High fat diet causes depletion of intestinal eosinophils associated with intestinal permeability

Abstract

The development of intestinal permeability and the penetration of microbial products are key factors associated with the onset of metabolic disease. However, the mechanisms underlying this remain unclear. Here we show that, unlike liver or adipose tissue, high fat diet (HFD)/obesity in mice does not cause monocyte/macrophage infiltration into the intestine or pro-inflammatory changes in gene expression. Rather HFD causes depletion of intestinal eosinophils associated with the onset of intestinal permeability. Intestinal eosinophil numbers were restored by returning HFD fed mice to normal chow and were unchanged in leptin-deficient (Ob/Ob) mice, indicating that eosinophil depletion is caused specifically by a high fat diet and not obesity per se. Analysis of different aspects of intestinal permeability in HFD fed and Ob/Ob mice shows an association between eosinophil depletion and ileal paracelullar permeability, as well as leakage of albumin into the feces, but not overall permeability to FITC dextran. These findings provide the first evidence that a high fat diet causes intestinal eosinophil depletion, rather than inflammation, which may contribute to defective barrier integrity and the onset of metabolic disease.

Fig 1. HFD causes intestinal permeability and eosinophil depletion.

Intestinal permeability was determined after 7 days HFD by FITC dextran assay (A) or at day 0, 1, 2, 3 HFD by fecal albumin (B). Each data point represents an individual mouse from one (FITC dextran) or two experiments (fecal albumin). (C) Plasma FITC dextran concentration was measured at time 0, 1, 2 and 5 hours after oral administration to indicate the peak of permeability. Each data point is the mean (+/-SD) of 4–8 mice. (D) Gating strategy to identify eosinophils and macrophages. Lamina propria cells were pre-gated as live, CD45⁺. CD11b⁺ cells were compared for expression of Siglec F and MHC-II revealing two distinct populations. Siglec F⁺ MHC-II^- eosinophils (red) show high SSC, intermediate F4/80 and CD11c expression whilst Siglec F^- MHC-II⁺ macrophages (blue) show low SSC, high F4/80 and CD11c expression. (E) Proportions and total number of eosinophils and macrophages in normal chow and 7 days HFD mice. Each data point represents an individual mouse from two experiments. (F) The total number of Siglec F⁺ (red) eosinophils and CD11b⁺ (red) MHC-II (green) macrophages per villus shown by immunofluorescence microscopy. Each graph represents multiple villi from >30 sections from 3 mice per group. *p<0.05, **p<0.01, ***p<0.001 (Mann-Whitney U test).

Fig 2. Lack of detectable inflammation in the intestine of HFD mice

(A) CX3CR1^GFP/+ monocytes/macrophages were quantified per villus by immunofluorescence microscopy. Multiple villi were assessed from >30 sections with 3 mice per group. (B) Lamina propria cells were isolated after 1 week or 16 weeks HFD and monocyte populations were analyzed as live, CD45+, CD11b⁺Ly6C⁺ cells. Each data point represents an individual mouse from one experiment. (C) The trafficking of adoptively transferred PKH26+ monocytes to the intestine is not enhanced in mice fed HFD for 8 weeks. Representative flow cyometry plots of a control receiving PBS in place of monocytes, normal chow recipient and HFD recipient mice are shown. Bar graph shows from n = 6 mice per group from one experiment. D) Inflammatory gene expression in the small intestinal mucosa of 1 week HFD or normal chow mice. *p<0.05, **p<0.01, ***p<0.001 (Mann-WhitneyU test). E) Representative H&E histology from the small intestine or colon of 6 week HFD or Low fat diet fed mice. F) Eicosanoid quantification in the ileum of mice fed HFD for one week or maintained on normal chow. Each data point represents an individual mouse from one experiment. *p<0.05, **p<0.01 (One-way ANOVA with Bonferroni post test).

Fig 3. Evidence of an eosinophil trafficking defect in HFD mice.

(A) Ileal sections from normal chow or 7 day HFD mice were stained for in situ apoptotic cells using TdT-Fluor staining protocol. Minimal apoptosis was detected in either condition relative to nuclease treated positive control. (B) Bone marrow and (C) peripheral blood leukocytes were isolated from 5 week HFD mice and proportions of eosinophils determined by flow cytometry. D) Expression of CCL11 and CCL24 genes in the small intestinal mucosa of 1 week HFD or normal chow mice. In all graphs each data point represents an individual mouse from one experiment. *p<0.05, **p<0.01, ***p<0.001 Mann-Whitney U test.

Fig 4. Intestinal eosinophil depletion is caused by HFD not obesity.

(A) Lamina propria cells were isolated from 12 week old Ob/Ob mice and littermate controls or (B) HFD fed, normal chow fed or mice switched from HFD to normal chow (HFD-NC). Eosinophil populations assessed by flow cytometry as shown in Fig 1. Each data point represents an individual mouse from one experiment. *p<0.05, **p<0.01, ***p<0.001 Mann-Whitney U test.

Fig 5. Intestinal permeability compared in HFD and Ob/Ob mice.

(A) 4kDa Fitc-dextran flux across jejunum or ileum of normal chow (n = 4–6), HFD (n = 4) or Ob/Ob (n = 4) mice measured in Ussing chambers. (B) FITC-dextran concentration in the plasma of Ob/Ob mice and wild-type littermates following oral administration each data point represents an individual mouse (n = 8 per group) from one experiment. (C) Transepithelial conductance across jejunum or ileum of normal chow (n = 4–6), HFD (n = 4) or Ob/Ob (n = 4) mice measured in Ussing chambers D) Fecal albumin concentration in Ob/Ob mice, wildtype littermates, normal chow and HFD mice. Each data point represents and individual mouse from one experiment. *p<0.05, **p<0.01, ***p<0.001 by one way ANOVA (Ussing chamber analysis) or Mann-Whitney U test (in vivo FITC dextran or fecal albumin assays).