High-fat diet impairs duodenal barrier function and elicits glia-dependent changes along the gut-brain axis that are required for anxiogenic and depressive-like behaviors

Abstract

Background: Mood and metabolic disorders are interrelated and may share common pathological processes. Autonomic neurons link the brain with the gastrointestinal tract and constitute a likely pathway for peripheral metabolic challenges to affect behaviors controlled by the brain. The activities of neurons along these pathways are regulated by glia, which exhibit phenotypic shifts in response to changes in their microenvironment. How glial changes might contribute to the behavioral effects of consuming a high-fat diet (HFD) is uncertain. Here, we tested the hypothesis that anxiogenic and depressive-like behaviors driven by consuming a HFD involve compromised duodenal barrier integrity and subsequent phenotypic changes to glia and neurons along the gut-brain axis.

Methods: C57Bl/6 male mice were exposed to a standard diet or HFD for 20 weeks. Bodyweight was monitored weekly and correlated with mucosa histological damage and duodenal expression of tight junction proteins ZO-1 and occludin at 0, 6, and 20 weeks. The expression of GFAP, TLR-4, BDNF, and DCX were investigated in duodenal myenteric plexus, nodose ganglia, and dentate gyrus of the hippocampus at the same time points. Dendritic spine number was measured in cultured neurons isolated from duodenal myenteric plexuses and hippocampi at weeks 0, 6, and 20. Depressive and anxiety behaviors were also assessed by tail suspension, forced swimming, and open field tests.

Results: HFD mice exhibited duodenal mucosa damage with marked infiltration of immune cells and decreased expression of ZO-1 and occludin that coincided with increasing body weight. Glial expression of GFAP and TLR4 increased in parallel in the duodenal myenteric plexuses, nodose ganglia, and hippocampus in a time-dependent manner. Glial changes were associated with a progressive decrease in BDNF, and DCX expression, fewer neuronal dendritic spines, and anxiogenic/depressive symptoms in HFD-treated mice. Fluorocitrate (FC), a glial metabolic poison, abolished these effects both in the enteric and central nervous systems and prevented behavioral alterations at week 20.

Conclusions: HFD impairs duodenal barrier integrity and produces behavioral changes consistent with depressive and anxiety phenotypes. HFD-driven changes in both peripheral and central nervous systems are glial-dependent, suggesting a potential glial role in the alteration of the gut-brain signaling that occurs during metabolic disorders and psychiatric co-morbidity.

Keywords: Behavioral disorders; Enteric glia; Glial signaling; Gut-brain axis; High-fat diet; Intestinal hyper-permeability.

Fig. 1

HFD evokes an intestinal inflammation and impairs duodenal mucosa without enteric glia involvement. Mice were fed with a standard diet (SD) or 72% high-fat diet (HFD) for 20 weeks, alone or with a daily IP of 10 μmol/kg fluorocitrate (FC) to investigate the enteric glia involvement in HFD-induced neuropathology. Representative images of duodenal cross-sections stained by hematoxylin and eosin with relative total histological damage score at weeks 6 (a, b) and 20 (e, f) of the diet protocol are shown, respectively. Representative images showing infiltrated CD45-positive cells within the duodenal microvilli and crypts and relative quantification at weeks 6 (c, d) and 20 (g, h). Immunoreactivity for Dapi (blue), ZO-1 (green), and occludin (red) and immunolabeling quantification for ZO-1 and occludin are also shown at weeks 6 (i, j, and l) and 20 (k, j, and l) of the diet protocol, respectively. Inserts within the merge pictures are enlarged areas. Data were analyzed by 2-way ANOVA or 1-way ANOVA and Dunnett post hoc. Results are expressed as cumulative histological damage, the average number of CD45+ cells or the average relative fluorescence units (RFU) ± SEM per area unit of n assessments. *P<0.05 and **P<0.01 versus relative SD group. Scale bars = 10 and 30 μm

Fig. 2

HFD induces glial-dependent changes in the duodenal myenteric plexus and dendritic spines decrease in neurons. Representative images of duodenal myenteric plexus from mice fed with a standard diet (SD) or high-fat diet (HFD) for 20 weeks, in the absence or presence of daily IP of 10 μmol/kg fluorocitrate (FC). GFAP (green), TLR4 (red), and Dapi (blue) immunoreactivity and relative immunolabeling quantification at weeks 6 and 20 (a–c) of diet protocol, respectively. BDNF (green) and HuCD (red) protein expression were also assessed at 6 and 20 weeks (d–f), respectively. The number of spines (PSD 95 immunoreactivity, red) was measured along neuronal dendrites (MAP-2 immunoreactivity, green) of cultured enteric neurons isolated from the duodenal myenteric plexuses of the above-mentioned experimental groups at weeks 6 and 20 (g–i) of diet protocol. Data were analyzed by 2-way ANOVA or 1-way ANOVA and Dunnett post hoc. Results are expressed as average relative fluorescence units (RFU) ± SEM per area unit or average number of dendritic spines/10 μm of n assessments. **P<0.01, ***P<0.001, and ****P<0.0001 versus relative SD group. Scale bars = 10 and 20 μm

Fig. 3

HFD evokes glial activation in the nodose ganglia after 20 weeks on diet. Immunofluorescence images show the expression of GFAP (green), TLR4 (red), and Dapi (blue) in the nodose ganglia and relative immunolabeling quantification at weeks 6 and 20 (a–c) of the diet protocol. BDNF (green) and HuCD (red) expression in nodose ganglia and relative quantification is shown at the same time points (d–f). Data were analyzed by 2-way ANOVA and Dunnett post hoc. Results are expressed as average relative fluorescence units (RFU) ± SEM per area unit of n assessments. *P<0.05, **P<0.01, ***P<0.001, and ****P<0.0001 versus relative SD group. Scale bar = 20 μm

Fig. 4

HFD drives glial-dependent changes in the hippocampus and neuronal dendritic spines decrease. Representative images of hippocampal slices from mice fed with a standard diet (SD) or high-fat diet (HFD) for 20 weeks, in the absence or presence of daily IP of 10 μmol/kg fluorocitrate (FC). Triple-label immunofluorescence for GFAP (green), TLR4 (red), and Dapi (blue) and related fluorescence intensity quantification at weeks 6 and 20 (a–c) of diet protocol are shown, respectively. In parallel, BDNF protein expression (green) in the dentate gyrus was assessed at 6 and 20 weeks (d and e), respectively. The number of dendritic spines (PSD 95 immunoreactivity, red) was measured in hippocampal neurons’ dendrites (MAP-2 immunoreactivity, green) isolated from the abovementioned experimental groups at weeks 6 (f and g) and 20 (f and h) of diet protocol. Data were analyzed by 2-way ANOVA and 1-way ANOVA and Dunnett post hoc. Results are expressed as average relative fluorescence units (RFU) ± SEM per area unit or the average number of dendritic spines/10 μm of n assessments. *P<0.05, **P<0.01, and ****P<0.0001 versus relative SD group. Scale bars = 10 and 20 μm

Fig. 5

HFD mediates depressive- and anxiety-like behaviors by involving a glial signaling pathway. a Tail suspension and b forced swimming tests were performed at days 0 and 140 of the diet protocol to assess the depressive-like behavior. Mice were suspended by the tail on a horizontal bar or placed in a cylinder filled with heated water and immobility time (in seconds) was recorded during the last 4 min of the 6-min test. Open field test was carried out to assess anxiety-like behavior by recording the total c mobility and d immobility time (in seconds), and e the number of entries in the center of the open field during the last 10 min of the 20-min test. Data were analyzed by 2-way ANOVA and Tukey post hoc. Results are expressed as the average time in seconds or the average number of entries/min ± SEM of n assessments. ***P<0.001 and ****P<0.0001 versus SD group at day 140 (week 20), °°°P<0.001 and °°°°P<0.0001 versus HFD group at day 140 (week 20), #P<0.05 for day 0 versus day 140 within the same group

Fig. 6

Schematic model of the gut-brain signaling during metabolic diseases and related behavioral disorders. Chronic exposure to high-fat diet (HFD) alters intestinal physiology, inducing a low-grade inflammation, dysbiosis, and increased mucosal permeability. Microbe-derived products that violate the impaired intestinal barrier may activate enteric glia within the enteric nervous system (ENS) via TLR4 pathway, leading to reactive gliosis and ENS neurons dysfunction. This elicits a glia-dependent signaling between the gastrointestinal tract and the brain responsible for impairing the energy metabolism and behavior, including depressive- and anxiety-like symptoms