Gut Microbiota-Derived Trimethylamine N-Oxide Contributes to Abdominal Aortic Aneurysm Through Inflammatory and Apoptotic Mechanisms

Abstract

Background: Large-scale human and mechanistic mouse studies indicate a strong relationship between the microbiome-dependent metabolite trimethylamine N-oxide (TMAO) and several cardiometabolic diseases. This study aims to investigate the role of TMAO in the pathogenesis of abdominal aortic aneurysm (AAA) and target its parent microbes as a potential pharmacological intervention.

Methods: TMAO and choline metabolites were examined in plasma samples, with associated clinical data, from 2 independent patient cohorts (N=2129 total). Mice were fed a high-choline diet and underwent 2 murine AAA models, angiotensin II infusion in low-density lipoprotein receptor-deficient (Ldlr^-/-) mice or topical porcine pancreatic elastase in C57BL/6J mice. Gut microbial production of TMAO was inhibited through broad-spectrum antibiotics, targeted inhibition of the gut microbial choline TMA lyase (CutC/D) with fluoromethylcholine, or the use of mice genetically deficient in flavin monooxygenase 3 (Fmo3^-/-). Finally, RNA sequencing of in vitro human vascular smooth muscle cells and in vivo mouse aortas was used to investigate how TMAO affects AAA.

Results: Elevated TMAO was associated with increased AAA incidence and growth in both patient cohorts studied. Dietary choline supplementation augmented plasma TMAO and aortic diameter in both mouse models of AAA, which was suppressed with poorly absorbed oral broad-spectrum antibiotics. Treatment with fluoromethylcholine ablated TMAO production, attenuated choline-augmented aneurysm initiation, and halted progression of an established aneurysm model. In addition, Fmo3^-/- mice had reduced plasma TMAO and aortic diameters and were protected from AAA rupture compared with wild-type mice. RNA sequencing and functional analyses revealed choline supplementation in mice or TMAO treatment of human vascular smooth muscle cells-augmented gene pathways associated with the endoplasmic reticulum stress response, specifically the endoplasmic reticulum stress kinase PERK.

Conclusions: These results define a role for gut microbiota-generated TMAO in AAA formation through upregulation of endoplasmic reticulum stress-related pathways in the aortic wall. In addition, inhibition of microbiome-derived TMAO may serve as a novel therapeutic approach for AAA treatment where none currently exist.

Keywords: PERK kinase; aortic aneurysm, abdominal; endoplasmic reticulum stress; gastrointestinal microbiome; trimethylamine.

Figure 1.. The association of TMAO levels with AAA in the European and USA AAA case/control cohorts.

Box–Whisker plots of TMAO levels stratified by baseline abdominal aortic diameter and AAA status (A, B), Data are represented as boxplots: middle line is the median, the lower and upper boundaries to the boxes represent 25th and 75th percentiles, and the whiskers represent 10th and 90th percentile; P values were calculated using Kruskal-Wallis (K.W.) test with a Dunn post-hoc analysis, Jonckheere -Terpstrata test of increasing trend and Wilcoxon rank sum test; Forest plots indicating the odds of AAA according to the quartiles of TMAO levels, multivariable logistic regression model for odds ratio included adjustments for age, sex, current smoking, hypertension, diabetes mellitus, CVD, statins, aspirin and renal insufficiency in the European cohort(C), and for age, sex, HDL, LDL, current smoking, hypertension, diabetes mellitus, CAD, C-reactive protein, statins, aspirin and CKD in the USA cohort (D), symbols represent odds ratios and the 5–95% confidence interval is indicated by line length. The number of AAA cases and total number of participants in each quartile were included.

Figure 2.. Dietary choline increases plasma TMAO and augments AAA formation in a gut microbiota/TMAO-dependent manner.

(A) Aortic diameters from mice fed the indicated diets ± ABX and TMA/TMAO supplemented in the drinking water measured ex vivo 28 days after AngII-infusion (n = 10 – 28/group). Plasma TMAO levels displayed as mean ± SEM for each group (n = 9 – 16/group). Significance determined using a Kruskal-Wallis test and Dunn post-hoc analysis. (B) AAA incidence in mice from all groups with an aneurysm defined as an aortic diameter ≥ 1.2 mm or rupture. Total number of aneurysms observed in each group reported underneath. Significance determined by multiple pairwise Fisher’s exact tests. (C) Representative images of aortas ex vivo from the indicated groups after 28 days of AngII infusion. (D) Kaplan Meier curve representing survival and aortic rupture-induced deaths post AngII-infusion. Significance determined by log-rank test. (E) Representative aortic sections from the indicated groups stained with Picrosirius Red for type I and type III collagen imaged by brightfield (left) and polarized light (center). Representative aortic sections stained for CD68 (right) and imaged by fluorescent microscopy. (F) Quantification as percent of total section area of type I collagen in Picrosirius Red stained sections and visualized under polarized light (n=9–12/group). Significance determined by two-way ANOVA. (G) Quantification as percent of total section area of CD68 (n = 5 – 10/group). Significance determined by two-way ANOVA. Data represented as data points and mean ± SEM.

Figure 3.. FMC inhibits gut microbial production of TMA and TMAO from dietary choline and attenuates AAA in choline-fed mice.

(A) Abdominal aortic diameters from mice fed the indicated diets ± FMC and TMAO supplemented in the drinking water measured ex Vivo 28 days after AngII-infusion represented individual data point and mean ± SEM (n = 4 – 36/group). Plasma TMAO levels are shown as mean ± SEM for each group (n = 4 – 16/group). Significance determined using a Kruskal-Wallis test and Dunn post-hoc analysis. (B) AAA incidence in mice from all groups with aneurysm defined as abdominal aortic diameter ≥ 1.2 mm or rupture. Total number of aneurysms observed in each group reported underneath. Significance determined by multiple pairwise Fisher’s exact tests. (C) Kaplan Meier curve representing survival and aortic rupture-induced deaths post AngII-infusion. Significance determined by log-rank test. (D) Representative aortic sections from the indicated groups stained with Picrosirius Red for type I and type III collagen imaged by brightfield (left) and polarized light (center). Representative aortic sections stained for CD68 (right) and imaged by fluorescent microscopy. E) Alpha diversity (Shannon Index) analysis and (F) Principal component analysis (PCA) comparing the 5 groups in bacterial diversity of fecal 16S rRNA profiles (n = 6 – 15/group). (G) Stacked bar chart representing genera-level relative abundance of the 5 groups (n = 6 – 15/group). (H) Bar graph of relative abundance patterns for differentially abundant taxa (Random forest, p<0.05) representative genera-level taxa of the 5 groups (n = 6 – 15/group).

Figure 4.. FMC is protective against growth and rupture of established AAAs in choline-fed mice.

(A) Study schematic. Mice were started on choline diet 1 week prior to initial AngII-infusion. After 4 weeks, ultrasound was performed to identify mice with developed AAAs (aortic diameter > 1.2mm). Mice were then randomly assigned to either the FMC or placebo treatment groups and subsequently underwent a second round of AngII-infusion. Final aortic diameters were measured 2 weeks post the second round of AngII-infusion. (B) Abdominal aortic diameter of mice with established AAA receiving choline diet (n = 9) or choline diet + FMC (n = 9) at 28 days and 42 days. Significance determined by two-way ANOVA. (C) Aneurysm growth as determined by 28 day and 42 day ultrasound, significance determined by Mann Whitney rank sum test. (D) Kaplan Meier curve representing survival and rupture-induced deaths after randomization. Significance tested by log-rank test. Plasma TMA (E), TMAO (F), and Choline (G) measured at end of study by mass spectrometry in the indicated intervention groups (n = 6 – 9/group). Significance determined by Student’s t-test. Data represented as data points and mean ± SEM.

Figure 5.. Disruption of FMO3 production of TMAO protects mice from AngII-induced AAA.

(A) Aortic diameters in Ldlr^−/−/Fmo3^−/− vs. Ldlr^−/−/Fmo3^+/+ mice measured ex Vivo 28 days after AngII-infusion (n = 9 – 15/group). Plasma TMAO levels displayed as mean ± SEM for each group (n = 8/group). Significance determined by Students t-test. (B) AAA incidence with an aneurysm defined as aortic diameter ≥ 1.2 mm or rupture. Total number of aneurysms observed in each group reported underneath. Significance determined by multiple pairwise Fisher’s exact tests. (C) Kaplan Meier curve representing survival and aortic rupture-induced deaths post AngII-infusion. Significance determined by log-rank test. (D) Representative aortic sections from the indicated genotypes stained with Picrosirius Red for type I and type III collagen imaged by brightfield (left) and polarized light (center). Representative aortic sections stained for CD68 (right) and imaged by fluorescent microscopy. (E) Quantification as percent of total section area of type I collagen in Picrosirius Red stained sections and visualized under polarized light. Significance determined by Student’s t-test (n = 3 – 5/group). (F) Quantification of CD68 staining as percent of total section area of CD68 positive (n = 8/group). Significance determined by Student’s t-test. (G) Aortic diameters measured ex Vivo 28 days after AngII-infusion from mice given dietary DIM compared to placebo treatment (n = 10 – 12/group). Plasma TMAO levels displayed as mean ± SEM for each group (n = 10 – 14/group). Significance determined by Students t-test. (H) AAA incidence with an aneurysm defined as aortic diameter ≥ 1.2 mm or rupture. Total number of aneurysms observed in each group reported underneath. Significance determined by Fisher’s exact test. (I) Kaplan Meier curve representing survival and aortic rupture-induced deaths post AngII-infusion. Significance determined by log-rank test. Data represented as data points and mean ± SEM.

Figure 6.. RNAseq demonstrates increased apoptosis and ER stress and decreased autophagy with TMAO treatment in Vitro.

HAVMCs were prepared with three independent samples, conducted in triplicate, treated with either sterile saline or TMAO (100μM) for 5 hours, where RNA was processed and RNA sequencing was performed. (A) Principle component analysis and (B) heatmap clustering of RNA sequencing data comparing TMAO (100 μM) to saline control. (C) Volcano plot of differentially expressed genes between TMAO and saline treated HVSMCs and (D) fold change of UPR pathway related genes. (E) Representative western blot and (F) quantification of caspase 3 and activated caspase 3 protein levels in TMAO and saline-treated HVSMCs (saline n = 4, TMAO n = 4). Significance determined by Student’s t test. (G) Representative and (H) quantified results of flow cytometry analyses in for apoptotic cells identified by Annexin V and 7-AAD using the indicated gating strategy (n = 5/group). Circles represent individual data points, while diamonds are mean ± SEM. P values listed on figure, or *P < 0.01. Relative quantification and gated cells were analyzed using a Kruskal-Wallis test and Dunn post-hoc analysis. Significance is defined as fold change ≥ 1.5 and FDR P ≤ 0.05 with Cuffdiff. FDR, false discovery rate.

Figure 7.. RNAseq demonstrates short term choline feeding in the AngII AAA mouse model increased apoptosis and ER stress while decreasing autophagy.

(A) Experimental schematic for RNA sequencing analysis from whole abdominal aortic sections from the indicated groups for (B) pathway analysis and (C) gene expression changes in the abdominal aorta isolated from Ldlr^−/− mice fed either a control diet with saline infusion (n = 4), control diet with AngII infusion (n = 6), choline diet with AngII infusion (n = 5), or choline diet with AngII infusion plus Perk inhibitor (n = 5). (D) Spatial separation between indicated groups examined using principle component analysis.