Skip to main content

Biological News

Fructose In The Brain?

Let’s talk sugar, and how you know if you’ve eaten enough of it. Just in time for Halloween! This is a field I’ve done drug discovery for in the past, and it’s a tricky business. But some of the signals are being worked out.
Blood glucose, as the usual circulating energy source in the body, is a good measure of whether you’ve eaten recently. If you skip a meal (or two), your body will start mobilizing fatty acids from your stored supplies, and circulate them for food. But there’s one organ that runs almost entirely on sugar, no matter what the conditions: the brain. Even if you’re fasting, your liver will make sugar from scratch for your brain to use.
And as you’d expect, brain glucose levels are one mechanism the body uses to decide whether to keep eating or not. A cascade of enzyme signals has been worked out over the years, and the current consensus seems to be that high glucose in the brain inactivates AMP kinase (AMPK). (That’s a key enzyme for monitoring the energy balance in the brain – it senses differences in concentration between ATP, the energy currency inside every cell, and its product and precursor, AMP). Losing that AMPK enzyme activity then removes the brakes on the activity of another enzyme, acetyl CoA-carboxylase (ACC). (That one’s a key regulator of fatty acid synthesis – all this stuff is hooked together wonderfully). ACC produces malonyl-CoA, and that seems to be a signal to the hypothalamus of the brain that you’re full (several signaling proteins are released at that point to spread the news).
You can observe this sort of thing in lab rats – if you infuse extra glucose into their brains, they stop eating, even under conditions when they otherwise would keep going. A few years ago, an odd result was found when this experiment was tried with fructose: instead of lowering food intake, infusing fructose into the central nervous system made the animals actually eat more. That’s not what you’d expect, since in the end, fructose ends up metabolized to the same thing as glucose does (pyruvate), and used to make ATP. So why the difference in feeding signals?
A paper in PNAS (open access PDF) from a team at Johns Hopkins and Ibaraki University in Japan now has a possible explanation. Glucose metabolism is very tightly regulated, as you’d expect for the main fuel source of virtually every living cell. But fructose is a different matter. It bypasses the rate-limiting step of the glucose pathway, and is metabolized much more quickly than glucose is. It appears that this fast (and comparatively unregulated) process actually uses up ATP in the hypothalamus – you’re basically revving up the enzyme machinery early in the pathway (ketohexokinase in particular) so much that you’re burning off the local ATP supply to run it.
Glucose, on the other hand, causes ATP levels in the brain to rise – which turns down AMPK, which turns up ACC, which allows malonyl-CoA to rise, and turns off appetite. But when ATP levels fall, AMPK is getting the message that energy supplies are low: eat, eat! Both the glucose and fructose effects on brain ATP can be seen at the ten-minute mark and are quite pronounced at twenty minutes. The paper went on to look at the activities of AMPK and ACC, the resulting levels of malonyl CoA, and everything was reversed for fructose (as opposed to glucose) right down the line. Even expression of the signaling peptides at the end of the process looks different.
The implications for human metabolism are clear: many have suspected that fructose could in fact be doing us some harm. (This New York Times piece from 2006 is a good look at the field: it’s important to remember that this is very much an open question). But metabolic signaling could be altered by using fructose as an energy source over glucose. The large amount of high-fructose corn syrup produced and used in the US and other industrialized countries makes this an issue with very large political, economic, and public health implications.
This paper is compelling story – so, what are its weak points? Well, for one thing, you’d want to make sure that those fructose-metabolizing enzymes are indeed present in the key cells in the hypothalamus. And an even more important point is that fructose has to get into the brain. These studies were dropping it in directly through the skull, but that’s not how most people drink sodas. For this whole appetite-signaling hypothesis to work in the real world, fructose taken in orally would have to find its way to the hypothalamus. There’s some evidence that this is the case, but that fructose would have to find its way past the liver first.
On the other hand, it could be that this ATP-lowering effect could also be taking place in liver cells, and causing some sort of metabolic disruption there. AMPK and ACC are tremendously important enzymes, with a wide range of effects on metabolism, so there’s a lot of room for things to happen. I should note, though, that activation of AMPK out in the peripheral tissues is thought to be beneficial for diabetics and others – this may be one route by which Glucophage (metformin) works. (Now some people are saying that there may be more than one ACC isoform out there, bypassing the AMPK signaling entirely, so this clearly is a tangled question).
I’m sure that a great deal of effort is now going into working out these things, so stay tuned. It’s going to take a while to make sure, but if things continue along this path, there could be reasons for a large change in the industrialized human diet. There are a lot of downstream issues – how much fructose people actually consume, for one, and the problem of portion size and total caloric intake, no matter what form it’s in, for another. So I’m not prepared to offer odds on a big change, but the implications are large enough to warrant a thorough check.
Update: so far, no one has been able to demonstrate endocrine or satiety differences in humans consuming high-fructose corn syrup vs. the equivalent amount of sucrose. See here, here, and here.

22 comments on “Fructose In The Brain?”

  1. Ty says:

    Thank you for the great head’s up. Very interesting indeed.

  2. gecko says:

    But I was given to understand that high-fructose corn syrup is between 40%-60% fructose, and the rest glucose (i.e., it is “high fructose” relative to 100% glucose “plain” corn syrup). But HFCS is an industrial substitute for sucrose, which is broken down in the stomach and gut into 50% fructose and 50% glucose. Is this your understanding?

  3. woops says:

    “Glucose, on the other hand, causes ATP levels in the brain to rise – which turns down AMPK, which turns down ACC, which allows malonyl-CoA to rise, and turns off appetite.”
    I think you meant “turns up ACC”.

  4. Hap says:

    If fructose is altering appetite or the perception of such, then portion sizes could be an effect of the use of high-fructose corn syrup sweeteners rather than an independent occurence.

  5. Derek Lowe says:

    Fixed the up/down part, and added some more references. This is a messy field, to be sure.

  6. Timothy says:

    Gecko – To address your point, I remember researching a few years back that most industrial HFCS used in sodas as a sucrose substitute is 55/45 Fructose/Glucose. Interestingly, this is not that much different than the natural ratio present in things like unsweetened apple juice.
    Also interesting is that, IIRC, if you look at the nutition information for sugar sweetened cola vs HFCS sweetened cola it turns out that there’s more sugar in the former on a per-serving basis because sucrose does not taste as sweet as HFCS…this would imply, to me at least, that you’re getting roughly equal proportions of Fructose from either source.

  7. Jeramia says:

    It was my understanding that excess fructose does not “short circuit” glycolysis in most cells, as the main commitment step is fructose-6-P to fructose-1,6-bisP via phosphofructokinase, which is highly regulated. So while excess fructose would lead to excess F6P, it would not be further metabolized (unless needed). However, in the liver, fructose can be metabolized to F1P and then to Glyceraldehyde-3-P, which is further “down” the glycolytic cycle, bypassing the major metabolic control point. Still doesn’t explain why a mix of fructose and glucose would have a different effect than sucrose, though.

  8. Jose says:

    Isn’t there data showing a co-incidence of higher obesity rates and the widespread introduction of HFCS into food, beginning in the early 1980’s? Not to contradict the lab studies above, but I though that epidemiological data was available. (yes, I know correlation DNE causation).

  9. Rev. Howard Furst says:

    Another consequence of fructose ingestion is that the rapid depletion of hepatic ATP by fructose results in increased serum uric acid, with implications for the rising incidence of gout and cardiovascular disease – uric acid seems to contribute to vascular endothelial dysfunction.
    Am J Clin Nutr. 2007 Oct;86(4):899-906. Links
    Potential role of sugar (fructose) in the epidemic of hypertension, obesity and the metabolic syndrome, diabetes, kidney disease, and cardiovascular disease.
    Johnson RJ, Segal MS, Sautin Y, Nakagawa T, Feig DI, Kang DH, Gersch MS, Benner S, Sánchez-Lozada LG.
    Currently, we are experiencing an epidemic of cardiorenal disease characterized by increasing rates of obesity, hypertension, the metabolic syndrome, type 2 diabetes, and kidney disease. Whereas excessive caloric intake and physical inactivity are likely important factors driving the obesity epidemic, it is important to consider additional mechanisms. We revisit an old hypothesis that sugar, particularly excessive fructose intake, has a critical role in the epidemic of cardiorenal disease. We also present evidence that the unique ability of fructose to induce an increase in uric acid may be a major mechanism by which fructose can cause cardiorenal disease. Finally, we suggest that high intakes of fructose in African Americans may explain their greater predisposition to develop cardiorenal disease, and we provide a list of testable predictions to evaluate this hypothesis.

  10. gretchen says:

    With adaptation, the brain can run on up to 75% ketones. The idea that it uses only glucose is a myth.

  11. kynefski says:

    If I’m missing something, please forgive my physiological ignorance.
    High fructose corn syrup is a sucrose replacement. The first step in sucrose metabolism is sucrose hydrolysis, generating as much fructose as glucose. So if HFCS is 55% fructose, we’re not talking about a big difference in available fructose.
    Plus, there’s energy in that glycosidic bond of sucrose. I don’t know if we use it, but my understanding is that oral streptococci use it in dextran synthesis

  12. Linkr says:

    If you find yourself voting for Mccain next week, you’ve probably got fructose on the brain.

  13. UnLinkr says:

    If you find yourself voting for Obama next week, you’ve probably got fructose on the brain.

  14. Joshua says:

    Does the fructose in question include the nature fructose sugars found in fruits?

  15. Joshua says:

    Does the fructose in question include the natural fructose sugars found in fruits?

  16. Brooks Moses says:

    gretchen @10: Do you have a citation for that claim?
    Also, did you perhaps miss the sentence where Derek said, “Even if you’re fasting, your liver will make sugar from scratch for your brain to use”? And perhaps you’re thinking of data that shows that the brain works even when there’s insufficient glucose in food (which is, of course, a different thing from what’s actually delivered by the blood to the brain)?

  17. Morten says:

    @14: There is no difference between natural fructose and the synthetic fructose in HFCS but! the concentrations in fruit are usually lower and fruit is more filling than where HFCS is normally used. The sweetness of fruit is also augmented by various esters (which of course work through your nose rather than your tongue).
    My question though: I thought fructose was metabolized by liver and intestinal cells exclusively like so many other toxins with the difference being that the clearing mechanism feeds it into the glycolytic pathway rather than using cytochromes (in much the same manner as alcohol is cleared by conversion into energy or fat – and both alcohol and fructose leads to fatty, cirrhotic liver when consumed in excess).
    Does fructose really go everywhere in the body?
    Ok, I think I found the answer to my own question: GLUT2 also transports fructose and besides liver and intestine it is expressed in pancreatic beta cells, hypothalamus, and renal tubular cells. GLUT5 is an exclusive fructose transporter and is expressed in liver and intestine (and I think sperm).
    I don’t know if I need an answer now =]

  18. Doc Sucrose says:

    Thanks for writing about this very important topic – HFCS seems to be in just about everything nowadays.
    I find myself only drinking soda with real sugar – I get the mexican coke bottles and I just think those taste great. I’m seeing a lot of new interest in the non-HFCS market, just today I was at reading about a new product that’t made to speak against HFCS
    keep on blogging – spread the truth!

  19. It’s a red herring to argue over whether we are talking about HFCS or sugar – both contain similar amounts of fructose which does end up in our bloodstream as pure fructose.
    The big question with this research is whether the fructose makes it past the liver. And that seems unlikely to me given the extremely efficient hepatic take-up of fructose in preference to glucose.
    A mechanism which produces the same end result but without the leap of faith is the leptin inhibiting properties of fructose discussed in this paper:

  20. Morten says:

    But David, the article you quote suggests that fructose does indeed reach the hypothalamus. The anorectic response to leptin is mediated through the leptin receptors in the hypothalamus. The article you quote actually suggests that there is a longer-range obesity promoting effect of fructose – one that persisted after the mice were taken of fructose.
    Dr Sucrose, the last line of Derek’s post has 3 links to explain how there’s no difference between sucrose and HFCS. Please read. The short of it is that your body breaks sucrose into glucose and fructose in an extremely efficient manner (unless you have Congenital Sucrase-Isomaltase Deficiency (CSID) and the rampant gas and diarrhea would tip you of to that) so there is no metabolic difference between eating sucrose and 50/50 glucose/fructose. It only changes the taste and consistency of what you eat.

  21. > [activation of AMPK] may be one route
    > by which Glucophage (metformin) works.
    Could that be why Glucophage makes (some) people think they’re hungry? My dad became absolutely ravenous (his food consumption at least doubled, maybe tripled) when he was taking that stuff.

  22. anon says:

    I would just like to state that hfcs has msg and this is what is causing the obesity epidemic…
    Fructose is a good healthy and beneficial sugar found in fruits… It is that olympic athletes eat fruit and vegetables for teh meajoirty if not all of their diet…
    Fructose is a good sugar!

Comments are closed.