Search This Blog

Friday, 20 February 2015

Gluconeogenesis Drives Ketogenesis - role of the Nutritional Prometheus.

In trying to explain the findings of Newburgh and Marsh*, and of Karl Petren, from 1923 that switching to a high fat, restricted protein, and very low carbohydrate diet - a ketogenic diet - suppresses diabetic ketoacidosis (DKA) in diabetics without access to insulin, I can't help noticing that gluconeogenesis is a driver of ketogenesis. DKA is a dehydrating syndrome characterized by hyperglycaemia, due in large part to runaway gluconeogenesis, plus levels of ketone bodies, much higher than those seen in starvation or nutritional ketosis, which result in a lethal acidosis. And excess glucose and excess ketones are linked metabolically.

Remember the old saw, that fat burns in a carbohydrate flame? Laugh all you like, but this is true. And it is even more true when the flame is taken away - when carbohydrate (glucose) is being stolen from mitochondrial metabolism. Gluconeogenesis involves a direct loss of oxaloacetate from the citric acid (Krebs, TCA) cycle. Without this oxaloacetate, the fat-burning flame sputters; the smoke that escapes from incomplete combustion is the ketone bodies. I'm not a chemist, but this seems to me a pretty certain way of interrupting the TCA cycle. And a very convenient one in evolutionary terms; at times when you need endogenous glucose, you can use a few extra ketones as well.

Pyruvate from glucose can supply acetyl-CoA or oxaloacetate, fatty acids can only supply acetyl-CoA, if there's no oxaloacetate acetyl-CoA can't be converted to citrate and is converted to ketone bodies instead (not shown).

[In Starvation] degradation of fatty acids in the liver proceeds more rapidly than usual, with augmented production of of acetoacetyl-CoA and acetyl-CoA and their products.
In addition there is a deficit of oxaloacetate and thus a decrease in formation of citrate.
The low level of oxaloacetate is further accentuated because it is being utilized for gluconeogenesis.
This further impairs operation of the citric acid cycle.

Ketosis incident to starvation is most frequently encountered clinically in gastrointestinal disturbances in infancy or pregnancy. Other circumstances in normal individuals in which excessive lipid and diminished carbohydrate are being metabolised may also lead to ketosis, e.g. renal glycosuria and abrupt replacement of a normal diet by one low in carbohydrate and very rich in lipid.

Clinically, the most important cause of ketosis is diabetes mellitus. In the diabetic individual, in contrast to the above situations, glucose is present in excessive amounts in the fluids of the body; however, the metabolic defect, viz., insulin deficiency, prevents glucose utilization from operating at a normal rate. From the point of view of the effect upon lipid metabolism, diabetes and starvation resemble one another.

In diabetic individuals with severe ketosis, urinary excretion of ketone bodies may be as high as 5,000mg/24 h and the blood concentration may reach 90mg/100ml, in contrast to normal values of less than 125mg and less than 3mg respectively.

Ketogenesis, from Principles of Biochemistry, 5th Edn, White A, Handler P, Smith EL. McGraw Hill, 1973, p577-578.

[NB: acetyl-CoA is also a precursor for cholesterol;
"The data suggest that, although acetyl-CoA is channeled towards ketone body formation in both diabetes and fasting, augmented cholesterol synthesis is evident only in diabetes." This suggests that the closer the diabetic diet gets to a ketogenic diet, the less cholesterol synthesis will be augmented - as does seem to be the case in practice.]

So what happens when a diabetic without insulin eats carbohydrate or excess protein?
As we saw in earlier posts, glucagon is released from pancreatic alpha cells in response to carbohydrate and protein. This elevates gluconeogenesis in the liver. Blood glucose is elevated by the meal and by GNG, and hyperglycaemia itself increases hepatic GNG further. Lipolysis is increased by the glucagon, so the liver has additional fatty acids to metabolize. Perfect conditions for ketogenesis to be enhanced above normal levels, because oxaloacetate is being extracted from the TCA in record amounts as this fat is being burned.

What happened when diabetics, in acidosis and without insulin, were switched to the Newburgh and Marsh ketogenic diet in 1923?
With no glucose and minimal protein to trigger glucagon, hepatic GNG is lower. With no glucose to add its load to hyperglycaemia, there is less portal hyperglycaemia to additionally drive GNG.
Less GNG = less ketogenesis.
And, as a bonus, it is likely that dietary fat has an inhibitory effect on lipolysis that is independent of hormonal controls. As long as it's saturated, or not polyunsaturated - in 1923, endocrinologists favoured butter as a source of fat.

Beef tallow diet decreases beta-adrenergic receptor binding and lipolytic activities in different adipose tissues of rat.
Matsuo T, Sumida H, Suzuki M. Metabolism. 1995 Oct;44(10):1271-7.

The effects of dietary fats consisting of different fatty acids on lipolytic activity and body fat accumulation were studied in rats. Sprague-Dawley male rats were meal-fed an isoenergetic diet based on either beef tallow or safflower oil for 8 weeks. Lipolytic activities in epididymal and subcutaneous adipose tissues were lower in the beef tallow diet group than in the safflower oil diet group. Body fat accumulation was greater in rats fed the beef tallow diet versus the safflower oil diet. Norepinephrine (NE) turnover rates used as an index of sympathetic activities in adipose tissues were lower in the beef tallow diet group. beta-Adrenergic receptor binding was determined with [3H]dihydroalprenolol. Binding affinities of beta-receptors in adipose tissues were significantly lower in the beef tallow diet group. Membrane fluidities of adipose tissues were also lower in the beef tallow diet group. Membrane fluidities were correlated with the affinities of the beta-receptor. We believe from these correlations that the decreases in beta-receptor binding affinities are due to the changes in membrane fluidities. The results of the present study suggest that intake of the beef tallow diet promotes body fat accumulation by reducing lipolytic activities resulting from lower beta-receptor binding and sympathetic activity in adipose tissues.

Dr Bernstein describes the mechanism of DKA differently; he doesn't consider that the liver is the main source of ketones, or that gluconeogenesis drives ketogenesis. His description addresses the pathology of DKA well, but not I think the early links in the chain of causality. Perhaps the difference is that he is describing the failure of insulin to work, and I am describing the long-term absence of insulin. But we are both agreed; dietary carbohydrate is the cause of DKA in diabetics.

"Furthermore, the higher your blood sugars go, the more insulin resistance you will experience. The more insulin-resistant you are, the higher your blood sugars are going to be.

A vicious circle. To make the circle even more vicious, when you have high blood sugars, you urinate—and of course what happens then is that you get even more dehydrated and more insulin-resistant and your blood sugar goes even higher. Now your peripheral cells have a choice—either die from lack of glucose and insulin or metabolize fat. They’ll choose the latter. But ketones are created by fat metabolism, causing you to urinate even more to rid yourself of the ketones, taking you to a whole new level of dehydration."
See also

Edit: here's a bit more on starvation, from this book
After about 3 days of starvation, the liver forms large amounts of acetoacetate and d-3-hydroxybutyrate (ketone bodies; Figure 30.17). Their synthesis from acetyl CoA increases markedly because the citric acid cycle is unable to oxidize all the acetyl units generated by the degradation of fatty acids. Gluconeogenesis depletes the supply of oxaloacetate, which is essential for the entry of acetyl CoA into the citric acid cycle. Consequently, the liver produces large quantities of ketone bodies, which are released into the blood. At this time, the brain begins to consume appreciable amounts of acetoacetate in place of glucose. After 3 days of starvation, about a third of the energy needs of the brain are met by ketone bodies (Table 30.2). The heart also uses ketone bodies as fuel.

And diabetes:
Because carbohydrate utilization is impaired, a lack of insulin leads to the uncontrolled breakdown of lipids and proteins. Large amounts of acetyl CoA are then produced by β-oxidation. However, much of the acetyl CoA cannot enter the citric acid cycle, because there is insufficient oxaloacetate for the condensation step. Recall that mammals can synthesize oxaloacetate from pyruvate, a product of glycolysis, but not from acetyl CoA; instead, they generate ketone bodies. A striking feature of diabetes is the shift in fuel usage from carbohydrates to fats; glucose, more abundant than ever, is spurned. In high concentrations, ketone bodies overwhelm the kidney's capacity to maintain acid-base balance. The untreated diabetic can go into a coma because of a lowered blood pH level and dehydration.

Note for future research: Mammals can synthesise oxaloacetate from pyruvate, but what if this step depends on insulin (which suppresses ketogenesis) and the conversion of pyruvate to acetyl-CoA doesn't?
The diabetic hepatocyte is swamped with glucose, it can't resist metabolising it, and 65-85% of the carbon from this glucose is recycled as GNG glucose.
What if this glucose, without the guiding hand of insulin, is, like fatty acids, a poor source of oxaloacetate and a good source of acetyl-CoA? After all, its metabolism is not suppressing ketogenesis - the opposite seems to be true.
Ketone bodies for use by heart muscle in normal hepatic metabolism are produced from glycogen, according to the first text I quoted.
So - is glucose itself a ketogenic substrate under certain conditions?
The quest continues...


Peter said...

Yay! Excellent post George. Thanks for the legwork!


ItsTheWooo said...

Fabulous entry George. Sadly internet ate my comment much to say otherwise.

Interesting re scfa and sns depression. This report is seeming antagonist subjective experiences of keto dieter who generally report energy from scfa.

The bit about acetoacetate really makes me want to write mr blog entry about this ketone master minding keto adaption. It's only ketone that down regulates irs1/2 , and is most specific for ffa/, oxidatiob. Bohb gets all the credit when urine tests were most important all along.
A greater percent of ketones as acetoacetate is a marker for fat reliance, starvation. Reflected in acetoacetate programming the body to resist insulin.

Anyway, excellent. Thanks George.

raphi said...

That post made me think hard, forcing me to go over it twice (& take notes! lol)

Have you given some thought as to what role the malate-aspartate shuttle may play in this DKA scenario? I ask because of the parallels with how cancer cells exploit the shuttle to remain sugar (& to a much lesser extent, 'amino') fuelled. I'm pondering how it'd fit with the anabolic stimulus that insulin engenders: since the shuttle transports electrons from glycolysis for subsequent reduction in mitochondria, the break down (catabolism) of proteins is required. That's also something we tend to see to a pathological degree in diabetics. I think you rightly focus on CHO but I wonder how the aminos come into play since, as you say (in context) protein ==> glucagon ==> GNG.

Anyways, I'm thinking out loud - maybe (maybe not) there's something to go on here.

Again, good thought provoking post.

George Henderson said...

@ Peter,
thanks! I think this version of ketosis simplifies a few things; for example, mice are hard to get into ketosis because they have such a small obligate requirement for glucose.

@ itsthewoo
I think that the SFA> lipolysis picture is highly modifiable by insulin's effect on glucagon, it's only really big in its absence; and there HAVE to be other effects of dietary fat (chylomicrons?) on lipolysis, or feeding high fat diets without insulin would have been LETHAL. So this effect is probably just the tip of the iceberg regarding non-hormonal regulation of lipolysis.
With regard to oxaloacetate, once again we have lipid metabolism altered in the wake of carbohydrate metabolism. Carbohydrate is the tail that wags the dog.

@ Raphi,
there are going to be shuttles that supply minimal oxaloacetate to survival levels, from amino acids. I left the AAs out for simplicity but they drive GNG in the non-insulin model. ("protein is 58% glucose")
Anaplerosis and cataplerosis have always fascinated me; I think they are the keys to metabolic tone.
Think of the TCA as the Zodiac and the relative activity of the various anaplerotic and cataplerotic possibilities as the planetary aspects.

George Henderson said...

@ itsthewoo

Conversion of acetoacetate to B-OHB is a shunt to fix reductive stress in cytosol.
High levels of B-OHB relative to AcAc may indicate relative lack of ketoadaptation, or excessive ketosis.

Unknown said...

Hi George. I love your work. Thank you for doing it! Here is a newly published paper that you might find interesting:

George Henderson said...

@ Raphi,

protein catabolism supplies amino acid carbon to liver mainly in the form of ketoacids, propionate, and the few AAs directly metabolised there (the shuttles like alanine).
Those AA products that are anaplerotic at pyruvate are potentially supplying oxaloacetate or ketones, all the others are gluconeogenic (under these glucagonergic conditions).
(so I guess if pyruvate from AA can be ketogenic, so can pyruvate from glucose if it finds its way into the mitochondria during glucagonergic, gluconeogenic metabolism).

High fat feeding spares protein in this model, this is consistent with lower glucagon.
There are hormones other than insulin that inhibit glucagon to some extent, with lower CHO these perhaps exert more influence.

George Henderson said...

@ Unknown,

Brilliant link thanks!

"In vivo monotherapy with statins has been disappointing with either no change with statin therapy[75] or even a paradoxical increase in HCV RNA levels.[76] This may be related to dose effects as the serum concentration of statin with conventional dosing is 10-fold lower than what was found to be effective in in vitro study. Some have also speculated that the secondary effect of statins to upregulate LDL-R may conversely enhance HCV infectivity of hepatocytes and may explain the aforementioned study where a paradoxical increase in HCV RNA levels was seen with statin therapy in an HCV-HIV coinfected population.[77].

Their [77] says "Atherosclerosis has been described as a liver disease of the heart" - Yes!

Statins have very limited effect on HCV viraemia - in my experience ketogenic diet (<50g CHO/day) had a much larger, and consistent effect, reducing viral load by greater than 1 log compared to low carb diet (150gm).

And of course on the low carb diet my steatosis was completely reversed.

Peter said...


"mice are hard to get into ketosis because they have such a small obligate requirement for glucose".

So. How come the average diabetes/obesity researcher is capable ok ketosis? They have no proven brain function to need any glucose to support.....

OMG it's the Diabetes Paradox!


annlee said...


George Henderson said...

@Peter, the average diabetes researcher has indeed been hard to get into ketosis, but I think they are slowly beginning get into it again, at least more clinicians are, as they see more and more patients with normal HbA1cs.

George Henderson said...

Ketone metabolism summary

Ketone body synthesis occurs normally under all conditions. However, the formation
of ketone bodies increases dramatically during starvation. This seems to be due to a
combination of factors. Prolonged low levels of insulin result in both increased fatty
acid release from adipose tissue, and increased amounts of the enzymes required to
synthesize and utilize ketone bodies. In addition, in the liver, increased demand for
gluconeogenesis results in depletion of oxaloacetate, and therefore in decreased
capacity for the TCA cycle. This causes a rise in the levels of acetyl-CoA, the
substrate for ketone body production.

Both b-hydroxybutyrate and acetoacetate are released into circulation. The ratio
depends on the amount of NADH available in the liver mitochondria; if NADH concentration is high, the liver releases a higher proportion of b-hydroxybutyrate.

raphi said...

That summary was helpful.

Reading this ==> "Both b-hydroxybutyrate and acetoacetate are released into circulation. The ratio depends on the amount of NADH available in the liver mitochondria; if NADH concentration is high, the liver releases a higher proportion of b-hydroxybutyrate."

while considering a basic feature of the malate-aspartate shuttle ==> ( "Since the malate-aspartate shuttle regenerates NADH inside the mitochondrial matrix..."

suggests that the point ItsTheWoo made about a higher acetoacetate:BhB ratio reflecting the degree of reliance on fat oxidation, depends quite a bit on the enzymatic up/downregulation of this shuttle & on the ana/cataplerotic results of AA metabolism. Use more protein as 'fuel' will shift your NADH:NAD+ ratio & thus your AcAc_BhB.

More to ponder.

George Henderson said...

And, one of the functions of B-OHB is to carry that reducing equivalent from hepatocytes that have too much NADH to cells that can use more.

George Henderson said...

A very thought-provoking paper here from Raymund Edwards

Lipolysis drives GNG in obesity-related T2D here.
Acetyl-CoA is present in excess;
this upregulates Pyruvate Carboxylase, converting pyruvate that would go to to Acetyl-CoA to oxaloacetate instead.
Because glucagon is present (tho plasma levels aren't altered in the experiment, so it's not about alpha cell activity) and because there is a surfeit of ATP, this oxaloacetate doesn't all go into citrate, it spills into glucose.
Ketones aren't measured.

Peter said...

"And, one of the functions of B-OHB is to carry that reducing equivalent from hepatocytes that have too much NADH to cells that can use more". Like the brain's neurons. Which are complex I dependent and exclude FFAs....

Nice concept.


George Henderson said...

In the mouse paper lipolysis -> AcetylCoA -> GNG.
So is there increased FPG when lipolysis is increased by weight loss and the diet is high in fat?
No. Blood glucose is normalised.
"Our patient was overweight and was taking eight
medicines. When shifting towards the paleolithic
ketogenic diet she was able to discontinue all medications.
Her weight begin to decrease along with improving
glucose parameters and lowered blood pressure"

So what is the difference? Is it all about the WAT inflammatory cytokines, not the FFA at all?

Maybe here's a clue

Unknown said...

I'm glad you found the link helpful!

George, would you consider doing a focused post on Fatty Liver Disease and it's possible treatment with diet and nutrition? I have pieced together an approach based on several of your posts over the last several months but it would be great to see you apply your theory specifically to this problem.

Now that HCV is being successfully treated many people are left with resulting fatty liver disorders and no guidance from doctors on how to effectively manage it.

George Henderson said...

I think this, and the linked posts, is the best summary I've done of fatty liver disease, and I couldn't add much to it.

Basically, eat as close to ketogenic diet as possible, keep fats MUFA and SFA, but do add sources of DHA and EPA.
Undereat or if that's not possible, fast intermittently.
Exercise with something like sprints, something that exhausts you, regularly.
Don't eat huge amounts of dietary cholesterol because process driving NAFLD may have caused build up of this molecule in liver.
Take a good probiotic, and take ALCAR (acetyl-carnitine) if you like, it's good stuff.

Unknown said...

Thank you very much George. Off to do some aerobics!

nurul iman said...
This comment has been removed by a blog administrator.