Hyperlipid by Petro Dobromylskyj
16 February 2019You need to get calories from somewhere, should it be from carbohydrate or fat?
I don't do twitter or facebook (If you are a metabolic person and have friend-requested me and I've ignored it please don't take it personally, it's just not what I do with faceache) to any extent and almost never use them for metabolic subjects. By an accident of twittard I picked up this tweet by Chris Masterjohn (via raphi and Mike Eades). It made me laugh out loud and still has me giggling occasionally:
"No! Carbohydrate restriction is the stupidest approach to fatty liver ever devised. If it “works” in any case it is almost certainly by supplying more methionine and choline, not by lowering carbs. It is impossible to make more fat from carb than you get by eating fat"
I can't help but think "chylomicron", "thoracic duct" and "physiology".
Then I giggle some more.
I guess it ranks along side of "Masterjohn", "Martha" and "RQ 0.454".
Have a look at this:
Limits of aerobic metabolism in cancer cells
"To gain a better understanding of cell metabolism as a function of the growth metabolic demand we performed a back-of-the-envelope calculation focusing on the major biomass components of mammalian cells".
In these days of "shotgun metabolomics" two people appear to have sat down with one or more sheets of paper (possibly larger than an envelope, though you can get some quite large envelopes I guess) and have gone back to basic first principles. They then published in a Nature journal. I love this. I feel it rates alongside getting this image published in Cell Metabolism.
TLDR for the paper:
Glucose drops through glycolysis to lactate at a rate where ATP generation per minute massively outstrips that available from mitochondrial oxphos. In redox balance. It's fast.
Anabolism from glucose consumes pyruvate (and phosphoenolpyruvate) which then forbids the pyruvate -> lactate NAD+ regeneration step. This imposes a need to avoid or deal with a cellular NADH excess.
In the Cell surface oxygen consumption (2) post I hypothesised that the regeneration of NAD+ at the cell surface would be in direct proportion to anabolism derived from pyruvate (ie glucose/glycolysis anabolism), to maintain redox balance (ie get rid of excess NADH, cycling it back to NAD+). It is particularly a feature of highly glycolytic cancer cells.
These folks appear to be saying the same thing but looking at differing cellular techniques to avoid NADH cumulation.
There's lots of other good stuff in there too. Like the rate of mitochondrial ATP generation from HeLa cell mitochondria compared to that of normal cardiac myocyte mitochondria. The ATP production via oxphos is an order of magnitude greater in mitochondria from the cardiac myocytes.
Oh, and glutaminolysis as another NADH avoiding ploy. This is the quote:
"Glutamate can be converted to citrate via reductive carboxylation. In this pathway the NAD(P)H production by glutamate dehydrogenase is compensated by the reverse activity of the NAD(P) isocitrate dehydrogenase (Fig. 1). Glutamate can be taken from the medium or generated from glutamine by glutaminase. Interestingly, arginine and proline can be produced from glutamate with concomitant consumption of NADH (Fig. 4a). This could provide an additional mechanism for NADH turnover".
Note that the glutamate is not being oxidised, it is running a small section of the TCA backwards to generate citrate for lipid synthesis, ie anabolism. This is not glutamate turning the TCA in the normal direction toward oxaloacetate to generate ATP via NADH and oxphos, because the mitochondria of cancer cells don't seem to do oxphos very well. Somewhat Seyfried supportive.
Using RQ to track whole body substrate oxidation is pretty straight forward. An RQ of 1.0 means glucose oxidation and of 0.69 indicates fat oxidation. Mixtures come out in between. It is very simple to show that glucose is routinely converted to fatty acids because in the immediate post prandial period for any rodent fed standard low fat crapinabag the RQ becomes greater than 1.0. We would expect that during the later period when the rodent is asleep/not eating there would be a lower than expected RQ (lower than the calculated food quotient, FQ) while predominantly stored fat is oxidised. But on a high carb, very low fat diet we would expect the overall averaged RQ over 24h to be a little under 1.0, ie pretty much the same as the FQ. For an hypothetical "all glucose" diet part of the glucose diverts thus via fatty acids:
Eating: Glucose minus a little O2 -> fat RQ > 1.0
Sleeping: Fat plus lots of O2 -> CO2 + H2O RQ < 1.0
CO2/O2 = 1.0 on average over 24h.
If that 24h averaged RQ was all we had to work with we would not suspect that de-novo lipogenesis ever occurred. Nice and simple.
Much more difficult to pick up is the bulk conversion of fatty acids to glucose. This produces an unusually low RQ in the short term. But if the glucose is being produced to fuel the brain during starvation then its prompt oxidation would "correct" the unusually low RQ back upwards to a fatty acid RQ. The obvious exception was noted in a metabolically fat adapted and lactating young lady during extended fasting. She made glucose and galactose from fatty acids and gave them to her baby, rather than oxidising the sugars herself. End result was an RQ of 0.454 after just over three days of fasting with continued breast feeding.
She was making sugar out of fatty acids in bulk. She might or might not have been doing the same without lactation but in the absence of donating the sugars to her infant this would never show.
So the RQ and the FQ always average out to be the same unless something very specific is happening, ie as with Martha.
Much more difficult is to ask how do you tell whether glucose is being converted to pyruvate which then enters the mitochondria to join the TCA or whether glucose converts to lactate which is then shipped in to the mitochondria. And what if you have the absolutely crazy idea that glycolysis almost always leads to lactate and that this lactate is a transferable currency between cells? Glucose is then viewed as a one way gift from liver to tissues, to be shared out between cells/tissues as lactate.
That latter view has to use tracers to look at lactate or glucose flux. Label some lactate with carbon-13 and infuse it to steady state in the plasma of a mouse. Kill the mouse promptly and humanely and look where the C-13 atoms have ended up in glycolysis and/or TCA intermediates. Repeat the process with C-13 labelled glucose. Then glutamate. And then any other intermediary metabolite which might remotely shift bulk energy around the body.
It turns out that in starch fed mice glucose and lactate are the bulk plasma energy carriers, lactate slightly more so than glucose in the fed state and much more so in the fasted state. Certainly on a molar basis, bearing in mind that a mole of glucose has twice the carbon of a mole of lactate, which makes the situation slightly more complex. But lactate labels the TCA more strongly than glucose. Not surprisingly glucose labels glycolytic intermediates better than lactate.
Free fatty acids and ketones are a separate subject in high carbohydrate/low fat fed mice but they flux remarkably little energy, at least when fasting is limited to eight hours. Brain metabolism is also another separate subject.
Glucose feeds glycolysis to lactate. Most of this glycolytic lactate enters the plasma pool. Plasma lactate feeds the TCA in other cells.
Now the insightful bit from near the end of the letter:
"Among the many metabolic intermediates, why does lactate carry high flux? Lactate is redox-balanced with glucose. The rapid exchange of both tissue lactate and pyruvate with the circulation may help to equate cytosolic NAD+/NADH ratios across tissues, allowing the whole body to buffer NAD(H) disturbances in any given location. Nearly complete lactate sharing between tissues effectively decouples glycolysis and the TCA cycle in individual tissues, allowing independent tissue-specific regulation of both processes. Because almost all ATP is made in the TCA cycle, each tissue can acquire energy from the largest dietary calorie constituent (carbohydrate) without needing to carry out glycolysis. In turn, glycolytic activity can be modulated to support cell proliferation, NADPH production by the pentose phosphate pathway, brain activity, and systemic glucose homeostasis. In essence, by having glucose feed the TCA cycle via circulating lactate, the housekeeping function of ATP production is decoupled from glucose catabolism. In turn, glucose metabolism is regulated to serve more advanced objectives of the organism".
What I think this is saying is that lactate supplied to the TCA/OxPhos is for "housekeeping" ie ATP production. Glycolysis is for anabolism. Neither is absolute, but I find it an interesting point of view.
So the ultimate TLDR is:
Ox-phos = housekeeping
Glycolysis = anabolism
There is probably significant fudge-room.
You don't usually learn much from statements which you, personally, consider likely to be correct. Annoying statements are far more productive.
Working through Seyfried's paper
Mitochondrial Substrate-Level Phosphorylation as Energy Source for Glioblastoma: Review and Hypothesis
I came across this snippet which galled me a little:
"It is glucose and not lactate that primarily drives brain energy metabolism (Allen et al., 2005; Dienel, 2012; Nortley and Attwell, 2017), making it unlikely that lactate could serve as a major energy metabolite for neoplastic GBM cells with diminished OxPhos capacity".
Now, people will realise that the astrocyte-neuron lactate shuttle is more than a little inflammatory as a subject, to say the least. Currently it is not doing too well in the face of experimental data, which are not at all straightforward to obtain. I went to Nortley and Attwell as the most recent reference. As a rather pro-lactate shuttle sort of a person I found their straw-man setting up of the shuttle rather annoying but their data potentially convincing, though I am far from certain about this. Here is the link:
Control of brain energy supply by astrocytes
This left me wondering what more pro lactate-shuttle people might be thinking nowadays, so I went via the "see all" button to locate this commentary by Tang:
Brain activity-induced neuronal glucose uptake/glycolysis: Is the lactate shuttle not required?
which is a rather more circumspect but still accepts a decreasing probability that the lactate shuttle is in any way crucial to astrocyte-neuron energetic coupling. The silver lining was this link, used to point out that in Bl6 mice whole-brain lactate extraction from plasma is essentially zero under the reasonably normal physiological conditions studied:
Glucose feeds the TCA cycle via circulating lactate
The basic concept in the paper, that lactate is the predominant metabolic substrate for the TCA is fine to me but that the source of lactate is predominantly extracellular is very counterintuitive. But the data presented are quite convincing. So I'm interested. I think it needs a little aside before talking about the paper itself and the situation in the brain in particular, so I'll post some random thoughts before looking at the paper in more detail. The biggest down side to the paper is the authors' failure to mention Schurr, the main proponent of lactate as a redox-balanced product of glucose, a very deeply insightful and much neglected observation. But then Schurr is a serious proponent of the astrocyte-neuron lactate shuttle...
Macavity, Macavity, there's no one like Macavity,
He's broken every human law, he breaks the law of gravity.
His powers of levitation would make a fakir stare,
And when you reach the scene of crime - Macavity's not there!
You may seek him in the basement, you may look up in the air
But I tell you once and once again, - Macavity's not there!
I think people might have noticed over the years that I'm not a great fan of metformin acting clinically by blockade of complex I. Particularly over the last few months I have waded deep through layers of references looking at the morass of "paradoxes" about metformin. In particular that 250micromolar in your plasma will kill you but 4000 micromolar can be justified in cell culture because meformin "accumulates in the mitochondria" at up to 1000 times the level in the blood/cytoplasm, which is a deeply held belief structure on which an unimaginable amount of grant funding hangs. Does metformin cumulate in mitochondria?
It doesn't. I really enjoyed this review from last year. Talk about reinforcement of confirmation bias:
Metformin-Induced Mitochondrial Complex I Inhibition: Facts, Uncertainties, and Consequences
But when you have spent years slogging through papers thinking: That's crap! it comes as a huge relief to find that it's not just you who thinks this, no matter how politely the reaction is phrased.
So. Is metformin research all garbage? No, of course not. Anything involving a live animal on oral doses which do not cause death by lactic acidosis is worth thinking about. Any parallel cell culture research in the same paper using a 4millimolar concentration can be junked. In vivo effects are real, at real dose rates. Cell culture at 1000 times overdose is fiction.
Just to summarise my own speculations:
Under fasting the component of insulin signalling facilitated by the glycerophosphate shuttle can be replaced by saturated fatty acid oxidation via electron transporting flavoprotein dehydrogenase. This maintains insulin signalling at the "cost" of increased fat oxidation. Hence the weight loss.
In the peak absorptive period after a carbohydrate based meal the normal development of insulin-induced insulin resistance is blunted and glucose oxidation continues for longer than without the metformin. If you are eating the absolute crap suggested by any cardiologist or diabetologist this might be of some benefit. If you are already LC the increased fasting fatty acid oxidation is probably where the benefits accrue from.
Cancer benefits are likely to be real, off "target" and secondary to reduced insulin exposure.
Edit: These people seem to be looking at the real world too. Worth spending some time on this
Low metformin causes a more oxidized mitochondrial NADH/NAD redox state in hepatocytes and inhibits gluconeogenesis by a redox-independent mechanism