I needed to disable self-signups because I’ve been getting too many spam-type sign-ups lately. Please contact me directly if you want membership on this forum. Thanks.

Forum Navigation
Please to create posts and topics.

Expression of Vitamin A-related genes increases with BMI

Using RNA sequencing technology it’s possible to determine the amount of protein-coding mRNA in any given sample. What researchers normally do is sort the samples into two groups and compare the gene expression, for example one group will contain samples from NAFLD patients and those will be compared to a control group. However, I couldn’t find any studies where the samples were sorted by body weight instead of disease state. Therefore, I did my own analysis using software I programmed for this purpose, using the existing sample data from (Horvath, Erhart et al. 2014).

For this analysis there is no distinction between healthy, NAFLD or NASH. Instead, I divided the samples into groups with BMI > 25, > 30, > 35, > 40, >50 and >60, and compared the gene expression of each group with the control group containing samples with BMI < 24. This analysis shows that the expression of many retinol-related genes changes linearly as BMI increases.

It should be noted that mRNA expression often does not correlate with the amount of actual protein, and therefore care must be taken when drawing conclusions from this kind of data. To give an example, if the amount of free retinal within a cell increases, more proteins will be degraded by the retinal and the cell will produce more mRNA of these proteins to compensate. In this case, then, mRNA expression will be elevated even though the actual amount of functioning proteins is decreased.

Samples were taken from human livers of patients undergoing surgery. All results are expressed as log2 fold change compared to the BMI < 24 group.

Specific retinoic acid-metabolizing enzymes

The cytochrome P450 family 26 (CYP26) enzymes specifically metabolize retinoic acid. CYP26A1 predominantly forms (4S)-OH-RA, CYP26B1 forms both enantiomers equally and CYP26C1 predominantly forms (4R)-OH-RA. CYP26A1 is also believed to predominantly metabolize all-trans-retinoic acid (atRA) whereas CYP26C1 is believed to predominantly metabolize 9-cis-RA.

I have previously posted in this forum that (4S)-OH-RA gets converted to 4-oxo-RA, which is at least as potent as atRA and is more difficult to eliminate. 50% less 4-oxo-RA is formed from (4R)-OH-RA, which is why I believe that atRA must be converted to (4R)-OH-RA and not (4S)-OH-RA to be eliminated.

We can see that CYP26A1 expression decreases linearly with BMI, indicating that as the body gains weight, less and less retinoic acid is recycled. Concurrently CYP26C1 expression increases, indicating that until BMI 40 progressively more retinoic acid is eliminated. After a BMI of 40 is reached, the body begins limiting the production of all CYP26 enzymes. Additionally, CYP26B1, which is highly active during childhood and not so much in adulthood, is downregulated at BMI 50 and especially BMI 60, possibly indicating that the body is attempting to prevent fat accumulation by restricting any kind of growth. In my opinion, these results add some support to the theory that CYP26C1 is the main enzyme responsible for clearing retinoic acid.

Retinol-esterifying enzymes

LRAT is the enzyme mainly responsible for converting retinol to retinyl palmitate in humans. Additionally, the function of the LRATD2 protein is unknown, but it was found overexpressed in breast cancer cells. DGAT1 and DGAT2 are also capable of esterifying retinol, but their significance in relation to LRAT is less clear. Unlike DGAT1, DGAT2 is also expressed in lipid droplets, and in the space between the endoplasmic reticulum (ER) and cell nucleus. Therefore, one explanation could be that DGAT2 is the backup enzyme to catch any free retinol that was missed by DGAT1. The DGAT enzymes also help with the synthesis of triglycerides, which further distinguishes them from LRAT.

The data shows that expression of LRAT decreases as BMI increases. Concurrently, the DGAT enzymes increase. One interpretation of the data could be that the body attempts to prevent weight gain by downregulating LRAT since it specifically metabolizes retinol, and the non-specific DGAT enzymes assume its function. The role of LRATD2 is still unclear to me.

Retinol-hydrolyzing enzymes

A specific enzyme for hydrolyzing retinyl esters (converting retinyl palmitate to retinol) has not been identified yet. CES1, CES2 and PNPLA3 have been proposed to hydrolyze retinyl esters in humans. It appears that, similar to DGAT, CES1 is the primary enzyme of the CES, since it is widely expressed and found in the cytoplasm and elsewhere while CES2 is found in the ER. They hydrolyze many substances, including for example heroin and cocaine. PNPLA3, which hydrolyzes triglycerides, is mainly found in lipid droplets, and this enzyme is currently the most likely candidate for hydrolyzing retinol esters. 

The data shows that CES1 expression does not seem to change significantly, whereas CES2 expression decreases and PNPLA3 increases with BMI. Downregulation of CES2 is expected, since this prevents the movement of retinol into the ER. Progressively more REs are hydrolyzed by PNPLA3 in lipid droplets, which makes sense since the retinyl esters have to be cleared at some point, and CRBP proteins can transport them from the cytosol to the nucleus for catabolism in a controlled fashion.

Carotenoid-cleaving enzymes

Two carotenoid-cleaving enzymes have been identified in humans. BCO1 performs a symmetrical cleavage, producing for example two retinal molecules from beta-carotene. BCO2 performs an asymmetrical cleavage, producing apocarotenoids. It has been reported that BCO1 is downregulated after consumption of retinyl esters in mice.

BCO1 expression is constant until BMI 50, whereas BCO2 increases with BMI. This indicates that asymmetrical cleavage to products like trans-β-apo-8'-carotenal may be less detrimental than cleavage to retinal, even though apocarotenal has been reported to be mutagenic and genotoxic.

Nonspecific retinoic acid-metabolizing enzymes

Many CYP enzymes have been reported to show activity towards retinoic acid. Of interest to me were CYP3A4 and CYP1A1, which equally form both 4-OH-RA enantiomers, CYP2C8 and CYP3A7 (for which no data was available) which preferentially produce (4S)-OH-RA, and CYP3A5 and CYP2C9, which preferentially produce (4R)-OH-RA.

CYP1A1 decreases almost perfectly linear as BMI increases, with expression at BMI 60 being almost 75% less than at BMI 24. CYP3A4 begins decreasing at BMI 50. Of the (4R) enantiomer-producing enzymes, CYP2C9 is relatively unchanged whereas CYP3A5 expression consistently increases. CYP2C8 increases less than CYP3A5. Overall, these results are not conclusive enough to add support to the (4S)/(4R) theory.

Enzymes involved in cholesterol synthesis

CYP7A1 is generally considered to be the rate limiting enzyme in cholesterol biosynthesis, whereas CYP7B1 is also capably of forming cholesterol through an alternate pathway. SQLE is the overall master regulator of steroids and cholesterol by initiating the synthesis from squalene. Squalene is highly similar to beta carotene, and the different enantiomers 2,3-(S) and 2,3-(R)-epoxysqualene were confirmed to have opposite biological effects, which is why it was important for me to look at these enzymes as well.

Both SQLE and CYP7A1 increase significantly with BMI. SQLE is already expressed twice as much at BMI 30, and CYP7A1 is expressed twice as much at BMI 50. Since cholesterol is also a precursor of bile acids, this indicates to me that the body is attempting to clear retinol by synthesizing more bile acids. Alternatively, retinol metabolites may be substrates of some of these enzymes, especially CYP7B1, due to their chemical similarity to naturally occurring steroids. Additionally, CYP7B1 is believed to be responsible for synthesizing bile acids outside of the liver, and its downregulation may be indicative of increased bile acids in serum.

More results

Sometime soon, I’ll make another post analyzing the expression of more retinol-related genes like RBP1, RBP4 and STRA6, and downstream signaling proteins and transcription factors like STAT5A/B, PPARA/D/G and SIN3A/B. I’ll also do the same analysis on other samples to confirm the results. If anyone has any theories involving certain proteins, reply with a list of gene symbols and I’ll make sure to include them in the next analysis.

Overall, I think the data from this analysis adds some good support for the 4S/4R theory, and more importantly shows that the expression of many retinol-related genes increases or decreases linearly as BMI increases, which has not previously been reported. If anyone spots something in the data that I’ve missed, be sure to reply and let us know!


Horvath, S., W. Erhart, M. Brosch, O. Ammerpohl, W. von Schönfels, M. Ahrens, N. Heits, J. T. Bell, P. C. Tsai, T. D. Spector, P. Deloukas, R. Siebert, B. Sipos, T. Becker, C. Röcken, C. Schafmayer and J. Hampe (2014). "Obesity accelerates epigenetic aging of human liver." Proc Natl Acad Sci U S A 111(43): 15538-15543.

ggenereux, Jenny and 3 other users have reacted to this post.

Thank you for information. How do you think, based on this, is it possible to draw the opposite conclusion that the higher the BMI, the more vitamin A? And what about anorexic people, what do they have with the expression of their VA-related genes?

Retinoicon has reacted to this post.
Quote from Даниил on October 7, 2021, 1:48 am

Thank you for information. How do you think, based on this, is it possible to draw the opposite conclusion that the higher the BMI, the more vitamin A? And what about anorexic people, what do they have with the expression of their VA-related genes?

One of the reasons I did this analysis was to add some more support to the theory that Vitamin A directly causes weight gain, which has been demonstrated pretty conclusively (in my opinion) in animals but not humans. For the reasons described in the original post, I don’t think gene expression is a tool that can conclusively prove anything except in very specific, controlled experiments. I do believe, however, that this analysis is a solid circumstantial piece of evidence supporting the theory.

I have started writing a book where I try to explain biological concepts related to Vitamin A and its effect on gene expression for people unfamiliar with these topics. I’ve only finished the first chapter so far, in which I cover weight gain and the evidence supporting it, and I’ve attached it to this post. I think it answers your question with significantly more details.

I haven’t looked at anorexia at all, but I think it might be more related to processes within the brain involving how people judge their own appearance. The origin of Vitamin A-induced weight gain, on other hand, is the liver, most likely due to simple oversupply of retinol.

Uploaded files:
Jenny, Beata and 3 other users have reacted to this post.

Anorexia is definitely not something mental. I knew one girl who was undergoing inpatient treatment for anorexia. She really wanted to "get fat," but she couldn't. Despite the fact that she ate tons of meat, sugar, potato, etc. She really looked very thin and emaciated.

Also, my friend's mom used to be anorexic, then she took some kind of prebiotic. But it, on the contrary, made her too fat.

Ourania has reacted to this post.
Quote from Даниил on October 7, 2021, 2:21 am

Anorexia is definitely not something mental. I knew one girl who was undergoing inpatient treatment for anorexia. She really wanted to "get fat," but she couldn't. Despite the fact that she ate tons of meat, sugar, potato, etc. She really looked very thin and emaciated.

Also, my friend's mom used to be anorexic, then she took some kind of prebiotic. But it, on the contrary, made her too fat.

I guess I have should have distinguished between voluntary and involuntary anorexia. Involuntary anorexia can be the result of mutations in various genes, the most likely candidate being PPARG, as PPARG knockout mice were unable to build adipose tissue. LRAT-null mice were also reported to be unable to build adipose tissue, as well as certain RDH-null mice. I don’t remember which RDH it was exactly but it might have been RDH11 or RDH12. If both your friend and her mother were anorexic at some point, it’s very likely that this was caused by a mutation in some gene.

Jenny, Ourania and Даниил have reacted to this post.

Very interesting, but I must admit that Im probably not smart enough to understand it all.


But based on what you wrote here, would you say that overweight people or people that gain weight easy are more likely to be Vitamin A toxic than underweight people or people that have a hard time gaining weight?



I would say no. 

People who put on weight protect themselves from the worst damage imo. Adipose tissue is a great store for toxicity. My body put on lots of weight as I became toxic. I had annoying symptoms but was never really ill. I think the fat protected me. That’s why I’m only loosing weight slowly as I don’t want to flood my system with toxicity as the storage empties out. 

People who can’t make use of the protection of fat, for whatever reason, suffer more from the immediate effects of the toxicity I think. 

Hermes, kathy55wood and Retinoicon have reacted to this post.

I have heard a lot about fat being a protection but is there any evidence for this? Why do fat people die much sooner then? Why do most people find fat unattractive (other than in third world starving countries) if they are healthier? This would make no sense in a evolutionary way.


If fat indicates toxicity then it’s NOT a good thing. Being a non toxic slim person is what we want. 

However, it can offer certain protection from toxicity. Stephanie Seneff talks about some fat being a good thing as we age. Fat is known to store toxicity. 

The best thing is to not need extra fat. Piling on fat for toxicity reasons (like I did) means you are being poisoned!! Clearly this is very bad. However, what I’m saying is that if you are being poisoned and you can’t store any away in fat then you may get sicker. 

Fat offers only a temporary toxicity holding solution. I’m now having to slowly get rid of all the extra toxicity my body hid away in fat. I just saved it to deal with later!! Fat is not healthy. 

kathy55wood has reacted to this post.

If anyone has any evidence that significant amounts of vitamin A are stored in human adipose tissue please post it. Both tallow and lard are low in vitamin A despite cattle and pigs consuming diets that are not low in vitamin A.

Obesity and fatty liver disease is associated with decreased liver retinol stores and higher serum retinol levels in mice. Changes in the expression of important retinol metabolizing enzymes are possibly due to the effects of metabolic dysfunction on vitamin A metabolism. Weight loss and lack of appetite are two of the most important symptoms of Hypervitaminosis A.


Retinoicon has reacted to this post.