Inflammatory injury of the vascular endothelium is widely regarded as the initial stage of atherosclerosis [32]. In vitro studies have shown an up-regulated IL-6, CRP, TNF-$\alpha$ and reactive oxygen species (ROS), and reduced Nitric Oxide (NO) production in TMAO-treated endothelial cells (ECs), suggesting that TMAO can directly impair NO-mediated endothelial function [33]. TMAO can also up-regulate the expression of vascular cell adhesion molecule 1 (VCAM-1), promote the adhesion of monocytes, and inhibit the self-repair capability of ECs through protein kinase C (PKC) and NF-$\kappa$B signaling, while enhancing macrophage adhesion [14, 34]. Furthermore, TMAO reduces the expression of the anti-inflammatory cytokine IL-10, which can protect ECs from damage caused by inflammation and increased oxidative stress [35] (Fig. 2).

Fig. 2.

Putative molecular mechanisms underlying TMAO-induced atherogenic effects. Annotation: High concentration of TMAO plays a vital role in endothelial hyper permeability and inflammation, oxidative stress, cholesterol metabolism, plaque formation, activation and thrombosis.

In young mice, six months of dietary supplementation with TMAO can induced aging of the arterial endothelium. This effect was accompanied by increased vascular nitrotyrosine, a marker of oxidative stress [36]. In a different study, aged rats exhibited higher plasma TMAO, TNF-$\alpha$, and IL-1$\beta$, and increased eNOS expression in the aorta when compared to young rats, all of these effects were restored to control levels by treatment with DMB (an inhibitor of TMA formation) [37]. These finding clearly indicated that the increase in circulating TMAO levels caused by aging can lead to endothelial dysfunction.

It has been reported that there are several mechanisms underlying the activation of inflammation, including lysosomal rupture, altered K${}^{+}$ channel gating, and activation of ROS [38]. Krishna proved that endotheliitis caused by TMAO is activated both in vitro and in vivo and in the presence of blockers or inhibitors of individual pathways, and found that the formation and activation of the NLRP3 inflammasomes induced by TMAO in ECs were markedly attenuated or abolished by the ROS scavenger, N-acetyl-L-cysteine and cathepsin B inhibitor. These results showed that TMAO can activate NLRP3 inflammasomes in ECs through at least two different pathways, involving increased ROS production, destabilized lysosomes, and enhanced cathepsin B activity [39]. Sun also confirmed that TMAO activates the thioredoxin interacting protein (ROS-TXNIP-NLRP3) signaling pathway through oxidative stress and promotes the expression of the inflammasome NLRP3 [40]. A different mechanism for TMAO action involves a reduction in Adenosine triphosphate(ATP) production and induced endothelial cell apoptosis by damaging the structure and function of mitochondria, and the succinate dehydrogenase complex subunit B (SDHB)/ROS signaling pathway has been found to play an important role in this process [41].

It is well known that loss of integrity of tight junctions found in the endothelium helps to enhance the permeability of the quasi-cellular endothelium. Stimulating ECs with TMAO can reduce the expression of the tight junction protein ZO-1 in the ECs monolayer and change ECs permeability, which was prevented by silencing NLRP3 in the ECs [39]. The high-mobility basal box protein 1 (HMGB1) is also an inflammatory mediator, which can destroy cell-to-cell connections, resulting in vascular endothelial hyper permeability [42]. In a study conducted by Singh, it was discovered that TMAO significantly reduced the expression of junctional proteins such as ZO-2, VE-kadrin, and Okrudin in the EC monolayer, while this effect was markedly attenuated by glycyrrhizin, an HMGB1 binding compound [42]. Furthermore, TMAO can upregulate HMGB1, and HMGB1 causing further activation of toll-like receptor 4(TLR4, an important receptor for HMGB1 binding where HMGB1 can activate inflammatory pathways). TLR4 siRNA can protect ECs from the interference by TMAO-related tight junction proteins [42] (Fig. 2).

High TMAO levels were also found to be accompanied by low HDL levels in patients with cardiovascular disease [43]. The accumulation of oxidized low-density lipoprotein (ox-LDL) in LDL cholesterol macrophages can be transported to the liver for recirculation or excreted in the form of bile acid [44]. Previous studies have confirmed that the RCT of choline-added APOE${}^{-/-}$mice was reduced by about 30%, and total bile acids in mice and the expression of key bile acid synthases Cyp7a1 and Cyp27a1 in the liver were also significantly reduced relative to normal chow (P 0.05) [6, 22], resulting in changes to the main mechanism of removing cholesterol in the body (Fig. 2).

Several experimental models have demonstrated a role for TMAO in the process of atherosclerosis. A TMAO-diet significantly promoted plaque progression in APOE${}^{-/-}$ mice fed with high-fat for 20 weeks when compared to C57BL/6J mice fed with a normal chow diet. However, the levels of blood sugar, cholesterol and triglycerides had no significant effect on the plaque area [6], and this effect could be reversed by reducing the level of TMAO [45]. TMAO can also increase the formation of foam cells by up-regulating the macrophage scavenger receptors CD36 and SR-A1, which promote the continuous development of plaques [6]. Furthermore, the CD36/MAPK/JNK pathway may play a crucial role in the TMAO-induced formation of foam cells [46]. In addition, it was also found that TMAO can induce the expression of heat shock proteins GRP94 and HSP70 in J774A.1 mouse macrophages, leading to endoplasmic reticulum stress [47] (Fig. 2). TMAO not only promotes plaque formation, but also contributes to plaque instability. In the tandem stenosis mouse model, which reflects plaque instability as typically seen in patients, TMAO levels correlated with several characteristics of plaque instability, such as markers of inflammation, platelet activation, and intraplaque hemorrhage [48].

Platelets contribute to the formation of thrombus and foam cells and play an important role in the occurrence and development of atherosclerosis [49]. After 2 months of choline supplementation, healthy subjects showed a significant $>$10-fold increase in plasma TMAO levels at both 1- and 2-month periods, with a corresponding enhanced platelet aggregation response to submaximal adenosine diphosphate [50]. Furthermore, large scale clinical association studies (n $>$4000 subjects) independently demonstrated that plasma TMAO levels are associated with risk of thrombotic events [15]. The exact mechanisms by which TMAO activates platelets however are still unclear.

TMAO can induce the release of calcium from platelets by changing the inositol 1,4,5-triphosphate (IP${}_3$) signal transduction pathway, and directly enhance the platelet response to a variety of different agonists (ADP, thrombin and collagen) and promote platelet hyperresponsiveness by enhancing stimulus-dependent Ca${}^{2+}$ release [51]. Tissue factors are essential mediators of hemostasis and trigger thrombosis [52]. Xiaoye Cheng demonstrated that TMAO promoted tissue factor (TF) expression via activation of the NF-$\kappa$B signaling pathway in primary HCAECs [53]. TMAO can also attenuate the inhibitory effect of clopidogrel on platelet aggregation by inhibiting P2Y12 receptor inhibitors [54] (Fig. 2). Recently, Sarah M Skye found that the expression of the CutC gene from intestinal microbes is sufficient to transmit enhanced platelet reactivity and thrombosis potential in a host via TMA/TMAO generation [55]. These studies implicate TMAO as a potential molecular target for the treatment of atherosclerosis and thrombotic.

High levels of circulating TMAO are associated with an increased risk of both type 1 and type 2 diabetes (T2DM) [56, 57, 58], as well as pre-diabetes prevalence [59], and diabetic complications such as retinal neuropathy [60]. Miao found that LIRKO mice suffered from severe hyperglycemia after 4 months Pigen diet, which was completely prevented by the knockdown of FMO3 (TMAO synthase) [61]. Similar results were obtained on obese/insulin resistant subjects, where FMO3 was significantly increased in the morbidly obese group (P 0.05). However, this effect was smaller than that seen in the mouse model, possibly because diabetic patients were treated with insulin and/or other drugs to increase insulin sensitivity [61].

The exact biological mechanism of how plasma TMAO concentration is involved in glucose metabolism remains to be clarified. PI3K is a key protein that transduces insulin signals into the regulation of glucose metabolism, while Akt is an important molecule downstream of the PI3K pathway, and its stimulation will in turn activate glycogen synthetase (a critical regulatory enzyme for the regulation of hepatic glycogen storage) [62]. In a study conducted by Gao it was shown that dietary TMAO reduces the mRNA levels of PI3K and Akt, indicating that dietary TMAO may exacerbate the blocked insulin signaling pathway [63].

Another mechanism involving TMAO promotes the regulation of transcription factor FoxO1 caused by PERK, and inhibition of FMO3 or manipulation of the gut microbiota can hinder metabolic dysfunction mediated by the PERK-FoxO1 axis [64]. Specifically, FoxO1 drives glucose gene expression and inhibited by insulin, the deletion of Foxo1 in the liver can reduce excessive glucose production caused by general ablation of insulin receptors [65]. PERK is a key sensor of intracellular stress and has previously been shown to induce FoxO1 by phosphorylation on serine 298 [66]. These data indicate that TMAO may be central to the pathogenesis of metabolic syndrome.

Research conducted by Yang observed a significant decrease in microbial richness, diversity, and reproducibility in the spontaneously hypertensive rat [67], suggesting that the intestinal flora may be involved in the occurrence of hypertension, although the mechanism has not yet been fully elucidated. Meta-studies have shown that for every 5 $\mu$mol/L and 10 $\mu$mol/L increase in TMAO, the prevalence of hypertension will increase by 9% and 20% respectively, revealing a significant dose-dependent positive correlation between TMAO concentration and the risk of hypertension [68]. TMAO is an important metabolite of enteral malnutrition and may be involved in the pathogenesis of hypertension as a mediator.

A different study demonstrated that patients with hypertension have an increased abundance of the cutC gene in their intestines, which in turn produces more TMAO [69]. The integrity of the intestinal barrier is essential for maintaining host health and preventing inflammation and atherosclerotic processes. Kinga Jaworska confirmed that the permeability of the colon to TMA in hypertensive patients is increased and promotes its penetration into the circulatory system [70], representing a novel point of view that the high concentration of TMAO in cardiovascular diseases may depend upon the increased permeability of the colon to TMA, and changes in colon permeability rather than TMAO levels are the indicators of cardiovascular risk [70].

The clearance of TMAO depends largely upon its excretion by the kidneys [71]. Serum TMAO concentrations substantially increase with deterioration of kidney function, and this effect is reversed by renal transplantation [72]. Meta-analyses of a total of 32 eligible clinical studies involving 42,062 participants showed that TMAO was strongly inversely correlated with glomerular filtration rate(GFR) and positively associated with the urine albumin-to-creatinine ratio, serum creatinine, and urine albumin excretion rates [73]. Jason R reported that the increase in TMAO concentration in CKD patients receiving coronary angiography was related to the load of coronary atherosclerosis and may be related to long-term mortality in CKD patients receiving coronary angiography [72]. Moreover, it was observed that TMAO predicts a poor overall survival rate in subjects with CKD [74, 75, 76].

TMAO is an independent predictor of CKD phase 3–5 mortality rate [77] and in animal models, TMAO levels significantly increase renal fibrosis compared with control (P 0.05), while this result can be reversed by TMA lyase inhibitors [78]. This model was also demonstrated in ApoE KO mice, where the animals developed chronic kidney disease with elevated TMAO levels within 14 weeks of adenine administration, after using TMA inhibitors, and many markers of kidney injury (creatinine, cystatin C, FGF23, and urinary microalbumin) were significantly reduced [79]. This suggests that TMAO may represent a novel modifiable risk factor and therapeutic target for improving atherosclerosis in patients with CKD.

TMAO is a significant predictor of carotid plaque formation which is more significant than gender, diastolic blood pressure, total cholesterol, or diabetes mellitus [80]. A study from southern Germany showed that higher TMAO levels can predict thickness of the carotid intima-media (cIMT) [81]. TMAO is also an independent predictor of new lesions as seen by MRI after carotid artery stenting (CAS) [82]. In patients with severe carotid stenosis ($>$70%) it was found that elevated plasma TMAO level were associated with the increased risk of new lesions on diffusion-weighted imaging(DWI) within 1–3 days of CAS1, and that a TMAO of 4.29 $\mu$mol/L can better predict ischemic brain damage which is secondary to CAS [82].

In patients with non-ST-segment elevation myocardial infarction (NSTEMI) and ST-segment elevation myocardial infarction (STEMI), TMAO was found to be independently related to the severity and prognosis of coronary atherosclerosis load [83, 84, 85]. TMAO is also significantly related to the severity of the culprit segment of calcification, including maximum calcification arc, maximum calcification thickness, and calcification length, and can be used as an indicator to the calcium load of the culprit plaque [86]. Studies have shown that repeated rupture or erosion of a plaque followed by healing, is related to the vulnerability of the plaque and a high level of inflammation, which helps to increase the high risk of coronary thrombosis [87, 88]. The study reported that in patients with STEMI, elevated TMAO levels indicated a higher SYNTAX score (response to coronary atherosclerosis load) and the presence of multivascular disease [89], and TMAO levels in patients with plaque rupture was significantly higher than those with no rupture [90]. While TMAO levels in patients with repaired plaques were significantly higher than those with unrepaired plaques, and can be used as a potential biomarker to predict healed plaque presence with a cutoff value of 2.9 $\mu$mol/L [91]. The author believes that TMAO levels in the circulation may reflect the vulnerability and progression of coronary plaques.

TMAO not only plays a role in the process of primary atherosclerosis, but also contributes to late stent thrombosis after revascularization. A recent study found that an area under the curve (AUC) value for plasma TMAO of 0.85 can distinguish patients with new atherosclerosis and old atherosclerosis in STEMI patients who underwent PCI surgery, by using OCT to evaluate the culprit plaque [92]. The above studies indicate that high TMAO levels can be used as a useful biomarker to distinguish between plaque instability and rupture and represents a promising diagnostic marker for the prediction of new atherosclerosis in a stent.

TMAO can cross the blood-brain barrier and cause oxidative stress, thus potentially leading to brain damage [93, 94]. Research has found that higher TMAO levels are associated with increased risk of a first stroke [95, 96, 97], and are closely related to a poor prognosis, such as post-stroke cognitive impairment (PSCI). When predicting moderate to severe stroke, a critical value for plasma TMAO levels has been found to be 4.95 $\mu$mol/L [98]. Other studies however, suggest an optimal value of 6.6 $\mu$mol/L [96]. PSCI is a common consequence of stroke, affecting approximately one-third of stroke survivors after one-year post stroke [99]. A study included 256 patients with ischemic stroke, suggesting that increases in plasma TMAO levels are related to poor cognitive function one year after ischemic stroke, and supports TMAO as a predictive biomarker of PSCI [100]. In patients with acute ischemic stroke, TMAO at a concentration of 5 $\mu$mol /L can predict early deterioration in neurological function [101], and can be used as an independent predictor of stroke severity and infarct volume in patients with acute ischemia [102, 103, 104]. Furthermore, TMAO is also associated with poor prognosis in patients with cerebral hemorrhage at 3 months [105]. Su H suggested that TMAO can promote an increase in reactive astrocytes and glial scar formation through the Smurf2/ALK5 axis, thus aggravating nervous system damage after ischemic stroke [106]. Therefore, the study of TMAO can provide important predictive insights into the development and prognosis in stroke patients.

Dr. Vichai Senthong first reported that elevated plasma TMAO level are an important prognostic marker for patients with PAD [107]. In these patients, the increase of TMAO levels was related to an increased risk of death by 2.7 times, after adjusting for traditional risk factors, such as inflammatory biomarkers and a history of coronary artery disease, where the highest quartile for TMAO can still predict 5-year mortality [107]. PAD is characterized by atherosclerotic stenosis of the lower extremity vessels, leading to ischemic muscle pain in the elderly. Compared to patients with intermittent claudication, patients with advanced chronic severe limb ischemia (CLI) showed higher serum carnitine and TMAO levels and poorer long-term survival [108], whereas PAD patients with TMAO $>$2.26 $\mu$mol/L exhibited higher risk of cardiovascular death [109]. To date potential mortality and pathophysiological predictors of PAD have not been clearly defined, this may be the reason why mortality and ischemic amputation rates in PAD patients are still very high [110]. These studies confirm the clinical prognostic value of TMAO levels in patients with PAD, and these findings may be used to develop a new prognostic indicator of risk stratification.

Taken together the experimental data on TMAO it might be suggested that TMAO is a potent atherogenic cytokine and increased risk of developing atherosclerotic diseases such as CAS, coronary atherosclerosis, stroke, and PAD (Table 1, Ref. [80, 81, 82, 83, 84, 85, 86, 89, 90, 91, 92, 95, 96, 97, 98, 99, 100, 102, 103, 104, 105, 107, 108, 109]).