Protons: de novo lipogenesis

Okay, time to think about whole body insulin sensitivity, adipocytes and insulin.

First the core process; adipocytes which are listening to insulin will post GLUT4s on their surface, accept glucose and do enough de novo lipogenesis to both store and release palmitoleate. The palmitoleate is a signal that there is plenty of glucose around, let's use it.

Adipocytes are accepting glucose for signalling purposes and DNL lipid formation, plus they are sequestering away whatever lipid is available from the diet. The core function of insulin is the storage of DIETARY fat under the influence of carbohydrate. Boy, that is an old post! But the fact that there is plenty of glucose around means that the body should maintain insulin sensitivity, to make use of that glucose. But DNL in adipocytes which are insulin sensitive makes you, err, fat. As Cao et al point out in their lipokine paper:


"Additionally, genetic or pharmacological manipulations that boost de novo lipogenesis in adipose tissue (even though this sometimes leads to expansion of the fat depot) are associated with improved metabolic homeostasis (Kuriyama et al., 2005; Waki et al., 2007)."


I think this is a long winded paraphrase of the Hyperlipid concept "Getting fat is bad when you stop".


Increased insulin sensitivity in adipocytes makes you fat. That's as you would expect.

Back to the ice pick rats with their acute onset insulin hypersensitivity in adipocytes. During rapidly increasing bodyweight (on a low fat diet) there is a marked increase in obesity with excellent insulin sensitivity. Ended by six weeks.

Ditto MSG rats, but for in 4 weeks rather than 6 weeks. Can't tell from the gold thioglucose abstract, but at least a few weeks. Probably depends on all sorts of minutiae.

While ever these rodent models are gaining weight they maintain insulin sensitivity because they are doing DNL to get fat. On a low fat diet increasing obesity means DNL, palmitoleate and the ability to run metabolism on glucose. Logical.

Only once a brain-damaged rat becomes obese enough does hyperinsulinaemia set in, with attendant glucose intolerance. By this stage adipocytes are insulin resistant so have reduced ability to respond to insulin, reduced GLUT4 expression and, presumably, reduced palmitoleate synthesis. From the adipocyte's point of view there is not a lot of insulin around, whatever the blood concentration might be.

Lets look at the converse to obesity:

What does a lack of insulin signify? No food (or no carbohydrate, pax protein). Starvation requires insulin resistance as an obligate state for survival. How much good is palmitoleate going to do you under starvation or ketogenic dieting? Not a lot, unless you enjoy dropping precious glucose in to muscles until you brain falls to pieces.

An adipocyte which sees no insulin will not generate palmitoleate. If it generates anything at all (doubtful) it will be palmitate. Releasing some residual palmitoleate from adipocytes is fine for a few days, as long as there is glycogen hanging around. By three days this will be gone and so too should the palmitoleate. You are now in to hard core survival driven insulin resistance.

That's where I live.

Adipocyte distension induced insulin resistance is completely different. Here the adipocytes see low insulin when there is a ton of it around. There is a ton of glucose around too. But an adipocyte acts as per starvation and does what would be absolutely the correct thing under starvation circumstances. It releases palmitic acid and stops generating palmitoleate. Doing this while the macroscopic organism is eating bagels and french fries is bad. It's bound to generate massive hyperinsulinaemia to normalise glucose in the face of a ton of palmitic acid.



I'm just wondering whether there is time to look at the C57BL/6 mice. Just briefly as there is a lot of Mickey to be extracted on this subject when we get to idiots in detail. Briefly:

By an utter quirk of metabolism the VMH of C57BL/6 mice breaks under high dietary fat levels. So they have access to ample dietary fat when their VMH is injured, by definition. They store this fat because sympathetic tone to adipocytes is acutely lost and adipocytes become exquisitely insulin sensitive. Fat falls in to adipocytes as soon as the injury occurs, probably within hours of eating some butter, almost any amount of butter, however small.

But the fat they store is dietary fat. No DNL. They do not need to gain fat by DNL as it is there in the hopper. Dietary fat falls in to adipocytes. Palmitoleate synthesis? When fat is distending adipocytes so fast they are leaking FFAs despite losing lipolytic sympathetic tone? There is a ton of dietary fat dropping in to adipocytes, this is what gets released as they distend. By day three C57BL/6 mice are systemically insulin resistant. Their palmitoleate levels will be low and palmitic acid levels high. They are just a modification of the ice-pick/MSG/gold thioglucose family, but the process happens at warp speed due to the availability of DNL-free bulk fat.

Even on high fat plus high sucrose diets humans do not injure their VMH, at least not immediately. But C57BL/6 mice do and they have taught me a great deal over the years, made me think a great deal too. But they are still just a model, as explicable as the rest of the models, from the insulocentric point of view.

Once you have enough data.

To summarise: Palmitoleate is released by adipocytes when glucose and insulin are plentiful. Palmitate is released when glucose is sparse and insulin is low.

The sh!t hits the fan when glucose and insulin are plentiful but adipocytes are so distended that they THINK glucose and insulin are low. When both insulin and glucose are high you want palmitoleate. If your adipocytes give you palmitate under these circumstances you had better have a pancreas of steel or diabetes here you come.


I think we might go to PUFA and SCD1 in adipocytes before hepatic DNL in this series.


BTW It's nice to see people in comments being a post or two ahead! At least this isn't complete gobbledegook to everyone!

Peter

Protons: Palmitoleate

I think we have to start with the results section of Cao et al's very interesting (and free to study if you want all the detail) paper:

Identification of a Lipokine, a Lipid Hormone Linking Adipose Tissue to Systemic Metabolism

As always, the paper is a superb piece of detective work featuring a superabundance of genetically engineered mice from the C57BL/6J background fed an high fat diet, the nature of which doesn't make it in to the methods, but we can just assume that it's all the usual fare. They started from the protective effect of knocking out certain fatty acid receptors in the mice, which prevented the development of metabolic syndrome, and ran with this concept for the massive project detailed in the paper. It's big. It ended up with them doing the following to confirm that they got it correct. From the very end of the results:

"To define the effects of individual fatty acids on metabolic regulation, we prepared Intralipid with triglycerides composed of a single fatty acid, either TG-palmitoleate or TG-palmitate. Infusion of either lipid resulted in a two-fold increase in total plasma FFA levels with similar dynamics (Figure S13). While TG-palmitate suppressed the entire proximal insulin-signaling pathway including activation of insulin receptor and phosphorylation of insulin receptor substrate 1, 2 and AKT in liver, TG-palmitoleate strongly potentiated these insulin actions (Figure 7A). We observed similar effects of both lipids on muscle tissue where palmitoleate enhanced and palmitate impaired insulin signaling (Figure 7B)."

It's a switch, at the crude level of Intralipid infusions. Viewed macroscopically:

Palmitoleate = insulin sensitive
Palmitate = insulin resistant

I may have mentioned this before!

If you take a light switch apart, under the plastic there are some metal parts. The metal provides a sea of probability through which electrons can flow, provided the metal is continuous from light bulb to the powerstation (pax transformers). Or not flow, if we replace a few mm of metal with a few mm of room air.

If we accept that superoxide from complex I reverse electron transport is insulin resistance, then fatty acid binding proteins are a macroscopic overlay over this process, they are part of the plastic of the switch.

Superoxide never leaves the mitochondria, it probably converts to H2O2 to talk to the nucleus or acts locally to activate transcription factors which then talk to the nucleus. Adipocytes don't talk to muscles using superoxide either. The intermediary they use appears to be palmitoleate, probably the ratio of palmitoleate to palmitic acids, once you get away from bulk Intralipid infusions.

Why is it arranged this way? The body has to know what substrates are available. Ignoring protein, carbohydrate talks to the body through insulin, and through insulin transporting glucose in to adipocytes. That's the next post.

There: Not a mention of FADH2 or NADH. Even if I'm thinking about them, as per the last post...

Peter

BTW, Charles commented on the depressing amount of superoxide associated with a high fat, low carb diet. True, but about as scary as going for a walk at the brisk-but-not-excessive pace which is reputed to burn fat best. Burning fat is what LCHF eating is all about. Useful if you don't have the hours a day to walk for health purposes. Walking seems to be quite good for you!

Protons: FADH2:NADH ratios and MUFA

A few more thoughts building on F:N ratios of differing metabolic substrates:

Each cycle of beta oxidation (assuming an even numbered carbon chain fully saturated fatty acid) produces one FADH2, one NADH and one acetyl-CoA. This gives a total of 2FADH2 inputs and 4 NADHs per cycle of beta oxidation. But the very last pair of carbon atoms in a saturated fat do not need to go through beta oxidation as they already comprise acetate attached to CoA, so they can simply enter the TCA as acetyl-CoA. This last step only produces 1 FADH2 and 3 NADHs, with no extras.

So the shorter the fatty acid, the less FADH2 per unit NADH it produces. Short chain fatty acids like C4 butyric acid have an F:N ratio of 0.43 while very long chain fatty acids, up at 26 carbons, have an F:N ratio of about 0.49.

As Dr Speijer points out, differing length fatty acids are dealt with differently. Very short chain fatty acids head straight for the liver and get metabolised by hepatic mitochondria immediately. Any excess acetyl-CoA gets off-loaded as ketones.

Very long chain fatty acids end up in peroxisomes for shortening, usually to C8, which is then shunted to mitochondria for routine beta oxidation. Of course peroxisomal beta oxidation generates zero FADH2, except that from acetyl-CoA, because peroxisomal FADH2 is reacted directly with oxygen to give H2O2. And heat, of course.

Bear in mind that the ratio of F:N generated by a metabolic fuel sets the ability to generate reverse electron flow through complex I and subsequent superoxide production, macroscopically described as insulin resistance.

So fatty acids up to C8 are cool, dump them to the liver and make a few ketones. Very long chain fatty acids over C18, shorten to C8 in peroxisomes, shift them to mitochondria and make some ketones if needs must. The F:N ratio of C8 is about 0.47, a value chosen by metabolism as the end product of peroxisomal shortening. The number is important. Actually the number is even lower as peroxisomal beta oxidation generates the NADHs of beta oxidation, just not the FADH2s, but why allow facts like this to spoil a great argument. C8 from breast milk and/or coconuts seems fine and has that F:N ratio of 0.47.

Now the area of interest is, of course, C16, palmitic acid. This has an F:N ratio of about 0.48, almost as superoxide generating as a C26 fatty acid up at 0.49. And palmitic acid does, without any shadow of a doubt, produce macroscopic insulin resistance. That's 15 FADH2s and 31 NADHs.

So an F:N of 0.47 is not a serious generator of superoxide and an F:N of 0.48 is.

What happens when we drop a double bond in to palmitic acid? Mitochondrial beta oxidation generates FADH2 as it drops a double bond in to the saturated fat chain. If the double bond is already there, hey, no FADH2!

Palmitoleate has one double bond. This of course gives 14 FADH2s and 31 NADHs, an F:N ratio of 0.45.

Palmitate 0.48
C8 caprylic 0.47, chosen by peroxisomes to hand to mitochondria
Palmitoleic 0.45

Adding a single double bond to palmitic acid drops its F:N ratio from significantly superoxide generating to minimally superoxide generating. It looks like a switch to me.

I just love the way the numbers pan out. Of course we can now go on to what these number signify and what determines unsaturation. And uncoupling too, I guess. We are then back to insulin and stearoyl-CoA desaturase and also de novo lipogenesis. It might be worth an aside to PUFA and how these behave too, especially in adipocytes.

Peter

Protons: Lactate

I've been aware for some time that there is a reasonable idea that the brain runs on lactate. Dr Speijer emailed me a link to a very recent paper which supports this concept at the cutting edge of modern research, without having to go back to that old stuff from over five years ago which no one ever reads because it has no lovely photomicrographs and no ultracool transgenic mice.

The editorial has this nice diagram which sums up what might be going on:




Let's get back to electron donors. The brain hates superoxide. It hates fatty acids. It's a bit ambivalent about glucose (gasp). I don't think I would say it rejects glucose, just there are better fuels.

Is there anything the brain does like? Well, ketone bodies seem to be okay, but what the brain really seems to like is lactate. Perhaps I should rephrase all of this and say that the neurons of the brain love lactate. The rest of the brain seems fine on glucose and will even dabble with fatty acids at a pinch. But glucose is fed to neurons, pre digested by the glial cells, as lactate. The FFAs are fed as ketones, yes the glial cells in the brain are ketogenic, it's not just the liver that does this. I suppose the neurons might use glucose directly, but they become quite sick if you knock out lactate supply by eliminating MCT1 (mono carboxylic acid transporter 1).

Neurons are irreplaceable, more or less. They aim for zero superoxide production. This means behaving like a mitochondrial preparation which is being fed on glutamate, a provider of NADH only. Near zero free radical production is the closest you can come to having no mitochondria at all, yet still have the powerhouse of the electron transport chain at your command. When thinking about apoptosis that is. Apoptosis is not a good idea in non-replaceable cells...

This means minimising FADH2 utilisation. Fatty acids, with their beta oxidation derived FADH2, are out. No way in neurons.

Glucose is not ideal either. Why not? Well glucose supplies the best possible neuronal FADH2:NADH (F/N) ratio of 0.2, ie it gives one FADH2 for 5 NADHs. Usually. This is superb for minimising superoxide production (and maintaining insulin sensitivity). But not always. What about glycerol-phosphate dehydrogenase or glycerol-phosphate oxidase? Both of these, in much the same manner as the FADH2 moiety within electron-transporting flavoprotein dehydrogenase, can reduce the CoQ couple and promote superoxide production. That's without thinking about simply over driving the TCA with pathological hyperglycaemia. There is absolutely no doubt that hyperglycaemia generates superoxide production. Unfortunately most of the people discussing this on pubmed have no real concept of F:N ratios or what exactly goes on in the respiratory chain to generate superoxide. There is no nice neat diagram to copy paste. My own assumption is that massive enough inputs of glucose drive huge amounts of NADH production which cannot be accommodated once FADH2 from succinate dehydrogenase reaches a critical level or is supplemented by glycerol-phosphate dehydrogenase based FADH2. At this point a cell says no to glucose calories, ie it makes superoxide and becomes insulin resistant. As has been observed, insulin resistance is an antioxidant defence mechanism, you need it. If pushed hard enough to overcome insulin resistance a cell will take one step closer to apoptosis.

Not so with lactate. Lactate supplies acetyl-CoA (which itself has an F:N ratio of 0.25) along side a couple of extra NADH molecules (one each from lactate dehydrogenase and pyruvate dehydrogenase) which reduce the overall F:N ratio to 0.2, the same as glucose). Yet pre-prepared lactate does not need any glycolysis to take place in the neurons themselves. It has no possibility of supplying ANY FADH2-like input to the CoQ couple, outside of succinate dehydrogenase (complex II) activation in the turning of the TCA. It's the purest of complex I inputs available to any intact organism. No wonder the brain loves lactate. Lactate usage appears to be the best way of postponing apoptosis, short of abandoning mitochondria altogether. Glucose comes second.

I had always though of lactate in the brain as a sort of direct mitochondrial fuel injection system. Lactate dehydrogenase then mitochondrial uptake of pyruvate. Just a fast response time. But looking at FADH2 to NADH ratios gives a much deep insight in to what is going on.

What about fat????? Not for the brain.

But for the rest of the body? What makes mitochondria happy? Hint: It's not glucose.

Peter

Mmmmmm eggs!

Eggs will kill you!!!!!

As a UK resident: Thank god it's not London, London but London, Ontario. Phew. Thought the goons in epidemiology at Imperial College had been at it again. Happily the shame for this has to go to Canada. Oh dear, sorry Canada.

Peter

We are not alone

Obviously anyone with even a basic interest in origin of life questions will be watching the progress of Curiosity on Mars. Those of us who buy in to the serpentine and alkaline hydothermal vents concept will be interested in whether the crustal chemistry of Mars is olivine based and whether major water bodies were even present. Or equally, whether a semblance of white non-smokers might be present when acidic ground-water interacts with olivine, without needing an ocean and vents... An interesting time for testing hypotheses about whether there is life "out there", in our own back yard...

EDIT: A quick google shows olivine, serpentine and methane plumes are all present on Mars. The methane could easily be abiotic in origin, the question is whether it actually is or not...

On the more down to Earth front, if anyone thinks my basic ideas about the ratio of FADH2 based input vs NADH input to the ETC determining superoxide production are not totally incomprehensible, we are definitely not alone. I had a very nice email from Dr Speijer in Amsterdam, a fellow thinker along these lines. He has come to exactly the same conclusions and published an hypothesis paper in Bioessays back in 2011. The first section is just excellent. We may diverge in interpretation (but not FADH2:NADH ratios) very slightly late in the essay on PUFA, but it really is full of very good thinking and an excellent paper.

His ideas about peroxisomes (a very early eukaryotic invention) of course addresses that age old question of "Who's (macroscopic) fat is it anyway?", the answer being that the gut bacteria own it. On the sub cellular front, fat is primarily made in cytoplasm but at the behest of the mitochondria, only secondarily in peroxisomes and, as peroxisomes are probably a response to deal with overly long (ie excessively high FADH2 generating) fatty acids, the answer would seem to be mitochondria order fatty acid production, they own them and they have their own agenda for them. It's a sort of intracellular parallel the the fiaf series on gut bacteria and adipocytes. Very interesting concept.

If mitochondria own fatty acids I would expect them to enjoy burning fatty acids. Whatever the generation of controlled superoxide is, it's what keeps mitochondria happy. Then there is the brain to think about, its avoidance of fatty acids, it's love of ketones for an occasional fling and its very probable long term love affair with lactic acid. All based on FADH2 to NADH ratios of course.

There's a lot to post about. Back to the Protons series next (I think).

Peter

Insulin in the brain: Hyperphagia?

Let's start with this quote from Brain insulin controls adipose tissue lipolysis and lipogenesis:

"Insulin is considered the major anti-lipolytic hormone. Its anti–lipolytic effects are thought to be exclusively mediated through insulin receptors expressed on adipocytes (Degerman et al., 2003). Cyclic–AMP (cAMP) signaling represents the major pro–lipolytic pathway in WAT, which is chiefly regulated by the sympathetic nervous system (SNS)."

and then go on to this one from the discussion:

"We draw this conclusion from the finding that denervation of WAT leads to no change in lipogenic protein expression, but completely abrogates Hsl activation leading to increased adipose depot mass (Buettner et al., 2008)".

OK, got that? Brain insulin makes you fat by damping down lipolytic neurotransmission to adipocytes. Turning off your sympathetic nervous system supply to your fat cells allows insulin to go on an obesity spree.

Then there is this quote (MBH is medial basal hypothalamus, better known as VMH, ventro medial hypothalamus):

"Our studies raise several questions. One is which neuronal subtype within the CNS and the MBH mediates the effects of insulin on the regulation of WAT metabolism".

That first one really is an interesting question, one which we can go some way towards answering. We know that the cell type is, as already noted, part of the sympathetic nervous system. In the paper they found either surgical or chemical sympathectomy of adipose tissue increases both lipogenesis and inhibits hormone sensitive lipase in that tissue. I think this is straight forward. The sympathetic nervous system is tonically opposing insulin's lipogenesis effect and insulin's inhibition of hormone sensitive lipase.

The next thing we can say is that these cells sport glutamate receptors. We can safely assume this because, if neonatal rats are injected with the excitotoxin MSG, these are some (among many) of the cells which actually die. That is, the sympathetic nervous system supply to adipose tissue dies. Lipogenesis is unrestrained. Hormone sensitive lipase shuts down. Carbohydrate easily pours in to adipocytes and stays there. Blood glucose levels are low, free fatty acid levels are low, insulin sensitivity is excellent. While ever adipocyte expansion is on going that is. As the adipocytes stretch they eventually become insulin resistant. Here's the table of metabolic parameters from pre-obese MSG injured and control rats, from a previous post:



We know you can do exactly the same by killing these cells with gold thioglucose. This neurotoxin kills those nerve cells which inhibit lipogenesis.

As the authors say: After gold thioglucose injection "systemic insulin sensitivity is preserved [actually it's increased, but these are obesity researchers, so don't quibble] during the early phase of the obesity syndrome, resulting in extensive fat production".

These hypothalamic cells don't seem to take too kindly to the application of an ice pick either:

"In this study, we have measured the expression of the insulin-sensitive glucose transporter, Glut 4 and the activities and expression of key lipogenic enzymes (fatty-acid synthase and acetyl-CoA carboxylase) in white adipose tissue, one and six weeks after the lesion. All these parameters, as well as glucose transport and metabolism determined in white adipocytes, were markedly increased one week after the lesion. They returned to control values within six weeks in VMH-lesioned rats".

All of these interventions allow calories to pour in to adipocytes and stay there. So what does the poor rat do? It's losing a sh*t load of calories in to its adipocytes but, luckily, it has access to a massive hopper of crapinabag in its cage. It simply has to eat enough calories to supply the loss in to adipocytes, plus enough to run its metabolism on. This can be described, by non comprehending people, as hyperphagia. Metabolically it is normophagia.

These rats are calorically neutral or even in mild energy deficit. They have to be running "hyperphagic" just to stand still, metabolically. They are NOT showing "voluntary" overeating. They DO NOT have an injury to any sort of "satiety" centre. They have low insulin, low FFAs and low glucose. They are NOT being paid to over eat! They will NOT be producing a ton of superoxide, despite having a hugely increased caloric intake. Until...


When does this stop? It stops when FFA leakage due to the resistance to insulin induced by adipocyte distention exactly matches the FFA releasing effect which the (now non-existent) sympathetic nervous system would have been having on non distended adipocytes. Sorry for the convoluted sentence, can't simplify it! The distension process was complete by six weeks in the ice pick rat study cited above. Obese rodents then end up with a crudely normal metabolic rate. But this injured system is a complete bodge. We are looking at the replacement of a finely tuned fuel switching system which exactly matches fuel availability to fuel needs with a system where broken adipocytes are simply leaking FFAs at a level which constantly supplements glucose use, without any semblance of fine tuning to metabolic needs. The chronic elevation of fatty acids drives, through the NADH/FADH2 ratio, superoxide production and insulin resistance. Eating glucose then becomes unacceptable because there is inappropriate whole body insulin resistance from excess and inappropriate FFAs. A large amount of insulin is need to control hyperglycaemia under these conditions. Failure to supply adequate insulin to do this, for any reason, is labelled diabetes.

What has this to do with the current obesity epidemic? If you are overweight I would suggest you should take the ice pick out of your brain. No ice pick? Hmmmmm, damn! Back to the drawing board on that one then.

Ah, but maybe you are a C57BL/6J mouse?

Before we can tackle such a stupid question I think we need to go back to superoxide and fatty acids, to about where we were before this digression began.

Peter

Protons: Metformin

I'll just stick this post up to get it out of the way. I was going to go on to acute uncoupling next but the link from O Numnos in the last post comments is too good not to post about. It goes some way to tying weight gain in to LACK of superoxide, so brings the thread of insulin as a "satiety" hormone and this thread on weight gain as a failure to generate superoxide in adipocytes (good and bad) together. Might take more than one post... The summary of what's coming: Is insulin a satiety hormone? Only in so far as becoming stable-obese limits your hunger. Anyway, here are a few more thoughts on superoxide first.

Metformin is generally considered to be a Good Drug.

Interestingly it is an inhibitor of complex I of the respiratory chain, which is almost certainly its primary site of action. It aborts glycolysis to lactate because pyruvate is not much use to mitochondria with blocked complex I. Acute exposure to metformin in tissue culture generates a ton of superoxide. Just what you would expect to benefit someone with T2DM!

Let's have a look at this rather nice paper.

They are using differentiated 3T3-L1 adipocytes, a strange beast if ever there was one, but "everyone does it".

They are working under room air with 25mmol/l glucose, supplemental pyruvate, glutamine and 1000pmol/l insulin. These cells are being driven, hard, generating NADH which works through complex I. Complex II will be supplying some FADH2 but there is zero beta oxidation, unless the fatty acids in the adipocyte stores are being accessed. With insulin at 1000pmol/l this is not going to be happening.

Here is the effect of metformin on oxygen consumption:



A dose dependent fall, exactly what you would expect when blocking complex I. Here is the effect of 1.0 mmol/l metformin on oxygen consumption with time:



Nice curves! And here is the effect on ECAR, a surrogate for lactate generation, over 24 hours:



Metformin is only a relatively weak inhibitor of complex I, the incidence of life threatening lactic acidosis is very low. Not so for the more effective biguanides, phenformin and buformin. Obviously the latter two are no longer used clinically, there were too many hiccups.

Now, here is the level of DHE fluoresence, a specific marker of superoxide production. It's being compared to rotenone (remember Coopers Demodectic Mange Dressing? Thought not!), a serious complex I inhibitor.



Metformin is pretty good at generating superoxide. A bit counter intuitive for a drug which is the best treatment, short of insulin, for managing T2DM, a condition essentially defined by failure to overcome insulin resistance (aka superoxide production).

Hmmmmmmmmmmm.

Now, do 3T3-L1 adipocytes like being in forced to live on ATP from glycolysis plus whatever oxphos can be squeezed through metformin inhibited complex I? Annexin V is a marker of very, very unhappy cells. This is what metformin does to the % of cells which are moribund in culture:



So what is going on? Is metformin going to kill our fat cells in vivo?

It's all back to tissue culture conditions. Glucose at 25mmol/l makes the cells utterly dependent on a combination of glycolysis and NADH oxidation at complex I, plus a little FADH2 from succinate metabolism. Our adipocytes are not in this situation.

Now look at this graph:



This is in starvation medium. Only 2.5mmol/l glucose, no pyruvate, no glutamine. I think insulin is still supramaximal, but who cares about insulin when glucose is down at 2.5mmol/l in tissue culture? But here is the really interesting bit: They had also added 0.3mmol/l of palmitic acid to both the control cells and to the metformin cells. Compare it to the graph below, which is the same situation but with glucose and NADH drivers replacing palmitate:



So: Starvation medium plus palmitate completely reverses the fall on oxygen consumption produced by metformin. Palmitate plus starvation medium, even with metformin, actually allows more oxygen consumption that cells running flat out on glucose in the absence of metformin. It's what you would expect, the respiratory quotient is lower for fatty acids than for carbohydrate.

Metformin does not stop fatty acid oxidation. You do need some complex I activity to provide the NAD+ for beta oxidation, but no one is suggesting there is a complete block of complex I by metformin, it's not mange dressing.

So where do the free radicals come from with metformin? I would guess that the citric acid cycle still cycles, there is a build of of NADH due to complex I inhibition and complex II still reduces the CoQ couple. This could allow reverse electron transfer through whatever complex I functionality is left. There are absolutely no data on this, but I like the idea.

The group didn't look at superoxide production under starvation conditions or under starvation plus palmitate. I had a nice email reply to my query from the corresponding author along these lines, there are no data about this, as yet. I would expect the levels of superoxide to be comparable, with metformin being able to mimic palmitate based metabolism in the face of massive fat-free glucose supply, certainly for superoxide generation.

So, superoxide is insulin resistance. Adipocytes under metformin make a ton of superoxide. Are they insulin sensitive or resistant? Resistant of course.

Does an adipocyte which is insulin resistant listen to insulin's orders to store fat? Of course not. "Normal" insulin resistant adipocytes spew free fatty acids to the limit of albumin's transport provisions, with a few other moderating factors.

A metformin poisoned adipocyte is desperate for proton pumping substrate and complex I is doing bugger all to help. But electron-transfering flavoprotein dehydrogenase works perfectly well to allow an alternative electron supply...

Adipocytes under metformin have no choice but to burn fat. In vivo they have a barrel load of the stuff available as soon as they stop listening to insulin. They appear to use fatty acids for metabolism rather than dumping them as FFAs to plasma. Sounds like a recipe for treating metabolic syndrome to me.

Oh, that's what metformin is used for! Well I never...

So, do I think metformin causes adipocytes to become insulin resistant? Of course I do. Is this a Good Thing? You decide.

Peter

BTW Want an opposite to metformin? You can make adipocytes more sensitive to insulin with the thiazolidinediones. They allow insulin to become more effective on already over-distended adipocytes and generate lots of extra, nice, new, ready-to-stuff-with-fat adipoctes. They make you fatter. What would you expect?

Insulin in the brain: off topic giggle

I had my septic tank emptied a fortnight ago. The contents were a load of crap, but less crappy that the paper purporting to show that insulin is a satiety hormone as quoted by some obesity researcher.

What REALLY happens when you infuse insulin in to the cerebro spinal fluid of a mouse? You know, the satiety hormone... Just in to the brain, nothing systemic, no hypoglycaemia.



Insulin = big fat adipocytes. Big fat mice. Lovely micrographs.

http://www.jci.org/articles/view/31073/figure/5 will give you the legend.

http://www.jci.org/articles/view/31073 will give you the full text. Might discuss the paper better in a few months time!

But main conclusion:

The brain fine tunes the storage of lipid under the influence of insulin (by increasing fat storage via lipoprotein lipase and also by DNL from glucose). It uses the sympathetic nervous system outflow from the ventromedial hypothalamus to do this. Interpret with caution as these are C57BL/6, mice who may well have some very specific weakness in their ventromedial hypothalamus.

OMG did I laugh when I found this one.

Wanna loose some weight, go eat some potatoes. LMFAO!

Sorry for the crudity. Been on call too long this weekend!

Peter

Protons: Fasting

OK, this is another slightly sideways look at the paper on insulin resistance as an antioxidant defence mechanism.

The basic finding is that manipulating superoxide levels as close as possible to the ETC suggests that it is THE mediator of insulin resistance. Again, I'll skip a large amount of the extreme cleverness utilised and look at the bottom line and its implications. BTW the cleverness was very, very clever. How superoxide controls responsiveness to insulin, nobody knows (though George has some interesting ideas). But it appears to be a generic finding. They looked at steroids, they looked at TNF alpha, excess insulin (good old Somogyi) and, as you might expect, palmitic acid (as in the last post, on a background of 25mmol/l glucose). All cause insulin resistance in the models used. Also bear in mind that they are looking at myotubules and rather peculiar adipocyte-like cells. But I think they are probably correct in this basic conclusion.

Superoxide is core to insulin resistance.

It is very interesting to take this concept and look at various insulin resistance syndromes over the next few weeks.

Of course these folks are in obesity research so you have to be quite cautious when looking at their models and results. You also have to be very, very wary about their conclusions. This is the last sentence of the abstract:

"These data place mitochondrial superoxide at the nexus between intracellular metabolism [tick, agree] and the control of insulin action [tick, agree] potentially defining this as a metabolic sensor of energy excess [woaaaaah, care here]."

This is a slightly tricky sentence. It's that "excess" which bugs me. Look at section L from Fig 4 in the discussion to see how they are thinking:



Here we have a schematic of inactivity and overnutrition causing increased mitochondrial superoxide production. This clearly relates to the Denmark paper where people were paid to eat to excess while deliberately reducing their exercise. Fasting insulin spiked from 35pmol/l to 74pmol/l in 3 days. You can say that overnutrition certainly generates superoxide production. But is this what is happening in weight gain outside of paying people to over eat? That is not how most obese people become obese!

Inactivity and over nutrition are macroscopic changes and superoxide generation is a sub cellular mitochondrial effect. You have to be very careful in how you link the two features together. Superoxide may always signal insulin resistance but are there other drivers of superoxide production in addition to caloric excess?

The situation which keeps coming back to me is starvation.

There is no over nutrition during starvation. There is plenty of superoxide production. Why?

Humans have a brain which is rather dependent on glucose. Using glucose for non brain purposes during starvation would be potentially fatal. All tissues which can become insulin resistant should do so under these conditions.

Superoxide is utterly essential to the survival of starvation. Insulin resistance is a complete necessity.

It looks very much as if fat oxidation (especially palmitate) is directly set up to ensure this happens. It's the reason I was blogging about beta oxidation and FADH2 here. Fat supplies only two molecules of NADH for each of FADH2 and the beta oxidation derived FADH2 enters the electron transport chain through electron-transferring flavoprotein dehydrogenase, directly to the CoQ couple. This is a good situation to generate reverse electron transport, subsequent superoxide and trigger a specific refusal to process insulin. An overnight fasted human has total FFAs of around 0.5mmol/l and they stabilise at around 1.5mmol/l by four days of starvation. They stay there until some food, especially carbohydrate, is eaten.

This level (1.5mmol/l) should, by necessity, develop enough insulin resistance to stop GLUT4 dependent tissues from using glucose, to spare it for brain tissue.

Survival during starvation does not just necessitate using stored fat for energy. It necessitates the near complete abrogation of glucose usage for anything other than brain function. Not after that mere 14 hour fast before an oral glucose tolerance test, but certainly by four days without food. This abrogation cannot be reversed in a couple of hours during an OGTT. This is the "diabetes of starvation".

Superoxide is not always a marker of excess, though this is certainly one way of generating it. It is more accurately a marker of any situation in which insulin resistance is beneficial to survival.

Peter

And I really will get to emails some time soon (mea culpa!)

Protons: 25mmol/l

This post is a slight aside based on minor details in the paper "Insulin resistance is a cellular antioxidant defense mechanism". Which shows, quite clearly, that palmitate at levels as low as 0.05mmol/l causes some degree of insulin resistance. By 0.15mmol/l it's significant and by 0.5mmol/l it's worse. The graphs are from myocytes in cell culture.



This graph show GLUT4 count on the surface of myocytes. Left is control, next is the count after an acute exposure to insulin, taken as the 100% response. Adding more and more palmitate decreases the percentage response.

No one, not even a fatphobic vegan, has palmiatate levels in the FFAs of their blood as low as 0.05mmol/l.



This graph shows the effect of exposure time to 0.15mmol/l on GLUT4 translocation. Things get worse by the hour. Is this real?

Let's have a look in the methods:

"Palmitate (PALM) treatment was performed essentially as described in ref. 5"

So let's go to ref 5:

5. Hoehn KL, et al. (2008) IRS1-independent defects define major nodes of insulin resistance. Cell Metab 7:421–433.

In results we get this superb snippet:

"In our preliminary investigations, we observed that high (>300 μM) palmitate doses were toxic to cells, resulting in morphological changes and even detachment from the substratum."

Palmitate is clearly pretty nasty stuff. And I feed it to my daughter!

In the methods section under "Oxidative stress" we get a description of "stepdown medium", as used in both of the studies discussed. The composition of cell culture medium may be common knowledge to people using cell culture for a living but it was news to me. This is virtually a throw away comment:

"...while total glucose levels (measured with an Accu-Chek II glucometer [Roche]) decreased slightly from 24.7 ± 1.6 mM to 23.3 ± 1.9 mM".

The DMEM cell culture medium used here contains 25mmol/l of glucose!

So let's rephrase that toxicity of palmitate:

"Palmitate at 0.3mmol/l is severely toxic to cultured cells in the presence of 23mmol/l of glucose".

That I can certainly believe.

Aside: It is very interesting to note that palmiate at 0.15mmol/l is low-physiological for a human and yet is described as toxic, without mention of the grossly pathologic 25mmol/l of glucose in the culture medium. I suspect that whenever you look at a cell culture based study demonstrating palmitate toxicity this will apply.

So, does palmitic acid cause insulin resistance (aka superoxide production) under low glucose conditions? Where there is no caloric overload?

Studies looking at palmitic acid in the presence of low glucose are as common as hen's teeth...

Peter

Protons: Superoxide

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

is today's obligatory reading. What is it all about? They are trying to tease out what is happening directly within the mitochondria when different metabolic substrates are offered to the citric acid cycle. This is not easy. Their mitochondrial model is far from reality but is the best we can do at the moment, certainly in 2008. The authors are well aware of this and discuss the issue in some detail. I feel, personally, that what they have found makes perfect sense and can be extended, as is my tendency, to make a few more links in to basic physiology and, eventually, insulin resistance.

It's also worth pointing out that they are dealing with isolated mitochondria. No cells, no insulin, no adipocytes, not even any cytoplasm to support glycolysis. You can't run mitochondria on glucose! A model.

Here is their starting point:

"While it is generally accepted that mitochondria are the main site of cellular ROS production, studies in isolated mitochondria have shown that the amount of H2O2 released by mitochondria (H2O2 originates from the dismutation of O2•− [6], and is much easier to measure than O2•−) undermost conditions is rather modest".

And what is "needed" for this spewing of electrons on to molecular oxygen?

"In isolated mitochondria, reverse-electron transfer through complex I occurs when the ubiquinol pool is in a highly reduced state and a strong membrane potential is present, i.e. the energy of the membrane potential drives the ubiquinol (with electrons provided by succinate)-dependent reduction of NAD+ to NADH with electrons passing in the reverse direction through complex I [18]."

We talked about the CoQ couple in the last post. Here it is being referred to as ubiquinol, the reduced form of CoQ. A reduced CoQ couple is essential for reverse electron transport.

Also note the necessity of "a strong membrane potential".

Back to the study:

They isolated mitochondria and studied them asap, in the short time-window during which they remain remotely functional. They fed them with various components of the citric acid cycle and looked at H2O2 production, a reasonable surrogate for superoxide.

Adding in glutamate mixed with malate is a classic combination for driving NADH utilisation through complex I. The mitochondria generate about 30pmol/min of H2O2 under these conditions. This is not a lot and some labs report the amount as being close to zero.

Driving complex II (succinate dehyrdogenase) directly, using succinate but not supplying any NADH generators, produces rather more H2O2, around 400pmol/min. A ten fold increase.

Driving both complex I and complex II with a combination of all three of the above substrates can produce over 2000pmol/min H2O2. Sometimes but not always, as we shall see.

All of these findings where achieved at physiologically plausible concentrations of substrate. However oxaloacetate, formed in-situ from the malate supplied, turned out to be a confounder for that last result. Oxaloacetate is an inhibitor of complex II, so reduces reverse electron transport through complex I. With succinate dehydrogenase partially blocked the CoQ pool is less reduced, so more ready to accept electrons from Complex I, rather than driving them the wrong way through it to generate superoxide.

The end conclusion the paper came to is that anything which depletes oxaloacetate will disinhibit succinate dehydrogenase, reduce the CoQ pool and at the same time increase the likelihood of reverse electron transport through complex I, leading to superoxide generation.

The follow on from this is that anything supplying large amounts of acetyl-CoA will automatically deplete oxaloacetate because citrate synthetase consumes oxaloacetate as it combines it with acetyl-CoA to start the citric acid cycle with, err, citric acid. The group did it with pyruvate. They did it with palmitoyl carnitine. Pyruvate = carbohydrate. Palmitate = fat.

In the real world the cycle turns and the acetyl-CoA source which initially depleted oxaloacetate eventually restocks the oxaloacetate supply (except in the ketogenic liver of course). But the initial oxaloacetate depletion sends a signal. The mechanism is the activation of succinate dehydrogenase, which both allows the citric acid cycle to cycle and promotes a significant reduction of the CoQ couple which can generate superoxide.

This is a basic physiology paper. There is no suggestion or mention of gluttony, however coded. But excess acetyl-CoA might be suggestive of freely available metabolic substrate. But no comment in this paper, except my me. You could simply say that overeating supplies too much acetyl-CoA. Hmmm, maybe I should go in to obesity research? But there are additional considerations, like this one:

I would just like to point out that beta oxidation, through electron-transferring flavoprotein dehydrogenase (previously described in Peter terms as complex II-like), reduces the CoQ pool independent of succinate dehydrogenase (genuine complex II). Would you expect a reduced CoQ pool from beta oxidation to predispose to superoxide generation? Might this be a Good Thing or a Bad Thing? Think carefully about the semantics here. What do we mean by "good" and "bad"?

Superoxide is important. It speaks to tissues far away from the mitochondria which generated it in addition to the cell containing them. We need to translate this in to more familiar terms, which has been hard work avoiding slipping in to in this post!

The next post is about insulin resistance. You would be dead without it.

Peter

Hazel eats butter

Butter, rich source of palmitic acid. Uh oh, I think superoxide may be replacing palmitic acid as my favourite molecule. One good thing leads to another. Must get on with the next two posts!

Peter

BTW the mess on her face is, as always, 90% cocoa chocolate.

Protons: Where's the bias?

Executive summary: Complex I and Complex II are separate routes in to the electron transport chain. Glucose favours Complex I, fat favours Complex II. Now the extended version:

Here we have a nice schematic of the electron transport chain in a diagram of a mitochondrion taken from Wiki images.



The ATP Synthase complex shown on the upper left of the mitochondrial diagram allows protons from outside the inner mitochondrial membrane to pass back in to the mitochondrial matrix, generating ATP in the process. Under White Non Smoker conditions this electro chemical gradient might well have been maintained for free by the geochemistry of serpentinisation plus an acidic ocean. Nowadays the combined pH and electrical gradient which drives this ATP factory is maintained by the electron transport chain. This transports positively charged protons out of the mitochondrial matrix to maintain the gradient which is dissipated during ATP production.

In the diagram you can see two versions of the ETC being driven off of the citric acid cycle. On the upper right hand side a molecule of NADH provides electrons to Complex I. Complex I pumps some protons, hands the electrons to the Coenzyme Q pool (CoQ, marked as Q on the diagram) of electron transporters which then hand them on to Complex III. Complex II is not involved. The CoQ pool is a mobile reservoir of redox shuttles (electron transporters) which hands electrons to Complex III.

The second version, shown on the lower area, has succinate feeding in to Complex II. Complex II is actually the succinate dehydrogenase enzyme of the citric acid cycle. It is built in to the wall of the inner mitochondrial membrane and hands its electrons to the CoQ pool directly, no Complex I involved. Another difference is that Complex II doesn't pump any protons.

The proton pumping done by electrons passing through Complexes III and IV is independent of their route of entry to the ETC. Anything feeding in to the CoQ pool feeds onwards through Complexes III and IV. Mostly.

So we have the citric acid cycle processing acetyl-CoA to a ton of NADH for Complex I and a smidge of FADH2 within Complex II.

The FADH2 is quite tricky. It is embedded deeply within the succinate dehydrogenase enzyme and never, as far as I can make out, goes anywhere. It flicks between the FAD and FADH2 state as the citric acid cycle turns and basically acts as a bridge to transfer the effective oxidation of succinate to the reduction of the CoQ couple.

Another route in to the ETC, which seems sorely neglected, is Electron-Transferring-Flavoprotein Dehydrogenase, which sadly has no handy name. ETFD sits in the inner mitochondrial membrane and passes electrons to the CoQ couple, much as Complex II does, also without puming protons. ETFD gets its electrons from the FADH2 of an electron transfer flavoprotein which, thankfully, gets its electrons from the FADH2 of acyl-CoA dehydrogenase, the first enzyme of beta oxidation. Back on home territory.

Phew.

So fatty acid beta oxidation feeds in to the ETC at a "Complex II-like" membrane enzyme. It uses FADH2 to do this. It generates a small amount of NADH as well.

So we have two non-Complex I inputs in to the CoQ couple.

Aside: There are three if we include glycerol-3-phosphate dehydrogenase. Four if we include glycerol-3-phosphate oxidase Probably more. But let's keep it simple and stop at two... Actually glycerol-3-phosphate oxidase is really interesting as it specifically generates H2O2 enzymically. H2O2 production is generally considered to be a Bad Thing. Now what might the deliberate generation of H2O2 be signalling? Very interesting! Maybe another day.

So the citric acid cycle inputs just a few electrons through FADH2 at Complex II compared to the number it supplies using NADH at Complex I. Glycolysis is even more Complex I focused as it only adds NADH to its acetyl-CoA generation. However beta oxidation markedly inputs through the FADH2 of ETFD, with relatively little input using the NADH from the beta oxidation process, again in addition to generating acetyl-CoA. Obviously all acetyl-CoA generates the same ratio of NADH to FADH2.

The actual biases can be seen from these numbers, nicely posted by Lucas Tafur here. A direct quote:



As you can see glucose produces 5 molecules of NADH for each FADH2 where as fat produces only 2 molecules of NADH for each FADH2.

Glucose drives complex I significantly harder than fat does. Fat drives with a "Complex II-like" bias, supplying FADH2 from ETFD much as succinate dehydrogenase supplies some FADH2 from acetyl-CoA.

Both FADH2 inputs do exactly the same thing to the CoQ couple, they reduce it. A reduced CoQ pool has major implications for electron transport and free radical generation.

I rather like eating fat. What does that do to Complex I?

It's probably not the obvious answer.

Peter

Protons: Where's the pump?

This post, pictures excepted, is largely based on the core ideas presented in this paper by Nick Lane, John Allen and William Martin. It's downloadable as a pdf from Nick Lane's website and gives great pleasure in return for careful reading. There are more details on the nature of catalysis in pre protein conditions and the acetyl-CoA pathway in Michael Russell's paper (unfortunately PPV, I have the text) co-authored with William Martin here. Some ideas make a great deal of sense. These are in that category. Enough preamble, on to the post:










The Lizard Peninsula in Cornwall is an interesting place. For a variety of geological reasons a chunk of deep ocean mantle is available to visit on the Earth's surface, without getting too wet or borrowing a deep ocean submersible. We visited Kynance Cove about 10 years ago to pick up a few pebbles of serpentine. We were LC beginners at the time.



Serpentine is formed during the hydration of olivine by sea water, as it percolates in to the earth's crust. The process generates heat, produces molecular hydrogen and increases the volume of the rock by about one third, massively fragmenting it. Sorry for the lack of a hammer.



Large amounts of warm, hydrogen rich fluid are produced under pressure and enter the ocean at hydrothermal vents. The chemistry of serpentinisation also means that the fluid is alkaline. The process is continuous and occurs over geological time scales. These are alkaline hydrothermal vents. I prefer the term of White Non Smokers (rejected by Nick Lane).

Unlike the well known Black Smokers of the mid oceanic ridges/troughs, White Non Smokers generate temperatures and mineral concentrations which are not particularly aversive to the abiotic chemistry which might be considered as pre biotic.

The early Earth is thought to have had an atmosphere, like that of present day Mars, which was CO2 rich. This would have made the early oceans mildly acidic.

White Non Smokers are also structurally full of microporous vents. These have vesicle structures which have warm alkaline fluid within and cool acidic fluid without. There is, intrinsically, a pH gradient across their wall. The difference in positively charged hydrogen ions across the vesicle wall is comparable to the proton gradient across microbial, and of course mitochondrial, surface membranes.

This might be where life started.

If it is, here is the proton gradient, nowadays maintained by the electron transport chain pumping protons, pre dating the development of that chain. Under WNS conditions it is possible to generate high energy molecules using the geochemical proton gradient intrinsic to the vent vesicles. For life as a more distinct entity to leave the WNS suburb simply requires a method to maintain the proton gradient away from the geochemical reactor which initially sustained it.

ATP or, in all probability, a simpler high energy molecule could be made for free in the WNS environment. Away from any "free" proton gradient you need to do work to sustain one. Acetate (like methane) is one of the few products of the exergonic combination of molecular hydrogen (from the vent) with CO2 (from the ocean) which supplies enough energy to maintain a proton gradient in an ATP producing state, without any other energy input, no geochemistry, photons or complex organic chemicals.

There are modern, highly evolved and sophisticated bacteria thriving on this utterly primordial pathway even today. Acetate, to them, is waste. But it can be used if you are so inclined.

As you might expect, activated-acetate (nowadays in the form of acetyl-CoA) forms the basis of most modern metabolism, generating ATP in large amounts through the electron transport chain. But it's probably less primordial than the proton gradient itself.

I think that the mitochondrial inner membrane potential both pre dated life and is possibly rather important to on going life. And you can adjust it, under modern conditions, by what you do or don't eat off of your plate...

Which has some bearing on health.

Peter

Are you free, T3?

Measuring thyroid hormone level is a very simple matter in clinical practice. For total T4 you can do it in-house if you don't care too much about accuracy. A commercial lab is better. Very occasionally you meet a patient with very clear cut markers of thyrotoxicosis which has a T4, as measured by a commercial lab, which is persistently within the upper end of the lab reference range for normality.

For these (usually cats) we check the free T4 level. Free T4 is not cheap, despite the name, and takes some time to come through from a referral lab. We use this test because almost all of the total T4 in plasma is bound to albumin and assorted other plasma proteins. We need the "unbound" or active concentration thyroxine because this is what does what it does. Of course no one really wants to measure free T4 anyway, what we want is the actual active hormone, T3. Preferably free T3. However, for many cases, free T4 is good enough.

Measuring T3 or (gasp) even free T3 is another ball game and is something I only request occasionally. Usually when trying to get to the bottom of apparently hypothyroid dogs when all lab results come back "borderline low". Unfortunately free T3 is not available in the UK and sample gets couriered to the USA. I think the courier must swim the Atlantic judging by the time taken.

But ultimately even the free T3 is only a surrogate for the level of T3 which is actually bound to its receptor within the nucleus of each cell, including those of the brain.

Measuring receptor occupancy this is neither easy, clinically appropriate nor commercially available. But fortunately there is a surrogate.

You can get an idea of whether the brain thinks there is enough T3 sitting on its receptors by whether it is asking for more. It asks for more using (eventually, after several steps) TSH, thyroid stimulating hormone. This is released from the pituitary as a signal to the thyroid to increase production.

So the rule of thumb with a suspected hypothyroid patient is to ask whether the TSH level is elevated, ie is the brain unhappy with the current thyroid level. When you don't have the time or finances available for that courier to swim the Atlantic, this is what we use. It's a surrogate, but useful.

As so often, this is just basic clinical chemistry. It defines how I view hypothyroidism.



So let's put some folks on a diet, get them down to 10-15% below their start weight and look at their thyroid status. Keep them as weight stable as you can and look at total T3 levels on three different diets.

Not surprisingly the level of thyroid hormone falls with weight loss. The run-in diet provides weight stability before weight loss and the T3 is 137ng/dl.

The same folks after weight loss, and on 310g/d of carbs, have a T3 of 121ng/dl.

On 205g/d of carbs the T3 is about the same at 123ng/dl. But with carbs restricted to 50g/d of it drops a whopping 29ng/dl to 108ng/dl, twice the drop of the more moderate carb diet phases.

There you have it. Eat LC and thyroid deficiency, here you come.

OK, so the next question is: What does the brain think about all of this? Remember T3 is not free T3 and certainly not nuclear bound T3, so we have to look at the surrogate. What is the message from the brain to the thyroid gland concerning the adequacy (or not) of current thyroid levels? Which way does the TSH, our crude surrogate for effective neuronal nuclear bound T3, shift?

The run-in TSH is 1.15microIU/ml, this is on obese weight stability. It goes up (the Badness direction) to 1.27microIU/ml on high carb, 1.22microIU/ml on moderate carb and it drops (the Happy direction) a gnat's whisker to 1.11microIU/ml on LC.

Summary: Despite the limited fall in T3 on higher carb diets, the brain is not happy with thyroid status. TSH goes up. Gimme gimme gimme, more more more.

However, even with the greater fall in total T3 under LC eating, the brain is happy with whatever level of free T3 it is "seeing", as judged by TSH level. Should the brain be happy?

There are hints. In particular the TEE was reduced least in the LC phase of the study. There was a reduction in TEE of course. But less than for either of the other two phases imposing weight stability at reduced BMI. Despite the largest drop in total T3. It seems like a reasonable idea that both free T3 and receptor bound T3 might actually be higher under LC eating. As so many times, we will never know.

Another way of looking at the change would be to consider whether as much free T3 is needed on a LC diet. Sam Knox provided this rather nice link in the comments to The lost 300 post. It's certainly worth thinking about. Of course, I quite like the idea. But then I would!

So will low carbohydrate eating lead to thyroid deficiency? Who knows, in the long term. This was a very short study. But in this paper the brain seems quite happy with 108ng/dl of total T3 as judged by a TSH of 1.11microIU/ml.

This does not look like hypothyroidism to me.

But then I'm just this clinician see...

Peter


And here's an aside on LC eating and 24h urinary cortisol. I'll just stick the key quotes from the discussion:


"As in previous studies, discrepancy between cortisol regeneration measured during dynamic testing and the more conventional index of 24-h urinary endogenous cortisol/cortisone metabolite ratios (Table 2) reflects the confounding effects of 5 alpha- and 5 beta-reductase activities on ratios of steroids excreted in urine."

Translation: Relying on 24h urinary cortisol may mislead you. That might help with LC bashing, but you're still misled.


"Low-carbohydrate intake appears to be the key factor responsible for alterations in glucocorticoid metabolism"

Translation: LC eating is what is KEY to IMPROVING glucocorticoid metabolism.


"...extraadrenal regeneration of cortisol is responsive to the macronutrient content of the diet. In these obese men, a low-carbohydrate diet reversed the increase in metabolic clearance of cortisol (3), increase in 5 alpha- and 5 beta-reductase (4), and decrease in hepatic 11 beta-HSD1 (5, 6) previously described in obesity.

Translation: LC eating reverses the nasty effects of obesity.


"The increase in 11 beta-HSD1 activity, and hence intrahepatic cortisol concentrations, caused by a ketogenic low carbohydrate diet has implications for the efficacy of different dietary strategies in reversing the metabolic consequences of obesity."

Translation: LC eating wins hands down for correcting the metabolic consequences of obesity.



24h urinary cortisol? Pah.

The lost 300

Richard over at Free the Animal has done all of the donkey work on the latest TEE study. I'd just like to add a happenyworth.

Dr Micheal Eades in 2007:

"...what we’re talking about as a metabolic advantage is at the max about 300 kcal per day."

Ludwig's group using stable isotope doubly labeled water for Total Energy Expenditure assessment in 2012:

"During isocaloric feeding following weight loss, REE was 67 kcal/d higher with the very lowcarbohydrate diet compared with the low-fat diet. TEE differed by approximately 300 kcal/d between these 2 diets..."



I'm no great fan of metabolic advantage arguments. I like uncoupling proteins and the way that feeding electrons in to the respiratory chain at the FADH2/CoQ couple is significantly less efficient than feeding them in as NADH at complex I. Calories out can be in to heat (or in to adipocytes if you are so inclined). Your body can't harvest heat from the respiratory chain. We radiate that. There is a modest emphasis on NADH production from glucose and on FADH2 generation from beta oxidation... They feed in differently.

There have been some ugly arguments on the net over the years about metabolic advantage. Eventually the numbers give you some sort of idea as to who is correct and who is talking bollocks.

Quite why fat metabolism should be intrinsically more thermogenic than glucose metabolism is very interesting. Maybe there will be time to go in to this some day. But I live with a core body temperature at well above ambient, most of the time.

But for now, I simply find the number match between 2007 and 2012 rather gratifying.

Peter

The Flatline Days

Almost done with insulin infusions, thankfully. This post follows on from the initial post here.

It's time to discuss the discussion and then leave this paper for ever. Here's your starter for 10, and I quote:

"These three studies suggest the following: (1) insulin limits meal size when blood levels are modestly elevated for prolonged periods of time in the rat, (2) this decrease in meal size is not compensated for by an increase in meal frequency and, hence, total daily food ingestion and body weight gain are reduced, and (3) this effect appears to be a heightening of satiety rather than an induction of illness."

and at the end of the discussion:

"...it seems probable that our prolonged, modest elevations of insulin resemble the elevated basal plasma insulin induced by prolonged overfeeding and perhaps, obesity."

Let's combine suggestions (1) and (2) and drop down to our local McMuffin restaurant to watch the people eat. I've never tried this, so you have to realise I'm making all of this up, in its entirety.

In comes Jo Blob at 400kg, fasting plasma insulin at 100microIU/ml. Obviously this fasting hyperinsulinaemia blunts his appetite and he turns down the "you wanna supersize that?" offer, sits picking at his fries and soda for half an hour and eventually pushes the burger-in-a-bun away after three mouthfuls as his satiety hormone has kicked in, to even higher levels than it was when he was fasting.

Across the aisle sits Dr Guyvernment at 55kg with a fasting insulin of 5microIU/ml. Where is his satiety? Obviously he is ravenous and after the double baked potato with a baked potato on the side and three baked potatoes to follow, he's still ravenous because he is so insulin sensitive that he can't get his satiety hormone level over diddly squat.

It's the age old story. Skinny people overeat because their insulin levels are low and and fat people are chronically over sated so refuse food. Have you noticed anything along these lines? No? Somewhere along the line we do have to have a reality check!

The next statement from the discussion which caught my attention was this one:

"As pointed out earlier, some animals which received the higher dosage of insulin showed hyperphagia, as has been reported in numerous other studies [3. 8, 9, 11, 15]. It is probable that animals which became hyperphagic were more insulin sensitive and perhaps increased food intake to counteract hypoglycemia."

Okaaaay. This suggests that the animals on 6iu/24h over-ate. On average. I hate to query the obvious but does this imply the animals on 1iu/24h didn't overeat? So of course this means that the 6iu/24h animals must have gained more weight due to their hyperphagia. So did the 6iu/24h group really gain extra weight compared to the 1iu/24h group? Go look at Fig 1. Here it is again if you've forgotten:



Duuuuh. The 1iu/24h group gained more weight than the (partially hyperphagic) 6iu/24h group, even if p never got below 0.05. The people who wrote the quoted text are the same people who drew the graph... They have the daily food intakes and weight gains for each individual rat...



And finally, before I leave this execrable paper for ever, are the animals on insulin pumps just ill from their insulin infusion? Let's quote the authors again:

"We realize that a simpler explanation for our results might be that the animals become sick following the release of the insulin, however, we offer two arguments against this. First, water intakes were not decreased by insulin infusions from the Minipumps but were elevated by an average of 47.3% during the first 2 days following pump implantation and then returned to normal. Following this initial elevation, water intakes were not significantly different from controls (0 U/day animals) and the pump-implanted animals’ own baselines..."

Never mind the second argument. Let's think about polydipsia (and presumably polyuria because weight didn't increase) as markers of robust good health in a patient, any patient. I'll use a make-believe, utter fantasy, clinical setting:


Dr Insulin: Ah, hello Mrs Ratty, how are you since I implanted your insulin infusor pump two days ago, to help control your appetite?

Mrs R: I can't seem to stop drinking. I've always got to have a bottle of Evian by my side and I'm spending 47.3% more on the stuff. I wake up in the night to have a drink and I always seem to be spending a penny.

Dr I: Excellent, a good thirst is always my first maker of robust health.

Mrs R: Oh, so you won't need this urine sample I've brought?

Dr I: Oh no, no need to check your urine if you have a healthy thirst.

Mrs R: So I can throw it away?

Dr I: Of course. By the way, is that your sample in the five litre container? Could I oblige you by assisting with its disposal?

Mrs R: Thank you so much, it is quite heavy. You are so helpful. Goodbye.

Dr I: Goodbye (and after Mrs R has left): Igor, IGOR! Come, never mind the LIRKO mice, we have jam a-plenty tonight. Boil down this sample at once...



Okay. If a rat on a pump giving a constant rate infusion of insulin gets Somogyi overswing, how long does it take for the overswing to correct itself, while ever insulin levels are held constant?

I would guess two, at the most three, days. The Flatline Days. When glucose is high and appetite is consequently low. Pure speculation. It would have taken 30 seconds on a urine glucose test stick to check this. They had the sticks.

Peter

Insulin, are you hungry?

An apology. This is a dry post, I had to edit the zombies out as it was getting way too long, maybe another day. It's a bit difficult to know where to start on quite how bad this paper is. Obviously, having read the abstract, we can flick down pretty well immediately to Fig 1 in the full text.



There are a few oddities. First is the flat line in weight gain on days 1, 2 and 3. This is the suppression of hunger by insulin, maybe. There was a full seven days on insulin. This I will return to in the next post.

Next is the sudden increase in weight gain through days 4, 5, 6 and 7 in the insulin infused groups, giving a final set of weight gains on day 7 which are not statistically distinguishable from controls. Except in the group on 2iu/24h of course. The group receiving 2iu/24h is special.

Then there are the data from days 11, 12, 13 and 14. By this time the insulin infusion had stopped (which occurred around day 7ish). Look at the 2iu/24h group. Waaaay after the insulin infusion had stopped their weight gain was still much slower per day than the other three groups. Oddly this didn't reach p < 0.05, despite standard errors which were far from overlapping those of the other three groups. But trying to see what the final weights gains were is difficult because these "post pump" weight gains have been, err, umm, sort of, err. I'm not sure what the word I need is...

You see the data from these last four time points are slightly moved. Each plot has been pulled down, and by a different amount each. No one is going to say by how much. It's pretty obvious that the control line can simply be moved back up to show a linear increase in weight from the insulin infusion period as these rats never got any insulin. But all lines have been shifted down so their day 11 values are set to their day 7 values, whatever the intermediate weight gain on days 8, 9 and 10 was. It is quite likely that the 6iu/24h and the 1iu/24h rats gained weight fairly linearly and so possibly ended up on day 14 at exactly the same weight as the control group. Or heavier.

It's also very likely that the 2iu/24h group also gained weight fairly linearly but slowly, ie their "pulling down" of day 11 values to those of day 7 didn't involve much of a drop compared to the other three groups. Who knows outside the lab?

Here are the data from Fig 1 in tabular form:




Anyhoo, the 2iu/24h rats, however much they did or didn't eat/gain on days 8, 9 and 10, only gained 1.39g/d on days 11, 12, 13,and 14. Food intake per day was down significantly through this later period, 27.7g/d vs at least 30g/d in all other groups. This is very important. The implication is that if you get yourself set up with just the right insulin infusion for a week, then you still won't be hungry a week later! Wow. Insulin is a satiety hormone blah blah blah.

But if you under-dose at 1iu/24h then it's, oh-oh, back up to pre-infusion weight gain rate, or possibly slightly more. Ditto if you over-dose at 6iu/24h, just the same thing happens. Fascinating. Do you think there might be something odd about this 2iu/24h group? Perhaps someone should repeat the experiment at this infusion rate? Then we might see if the result for these rats, on which the whole concept of suppression of weight gain over 7 days rests, was a quirk. No stats were done on the zero weight gain days, ie days 1-3 on insulin. The only p< 0.05, on which the title of the paper rests, was the 2iu/24h group at day seven.

If we lose the 2iu/24h group all we can say is that an insulin infusion reduces weight gain for three days, with complete restoration of any lost weight gain by the seventh day of a continuing infusion.


So, has the experiment been repeated? Luckily it has. By this very group. And the results are in this very same paper! But well buried. You have to be a dissonant pedant to find it. It's all in Figure 4.





This not quite the same experiment as Fig 1, the timings are slightly changed, but the basic design with insulin at 2iu/24h for seven days is identical.

In the main experiment time "on pump" was 7 days and they looked at all of these days, averaging everything over this time.

In Fig 4 they did the same 2iu/24h pump for seven days but only analysed days 3, 4 and 5 as time "on pump". Go figure. They also chose days 8, 9 and 10 as their "post pump" days vs days 11-14 in the first part of the study. Again, go figure. But eyeballing the graphical weight changes in Fig 1, I doubt this matters.

The data in Fig 4 look at meal size and meal frequency because that's how you bury data. But we can reverse engineer Fig 4 to get total food intake per day. Take a ruler to the graph. Multiply meal size by meal frequency and you get food intake per day, neat huh?

The rats on 2iu/24h ate 25.5g/d during "on pump" days 3, 4 and 5. This is pretty much the same as the total 7 day value from Fig 1 and Table 1. Happy researchers? Well done for correct choice of days. But...

Does the depressed food intake continue even after insulin has finished? Do you get sustained appetite control if you get the insulin infusion "just right" for a week? Eyeballing Fig 4's "post pump" values, these are about 3.4g/meal, 9.8 meals/day giving over 33g/d food intake...........

My, those are bloody hungry rats! This is the highest food intake per day in any group in the whole paper. It's the direct opposite of the findings presented in Fig 1 "off pump" section. The sustained depression of food intake shown in both Fig 1 and Table 1 could not be repeated in the Fig 4 experiment.

It doesn't happen.

The 2iu/24h group are no different to any other infusion rate when you look at Fig 4 "post pump" section. Quite why the rats on 2iu/24h used to generate Fig 1 data showed depressed weight gain long term is a complete mystery. Personally I'd want to have had a pathologist check out the pumps in the lowest food intake rats in this group, looking for low grade peritonitis. The pumps are in the abdominal cavity. Maybe some surgeon dribbled in to the wound during implantation. I've worked with surgeons. Ultimately we'll never know.

But ANYONE quoting the data presented of Fig 1 to you WITHOUT even mentioning the results of Fig 4 to you is, well, hmmmmm..... probably in obesity research.

I was going to go on to discuss the flat line of weight gain on days 1, 2 and 3 (at all insulin infusion rates) next but I'll leave that to another post as it has nothing to do with the "insulin at 2iu/24h causes sustained decreased food intake" claim.

Which is complete bollocks.

Peter

The Zombie paper

Just a brief thank you to Julianne and Beth for the full coffin nail paper. I was going to leave zombie rats alone after the last post and didn't think the paper itself was needed to see what was going on. But the quick scan I've had of the coffin nail shows it to be very interesting. Want to read some execrable science? I'll deconstruct it soon but there is an on call night tonight and a family weekend coming up, so I might not be as quick as it deserves. But don't worry, there are still plenty of zombies around...

Perhaps best not comment on this post, comments can stay on the last one all grouped together. I'll take this one down as the next one comes through.

Oh, Hi Melchior. Nice to see you about. It's good when facts plus logical consistency ultimately win through. As they must.

Peter