Enhanced Bacteremia in Dextran Sulfate-Induced Colitis in Splenectomy Mice Correlates with Gut Dysbiosis and LPS Tolerance

Abstract

Because both endotoxemia and gut dysbiosis post-splenectomy might be associated with systemic infection, the susceptibility against infection was tested by dextran sulfate solution (DSS)-induced colitis and lipopolysaccharide (LPS) injection models in splenectomy mice with macrophage experiments. Here, splenectomy induced a gut barrier defect (FITC-dextran assay, endotoxemia, bacteria in mesenteric lymph nodes, and the loss of enterocyte tight junction) and gut dysbiosis (increased Proteobacteria by fecal microbiome analysis) without systemic inflammation (serum IL-6). In parallel, DSS induced more severe mucositis in splenectomy mice than sham-DSS mice, as indicated by mortality, stool consistency, gut barrier defect, serum cytokines, and blood bacterial burdens. The presence of green fluorescent-producing (GFP) E. coli in the spleen of sham-DSS mice after an oral gavage supported a crucial role of the spleen in the control of bacteria from gut translocation. Additionally, LPS administration in splenectomy mice induced lower serum cytokines (TNF-α and IL-6) than LPS-administered sham mice, perhaps due to LPS tolerance from pre-existing post-splenectomy endotoxemia. In macrophages, LPS tolerance (sequential LPS stimulation) demonstrated lower cell activities than the single LPS stimulation, as indicated by the reduction in supernatant cytokines, pro-inflammatory genes (iNOS and IL-1β), cell energy status (extracellular flux analysis), and enzymes of the glycolysis pathway (proteomic analysis). In conclusion, a gut barrier defect after splenectomy was vulnerable to enterocyte injury (such as DSS), which caused severe bacteremia due to defects in microbial control (asplenia) and endotoxemia-induced LPS tolerance. Hence, gut dysbiosis and gut bacterial translocation in patients with a splenectomy might be associated with systemic infection, and gut-barrier monitoring or intestinal tight-junction strengthening may be useful.

Keywords: dysbiosis; endotoxemia; gut barrier defect; splenectomy.

Figure 1

Characteristics of mice after sham or splenectomy (Splx) as determined by leaky gut (FITC-dextran assay and endotoxemia) (A,B), bacterial burdens in mesenteric lymph nodes, and mean fluorescent intensity of intestinal tight-junction molecules, occludin and zona occludens-1 (ZO-1), with the representative immune-fluorescent staining pictures (C–E), systemic inflammation (serum IL-6), and organ injury markers (serum creatinine and alanine transaminase) (F–H) are indicated (n = 7/group or time-point).

Figure 2

Characteristics of sham or splenectomy (Splx) mice (at 2 weeks post-surgery) after 1 week administration by dextran sulfate solution (DSS) or control drinking water (water) as indicated by the schema of experiments (A), survival analysis (B), stool consistency index (C), body weight (D), colon histological score (E), gut barrier defect (endotoxemia and FIT-dextran assay) (F,G), and serum cytokines (TNF-α, IL-6, and IL-10) (H–J) are demonstrated (n = 15/group for B-D and n = 8/group for others).

Figure 3

Representative pictures of colon histology from hematoxylin and eosin (H&E) staining of sham or splenectomy (Splx) mice (at 2 weeks post-surgery) after 1 week administration by dextran sulfate solution (DSS) or control drinking water (water) are demonstrated. Arrows indicate the loss of mucosal epithelium in the lesions.

Figure 4

Characteristics of sham or splenectomy (Splx) mice (at 2 weeks post-surgery) after 1 week administration by dextran sulfate solution (DSS) or control drinking water (water) as indicated by blood bacterial burdens (bacteremia) and the list of identified bacteria based on bacterial colony characteristics (mass spectrometry analysis) (A), and the fluorescent intensity of green fluorescent protein (GFP)-expressing Escherichia coli (E. coli) in several organs (mesenteric lymph nodes, kidneys, livers, and spleens) with representative pictures (B,C) are demonstrated (n = 10/group for A, B, and n = 5/group for C). Of note, GFP E. coli were detectable only in mesenteric lymph nodes, but not other organs, of splenectomy mice with drinking water (Splx water) and non-detectable in all organs of sham mice with drinking water control (sham water).

Figure 5

Fecal microbiota analysis of sham or splenectomy (Splx) mice (at 2 weeks post-surgery) after 1 week administration by dextran sulfate solution (DSS) or control drinking water (water) as indicated by the relative abundance of bacterial diversity at phylum and genus with the average calculation (A,B), the graph presentation of some analysis at phylum and at the genus level (C,D), total Gram-negative bacteria in feces (E), and the beta diversity analysis (Chao-1 and Shannon analysis) (F) are demonstrated.

Figure 6

Characteristics of sham or splenectomy (Splx) mice (at 2 weeks post-surgery) with lipopolysaccharide (LPS) intravenous administration (tail vein) in a single dose, using phosphate buffer solution (PBS) followed by LPS (PBS/LPS), or two doses (LPS tolerance) (LPS/LPS) as indicated by the schema of experiments (A), time-points of serum cytokines (TNF-α, IL-6, and IL-10) (B–D), and the graph presentation of peak cytokine levels (15 min post-injection) (E–G) are demonstrated (n = 7/group or time-point).

Figure 7

Characteristics of bone marrow-derived macrophages at 24 h after 2-doses stimulation (24 h dose separation) by two doses of phosphate buffer solution (PBS/PBS) (A–L), a single dose of lipopolysaccharide (LPS), or twice LPS (LPS/LPS) as indicated by the schema of the experiments (A), time-points of serum cytokines (TNF-α, IL-6, and IL-10) (B–D), gene expression of pro-inflammation (iNOS and IL-1β), and anti-inflammation (Arginase-1 and TGF-β), (E–H) with cell energy status analysis using oxygen consumption rate (OCR) of the mitochondrial stress test, extracellular acidification rate (ECAR) for the glycolysis stress test with graph presentation of respiratory capacity (maximal respiration), and glycolysis capacity (maximal glycolysis) are demonstrated (independent triplicate experiments were performed).

Figure 8

Illustrations of mapped proteins that are associated with cell energy metabolisms from proteomic analysis of bone marrow-derived macrophages after stimulation with two lipopolysaccharide (LPS) administrations (LPS tolerance) relative to single LPS stimulation are demonstrated. The proteins in red- and green-colored texts represent upregulated and downregulated proteins from proteomic analysis, respectively.

Figure 9

Illustration of the working hypothesis indicating (i) the effective organismal control by spleen after transient bacteremia from physiologic gut barrier defect [45,46], (ii) the loss of microbial control after splenectomy causes minimal gut bacterial translocation but in the level that is enough to control by non-splenic immune responses, as indicated by chronic endotoxemia (splenectomy-induced endotoxemia) without bacteremia [10] (the systemic LPS responses induce stress that additionally facilitates gut dysbiosis [33,47]), (iii) splenectomy-induced gut-dysbiosis enhanced susceptibility to mucositis from dextran sulfate solution (DSS)-induced injury with the higher level of gut bacterial translocation than non-DSS mice (more prominent bacteremia), partly due to the limited cytokine responses from asplenia and chronic endotoxemia-induced macrophage LPS tolerance. Picture is created by https://biorender.com/ (accessed on 10 December 2021).