Vagal Regulation of Group 3 Innate Lymphoid Cells and the Immunoresolvent PCTR1 Controls Infection Resolution

Abstract

Uncovering mechanisms that control immune responses in the resolution of bacterial infections is critical for the development of new therapeutic strategies that resolve infectious inflammation without unwanted side effects. We found that disruption of the vagal system in mice delayed resolution of Escherichia coli infection. Dissection of the right vagus decreased peritoneal group 3 innate lymphoid cell (ILC3) numbers and altered peritoneal macrophage responses. Vagotomy resulted in an inflammatory peritoneal lipid mediator profile characterized by reduced concentrations of pro-resolving mediators, including the protective immunoresolvent PCTR1, along with elevated inflammation-initiating eicosanoids. We found that acetylcholine upregulated the PCTR biosynthetic pathway in ILC3s. Administration of PCTR1 or ILC3s to vagotomized mice restored tissue resolution tone and host responses to E. coli infections. Together these findings elucidate a host protective mechanism mediated by ILC3-derived pro-resolving circuit, including PCTR1, that is controlled by local neuronal output to regulate tissue resolution tone and myeloid cell responses.

Keywords: infections; inflammation; lipid mediator profiling; macrophages; omega-3; protectins; vagal reflex.

Copyright © 2017 Elsevier Inc. All rights reserved.

Figures

Figure 1. Vagotomy diminishes PCTR1 concentrations and elevated eicosanoids in peritoneum

Mice were subjected to right unilateral vagotomy or sham surgery, at the indicated intervals the peritoneal cavity was lavaged and lipid mediators profiled using LC-MS-MS based LM metabololipidomics (see methods for details). (a) Representative multiple reaction chromatograms for identified lipid mediators, and (b) MS-MS spectrum employed in the identification of PCTR1. (c) Lipid mediator profiles were then interrogated using Partial least square discriminant analysis (PLS-DA). Top panel: 2D score plot bottom panel: 2D loading plot. (d) Peritoneal acetylcholine levels. Results for a-c are representative of n=5–6 mice per group from two distinct experiments. Results for d are mean± s.e.m. n=4–5 mice per group **p<0.01 vs Sham mice. m/z = mass to charge ratio.

Figure 2. Resident peritoneal ILC3 numbers are reduced and peritoneal macrophage phenotype is altered in vagotomised mice

Mice were subjected to unilateral vagotomy or sham surgery, and peritoneal cells were collected at the indicated intervals. (a) ILC3 and (b) macrophage numbers were determined using light microscopy and flow cytometry. (c) Peritoneal macrophage phenotypic marker expression was determined using flow cytometry and results interrogated using PLS-DA. Top panel: 2D score plot bottom panel: 2D loading plot. (d) Peritoneal macrophages were isolated and lipid mediators determined using lipid mediator profiling. Results were then interrogated using PLS-DA. Top panel: 2D score plot bottom panel: 2D loading plot. Results for a, b are mean ± s.e.m. 4–5 mice per group from two distinct experiments. Results for c, d are representative of 3–4 mice per group. Related to Figure S2, S3 and Table S1, S2

Figure 3. Vagotomy delayed resolution of infections

Mice were subjected to unilateral vagotomy or sham surgery; after 7 days E. coli (1×106 CFU/mouse) were administered via i.p. injection. Inflammatory exudates were collected at the indicated intervals: (a) cell counts determined using light microscopy and flow cytometry and resolution indices calculated. (b) Exudate leukocyte phagocytosis of E. coli (CD11b+E. coli+ cells) was determined by flow cytometry in 12h exudates (c) 12h exudate bacterial loads. Results are mean ± s.e.m. n= 4 mice per group from two distinct experiments. Related to Figure S3

Figure 4. PCTR1 restores resolution responses in vagotomised mice during bacterial infections

(a–c) Mice were subjected to unilateral vagotomy or sham surgery, after 7 days vagotomised mice were administered vehicle (saline containing 0.01% EtOH) or PCTR1 (75ng/mouse) via i.p. injection. After 16h (a) peritoneal lavages were collected and macrophage phenotype was determined using flow cytometry and (b) peritoneal LM profiles were determined using LM metabololipidomics. (c) peritoneal cells were collected, plated in 96-well plates (5×104 cells/well; DPBS; 2h; 37°C) then incubated with fluorescently labeled E. coli (2.5×106 CFU/well; 37°C, PBS+/+, pH 7.45, 40 min) and phagocytosis assessed using a fluorescence microscopy. Results are representative of n= 5–6 mice per group from two distinct experiments. (d–g) Mice were administered E. coli (1×106 CFU/mouse, i.p.), (d) exudates collected at the indicated intervals and neutrophil count determined using flow cytometry and light microscopy. (e) 12h Exudate leukocyte E. coli phagocytosis determined as CD11b+E. coli+ cells, (f) 12h exudate bacterial loads and (g) 24h exudate macrophage efferocytosis determined as F4/80+ Ly6G+ cells. Results for c are mean±s.e.m. n= 5–6 mice per group from two distinct experiments. Related to Table S3 and S4

Figure 5. ILC3 regulate peritoneal macrophage phenotype, PCTR1 production and resolution of infectious-inflammation

Mice were subjected to right unilateral vagotomy or sham. Vagotomised mice were then administered vehicle (Vagotomy; PBS) or Carbachol (Vagotomy+Carbachol; 0.1mg/Kg/day for 7days). Peritoneal lavages were collected (a) expression of macrophage phenotypic markers (b) CD335+ ILC3 and (c) PCTR1 levels determined. Results for a are representative of n=3 mice per group. Results for b,c are mean±sem of n=3 mice per group. * p<0.05. (d–e) Rag1−/− mice were administered an anti-CD90.2 (250µg/mouse) or Isotype control (250µg/mouse) antibody, (d) peritoneal lavages were collected and lipid mediator profiles determined using LC-MS-MS based LM metabololipidomics and PLS-DA. Left panel 2D score plot, right panel 2D loading plot. (e) Peritoneal macrophages were collected and phenotypic marker expression determined using flow cytometry. Left panel 2D score plot, right panel 2D loading plot. Results are representative of n= 4 mice per group from two distinct experiments. (f–h) Rag1−/− mice were administered an anti-CD90.2 (250µg/mouse) or Isotype control (250µg/mouse) antibody. After 3 days they were given E. coli (1×105 CFU/mouse) via i.p. injection, exudates collected at the indicated intervals and (f) cell counts were determined using light microscopy and flow cytometry and resolution indices calculated. (g) Exudate leukocyte phagocytosis of E. coli (CD11b+E. coli+ cells) was determined by flow cytometry in 12h exudates. (h) 12h exudate bacterial loads. Results are mean ± s.e.m. n= 4 mice per group from two distinct experiments. Related to Table S6

Figure 6. Acetylcholine upregulates SPM biosynthetic pathways in human and mouse ILC3 that restore peritoneal host responses to infections in vagotomised mice

(a) Expression of mouse 12/15-LOX in mouse CD335+ ILC3 (b) mouse ILC3 were isolated using FACS sorting (see methods) then incubated with DHA (300ng/ml) then with ACh (1nM, 37°C, PBS+/+, pH 7.45, 60 min). Incubations were then quenched and 17-HDHA levels assessed using LM metabololipidomics. Results for a are representative of n = 3 mice. Results for b are mean ± s.e.m. n=3 mice per group. (c) peritoneal macrophages (Mφ; 1×106 cells/well) from vagotomised mice were incubated with mCD335+ ILC3 (7.5×102 cells/well, 6h, 37°C, PBS+/+, 2%FBS). Incubations were quenched and PCTR1 levels determined using lipid mediator profiling. Results are mean±sem n=3 mice per group. *p<0.05 vs Mφ; #p < 0.05 Mφ plus ILC3. (d) omenta were isolated from naive mice and stained for choline acetyltransferase (red), CD3 (green), F4/80 (cyan) and RoRγt (white) and staining was imaged using confocal microscopy (see methods for details). Magnification × 40. Results are representative of n-4 mice. (e) Mice were subjected to unilateral vagotomy surgery, after 7 days they were administered vehicle, CD335+ ILC3 (5×102 cells), DHA (1µM) plus ACh (1nM) or CD335+ ILC3 (5×102 cells), DHA (1µM), ACh (1nM) plus Lipoxygenase Inhibitor (10µM; L.I.; see methods for details) 3 h prior to E. coli (1×106 CFU/mouse). At the indicated intervals leukocytes were collected and neutrophil counts determined by light microscopy and flow cytometry. (f) Exudate bacterial loads 12h post E. coli inoculation. Results are mean ± s.e.m. n= 4 mice per group from two distinct experiments. (g) Human ILC3 expression of 15-lipoxygenase type I (15-LOX) was determined using flow cytometry. Results for a are representative of n = 4 donors. (h) Mice were subjected to unilateral vagotomy surgery, after 7 days they were administered vehicle, ILC3 (1×103 cells), DHA (1µM), plus ACh (1nM) or ILC3 (1×103 cells), DHA (1µM), ACh (1nM) plus Lipoxygenase Inhibitor (10µM; L.I.; see methods for details) 3 h prior to E. coli (1×106 CFU/mouse; see methods for details). At the indicated intervals leukocytes were collected and neutrophil counts determined by light microscopy and flow cytometry. Results are mean ± s.e.m. n=6 mice per group from two distinct experiments.