Article at a Glance
- trimethylamine-N-oxide, or “TMAO” is a gut metabolite that a New England Journal of Medicine study linked to heart disease.
- TMAO is produced by gut bacteria as they “digest” certain types of food, such as fish and eggs.
- There is a link between elevated TMAO and heart disease, but fish is a large source of TMAO, and many of the world’s longest lived people eat fish.
- Dangerous TMAO levels may be reached by a combination of gene expression, and environmental pollutants, although there are no definitive links as of yet in human models.
Updated February 26th, 2018
In a previous post (and several others) John covered the ongoing debate about dietary cholesterol, with a particular focus on eggs. As part of that post he introduces the molecule called trimethylamine-N-oxide, or “TMAO” for short. TMAO is of great interest in the health community as a New England Journal of Medicine study linked elevated TMAO to heart disease.
Some of the comments on John’s post referenced disagreement about TMAO and dietary cholesterol and how both impact the risk for heart disease. The controversy makes sense. For example, elevated TMAO is definitely linked to heart disease, but seafood is the biggest source of dietary TMAO and virtually every blue zone (regions where people routinely live to be 100) eats a good amount of fish.
This post came about as a result of my reading about TMAO metabolism and how genetics may play an important role in its activity. So, yes, as with almost every other topic related to diet, genes play a role when it comes to TMAO, the problem is we just don’t know how much of a role as of yet.
The above image provides a good overview of the current understanding of TMAO and its association with various diseases including CVD.
Sources of TMAO
There are two main sources of TMAO, from the diet directly and then via metabolism of other compounds. Targeting dietary sources is relatively straightforward as we know which foods are TMAO rich; however, targeting TMAO synthesis is rather more difficult as several key pathways are involved in its formation.
The major dietary source of TMAO is from saltwater fish and other seafoods.
It is thought to protect against urea buildup and also mitigate the effects of pressure (useful for fish under the sea, not so useful for us), and interestingly it is the breakdown of TMAO into TMA (trimethylamine) which gives rise to that distinctive seafood smell.
The other major source of TMAO is biosynthesis from trimethylamine (TMA). TMA is a gas generated by bacteria in our gut as they breakdown dietary choline and phosphatidylcholine (think eggs, and dairy). An additional pathway was later described where L-carnitine (think red meats) is converted into γ-butyrobetaine (γBB) and then into TMA again by gut bacteria.
Put simply, gut bacteria produce TMAO.
This was confirmed in the NEJM study where researchers gave volunteers broad spectrum antibiotics which depleted their stomach bacteria. The depleted gut bacteria then correlated with a significant drop in circulating TMAO levels (R).
So from a genetic point of view, the obvious target of interest will be the gene that converts TMA into TMAO. The flavin-containing monooxygenase (FMO) enzyme family catalyze numerous reactions involving xenobiotics (proteins generated outside the body) including TMA to TMAO and is comprised of five members FMO1-5.
All are expressed throughout the body but it is thought that FMO3, which is expressed mainly in the liver, is the most important in relation to TMA to TMAO conversion, due to its expression in the liver and increased TMA to TMAO activity.
So what is the genetic contribution to TMAO?
One point we keep coming back to here is the fact that a one size fits all approach to nutrition just doesn’t work anymore.
The debate about dietary fats (and by extension cholesterol, TMAO etc…) is so confusing because so many different effects are reported by different groups.
As a geneticist, when I see things like this happening the obvious angle for me is to look at the genetic level. As I discussed above, for TMA to TMAO processing the key target seems to be FMO3, so are there any SNPs or mutations associated with an increase in TMAO levels?
Low FMO3 activity and trimethylaminuria
Interestingly there is an entire disorder associated with a lack of FMO3 activity (R). Trimethylaminuria (TMAU) occurs when individuals have severely reduced or lacking FMO3 activity. TMA builds up in their body and is released through their sweat and breath. As TMA has a strong fishy odor, this release gives rise to the diseases major symptom and its other common name “fish odor syndrome.”
Outside of these symptoms, the disorder is not associated with any other health benefits, and although unpleasant, the disorder is generally regarded as benign (suffers may disagree with this, and avoiding choline or L-carnitine rich foods is a major dietary intervention).
40+ mutations in FMO3 are associated with the development of TMAU driven by lack or loss of FMO3 activity (R). Interestingly, in a mouse model where scientists prevented expression of functional FMO3 this loss of function was shown to attenuate atherosclerosis by decreasing TMAO levels and also by regulating both lipid metabolism and inflammation, further evidence that TMAO is a strong indicator of CVD risk (R).
But what about in humans? Well, there is currently nothing published, which the authors of this study attribute to the relatively low numbers of TMAU sufferers (R). However, searching did reveal this forum post (so treat accordingly) which mentions a Canadian research group investigating the effect… however the post is dated 2012 and further searches have not returned any results.
High FMO3 activity = elevated TMAO?
So a lack of FMO3 activity demonstrated some positive effects for CVD risk in a mouse model, and may be associated with a beneficial effect in humans (or at least no negative effect). What about if we look the other way, are there any SNPs associated with an increased level of FMO3 activity and hence elevated TMAO?
Well, after reading about every SNP in FMO3 that I could find, there’s nothing out there.
However, there are some animal studies which show a definitive link between FMO3 and disease states such as obesity and type 2 diabetes (R), and a strong correlation in man.
The role of dioxins
So nothing conclusive at the genetic level, but while searching the literature I did come across this interesting study which investigated the impact of dioxins on TMAO levels and FMO3 activity.
Dioxins are dangerous chemicals, forming part of the “dirty dozen” due to their persistence in the environment and potential to cause great harm. Typically produced by heavy industry, dioxins are found throughout the world and are readily absorbed into our fat tissues, where they linger for upwards 5 years. Due to this accumulation in fatty tissue they tend to accumulate as you move up the food chain… I’m sure you can see where this is going.
In a mouse model scientists saw that following a single small dose of dioxin, levels of FMO3 expression and activity were significantly increased, and circulating TMAO was increased 5x.
The authors didn’t perform any long term exposure studies or follow this administration through to see what if any health impacts the mice developed, but they did report on significant changes in the levels of inflammatory markers and genes which regulate lipid and cholesterol metabolism, all markers of an increased CVD risk.
Building on their work, the authors suggest that the dioxin may be targeting the gut microbiota and that this area may be worth investigating therapeutically.
The gut microbiota and TMAO
We know several key digestive processes result in the formation of TMAO including some very common commensal bacteria. What this means is that in otherwise healthy individuals, targeting the gut microbiota in order to reduce TMAO is probably a non-starter.
So what about potentially harmful bacteria such as C. difficile and H. pylori? Well there is only a single study investigating H. pylori and TMAO, and although the authors show an interesting synergistic effect between the two (TMAO makes the effects of H. pylori infection worse, and vice-a-versa) there is no data to show that one predisposes to the other (R).
As a highly novel therapy some researchers have proposed inoculating the gut with commensal bacteria which don’t convert TMA to TMAO such as those from the Methanobacteriales order. These bacteria process TMA into methane rather than TMAO, which while maybe not the most pleasant of gases, is likely to much less harmful than TMAO. However, this concept remains at the hypothetical stage and no work has been performed at the clinical level.
Take home message
When I started writing this blog I was sure there was going to be a nice clear cut story. The very first figure suggests two really apparent targets to limit TMAO production. However, FMO3 is tricky as a target, there is some evidence that blocking its activity can be beneficial in reducing CVD risk, but this study was only performed over a short period and in mice, long term human data is sadly lacking.
Similarly, the microbiota of the gut clearly plays a big role in determining TMAO levels, as can be seen when it is reduced following antibiotic treatment. However, as we know the longterm health impacts of sustained antibiotic treatment can be just as severe as those associated with TMAO so this isn’t really an ideal therapy. Modulation of the makeup of the gut microbiota is a promising idea, but remains that, an idea.
So what to do about TMAO?
Well to me it is clear that it has a role in various diseases when present at high levels. But it is also clear that there is a huge level of variation in peoples responses. We don’t currently have a good handle on what’s causing that variation so the best advice for now would be to check your own levels, and then make a decision based on those in conjunction with your healthcare provider.