Choline and trimethylamine N-oxide supplementation in normal chow diet and western diet promotes the development of atherosclerosis in Apoe –/– mice through different mechanisms

Abstract Trimethylamine N-oxide (TMAO), a gut microbiota-dependent metabolite, has been shown to aggravate cardiovascular disease. However, the mechanisms of TMAO in the setting of cardiovascular disease progress remain unclear. Here, we aim to investigate the effects of TMAO on atherosclerosis (AS) development and the underlying mechanisms. Apoe –/– mice received choline or TMAO supplementation in a normal diet and a western diet for 12 weeks. Choline or TMAO supplementation in both normal diet and western diet significantly promoted plaque progression in Apoe–/– mice. Besides, serum lipids levels and inflammation response in the aortic root were enhanced by choline or TMAO supplementation. In particular, choline or TMAO supplementation in the western diet changed intestinal microbiota composition and bile acid metabolism. Therefore, choline or TMAO supplementation may promote AS by modulating gut microbiota in mice fed with a western diet and by other mechanisms in mice given a normal diet, even choline or TMAO supplementation in a normal diet can promote AS.


Introduction
Cardiovascular disease (CVD) remains the leading cause of death worldwide, and atherosclerosis (AS) is the main pathological basis . Conventional cardiovascular risks factors, such as high blood cholesterol, triglyceride, hypertension, and diab etes, are well-established. Recently, trimethylamine-N-oxide (TMAO), has been identified as a novel and independent risk factor for promoting AS (Yu et al. 2019), but the definite role and exact mechanism of action for TMAO in AS remain unclear. TMAO generation depends on the gut microbiota, which first degrades dietary choline and carnitine to trimethylamine (TMA), after which TMA is converted to TMAO by the host liver enzyme flavin monooxygenase (Li et al. 2017).
The gut microbiota has been recognised as a key player in the pathogenesis of TMAO-induced AS, and it might be a new potential therapeutic target for the prevention and treatment of CVD. In addition, the gut microbiota is involved in the development of AS (Witkowski et al. 2020), not only because it functions as an endocrine organ, generating bioactive metabolites, but also because bacterial products translocate into the systemic circulation and promote an inflammatory state (Tang et al. 2017). In addition, significant interest in recent years has focused on the impact of gut microbiota on cholesterol and bile acid metabolism (Winston and Theriot 2020). There is no doubt that high blood cholesterol is the key risk factor for AS, and bile acid synthesis is a major pathway for cholesterol catabolism (Libby et al. 2019;Liu et al. 2020). Host hepatocytes synthesise primary bile acids using liver cholesterol as a raw material. Once these host-derived primary bile acids enter the gastrointestinal tract, the gut microbiota chemically modifies them into secondary bile acids via deconjugation, dehydrogenation, and dehydroxylation in the gut, which is more hydrophobic and could be excreted more readily in the faeces Yu et al. 2021). Thus, bacteria present in the gut play a crucial role in cholesterol and bile acid metabolism. A better understanding of the gut microbiota composition will broaden the path to discovering new targets for TMAO-induced AS.
TMAO is reported to regulate lipid metabolism, endoplasmic reticulum stress and inflammation (Warrier et al. 2015;Seldin et al. 2016;Chen et al. 2017), contribute to platelet hyperreactivity and enhanced thrombosis potential , and promote the formation of foam cells (Geng et al. 2018). TMAO has also been shown to induce AS by regulating bile acid metabolism via remodelling the gut microbiota (Chen et al. 2016). However, a previous study reported that TMAO supplementation does not influence AS development in Ldlr−/− and Apoe−/− male mice (Aldana-Hernandez et al. 2020).
In addition, most studies on TMAO and AS were conducted on animals fed with high-fat diets (Geng et al. 2018;Aldana-Hernandez et al. 2020), and little is known whether TMAO can act similarly when the animals are given a normal chow diet. In this study, we investigated the impact of TMAO on AS using mice fed with choline or TMAO in both a normal diet and western diet, with a focus on gut microbiota-related cholesterol and bile acid metabolism.

Animals and treatment
Nine-week-old male Apoe−/− mice were purchased from Vital River Laboratory Animal Technology Co. Ltd (Beijing, China). After one week of acclimation on a normal chow diet (ND), mice were randomly distributed into six groups and fed specific diets for 12 weeks: ND (1025), ND + CHO (ND supplemented with 1% choline, Sigma-Aldrich, St.Louis, MO, USA), ND + TMAO (ND supplemented with 0.12% TMAO, Cayman Chemical Company, AnnArbor, MI, USA), WD (western diet, synthetic diet supplemented with 0.15% cholesterol, w/w, H10141), WD + CHO (WD supplemented with 1% choline), and WD + TMAO (WD supplemented with 0.12% TMAO). The dosages of choline and TMAO were based on a previous study (Wang et al. 2011). All the diets were purchased from Beijing HFK Bioscience Co. Ltd (Beijing, China). During the whole experiment, the padding and water were changed once a week, and the high-fat contained diet was changed twice a week to avoid the oxidation of fat to produce odour which may prevent the mice to eat. Body weight and food intake were measured every two weeks. Mice were anaesthetized with pentobarbital sodium prior to cardiac puncture to collect blood. Then aorta samples were collected. In addition, caecal contents were aseptically collected and immediately stored at −80 °C until use. The study design and protocols were approved by Peking University Biomedical Ethics Committee Experimental Animal Ethics Branch (LA2022010).

Histology and immunohistochemistry
Briefly, for enface analysis of the aorta, aorta samples (from the proximal ascending aorta to bifurcation of the iliac artery) were dissected from mice after stripping off connective tissues and adipose tissue and fixed in 4% paraformaldehyde for 1 h. Then the aortas were opened longitudinally and stained with Oil Red O (ORO; Beijing Solarbio Science & Technology Co., Ltd) to determine the lesion area. Atherosclerotic lesions were expressed as a percentage of the total area, which was calculated by dividing the ORO-stained area over the total surface.
To analyse the aortic root, the hearts with ascending aorta were cut below the aortic root and embedded in paraffin. Sections were further stained with haematoxylin and eosin (HE), Masson and immunohistochemistry of F4/80 using standard procedures. Sections for immunofluorescence staining were incubated with antibodies targeting intercellular adhesion molecule-1 (ICAM-1) and vascular cell adhesion molecule-1 (VCAM-1) (Santa Cruz, Santa Cruz, CA, USA) and separately incubated with Dylight488 or 549-labelled rabbit anti-goat IgG (KPL, Gaithersburg, MD, USA). Hoechst33342 (Invitrogen, St.Louis, MO, USA) was used to stain nuclei. The sections were photographed under a laser scanning confocal microscope (TCS SP5, Leica, Mannheim, Germany). Aortic root frozen sections were stained with ORO as routine. All images were analysed and quantified using ImageJ software (U.S. National Institutes of Health, Bethesda, MD, USA).

Biochemical assays
The serum contents of total cholesterol (TC), triglyceride (TG), low-density lipoprotein cholesterol (LDL-C) and high-density lipoprotein cholesterol (HDL-C) were measured using a 7180 automatic biochemical analyser (Hitachi Ltd., Tokyo, Japan) according to the manufacturer's instructions.

Liquid chromatography-tandem mass spectrometry analysis
The concentration of bile acids in caecal contents was quantified in passive MRM mode using an liquid chromatography-tandem mass spectrometry system consisting of an ACQUITY UPLC I-CLASS (Waters, Milford, MA, USA) coupled with a QTRAP 6500+ LOW MASS (AB Sciex, Foster City, CA, USA). Bile acids were separated on a UPLC BEH Amide column (1.7 μm, 100 × 2.1 mm, Waters, Milford, MA, USA), and eluted with a mobile phase consisting of water/acetonitrile (10:1, v/v, containing 1 mmol/L ammonium acetate, phase A) and acetonitrile/isopropanol (1:1, v/v, phase B) at a flow rate of 0.26 mL/min. The final concentration was obtained using a standard curve method.

Statistical analysis
All data were expressed as mean ± standard error of the mean (SEM). Statistical analysis was performed using one-way ANOVA followed by a Tukey post hoc test. The correlations between the different filtered faecal bacteria and serum lipids, inflammatory factors and lesion area were analysed using the Spearman rank correlation. P values less than 0.05 were considered statistically significant. There were 6 -10 mice used for ORO staining of whole aortas and 4-6 mice for HE, ORO, Masson, F4/80, ICAM-1 and VCAM-1 staining of aortic roots. In addition, eight mice were used for body weight, food intake, LDL-C, HDL-C, TC, TG, TNF-α, IL-1β, MCP-1 and bile acids analysis, and 4-5 mice for gut microbiota analysis.

Choline or TMAO feeding promoted atherosclerosis development
We first measured the levels of TMAO in serum and found that choline or TMAO supplementation increased the concentrations of TMAO (Supplementary Figure 1). Then we observed atherosclerotic lesion areas in the whole aorta and aortic root. The ORO staining of the whole aorta and aortic root is shown in Figure 1. Unexpectedly, both choline and TMAO supplemented in a ND promoted atherosclerotic lesion development. The total plaque areas and lipid content of WD-fed mice significantly increased, compared with ND-fed Apoe-/-mice. Choline and TMAO supplementation in WD enhanced atherosclerotic lesion development and lipid content in mice more than in choline-free WD-fed mice (Figure 1(A and C)). Consistent with the above results, the HE and Masson staining also showed that dietary choline or TMAO supplementation in the ND group induced more atherosclerotic lesions and collagen content in the aorta. The lesion area and collagen content increased in the WD group, while no obvious atherosclerotic lesion and collagen were observed in the ND group. In addition, there were markedly increased atherosclerotic lesions and collagen in WD + CHO and WD + TMAO groups (Figure 1(B and D)).

Choline or TMAO feeding modulated food intake and serum lipid concentrations
The body weight of Apoe-/-mice increased after 12 -weeks of feeding with ND, WD and diets supplemented with choline or TMAO, but there were no obvious differences between the groups (Figure 2(A)).
The administration of choline or TMAO in ND did not alter food intake in the mice. However, WD greatly promoted food intake in the Apoe-/-mice during most feeding times. In addition, the mice consumed more food when supplemented with choline or TMAO in WD (Figure 2(B)). The presence of choline in ND also increased serum LDL-C and TC levels. The levels of LDL-C, HDL-C and TC in serum significantly increased in mice of three WD groups compared with those of the control group. In the three WD groups, choline-exposed mice displayed higher LDL-C and HDL-C than mice fed with WD ( Figure 2(C-E)). Serum TG concentrations did not change among these groups, and although there was a slight increase in the ND + CHO and ND + TMAO groups, it was not significant (Figure 2(F)).

Choline or TMAO feeding partly enhanced inflammation
We evaluated macrophage recruitment in plaque lesions since macrophages play a critical role in atherosclerotic plaque development Galle-Treger et al. 2020). Figure 3(A) shows representative pictures of macrophage recruitment (F4/80 positive area) in the aortic root. Notably, mice in the ND group supplemented with either choline or TMAO showed a greater F4/80 positive area than those in the ND group. In addition, WD feeding induced marked infiltration of macrophages into the plaque lesions, and the mice fed with WD supplemented with choline or TMAO also showed a large F4/80 positive area.
As previously reported, proatherogenic factors, such as ICAM-1, VCAM-1, TNF-α, IL-1β and MCP-1 are involved in atherosclerotic development (Zhou et al. 2018;Kim et al. 2021). We additionally investigated the effect of choline and TMAO on inflammation. Immunofluorescent staining of aortic root sections revealed that the administration of choline or TMAO-contained ND caused a marked increase in the expression of ICAM-1 and VCAM-1. There was an elevated ICAM-1 and VCAM-1 expression in the WD group, and the addition of choline or TMAO led to a higher expression of ICAM-1 and VCAM-1 than feeding with a WD alone (Figure 3(B and C)). The ELISA results showed that the addition of choline or TMAO in ND did not stimulate further production of TNF-α, IL-1β and MCP-1 in serum, while the serum expression of TNF-α, IL-1β and MCP-1 increased in WD groups with and without the addition of choline or TMAO. The addition of choline or TMAO showed an increasing trend in TNF-α and MCP-1 expression than the WD group but the increase was not statistically significant (Figure 3(D-F)). These findings indicated that macrophage recruitment and proinflammatory factors expression was partially enhanced by choline or TMAO supplementation.

Choline or TMAO feeding did not significantly change bile acid profiles
Bile acids are synthesised from hepatic cholesterol, which represents an important way to eliminate cholesterol and decrease the risk of AS (Porez et al. 2012). We measured the effects of choline or TMAO feeding on the caecal contents of bile acid composition. Supplementation with either choline or TMAO in ND did not change bile acid profiles. However, there was a significant increase in total conjugated bile acids, taurodeoxycholic acid (TDCA) and glycocholic acid (GCA), and total unconjugated bile acids, cholic acid (CA), chenodeoxycholic acid (CDCA), alpha-muricholic acid (α-MCA), beta-muricholic acid (β-MCA), deoxycholic acid (DCA), hyodeoxycholic acid (HDCA) and ursodeoxycholic acid (UDCA) in the WD group. Meanwhile, adding choline or TMAO in WD slightly decreased several types of unconjugated bile acids. However, the conjugated/unconjugated bile acids ratio between groups was unchanged (Figure 4(A-E)). WD feeding enhanced bile acid synthesis, including total, primary and secondary bile acids. Furthermore, adding choline or TMAO in WD may reduce loss in bile acids, but the changes were not significant (Figure 4(F-H)).

Choline or TMAO feeding modulated the composition of caecum gut microbiota
Since choline or TMAO feeding promoted AS development and reduced bile acids loss, we subsequently assessed whether this effect was related to the gut microbiota, which is essential for the metabolism of bile acids by analysing caecal contents. The alpha diversity and OTU did not change after different treatments, indicating that these treatments did not affect the caecal microbiota richness and alpha diversity ( Figure 5(A and B)). However, beta diversity (Unweighted uniFrac principal coordinate, PCoA) analysis of microbial genus revealed distinct clusters in different groups, indicating that choline or TMAO supplementation induced significant rearrangements in microbial composition ( Figure 5(C)).
increased in the WD group, and choline or TMAO supplementation of WD decreased the relative abundance of Bifidobacterium. In addition, choline treatment decreased the relative abundance of Streptococcus, and treatments with both choline and TMAO increased the relative abundance of Desulfovibrio. Gut microbiota plays an important role in bile acid metabolism, by mediating deconjugation, oxidation, epimerization and 7-dehydroxylation of primary bile acids. The main bacterial genera involved in bile acid metabolism include Bacteroides, Bifidobacterium, Clostridium, Enterococcus, Lactobacillus, and Listeria. WD induced elevated relative abundances of Bifidobacterium, Clostridium and Lactobacillus, and the addition of choline or TMAO partially decreased the relative abundances of Bifidobacterium, Clostridium and Lactobacillus, in accordance with a previous report (Chen et al. 2016). These results indicated that choline or TMAO supplementation might alter caecal microbial composition, by increasing pathogenic bacteria and decreasing beneficial bacteria, ultimately increasing AS (Figure 6).

Correlation between the gut microbiota and AS related parameters
We further analysed the correlation between the above faecal bacteria and AS-related parameters by the Spearman rank correlation. We observed significant correlations between lesion area, serum levels of TNF-α, IL-1β, MCP-1, LDL-C, HDL-C and TC and the relative abundance (at the genus level) of most filtered gut microbiota, among which, Bifidobacterium, Clostridium, Desulfovibrio, Lactobacillus and

Discussion
AS, characterised by endothelial dysfunction, inflammatory cell recruitment and foam cell formation, remains the single biggest contributor to global mortality. Emerging evidence suggests that gut microbiota and their metabolites are involved in AS development (Duttaroy 2021). In addition, TMAO is a diet-derived, gut microbial-host cometabolite that has recently gained much attention because of its potential adverse effects on AS (Chen et al. 2016;Yu et al. 2019). Many studies have focused on the pro-atherogenic role of TMAO and TMAO has been shown to regulate lipid metabolism, inflammation, thrombosis and foam cells formation (Warrier et al. 2015;Seldin et al. 2016;Chen et al. 2017;Zhu et al. 2016;Geng et al. 2018). However, it is not clear if this is a causal factor. Furthermore, it has been reported that dietary supplementation with choline or TMAO does not influence AS development in Ldlr−/− and Apoe−/− male mice (Aldana-Hernandez et al. 2020). Similarly, dietary choline supplementation in Apoe-/-mice expressing human cholesterol ester transfer protein did not promote AS (Collins et al. 2022). Another study showed no direct association between plasma TMAO and the extent of AS in mice and humans, despite TMAO plasma levels being associated with atherosclerotic plaque instability (Koay et al. 2021). In addition, there were different results on the association of TMAO and its precursors with CVD in different cohort studies (Wang et al. 2014;Guasch-Ferre et al. 2017;Heianza et al. 2017). Thus, we designed a mouse model of AS, adding extra choline or TMAO to the diet, to observe the effect of TMAO on AS.
Our research demonstrated that choline or TMAO treatment in both ND and WD promoted the development of AS plaque, elevated serum cholesterol levels, and enhanced aortic root inflammation. In addition, choline or TMAO administration in WD modulated gut microbiota composition and bile acid levels, although supplementing ND with choline or TMAO did not show similar results. Adding choline or TMAO to ND seems to promote AS in other ways.
In this study, there was obvious plaques formation in WD-fed Apoe-/-mice, indicating that the AS model had been successfully established. The HE, ORO and Masson staining results revealed that choline or TMAO addition in ND and WD treatment promoted AS progress, with increased plaque formation in the whole aorta and aortic root. Serum LDL-C and TC levels increased when extra choline was added in ND, while LDL-C and HDL-C levels increased with the supplementation of choline in the WD group. The body weight of the mice was not significantly different under all kinds of treatments, although choline and TMAO addition in WD promoted food intake. Unexpectedly, TMAO-exposed mice did not display increased AS plaque formation and serum lipids than choline-exposed mice, which may be because choline exposure increased food intake (Figure 1 and 2), in addition, serum concentration of TMAO seemed to be higher in choline-exposed mice than TMAO-exposed mice, but the difference is not significant (Supplementary Figure 1).
Experimental and clinical evidence has demonstrated that AS is a chronic inflammatory disease of Figure 6. effects of choline or tMao feeding on gut microbiota composition. (a) Heatmaps of gut microbiota composition (comparisons: nd vs nd + cHo; nd vs nd + tMao; nd vs Wd; Wd vs Wd + cHo; Wd vs Wd + tMao; q value ≤ 0.05 are represented; q value can be obtained after correction of p value) in caecum at the genus level. (B) the relative abundance of Bifidobacterium, Clostridium,Desulfovibrio,.
the vessel wall (Wolf and Ley, 2019;Zhang et al. 2021). In the presence of damaging stimuli, the endothelium responds by upregulating the transcriptional messenger and releasing a series of substances that enhance leukocyte adhesion on the endothelium E-selectin, VCAM-1, ICAM-1, and so on (Li et al. 2015). Additionally, immune cells and later foam cells all contribute to induce cytokine and chemokine production and the additional recruitment of circulating immune cells, setting off a cascade of an inflammatory response (Raggi et al. 2018;Li et al. 2021). TMAO was also reported to be involved in inflammation (Zhang et al. 2022). In this study, we evaluated the effect of choline or TMAO treatment on macrophage recruitment, ICAM-1 and VCAM-1 expression in the aortic root, in addition to TNF-α, Figure 7. correlation between gut microbiota and aS parameters. Spearman rank correlations between gut microbiota and lesion area, serum levels of tnf-α, Il-1β, McP-1, ldl-c, Hdl-c, tc and tG, red and blue indicates a positive and negative correlation, respectively. Significant correlations are indicated as *p < 0.05, and marked with asterisks (n = 4-5).
IL-1β and MCP-1 in serum. The representative pictures showed that WD, ND, and WD supplemented with choline or TMAO promoted macrophage recruitment, ICAM-1 and VCAM-1 expression in the aortic root. The ELISA results showed that WD supplemented with choline or TMAO induced higher expressions of TNF-α and MCP-1 than the WD group, but no statistical significance was found ( Figure  3). The above findings suggest that choline or TMAO may partly affect AS development by promoting inflammation.
Furthermore, our results found that choline treatment in ND and WD was associated with elevated serum cholesterol, and decreased or slightly decreased levels of total and unconjugated bile acids (only supplementation in WD) (Figures 2 and 4), which were presumably due to the reduced abundance of Bifidobacterium, Clostridium and Lactobacillus ( Figure  6). Bifidobacterium, Clostridium and Lactobacillus were reported as the bacteria that generated bile salt hydrolases. Bile salt hydrolases can deconjugate glycine or taurine from conjugated bile acids and produce unconjugated bile acids, which are more hydrophobic and easier to excrete via the faeces (Kriaa et al. 2019), by this way, cholesterol was excreted to the outside of the body.
Gut microbiota and its metabolites, such as short-chain fatty acids, lipopolysaccharides, and TMAO, are involved in the development and prevention of AS (Al and Backhed 2017;Verhaar et al. 2020). Alteration of the gut microbiota structure typically alters its functions. Patients with AS usually present with changes in gut microbiota composition accompanied by increased serum TC (Ma et al. 2021). In addition, the production of TMAO is closely linked with gut microbiota. In one study, antibiotics treatment or germ-free conditions significantly inhibited the activity of the intestinal microbiome and prevented dietary choline-driven TMAO generation and the development of AS (Chen et al. 2016). In addition, gut microbiota plays a key role in bile acid metabolism, as it mediates primary bile acid deconjugation and subsequent conversion to secondary bile acids (Fiorucci and Distrutti 2015). Thus, we analysed the gut microbiota composition, and the results showed that OTU and α-diversity did not change among the groups. However, the results of β-diversity demonstrated that the presence or absence of choline or TMAO for 12 weeks in WD altered the microbial composition ( Figure 5). We further identified changes in the faecal microbiota community after choline or TMAO treatments and performed correlation analyses.
Adding extra choline or TMAO in WD but not in ND partly increased the relative abundance of Desulfovibrio and decreased the relative abundance of Bifidobacterium, Clostridium, Lactobacillus and Streptococcus compared to WD ( Figure 6). Therefore, although OTU and α-diversity did not change, the gut microbiota composition changed substantially. Choline or TMAO supplementation in WD seemed to increase typical pathogenic bacteria, and decrease beneficial bacteria, which helped to promote AS. For example, the relative abundance of the typical pathogenic bacterium, Desulfovibrio, was increased . Moreover, the Streptococcus and Lactobacillus genera were decreased, which were reported as beneficial bacteria and demonstrated to have considerable anti-obesity effects in the treatment of metabolic disease (Ma et al. 2021). Lactobacillus, known as traditional probiotics, has previously been shown to exert antidiabetic and cholesterol-lowering effects in rodents (Su et al. 2021). In addition, as mentioned earlier, Bifidobacterium plays an important role in bile acid metabolism (Fiorucci and Distrutti 2015;Wang et al. 2020) and is associated with irritable bowel syndrome, and obesity (Genoni et al. 2020). However, Bifidobacterium was also reported as beneficial bacteria and showed anti-obesity activities (Cuevas-Sierra et al. 2019). Therefore, Bifidobacterium was referred to participate in multiple processes. In this study, the relative abundance of Bifidobacterium decreased in WD groups with choline or TMAO supplementation. The potential action of Bifidobacterium in TMAO-induced AS development is not clear and needs to be further studied. Similarly, the function of Clostridium which was decreased by choline or TMAO supplementation in WD, is conflicting, as Clostridium can help to convert primary bile acids into secondary bile acids (Kriaa et al. 2019) and was also shown to be positively correlated with several lipids in a mouse AS model (Wang et al. 2020). In our study, Bifidobacterium, Clostridium, Desulfovibrio, Lactobacillus and Streptococcus positively correlated with most AS parameters, such as lesion area, the levels of inflammatory factors, and serum lipids. In addition, other bacteria correlated with AS parameters and may be involved in the new pathways of TMAO induced AS, although this needs to be studied further.

Conclusion
This study found that adding choline or TMAO in WD for 12 weeks increased AS plaque formation through integrative effects involving inflammatory response, modulation of intestinal microbiota dysbiosis and microbiome-mediated functions on cholesterol and bile acid metabolism. Choline or TMAO supplementation in ND also promoted plaque progression in Apoe-/-mice through other mechanisms which need further study. In addition, TMAO may act differently in Apoe-/-mice when fed with ND or WD.