Sunday, November 25, 2018

More on insulin and the glycerophosphate shuttle

Raphi tweeted this paper recently

Nutritional Ketosis Increases NAD+/NADH Ratio in Healthy Human Brain: An in Vivo Study by 31P-MRS

which is nice provided, as he comments, it can be replicated. There is absolutely no possible conflict of interest anywhere so long as you accept it looks like an in-house Nestlé study. I haven't knowingly bought a Nestlé product in over 30 years.

Anyway. The study looks at healthy brain biochemistry under MCT induced ketosis. The ketone oxidation (or possibly the CNS oxidation of MCTs) increases the NAD+:NADH ratio, ie moves it in the Good direction.

There is a lot of talk about the NADH generation and NAD+ depletion during glycolysis to pyruvate, shifting the ratio in the Bad direction. The assumption (with which I disagree) is that the glycerophosphate shuttle is a rescue mechanism to regenerate essential NAD+ to allow glycolysis to continue, to which I will return in a moment.

The beauty of ketones is that they do not deplete cytoplasmic NAD+ at all and only consume one mitochondrial NAD+ during the conversion of BHB to AcAc. Because this happens within the mitochondria this, plus any NADH generated at the pyruvate dehydrogenase complex, is sitting next to complex I, the most prolific re-generator of NAD+ in the cell...

All well and good and bully for ketones and the manufacturers of Peptamen®1.5 Vanilla (Nestlé Health Science SA).

This got me thinking.

Of course no one in their right mind would expect glycolysis to be arranged in such a manner as to require the glycerophosphate shuttle for simple NAD+ regeneration. This is a wasteful loss of four pumped protons and this energy will appear as heat. Think of brown adipose tissue, full of mtG3Pdh, assuming insulin is plentiful.  The correct pathway for the metabolism of glucose without insulin is to lactate without any overall depletion of cytoplasmic NAD+. Lactate can then be taken up by mitochondria exactly as ketones are. Lactate will, in the mitochondria, be reconverted to pyruvate, depleting mitochondrial NAD+ in exactly the same way as the conversion of BHB to AcAc does. Equally this happen right next door to complex I, just waiting to regenerate NAD+ and keep that NAD+:NADH ratio nice and high.

The whole point of the glycerophosphate shuttle (in Protons terms) is to facilitate insulin signalling.  Insulin is the hormone of plenty, used to encourage caloric ingress in to cells. Loss of the four pumped protons due to bypassing complex I and using mtG3Pdh instead as part of insulin signalling appears perfectly reasonable under conditions of active caloric ingress. Sustained insulin signalling causes sustained loss of cytoplasmic NADH, which generates NAD+. Once this has happened there is no longer the surfeit of cytoplasmic NADH over NAD+ from glycolysis, which is essential to drive lactate formation. Glycolysis must therefor stop at pyruvate under insulin.

Summary: For insulin signalling the glycerophosphate shuttle is active and loss of NADH requires glycolysis to abort at pyruvate.

Without insulin signalling glycolysis runs to lactate which enters mitochondria without any depletion of cytoplasmic NAD+. The lactate should enter the mitochondria, under normal physiology.

Sooooooo. This had me thinking about what would happen if, in the presence of copious glucose and copious oxygen, there was to be a sudden profound fall in absolute insulin levels. I was particularly interested in systemic lactate levels.

A sudden, profound fall in insulin levels in the presence of glucose is pathology. It generates ketoacidosis, classically from acute beta cell destruction during the onset of DMT1. There is always a profound metabolic acidosis from the failure to suppress glucagon-induced lipolysis and subsequent massive acidic ketone generation. Under the canonical view the absence of insulin should not stop NAD+ regeneration by the glycerophosphate shuttle.

What I wanted to know was whether the Protons predicted shutting down of the glycerophosphate shuttle due to hypoinsulinaemia would result in diversion past pyruvate to lactate as the end result of glycolysis. In the presence of massive levels of ketones I would also expect this lactate to appear in the systemic situation.

Does it?

Yep. Ten seconds on Google says so.

Lactic acidosis in diabetic ketoacidosis

Very nice. I had no idea this was the case because it has no direct influence on treating DKA clinically...


Of course you have to think about the chicken and egg situation with insulin and mtG3Pdh activation (I have been for years!). Which comes first? I think insulin appears to be essential, as above. I do wonder if the insulin receptor will turn out to dock with the glycerophosphate shuttle in some way...


cavenewt said...

This is fascinating, although I struggle to keep my head above water in terms of understanding the finer points. Also the medium points. As well as some of the larger points.

So it may help if you could give an example of what might cause such a sudden profound fall: "This had me thinking about what would happen if, in the presence of copious glucose and copious oxygen, there was to be a sudden profound fall in absolute insulin levels."


Bob said...


I'm with you on the physiology. I have to read very carefully just to keep up, and even then it's quite a brain-teaser. My excuse of course is that I'm a lay reader of Peter's thinking. And he will always know more than I can ever learn from him.

I think the example was the 22-year-old T1 diabetic in the second paper. He either missed his insulin shot or gave himself too little and went into DKA. They found lactic acidosis and speculated as to why.

The post demonstrates the predictive power of the Protons thread concepts, at least concerning the glycerophosphate shuttle, which indicates Peter's been on the right track.

raphi said...

Schurr proposes that the end-product of glycolysis is always lactate and that pyruvate is obtained outside of the glycolytic pathway with LDH (or mLDH) doing the conversion. I find his model of glycolysis more convincing than the textbook 'branching' model giving either lactate or pyruvate. see here for his paper "Cerebral glycolysis: a century of persistent misunderstanding and misconception"

I bring this up because, I wonder, does it change anything to what you say here?

"Once this has happened there is no longer the surfeit of cytoplasmic NADH over NAD+ from glycolysis, which is essential to drive lactate formation. Glycolysis must therefore stop at pyruvate under insulin"

I think not, but im not sure...

I also think Schurr's model of glycolysis might be more congruent with the Proton's line of thinking to explain the situation where there's high glucose + high oxygen + super low insulin = lactic acidosis + ketoacidosis

if lactate is always the end-product of glycolysis then this circumstance is arrived at more directly than with a branching model of glycolysis (i think...)

Peter said...

Ah raphi,

What an enjoyable read! Your comment neatly ties together the primacy of lactate and the evolutionary context of insulin.

So. Glycolysis should run through from glucose to lactate without any branch to mtG3Pdh and so never give a requirement to accept pyruvate as the core mitochondrial fuel. Mostly.

Insulin. Insulin should be used primarily as cross talk between the pancreas and the liver. I’d have to go to Kraft’s work on what level of systemic insulin he considers pathology and I’ve loaned his book out to somebody. Insulin has no right entering the systemic circulation at 10,000picomoles. Probably not even 1000picomoles. If we accept that carbs were commonly consumed during evolution we would expect those carbs to be fibrous and never to deliver bulk glucose/fructose to the liver in quantities or speeds which allow penetration past the liver and in to the systemic circulation, requiring an emergency insulin dump from the pancreas to re-establish normoglycemia. This is pathology. Maybe honey is the exception for which a pathological level of insulin might be needed to deal with a human gather-able foodstuff…

So insulin, above basal levels, has limited role in the systemic circulation on an ancestral diet. Insulin signalling should be a minor player in a diet based around large herbivore fat. The glycerophosphate shuttle should only be as active as the low levels of insulin require, ie not very active. Pyruvate supply to the mitochondria should be low and Schurr’s beloved lactate should be the core mitochondrial fuel. Mostly.

No one should ever secrete so much insulin to induce insulin-induced-insulin resistance. I would think of the glycerophosphate shuttle as a sensor or emergency pressure valve. Not a bulk throughput but important for regulation of calorie ingress never the less. Nowadays it is so active that it has, to Schurr’s dismay, taken pride of place as “normal” glycolysis. Today it is normal. Today we have pathology.



Peter said...

Bob and cave, be back soon


Passthecream said...

Deeply interesting although I am struggling to understand. Peter and/or Raphi, any thoughts about how how Schurr's model might relate to the Warburg hypothesis eg the nice diagrams to base pondering on at Richard F.'s blog

raphi said...


i don't know if it really makes a difference to the metabolic theory of cancer - i don't think it does but i can't say i've spent a lot of time thinking about the implications of Schurr's model on this point. Schurr's model of glycolysis is more elegant though :) ... if you want to understand more about cancer and the metabolic theory i recommend Peter Attia's recent podcast with Thomas Seyfried

Seyfried is decades ahead of 99% of cancer researchers out there. even if his particular view of the metabolic theory turns out to be wrong he will still have been invaluable in mounting an air-tight case against the somatic mutations theory (SMT) of cancer simply belongs to the scientific graveyard.

Passthecream said...

At the risk of exposing my ignorance: if glycolysis always proceeds to lactate with Ox or without it (Schurr), and lactate is the tca input rather than pyruvate, it seems that the picture shifts to the underutilisation of oxygen and lactate by mitochondria rather than the overproduction of lactate by glycolysis. Faulty respiration, which is Warburg's main idea anyway.

T said...

"If you are post-obese, via low carb eating, there is every likelihood that repeatedly consuming the chips (to avoid the pancreatitis, don’tchano) will cure you of the post-ness of your obesity. Enjoy the chips by all means." -- Peter, on Ketogenic diet: Eat food

Peter, you are getting increasingly acerbic, and thus ever-more-devastatingly funny. I guess we need a bit of bile to digest all the fat. At any rate, better to fulminate against carbs than to have fulminating pancreatitis.

Just wanted to say thank you for your continuing higher-level cranky-man analysis, it is a real treat, even if most of it goes over my head and I have nothing of biochemical import to impart. Either way, your blood's worth bottling. Have a great Christmas break, and a wonderful 2019.

raphi said...


Glad you enjoyed. You can tell Schurr has a chip on his shoulder - I love it!

It's a really interesting point about mtG3PDH being an energy sensor/pressure valve rather than a bulk substrate delivery machine. I need to think through some of the implications...

raphi said...


lactate doesn't get into the TCA cycle but it does get into the mitochondria where it can then be turned into pyruvate. that lactate to pyruvate conversion can also happen in the cytosol (outside the mitochondria). there's a few nice diagrams in the Schurr paper I linked to if you want to have a graphical explanation of this

indeed, this fits nicely with the Warburg Effect. think of it this way: if you have faulty mitochondria and can't couple the electron transport chain to ATP production, but you still have 'activity' in the mitochondria, you're going to get
(a) a ton of lactate build up (since glycolysis proceeds to lactate)
(b) the ETC is still 'moving along' and oxygen is still flowing through the mitochondria, giving a false impression of normal ox phos
(c) lots of ATP is still available, and coming from mitochondria, but actually through ox phos but through substrate-level phosphorylation (SLP) instead

=== SLP depends on glutamine being sucked up through the alpha-ketoglutarate port of entry into the TCA cycle

this is what cancer researchers are missing. they don't understand SLP and don't design experiments that adequately control for it. they just say "oh hey lots of oxygen in the mitochondria, lots of ATP, oxphos is fine, nothing to see here move along. no faulty mitochondria"

and more expensive and useless "genotyping" and chemo and radiation is done without any real progress

a tragedy if there ever was one

Passthecream said...

Raphi I am slowly working my way through Seyfreid's book atm. Very slowly. He has a bigger chip on his shoulder than Schurr for good reasons I think. It's all way above my pay grade anyway. :/

You might find this older work interesting, nice differential equations and loads of lovely reaction constants in a very dense paper which combines spahetti with alphabet noodles. (You could install Modellica and just plug them all in.)

The effect of 5mM salicylate on liver mito. TCA is quite dramatic in their model, turns it into a short loop powered by transamination via glutamate, looping back from a-ketoglutarate, disables succinate dehydrogenase also.

Trying to digest that idea together with Seyfreid's ideas is interesting. In some ways it resembles his idea of faulty ox-phos but it seems to sidestep the extra atp from fermentation. Is that good or is it bad?