrobial sepsis by increasing neutrophil recruitment and function.

Figure 1.

Type I IFN deficiency worsens CLP but not endotoxicosis survival. (A) SEV129 and IFNAR^−/− mice underwent CLP surgery with a 27-gauge needle, and survival was monitored. The figure is the combination of two separate experiments with similar results (n = 24 per group; *, P = 0.0002 using Fisher’s exact test). (B) SEV129 and IFNAR^−/− mice were given an i.p. injection of 500 µg E. coli O111:B4 LPS, and survival was monitored for 7 d. The experiment was performed once (n = 12 per group; *, P = 0.0006 using Fisher’s exact test).

Figure 2.

Type I IFN signaling does not play a role in inflammation associated with CLP. (A) SEV129 and IFNAR^−/− mice underwent CLP surgery and were sacrificed at 0, 6, and 12 h after surgery. Serum cytokine levels from peripheral blood were determined by MILLIPLEX MAP Mouse Cytokine/Chemokine–Premixed 22 Plex kits. Select cytokines in this figure include TNF, KC, IL-6, IL-1β, and IP-10. Each time point was performed once (n = 3 per group per time point; *, P < 0.05 using the Student’s t test). Error bars indicate SD. (B) SEV129 and IFNAR^−/−mice underwent CLP surgery and were sacrificed at 12, 48, and 96 h after surgery. Bacteremia was determined from peritoneal lavage fluid plated on sheep blood agar. Each point represents CFUs from one mouse. The experiment was performed three times (n = 3 per group; P < 0.05 using the Student’s t test). Horizontal bars indicate means.

Figure 3.

Type I IFN signaling in the hematopoietic system is needed for survival to CLP. SEV129 mice irradiated and reconstituted with BM from SEV129 or IFNAR^−/− mice underwent CLP surgery with a 27-gauge needle, and survival was monitored for 12 d. The figure is the combination of two separate experiments with similar results (n = 26 for the SEV129 group and 27 for the IFNAR^−/− group; *, P = 0.0006 using Fisher’s exact test).

Figure 4.

Neutrophil and macrophage recruitment and phagocytic function require intact type I IFN signaling. SEV129 wild-type and IFNAR^−/− mice underwent CLP surgery and were sacrificed at 12, 48, and 96 h after surgery. Peritoneal cells were stained extracellularly for flow cytometry. (A) Total neutrophil counts were determined by gating on Gr1⁺CD11b⁺ cells. (B) The total number of macrophages was determined by gating on CD11b⁺F4/80⁺ cells. The dashed lines indicate the total number of neutrophils (A) or macrophages (B) found in sham animals. (C and D) 100,000 cells were incubated with FITC beads for 30 min at 37°C. Cells were washed two times with PBS and stained extracellularly for flow cytometry. (C) Cells were first gated on neutrophils (Gr1⁺CD11b⁺ cells), and then FITC⁺ cells were considered phagocytic. (D) Cells were first gated on macrophages (CD11b⁺F4/80⁺ cells), and then FITC⁺ cells were considered phagocytic. The dashed lines indicate sham phagocytic levels. *, P < 0.05 using the Student’s t test. Error bars indicate SD. The data shown for all panels were obtained with 10 mice per group performed over three independent experiments.

Figure 5.

CXCL10 improves outcome by decreasing bacteremia in IFNAR^−/− mice. (A) SEV129 wild-type mice (n = 10), IFNAR^−/− mice (n = 11), or IFNAR^−/− mice with 100 ng CXCL10 6 h and on day 3 after CLP (n = 11). Survival was monitored for 7 d. There was a 50% survival advantage in SEV129 wild-type compared with IFNAR^−/− mice, and a 54% survival advantage in IFNAR^−/− + CXCL10 compared with IFNAR^−/− mice. *, P = 0.01 using Fisher’s exact test. Data are from a single experiment with 10 mice in the SEV129 group and 11 mice in the IFNAR^−/− and IFNAR^−/− + CXCL10 groups. Two independent survival experiments were performed with consistent results. (B) IFNAR^−/− mice underwent CLP surgery, and 6 h after surgery mice were injected with PBS or 100 ng of recombinant IP-10. Mice were sacrificed 48 h after CLP, and bacteremia was determined from dilutions of peritoneal lavage fluid obtained aseptically. Each point represents CFUs from one mouse. P < 0.05 using the Student’s t test. Horizontal bars indicate means. The figure represents data from two independent experiments using four mice per group with similar results. (C) SEV129 wild-type and IFNAR^−/− mice were injected with 100 ng CXCL10. Peritoneal cells were harvested 18 h later and were examined by flow cytometry for phagocytic neutrophils (Gr1⁺CD11b⁺ cells containing FITC⁺ latex beads). *, P = 0.001 using one-way ANOVA; *, P < 0.05 using the Tukey post hoc analysis. Error bars indicate SD. The figure represents data from three independent experiments using at least four mice per group with similar results.

Figure 6.

CXCL10 treatment improves human neutrophil phagocytic function in vitro. Whole-blood leukocytes were collected using Histopaque 1119. 10⁵ unstimulated cells, cells treated with 100 ng CXCL10 for 4 or 2 h followed by an 18-h stimulation with 1 µg/ml LPS, or cells stimulated with LPS alone for 18 h were plated in round-bottom 96-well plates. FITC beads were added and cells were incubated at 37°C for 30 min. Cells were washed and stained for neutrophil markers (CD66b⁺CD16^hi). (A) Representative histograms of the percentage of phagocytic neutrophils from unstimulated (top) or CXCL10 treatment for 4 h (bottom). (B) Graphical analysis of the percentage of phagocytic neutrophils from one healthy control (P = 0.019 using the Student’s t test). (C) Representative histograms of the percentage of phagocytic neutrophils from unstimulated (top), LPS-stimulated (middle), and CXCL10-treated + LPS-treated leukocytes (bottom). (D) Graph of the percentage of phagocytic neutrophils from one healthy control (P < 0.001 using one-way ANOVA). For all panels, treatment was performed in triplicate and the experiment was performed on three healthy controls with similar results. Error bars indicate SD.