Suppression of gut dysbiosis reverses Western diet-induced vascular dysfunction

Abstract

Vascular dysfunction represents a critical preclinical step in the development of cardiovascular disease. We examined the role of the gut microbiota in the development of obesity-related vascular dysfunction. Male C57BL/6J mice were fed either a standard diet (SD) ( n = 12) or Western diet (WD) ( n = 24) for 5 mo, after which time WD mice were randomized to receive either unsupplemented drinking water or water containing a broad-spectrum antibiotic cocktail (WD + Abx) ( n = 12/group) for 2 mo. Seven months of WD caused gut dysbiosis, increased arterial stiffness (SD 412.0 ± 6.0 vs. WD 458.3 ± 9.0 cm/s, P < 0.05) and endothelial dysfunction (28% decrease in max dilation, P < 0.05), and reduced l-NAME-inhibited dilation. Vascular dysfunction was accompanied by significant increases in circulating LPS-binding protein (LBP) (SD 5.26 ± 0.23 vs. WD 11 ± 0.86 µg/ml, P < 0.05) and interleukin-6 (IL-6) (SD 3.27 ± 0.25 vs. WD 7.09 ± 1.07 pg/ml, P < 0.05); aortic expression of phosphorylated nuclear factor-κB (p-NF-κB) ( P < 0.05); and perivascular adipose expression of NADPH oxidase subunit p67phox ( P < 0.05). Impairments in vascular function correlated with reductions in Bifidobacterium spp. Antibiotic treatment successfully abrogated the gut microbiota and reversed WD-induced arterial stiffness and endothelial dysfunction. These improvements were accompanied by significant reductions in LBP, IL-6, p-NF-κB, and advanced glycation end products (AGEs), and were independent from changes in body weight and glucose tolerance. These results indicate that gut dysbiosis contributes to the development of WD-induced vascular dysfunction, and identify the gut microbiota as a novel therapeutic target for obesity-related vascular abnormalities.

Keywords: arterial stiffness; endothelial function; microbiota; obesity.

Fig. 1.

Antibiotic treatment suppresses the gut microbiota following Western diet-induced dysbiosis. A: changes in total bacteria count following standard diet (SD), Western diet (WD), and WD with antibiotic (Abx) treatment. B: cecal mass following WD and antibiotic treatment. C: principal component analysis (PCA) biplot of mouse fecal microbial communities colored by diet (closed circles = WD; open circles = SD) and indicating bacterial species driving variability in each quadrant. D: histogram of significantly differentially abundant bacterial taxa in WD relative to SD (q = 0.05) represented as log fold-change. Statistical analysis was performed using one-way ANOVA with Tukey’s post hoc test. Data are expressed as means ± SE; n = 10–12/group. *P < 0.05 vs. all other groups.

Fig. 2.

Antibiotics reverse Western diet-induced arterial stiffness. A: arterial stiffness was serially measured by in vivo aortic pulse wave velocity (aPWV). Antibiotic treatment in a subset of WD mice was initiated following 5 mo of WD feeding as indicated. B: media thickness was determined in segments of proximal thoracic aorta. Statistical analysis was performed using repeated-measures ANOVA (aPWV) and one-way ANOVA (media thickness) with Tukey’s post hoc test. Data are expressed as means ± SEM; n = 10–12/group. *P < 0.05 vs. all other groups.

Fig. 3.

Antibiotics reverse Western diet-induced endothelial dysfunction and restore l-NAME-inhibited dilation. Endothelium-dependent dilation (EDD) to acetylcholine (ACh) alone (A and B) and in the presence of nitric oxide synthase inhibitor l-NAME (C) was determined in mesenteric arteries. D: l-NAME-inhibited dilation, a measure of NO-dependent dilation, was calculated as the percent reduction in maximal EDD in the absence vs. presence of l-NAME. E and F: endothelium-dependent dilation (EID) to sodium nitroprusside (SNP). Statistical analysis was performed using a repeated-measures ANOVA (EDD and EID dose responses) and one-way ANOVA (all other outcomes). When a significant main effect was observed, Tukey’s post hoc test was performed to determine specific pairwise differences. Data are expressed as means ± SE; n = 8–12/group. *P < 0.05 vs. all other groups; #P < 0.05 vs. SD.

Fig. 4.

Antibiotic treatment does not alter body composition or glucose tolerance. Body weight (A), epididymal fat mass (B), and subcutaneous fat mass (C) following WD feeding and antibiotic treatment. Plasma leptin (D) and insulin (E) were determined via multiplex ELISA. F: area under the curve (AUC) for ip glucose tolerance test (GTT). Statistical analysis was performed using one-way ANOVA with Tukey’s post hoc test. Data are expressed as means ± SE; n = 10–12/group. *P < 0.05 vs. all other groups. #P < 0.05 vs. SD.

Fig. 5.

Antibiotic treatment attenuates Western diet-induced increases in circulating markers of endotoxemia and inflammation. Circulating LBP (A), LBP/sCD14 (B), IL-6 (C), PAI-1 (D), and MCP-1 (E) were determined in plasma via ELISA. Statistical analysis was performed using one-way ANOVA with Tukey’s post hoc test. Data are expressed as means ± SE; n = 5–12/group. *P < 0.05 vs. all other groups, #P < 0.05 vs. SD, ^P < 0.05 vs. WD.

Fig. 6.

Antibiotic treatment attenuates Western diet-induced increases in proteins related to inflammation and oxidative stress. A: the ratio of phosphorylated to total NF-κB was determined in aorta via Western blotting. Abundance of NADPH oxidase (NOX) 2 subunit p67phox (B), NOX4 (C), and advanced glycation end products (AGEs) (D) were determined in perivascular adipose tissue (PVAT) via Western blotting. GAPDH expression was used to account for protein loading differences. Representative blots are shown below each graph. Statistical analysis was performed using one-way ANOVA with Tukey’s post hoc test. Data are expressed as means ± SE; n = 7–12/group. *P < 0.05 vs. all other groups, #P < 0.05 vs. SD, ^P < 0.05 vs. WD.