Metabolic activation and colitis pathogenesis is prevented by lymphotoxin β receptor expression in neutrophils

Abstract

Inflammatory bowel disease is characterized by an exacerbated intestinal immune response, but the critical mechanisms regulating immune activation remain incompletely understood. We previously reported that the TNF-superfamily molecule TNFSF14 (LIGHT) is required for preventing severe disease in mouse models of colitis. In addition, deletion of lymphotoxin beta receptor (LTβR), which binds LIGHT, also led to aggravated colitis pathogenesis. Here, we aimed to determine the cell type(s) requiring LTβR and the mechanism critical for exacerbation of colitis. Specific deletion of LTβR in neutrophils (LTβR^ΔN), but not in several other cell types, was sufficient to induce aggravated colitis and colonic neutrophil accumulation. Mechanistically, RNA-Seq analysis revealed LIGHT-induced suppression of cellular metabolism, and mitochondrial function, that was dependent on LTβR. Functional studies confirmed increased mitochondrial mass and activity, associated with excessive mitochondrial ROS production and elevated glycolysis at steady-state and during colitis. Targeting these metabolic changes rescued exacerbated disease severity. Our results demonstrate that LIGHT signals to LTβR on neutrophils to suppress metabolic activation and thereby prevents exacerbated immune pathogenesis during colitis.

Fig. 1. LTβR expression in neutrophils protects from DSS-colitis.

Colitis was induced in Ltbr^fl/fl × Mrp8-Cre (LTβR^ΔN) and control littermates by cycles of 2.5% DSS for 5 days followed by 2 days of water and disease progression was monitored. a weight loss was monitored daily as % starting body weight. b Colon length on day 9. c H&E stained representative histology sections of distal colon at 20× (top; scale bar, 200 μm) and 40× (bottom, scale bar, 100 μm) magnification. d Blinded total histological score from distal colon. Representative immunohistochemical MPO staining (e) and quantification (f) of MPO-positive neutrophils from sections of distal colon tissue. Scale bar, 50 μm g Kaplan–Meier survival plots of the indicated groups during two cycles of DSS. Data representative of >3 independent experiments (a, c, e-g) or pooled data from three experiments (b, d). Student’s t test (b, d, f), two-way ANOVA (a, g). Error bars represent SEM.

Fig. 2. LIGHT signaling represses genes associated with mitochondrial function and ROS.

RNA-Seq analysis of sorted blood neutrophils from WT mice stimulated with a DSS-associated cytokine cocktail and with or without recombinant mouse LIGHT. a The top 50 most differentially expressed genes with respect to P value. Genes in bold font are associated with mitochondrial and/or ROS pathways. b Gene expression pathways upregulated in the absence of LIGHT stimulation were determined by the ToppGene Suite. c Ingenuity Pathway Analysis (Qiagen) identified oxidative phosphorylation as a pathway highly upregulated in the absence of LIGHT stimulation. Significantly upregulated genes and complexes of the electron transport chain are depicted in red. Gray: not significantly different. White: not determined in dataset. d, e GSEA of the comparison of WT neutrophils and WT stimulated with LIGHT. d GSEA for genes associated with mitochondrial function from the molecular signature database list GO_mitochondrial_gene_expression. ***P < 0.001; NES = 2.1; FDR < 0.001. e GSEA for genes associated with cellular ROS pathways from the molecular signature database list Houstis_ROS. **P < 0.01; NES = 1.7; FDR < 0.01. GEO accession number GSE150243.

Fig. 3. Energy-metabolic pathways are suppressed by LTβR.

Neutrophils were detected from BM of Ltbr^fl/fl × Mrp8-Cre (LTβR^ΔN) and control littermates by flow cytometry (a, f, g) or by negative magnetic enrichment (STEMCCELL technologies) (b–e) and metabolic parameters were measured. Mitochondrial mass was measured as mean intensity of Mitotracker staining (a) and mean Tom20 immunofluorescence intensity by confocal microscopy (b). c, d Tom20 foci indicative of mitochondria were quantified from z-stacked confocal micrographs of Ly6G+ cells (red) stained with Tom20 antibody (green) and quantified by Imaris. Organization of Tom20 in clusters and the 3D-shape reflected in z-stacks can lead to multiple individual foci recorded for one mitochondrium. Scale bar, 500 nm e Representative TEM images (right) and blinded quantification of granule and mitochondrial content (left). Red arrows indicate individual mitochondria. Scale bar, 500 nm. f Neutrophil mitochondrial superoxides were measured by MitoSox and quantified by flow cytometry as mean fluorescence intensity. g Representative gating and histograms (left) and quantification (right) of neutrophil oxidative burst assay at baseline and in response to fMLP (fMLP N-formylmethionine-leucyl-phenylalanine; DHR dihydrorhodamine-123). h Fluorescent glucose uptake was quantified using the fluorescent glucose analog 2-NBDG (Invitrogen) by flow cytometry. Student’s t test, data from 3–6 mice/group from 2 or more independent experiments. Error bars represent SEM.

Fig. 4. LTβR dampens neutrophil respiration and glycolysis.

CD11b+ Ly6G+ Neutrophils were isolated from BM of Ltbr^fl/fl × Mrp8-Cre (LTβR^ΔN) and control littermates (pooled from 5 mice/group) by magnetic sorting and their metabolic parameters were measured. a Oxygen consumption rate (OCR) and b extracellular acidification rate (ECAR) were measured in real time at baseline and in response to the indicated mitochondrial inhibitors. See “Methods” for details. c Basal and maximal mitochondrial respiration and glycolysis are mapped to illustrate pathway preference. ATP-linked respiration (d) and maximal respiratory capacity (e) are quantified as ΔOCR from flux analysis. Data from 5 mice/group, measured as mean of 5 wells for each timepoint, representing one of three independent experiments. Error bars represent SEM.

Fig. 5. Colon inflammation is associated with metabolic neutrophil activation in the absence of LTβR.

Colitis was induced in Ltbr^fl/fl × Mrp8-Cre (LTβR^ΔN) and control littermates by 2.5% DSS and disease and metabolic parameters were measured in colonic neutrophils. a Relative weight loss expressed as % of starting body weight. Quantification of neutrophil Mitotracker mean fluorescence intensity by flow cytometry (b) and representative histograms of colon neutrophil mitochondrial mass (c). d Mitochondrial superoxide production quantified by MitoSox by mean fluorescence intensity (left) and percent positive cells (right). Fluorescent glucose 2-NBDG uptake in neutrophils isolated from the colon quantified by flow cytometry (e) and representative histogram of 2-NBDG uptake (f). Student’s t test. Data from 3–4 mice/group, representative of 3 experiments. Error bars represent SEM.

Fig. 6. LTβR driven neutrophil metabolic activation causes exacerbated DSS-induced colitis.

Colitis was induced in Ltbr^fl/fl × Mrp8-Cre (LTβR^ΔN) and control littermates by 2.5% DSS. Metformin (Mtf, 200 mg/kg) or PBS was administered every other day by oral gavage and disease parameters were measured over time. a Weight loss was monitored daily as % starting body weight. b Colon length at endpoint of experiment. c H&E stained representative sections from distal colon with (d) blinded total histological score. Data represents three independent experiments (a, c) or is combined from three independent experiments (b, d). Significance tested by two-way ANOVA (a) or student’s t test (b, d). Scale bar, 200 μm. Error bars represent SEM.