Saturday, August 29, 2020

Ultra processed food (2) Haub vs Hall

I’ve taken this post down. With the information that Haub specifically restricted calories none of it makes any sense, however you discuss around the edges. Apologies to anyone who as been embarrassed, if it’s any consolation it has been much worse for myself!

Peter

Thursday, August 27, 2020

Protons (59) The NDI1 guys and gals are good

This little chap:























featured in this paper

Mitochondrial ROS Produced via Reverse Electron Transport Extend Animal Lifespan

which I discussed here. Obviously a group which can get the above image in to a Cell Metabolism paper has an admirably relaxed outlook on their own work and probably on science in general. You have to be good to have that mindset.

So now they have given us this review:

Role of Mitochondrial Reverse Electron Transport in ROS Signaling: Potential Roles in Health and Disease

which really summarises, at the most basic level, the nuts and bolts of what is happening to drive RET in the ETC under assorted inputs. The Protons starting point.




Edit: I've just fixed a broken link in Protons (03) Superoxide. Back in 2008 I was just starting to tease out the differences between glucose oxidation and lipid oxidation and the initial paper which started me on superoxide was this one from Muller et al

High rates of superoxide production in skeletal-muscle mitochondria respiring on both complex I- and complex II-linked substrates

Once you twig that palmitate always drives complexes I and II but linoleate doesn't drive complex II so much... The NDI1 people make this soooooo much easier that it was back then. End edit.


















They even found DHODH as another input (dihydroorotate dehydrogenase, I had to look it up too). Section A shows high ATP demand, low delta psi, minimal RET. Section C shows what happens when supply of nutrients exceeds ATP demand, delta psi rises and RET increases.

They've also got the TCA and beta oxidation working in parallel, as they do:














and have included the NADH:FADH2 ratios (admittedly upside down but I'm not complaining).

I hope their next move is in to subtleties of chain length and saturation to start to see how fatty acids have different ROS generating potential.

Then to relate ROS to insulin secretion/signalling. Insulin to obesity. Physiological vs pathological insulin resistance. Maybe metformin too.

But it's a great start. These people will go far.

Peter

Tuesday, August 25, 2020

So you want some DHA?

It seems like a very long time ago (only last year!) that George Henderson posted links in comments to the blog* about the absolutely crucial work done by Gibson and colleagues, documented in this paper

*Ooooh look, I just noticed how to link to comments. I'm so tech savvy!

Docosahexaenoic acid synthesis from alpha-linolenic acid is inhibited by diets high in polyunsaturated fatty acids

Another aside: Paywalled. If anyone has a few pence looking for a home Alexandra Elbakyan might be a good destination. I didn't say that. End aside.

It is impossible to say how good this work is. It's very good.

I'm no hyper-enthusiast for DHA. It's a tool. It does a job. Saturating yourself with the stuff is very likely to be a Bad Thing. This is perhaps best exemplified by the fierce negative feedback exerted by all of dietary C18 VLCPUFA precursors (omega 6 and omega 3s) on its synthesis (I would assume the same happens for arachidonic acid as well). The conversion of alpha linolenic acid to DHA is, for rats at least (and I would go with for humans too), very, very easily achieved by simply getting close to eliminating linoleic acid from the diet and also keeping ALA low, under 3% of calories. Here's my favourite figure from the paper, already tweeted and blogged by George:
























These are the DHA levels in phospholipids, presumably LDL and HDL secreted by the liver, extracted from plasma after three weeks of dietary intervention in Hooded Wistar rats.

"We conclude it is possible to enhance the DHA status of rats fed diets containing ALA as the only source of n-3 fatty acids but only when the level of dietary PUFA [ie all combined PUFA*] is low (less than 3% of energy)."

*My insert for emphasis.

Does anyone begin to recognise a pattern to PUFA requirements here?

Peter

Random aside. Rats. Have they been scavengers of the small amounts of edible tissue left on mammoth carcasses after humans had finished with them? Are rats evolved to be opportunist high fat, low PUFA adapted facultative carnivores? Now that's an interesting and useless thought but might help explain why they behave exactly as humans do on Surwit diets compared to low PUFA Surwit-like derivatives. Well, the idea entertains me. But then I like rodent studies...

Tweaks are ubiquitous

As so often happens you occasionally stumble over a gem by accident, like this one.

A neutral lipid-enriched diet improves myelination and alleviates peripheral nerve pathology in neuropathic mice

Ignore the TrJ mice, just look at the control mice.

They were being fed crapinabag chow or high sucrose (gasp), high starch (gasp), high anhydrous butter fat (mega gasp) based diets.

If I told you that the anhydrous butter fat diet was not supplemented with soya bean oil (OMG, these poor mice will develop life threatening PUFA deficiency in the six weeks of the study! Assume a sarcasm apology as being provided) would that make your ears prick up?

So the PUFA content of the Ultra Processed (sarcasm apology repeated) sugar/butter diet turned out to be 3.0% of the lipid calories, which makes PUFA under 1.5% of total calories...

Okay. You don't even need to read the results, you already know what the body weights are going to look like. In case you can't be bothered and would just like some confirmation bias, here they are:
















So. When researchers add "x" percent of soya bean oil to anhydrous butter fat diets to "prevent PUFA deficiency" you know that they are doing this, absolutely, because without the PUFA their butter based diets will not produce obesity. They know this, or at least the DIO manufacturers know this.

This insight is very, very important.

Over the years it has become clear to me that there are certain things which "everybody does" which are essential for getting the "desired" result. When Surwit specifies adding a PUFA based oil to an hydrogenated coconut oil based diet it is because he knows that it will NOT be obesogenic without it. He won't know why, but he will know that it is essential.

I think this is a general principle. I especially think it will apply to the intra cerebral injection of insulin behaving as a satiety hormone. There will be something which is routinely done which produces this effect and it won't be the insulin, it will be a "tweak", a normal lab procedure done for some plausibly justifiable reason. That's why it cannot be replicated in the hard nosed commercial lab of company which manufactures the insulin in question and which is looking to market the effect of the insulin, not of some dubious tweak (of which they are unaware).

I've no idea what the tweak might be.

But it will be there.

Peter

Monday, August 24, 2020

The miracle of fish oil (4) Adipocere

I learned a new word yesterday.

Adipocere.

Sadly, I find this word very fascinating. Here's why.

This is the first step of beta oxidation of a saturated fatty acid:












FADH2 derived electrons provide a large amount of energy as pumped protons on their route down the electron transport chain to oxygen. That's a standard part of life for any modern mitochondrion or aerobic bacterium.

But think of the old ways. Imagine you are an anaerobic microbe deep in an anaerobic peat bog in northern Siberia and that you would never even contemplate this new fangled oxygen based metabolism thing. Your core metabolic energy molecule might well be a very negative potential reduced ferredoxin molecule. There might be others but I rather like ferredoxins so I'll go with this one. If you want to do anything with reduced ferredoxin you need an electron acceptor. If you live in said peat bog with a dead mammoth you might just find a fatty acid molecule with a double bond. That's the electron acceptor you've been looking for. Bingo.
















Obviously you can only do this once with each double bond, and sometimes the fatty acid will break at the time of reduction of the double bond giving a pair of shortened molecules. Eventually everything ends up as pretty much a mess of saturated hydrocarbons which is termed "adipocere": fat-wax. A well recognised post mortem change in wet, anaerobic environments.

Which means that if you dig said mammoth out of the bog 40,000 years later you are not going to get a very pretty picture from your HPLC output when you go looking for the PUFA content of the mammoth adipose tissue.

So these guys have a problem:

The Fat from Frozen Mammals Reveals Sources of Essential Fatty Acids Suitable for Palaeolithic and Neolithic Humans

(Belated HT to Tucker for this paper, got carried away with adipocere!)

They need to reverse engineer the composition of the adipocere to make a best guess as to what was present in the mammoth while it was alive. Also adipocere formation is random. Sometimes a lot, sometimes a little, even within the same carcass. Tough call.

They used a combination of what they thought the mammoth might have eaten, what modern elephants have as their PUFA ratios and the output from the HPLC machine to do the best they could. I do not envy them in this task.

Here is their main results table after the reverse engineering process



















The top line (MY) is the mammoth, the others are horses and bison. The mammoth fat is calculated to have been composed of 7% linoleic acid and 18% alpha linolenic acid before adipocere.

Which looks preposterous to me.

The superscripts to the calculated percentages are the actually measured percentages. That would be 3.6% LA and 0.0% ALA, which were the basis for the modelling.

The superscript c to the MY identifier links to ref 19 which it claims specifies that "grass fed elephants" have similar values for PUFA to the mammoth values presented. Which sounds convincing.

Accumulation of polyunsaturated fatty acids by concentrate selecting ruminants

Until you find it is only one elephant.

And that there is absolutely no information in the paper about whether this one elephant was wild, ie grass fed, or was domesticated, ie concentrate fed. In fact none of the individuals have any information about grass fed, grain fed, hunted or slaughtered. There is no information as to what proportion of those 25% PUFA in the elephant's fat depot were LA vs ALA either. There is no information.

Who will bet it was a domestic working elephant fed on grains?

Me for one.

Especially because I've read this paper:

Molecular characterization of adipose tissue in the African elephant (Loxodonta africana)

All wild animals, culled as part of an elephant management operation.

How do their adipose tissue fatty acids pan out? In the absence of adipocere formation of course.


















That looks a bit like around 1% LA and 2% ALA.

I'd eat that.

Peter

Friday, August 21, 2020

Protons (58) When hydrogen peroxide becomes insulin

Preamble: I'm not going to discuss NADPH oxidase 4 or rho zero cells at this stage, not that these are unimportant or boring. For today's post it's just about some of the ROS from mitochondria.

Amber O'Hearn re-tweeted this paper,

Academic urban legends

with "Full disclosure: I didn't check the references" added. Which amused me greatly.

So you have to follow references back and back and back to be certain that the absolute fact that "X" causes "Y" is supported by more than someone's ad hoc hypothesis number 3297 as a one line throw away in a textbook from 1952. Or, worse, that they said the exact opposite! It happens.

I've spent an inordinate amount of time going through very old references in the past few weeks. The idea that hydrogen peroxide is an insulin mimetic turns out to be sound. It's not just an insulin mimetic for control of glucose uptake, it appears to be able to replace all of insulin's actions from initiation of signalling through to inhibition of signalling at high dose rates. The exogenous amounts needed in cell culture are compatible with the amounts generated by mitochondrial preparations under plausible conditions, as far as I am able to understand from the methods sections of isolated mitochondria papers. BTW For anyone who owns a MAGA hat you cannot replace parenteral insulin with parenteral hydrogen peroxide for diabetes management, undesirable effects will occur at the whole organism level.

I started out from this 2005 paper

Insulin Action Is Facilitated by Insulin-Stimulated Reactive Oxygen Species With Multiple Potential Signaling Targets

and went back in time to find out if it was true. This next paper is from 1974 when people were using transition metal ions to generate ROS, giving the realisation you could do the same thing with hydrogen peroxide alone, without the copper (or chromium) ion:

Evidence for Electron Transfer Reactions Involved in the Cu2+-dependent Thiol Activation of Fat Cell Glucose Utilization

This image is the rate of uptake of glucose into adipocytes under the influence of hydrogen peroxide in the culture medium:



















The effect was evident at 10micromol, peaked at 1mmol and was obtunded or eliminated by 4mmol. Bear in mind that these are the concentrations in the medium outside the cell. The concentration in the cytoplasm will be lower and within the mitochondria lower still. Catalase don'tchano. In isolated mitochondrial preps generating ROS in-situ we are talking nanomoles rather than micro or millimoles. But the pattern is there, where small amounts of peroxide get glucose in to adipocytes and larger amounts suppress this.

We can also look at the incorporation of glucose in to lipids and activation of the pyruvate dehydrogenase (PDH) complex using this paper, a jump forward to 1979:

The Insulin-like Effect of Hydrogen Peroxide on Pathways of Lipid Synthesis in Rat Adipocytes

where the pattern is repeated in the activation and deactivation by phosphorylation of the PDH complex at low and high hydrogen peroxide exposure (same pattern is seen for incorporation of glucose in to lipid too, graphs are in the paper):



















It's worth noting that the effect is present in the absence of glucose but  is enhanced when glucose is present at low levels. High levels of glucose swamp the effect (I didn't follow that particular ref) but I find this plausible because the glycerophosphate shuttle will be better able to generate supplementary ROS given a little glycolysis to work with.

I won't cite any of the many isolated cell culture papers showing that the oxidation of palmitate is good at generating ROS and that linoleic acid is poor at this, I've been through that too many times. They usually use high dose pure palmitate combined with hyperglycaemia and are aghast that cells die under these conditions. Palmitate is the devil incarnate. A deeper view allows more understanding.

Relating insulin signalling to mtG3Pdh activation and/or fatty acid oxidation ties ROS generation to insulin signalling and goes a long way to explaining many phenomena.

Peter

More lactate wars

I have a suspicion that lactate as a portable energy source might be going to become quite interesting. I'll hit publish on this post which has been lying around on the draft list for some time. Here goes.

Some groups of researchers have been interested in lactate as a fuel for oxidative metabolism for a very long time. My own biases rather like this approach, so beware.

Back in 2008 we have this paper:

Mitochondrial Lactate Dehydrogenase Is Involved in Oxidative-Energy Metabolism in Human Astrocytoma Cells (CCF-STTG1)

"Taken together, this study implicates lactate as an important contributor to ATP metabolism in the brain, a finding that may significantly change our notion of how this important organ manipulates its energy budget."

which is clearly preposterous if you are part of Fulghum's group in Kentucky. From 2019:

Mitochondria-associated lactate dehydrogenase is not a biologically significant contributor to bioenergetic function in murine striated muscle

"We find that cardiac mitochondria do not contain LDH ... These results indicate that cytosolic, and not mitochondrial, LDH promotes cardiac lactate oxidation."

"Our findings show negligible levels of lactate oxidation in isolated mitochondria from heart and skeletal muscle in sedentary, acutely exercised, and exercise-adapted conditions."

A finding which was promptly addressed in 2020 by Mailloux, now in Canada:

Lactate dehydrogenase supports lactate oxidation in mitochondria isolated from different mouse tissues

"Using the guide supplied by Passarella et al., we counter the conclusions drawn by Fulghum et al. and demonstrate that mitochondria oxidize lactate."

"Collectively, we can conclude lactate is a good fuel for mitochondrial bioenergetics in mammalian cells."

This is clearly an ongoing battleground and I doubt the exchange of half bricks is finished yet. It certainly brings to mind the astrocyte-neuron lactate shuttle

Lactate: the ultimate cerebral oxidative energy substrate?

which has largely been destroyed by

Control of brain energy supply by astrocytes

which I had a think about in this post. I do wonder if this declaration of destruction might be a little premature too. As was said in the Monty Python sketch: "I'm not dead yet!".

Ultimately, isolated mitochondria are very, very far away from anything physiological. I get the impression that the conditions they are studied under are utterly critical for the results you might like to get, or not get. The models are not useless per se but anything found needs to be considered very carefully, often in the absence of knowledge about what does and doesn't matter within the methods section and which may well have been tweaked to get the result desired. And in the context of what might be published next year.

I think abandoning lactate as a super-fuel might be a little premature. Beware of my biases.

Peter

Wednesday, August 19, 2020

Ultra processed food

This piece of epidemiological meta-analysis, hot off the press, is doing the rounds at the moment:

Consumption of ultra-processed foods and health status: a systematic review and meta-analysis

It illustrates yet another major error in nutrition research.

There are a few key words which flag a given publication for me as junk. If I see "reward" it signifies that the authors consider that certain foods force re-consumption and that such overeaten food has to be stored as fat. High "reward" overcomes the normal control of metabolism which has existed for millenia. This concept is junk to me.

The second phrase which alerts me is the "caloric density" of food. People really do think that you can trick metabolism in to overconsumption. That people and rats are programmed to (say) eat 100 mouthfuls per day. Put more calories in to each mouthful and you get fat. Another junk concept.

Now we have ultra processed food as the next junk term. Let's play a thought experiment.

Given a saucepan, a cooker, some milk, some rennet and a cheesecloth I think it's quite possible your granny might be able to put together something resembling a casein rich cheese-precursor. Somehow I doubt that she could produce a freeze dried pack of lab grade casein powder, so I think we can consider such a powder to be an ultra processed food component.

Sucrose can be extracted from beets or cane without too much technology but modern sucrose coming out of something resembling the Cantley sugar beet factory in Norfolk might be considered as ultra processed, never mind the smell. So might raw refined corn starch.

If you work at Sigma Aldrich you can take soya bean oil and convert it by an unknown (to me) and undoubtedly very, very clever technique in to tricaprylin, a triplet of octanoic acid molecules attached to a glycerol backbone. I challenge your granny to even extract the soybean oil from the soya beans, let alone convert it to tricaprylin. So I think we can suggest that this interesting oil is more than a little ultra processed.

Mix these components up and supply them to a lab in Japan to feed to some rats. We can merely look at the end weights from this paper:

Effects of Different Fatty Acid Chain Lengths on Fatty Acid Oxidation-Related Protein Expression Levels in Rat Skeletal Muscles

Feed one set of rats on crapinabag, which is about as un-processed as anything fed to a lab-rat ever gets.

Feed the next set on the tricaprylin mix, 60% of calories as this fat with generous casein, sucrose and cornstarch.

A final set can be fed with the same ultra-processed diet as the tricaprylin rats but with the soya bean oil left as soya bean oil.

Which rats get fattest? Okay, soya bean oil it is.

Which rats stay slimmest? Tricky. Whole food crapinabag or ultra-processed synthetic caprylic acid based syntho-food?

Well, I'd hardly be posting this if the ultra-processed food came out badly, now would I?

Here's Table 2













So, "whole food" SC crapinabag fed rats ended up at 239g bodyweight, seriously ultra-processed octanoate based MCFA at 216g, seriously ultra-processed soya bean oil based LCFA at 244g.

It's not the ultra processing. It's the effect on insulin, insulin signalling and the ability to resist insulin signalling when the resistance to that signal is physiologically appropriate. None of which was looked at in the paper, it was about something else.

Of course these are the PUFA levels:























The crapinabag was 11% fat, I think we can assume around just over half of that was linoleic acid, probably with a little alpha linolenic acid thrown in.

The 60% of calories as fat in the ultra processed diets both provided the same ratio of omega 3 to omega 6 but the absolute levels of total PUFA were around 3% for the MCFA fed rats and around 34% PUFA in the LCFA group.

The numbers speak for themselves.

What appears to matter is how capable adipocytes are to say "no" to extra in-coming calories. There are obviously a ton of down stream effects of distended adipocytes. Looking at PUFA combined with insulin shows how they get fat.

I'm the last person to suggest junk made of sucrose and starch are problem free but you have to be very careful of processed vs unprocessed as terminology when applied to foods. It's not likely to be as simple as it looks.

Peter

PS tricaprylin is interesting in its own right as it is weird stuff, but today I'm just looking at processed vs unprocessed. I hope no one would suggest that tricaprylin is an un processed food component.

Tuesday, August 11, 2020

Protons (57) When glucose becomes palmitate

 I'll just put this up as a brief post, there is a lot of background to it.

We all know that long chain fully saturated fatty acids yield approximately twice as much NADH as FADH2 giving an FADH2:NADH ratio just under 0.5 and that this high rate of FADH2 input at the CoQ couple facilitates superoxide generation by reverse electron transport through complex I.

Equally, we know that glucose oxidation, with five times the generation of NADH as FADH2, gives us a ratio of 0.2 and minimal reverse electron transport

We also know that, in order to balance the cytosolic NAD+:NADH ratio that NADH must be converted back to NAD+ to allow glycolysis to continue. This can be done using the malate-aspartate shuttle, conversion of pyruvate to lactate (both of which are redox neutral) or by using the glycerophosphate shuttle.

The latter is far from redox-neutral from the FADH2 input perspective. A cytoplasmic NADH is converted to an FADH2 within mtG3Pdh. This inputs at the CoQ couple. As far as the mitochondria are concerned that cytoplasmic NADH never existed. It behaves exactly as an FADH2. So, while the glycerophosphate shuttle is active, glucose presents to the mitochondria as two FADH2 and four NADH, giving us a nice, rather neat, FADH2:NADH ratio of 0.5. Slightly higher than palmitate or stearate.

I consider the glycerophosphate shuttle as generating essential ROS for insulin signalling. Small amounts of ROS generation facilitates insulin signalling. Large amounts inhibit it. Glucose, even hyperglycaemia, dose not generate ROS by the RET route. Adding insulin does do so because as the pyruvate dehydrogenase complex becomes more active then so the glycoerophosphate shuttle also becomes more active. The FADH2:NADH from glucose rises from 0.2 towards 0.5 and ROS increase to generate (given enough activation of the PDH complex) insulin resistance.

Insulin induced insulin resistance.

Peter