Tuesday, February 21, 2023

How can insulin resistance cause weight gain? (4) amplification

Okay. Here are the basics of insulin facilitated glucose oxidation:

Insulin -> Insulin Receptor (IR) -> G coupled protein -> NADPH oxidase 4 (NOX4) -> ROS

These ROS are the spark which triggers insulin signalling.

They act to inhibit assorted protein tyrosine phosphatases (PTPs). With PTPs disabled the normally suppressed auto-phosphorylation of the IR activates and its self activation allows calories to flood in to the cell. So NOX4 starts the process. It's a eukaryote system, there is nothing fundamental about it, just as insulin signalling is an eukaryote system. But it uses ROS because they are fundamental.

Once glucose calories are entering the cell they are supplied to the mitochondria which generate ROS by the rise in F:N ratio, triggered by the glycerophosphate shuttle (my opinion) converting glucose derived cytoplasmic NADH into mitochondrially targeted FADH2. My expectation is that this degree of ROS generation will keep the PTPs suppressed and so the insulin receptor phosphorylated/active while ever the cell is using glucose and so still wants glucose calories to enter.

These ROS are physiological and represent the normal control and maintenance of insulin signalling at peak levels. The next step is the use of ROS to disable insulin signalling.

There are times when it is necessary to actively suppress insulin signalling. This is most easily visualised by considering a cell which has received more than enough insulin mediated calories to meet its needs. Under these circumstances there is a surfeit of ATP which activates a negative feedback acting on ATP synthase. The inhibition of ATP synthase means that delta psi is no longer being dissipated so it will increase. At values above 140mV the rate of ROS generation increases exponentially and reaches levels that will directly act on the proteins of the insulin receptor and signalling cascade to de-activate them.

This is beautifully illustrated in this paper from 1974, previously discussed here

It's tracking the evolution of CO2 from radiolabelled glucose and using H2O2 as a direct replacement for insulin to facilitate glucose uptake/oxidation by adipocytes.

It is very clear that 0.01mM of H2O2 initiates glucose uptake/oxidation, ie with absolutely no insulin present then 0.01 mM of H2O2 has the signalling action of a low dose of insulin. I think of this as equivalent to the NOX4 action. There is a level of H2O2 at around 0.3mM which performs the function of peak insulin exposure. When H2O2 exposure is further increased to 4mM or 5mM then the H2O2 disables its own ability to replace the function of insulin.

Let's make this absolutely clear, insulin is in no way fundamental. It can be completely replaced by H2O2 which, in common with superoxide, is the core signalling system and is ubiquitous as a growth/reproductive signal going right back to bacteria, preserved in their behaviour today. H2O2 is not an insulin mimetic, it is the core signal. Suppressing this ROS signal in modern bacteria, which never use insulin, inhibits their growth and reproduction.

Aside: No one has yet worked out how bacteria generate superoxide or H2O2 today. It doesn't seem to be via NOX enzymes which are, sadly, purely eukaryotic. I say sadly because they are extremely simple enzymes with a six helix tube through the cell membrane containing two Fe moieties as a "wire" to carry electrons to extracellular O2 from a bolted-on NADPH oxidase within the cell. If we ignore the multiple control systems which can also be added then it's very simple. The core of NOX just looks primordial. But it isn't. Sigh. End aside.

I hope I have established that there is a primordial signalling system which uses low concentrations of ROS to initiate nutrient uptake and usage to a certain maximal effect, above which a negative feedback using the same ROS disables further nutrient uptake/utilisation.

I've left fatty acid oxidation out of this narrative to keep it simple.

All of which leads us back to 4-HNE.

We can accept that the NOX4 minor contribution and the mitochondrial major contribution to ROS are generated in absolute proximity to the lipid membranes. All lipid membranes have functional needs for PUFA for structural purposes and if these are directly abutted to the electron transport chain complexes which are producing the above ROS signals then we can expect a proportion of the ROS to interact with those PUFA. The best studied end product is 4-HNE.

Superoxide is a poorly mobile signalling molecule ideal for short distance signalling but with limited  ability to cross lipid membranes. H2O2 is more stable and better able to cross lipid membranes so makes a good intermediate distance signal within a cell. 4-HNE carries the same information about the state of the ROS generation from the electron transport chain (pax NOX4) but is stable enough to be transported through the blood stream in measurable and modifiable concentrations. Because it carries the same information as superoxide/H2O2 it should come as no surprise it elicits the same response. In fact I view it as an amplifier of the ROS signal. It seems that 0.1μM 4-HNE can give a similar effect on adipocyte lipid accumulation as 0.3mM (ie 300μM) H2O2. Roughly. And anything over 5μM 4-HNE disables insulin signalling, equivalent to 5mM H2O2.

Obviously Protons makes this simple story a little more complex as you add in fatty acids and the influence of linoleic acid which both reduces and increases ROS generation as well as being the core substrate for generating 4-HNE per se, but I think this will do for today.

The above description makes the whole system look static. It's not. All of it oscillates, delta psi and ROS generation. You can see why.


Throw away thoughts:

If ROS are a growth signal, so too should be 4-HNE. Anyone can Pubmed "cancer + 4-HNE".

Also H2O2 is a biological warfare molecule. Macrophages "throw" H2O2 at bacteria to kill them. Any PUFA in the area of the battleground will generate 4-HNE. This is a medium distance signal to recruit more macrophages to join the fight. These are not directly related to metabolism but clearly important. Actually, you could ask whether high levels of both H2O2 and 4-HNE kill bacteria metabolically. There's a thought.

Then you can get in to plasma 4-HNE and it's effects on lipoprotein PUFA components. 4-HNE is a direct generator of ROS in its own right (it's function is as an amplifier of the ROS signal after all) so filling lipoproteins with linoleic acid while elevating plasma 4-HNE will attract macrophages to the extra 4-HNE wherever those lipoproteins stick. Like arterial/arteriolar walls. Stroke? Heart attack?

Hmmm. All fun stuff. Blame your cardiologist.

Monday, February 20, 2023

How can insulin resistance cause weight gain? (3) AD-9308

This is a random, throw away post. A one-liner, don't laugh. I happened on this paper following a link in a link to a link during my 4-HNE reading:

 Disruption of the mGsta4 Gene Increases Life Span of C57BL Mice

This sort of comment makes me sit up and take notice

"Surprisingly, the opposite was true."

as in

"We expected that disruption of mGsta4, a murine gene encoding a major antielectrophilic enzyme, will parallel the effect of a similar intervention in C. elegans and curtail the life span of the knockout mice. Surprisingly, the opposite was true. In the present article, we report this observation and provide a possible explanation for this unexpected effect."

The enzyme from the gene mGsta4 is a membrane associated glutathione-S transferase which is important (like ALDH2 in previous posts) for the detoxification of ROS-derived substances such as 4-HNE (amongst others). The best explanation for the findings in the study is that 4-HNE accumulates in tissues of mGsta4 null mice which activates extra defence mechanisms against electrophilic molecules. The end result is near-normalisation of tissue 4-HNE combined with an extended lifespan. At least in Bl/6 mice. But not in C elegans, here it does the opposite. Choose your model wisely before tinkering with your metabolism.

Apart from the somewhat amusing concept that reducing the detoxification of 4-HNE might be a longevity ploy, it made me wonder about the ALDH2 mutations in the East Asian human populations. Full introduction of this mutated gene in mice exposed to an high linoleic acid plus high sucrose diet leads to supplementary obesity. This is unlikely to be quite so problematic on a more species appropriate diet for either mice or humans. The ALDH2 modified mice were normal weight on chow.

For a mutation to persist there has to be an advantage to the individual carrying that mutation which increases their probability of reproductive success. Null mGsta4 mice show no change to their maximum life span but have a 10% increase in median life span and have a marked increase of 36% in lifespan at the 10th percentile, ie during the peak reproductive period of life.

Does the same happen with the reduced efficacy gene for ALDH2 in people as happens in the null mGsta4 gene in Bl/6 mice? Would this show as an increased healthspan but without increased peak longevity? Provided you don't live on fudge made from sucrose and corn oil of course. I tend to think median lifespan extension probably correlates with increased healthspan but that peak longevity might be more of a luck-based failure to die. Hmmmm...

So the law of unintended consequences makes me think long and hard about ALDH2 activation using a drug such as AD-9308. You could use it to reduce 4-HNE from an obesogenic 1.0~ishμM to almost nothing and so reverse whatever benefit has been preserved by the ALDH2 down regulatory mutation in humans. After all 560 million East Asians carry the gene which appears to have been designed to make us fat and insulin resistant when fed a high linoleic acid, high sucrose diet. But not if eating real food. So lets develop a drug for all East Asians to be able to eat junk food without getting fat! Might this simply result in looking good in your coffin?

Evolution is not stupid. It leaves that to drug developers.


How can insulin resistance cause weight gain? (2) 0.1 μM

When you type adipocyte + 4-HNE to Pubmed, this is your first hit

4-hydroxynonenal causes impairment of human subcutaneous adipogenesis and induction of adipocyte insulin resistance

and it is, of course, just a reinforcement of my biases which suggest insulin resistance should limit lipid accumulation in adipocytes

"Exposure to multiple doses of 4-HNE led to reduction in growth and differentiation of preadipocytes in both groups (Figure 4A), marked by upregulation of anti-adipogenic FABP4 and down regulation of the adipogenic FASN and SREBF1 genes (Figure 4B). "

Here is the rather pretty part A of Figure 4. Look carefully and you will see two stains have been used. The hard-to-see dark blue stain is DAPI and stains nuclei, ie it gives you a cell count so you can compare equal numbers of cells in a given view (or count cells in a proliferation assay). The fluorescent green staining shows lipid and can be used to assess how much lipid a pre-adipocyte has accumulated during the differentiation process in to an adipocyte. All by AI assisted automated counting systems.

The top row are differentiated adipocytes from insulin sensitive obese ladies. On the left of the top row we have pre-adipocytes after "normal" differentiation, with lots of lipid. On the right of the top row we can see that differentiation under 10μM 4-HNE pretty well completely stops lipid accumulation.

The bottom row features adipocytes from ladies who are already insulin resistant. Their pre-adipocytes accumulate very little lipid even after "normal" differentiation and this is made a bit worse by exposure to 4-HNE. Basically these cells are already f*cked and adding 10μM 4-HNE only makes things a little bit worse.

Okay. Biases confirmed. But why did they use 10μM 4-HNE? They did a dose response experiment looking at pre-adipocyte proliferation during differentiation using that pretty blue DAPI nucleus stain. Their benchmark was cell growth without 4-HNE exposure and they then used increasing 4-HNE exposure to see what the dose response was like:

I'd guess they used 10μM because it had some effect but was not as catastrophic as 40μM. Seems fair. It looks like 2.5μM was harmless and 5μM was borderline in terms suppression of proliferation. So 10μM got used for most of the rest of the paper because they were primarily interested in insulin resistance.

None of which helps me explain this

or the other gem from Tucker in here

which is a toxicological study injecting rats with exogenous 4-HNE:

This gives a clue as to what is going on. At 10mg/kg of injected 4-HNE the rats gain weight, at higher doses they don't.

There are a wealth of studies showing 4-HNE causes insulin resistance and a fair few which show that this is demonstrable by decreased phosphorylation of AKT, usually using a (quite high) dose around 100μM. To see what happens at lower doses we need some more digging and it's not too hard to find this review:

From section 5.2 we have this snippet:

"Depending on the concentration, 4-HNE can be beneficial or detrimental to cells. Many studies in the literature suggest that a 4-HNE concentration below 2 μM leads to cell survival and proliferation. However, a concentration above 10 μM is detrimental to the cell, leading to genotoxicity and cell death"

and further down the same section is

"A study carried out investigating vascular endothelial growth factor (VEGF) in retinal pigment epithelial (RPE) cells showed that a low concentration of 4-HNE (0.1 μM) leads to increased secretion of VEGF, while its expression is blocked when the 4-HNE concentration is greater than 5 μM"

does confirm that 4-HNE is a proliferation agonist at low concentrations (0.1 μM) and becomes borderline at 5μM, so all of the work on growing adipocytes at 10μM in the first study I mentioned are way above the levels which might be expected to promote adipogenesis and lipid accumulation in adipocytes.


I would suggest that reducing the function of ALDH2 in mice as in

does not achieve sustained 4-HNE at or above 5μM, more likely the figure will be below 1μM, simply based on looking at the phenotype. I'd bet AKT phosphorylation would be enhanced.

Equally in

I would expect 10mg/kg exogenous 4-HNE to provide less than 5μM 4-HNE to adipocytes and anything over 30mg/kg to provide in excess of 5μM to adipocytes.

That would make sense.

In the next post I'll consider why this might be the case.


Sunday, February 19, 2023

How can insulin resistance cause weight gain? (1) ALDH2

Mea culpa. Sorry all for neglected comments else where on the blog and IRL. Blame Tucker for this link in here... Here's the next post.

There are three studies which have profoundly shaped my thinking about weight gain. First is this one

Insulin resistance is protective against future weight gain (but not metabolic syndrome!)

Second is this one

in which formerly obese women (ie at high risk of weight regain) are much more insulin sensitive than closely matched control women:

And finally this one:

Insulin sensitivity is increased and fat oxidation after a high-fat meal is reduced in normal-weight healthy men with strong familial predisposition to overweight

in which normal weight, young people with a marked predisposition to obesity are more insulin sensitive than a matched control group:

So, when Tucker presents us with this paper looking at mice engineered to mimic people with reduced ability to detoxify the known generator of insulin resistance, 4-HNE (amongst other molecules)

then I have to think about this graph:

which is very exciting from the Protons perspective. 4-HNE augmentation causes insulin resistance *and* causes obesity. Yet obesity is the result of excessive insulin *sensitivity*, and is clearly visible (in this study) between weeks four and 16, as weights diverge:

Obviously by week 28 both groups of mice are going to be solidly insulin resistant so will gain minimal extra weight. It's weeks 4-16 we're interested in. You don't get fat without storing fat. We use insulin signalling to store fat. Resisting insulin signalling resists fat storage. A paradox, sent to provide insight.  That's worth a post in its own right.

Fascinating stuff. Lots to dig in to.


Saturday, February 18, 2023

Delta psi and ROS

I'll post this now as I've gone off down a related but more concise rabbit hole for the next posts.

Okay. Here are a couple more building blocks for the future, snippets from the slightly dubious 

Here is the pattern of ROS generation from isolated mitochondria fed with increasing concentrations of palmitoyl carnitine:

Clearly there is a rise in ROS with increasing PCarn exposure, peaking at 18microM and decreasing to 72microM. There were too few functional preparations to include the 36microM value in statistical analysis. There is also evidence of uncoupling at 72microM. So 18microM gives maximum ROS generation. All measurements were taken under oligomycin to mimic state 4 respiration, ie fed with PCarn but conversion of ADP to ATP was not allowed.

And here is the delta psi developed by PCarn (plus extra carnitine) at increasing concentrations of PCarn. Note that there is a consistent delta psi estimated using between 3microM and 18microM PCarn with no suggestion of an upward trend:

Other places in the paper suggest that this delta psi is around 140mV (see below) and that exposure to PCarn at lower or higher concentrations will produce less ROS than the 18microM exposure does. To get more information about ROS generation they used oligomycin to establish a high delta psi then lowered it a little with 15pM FCCP (right hand end of the graph below) or a lot with 100pM FCCP (left hand end of the graph. As 18microM PCarn can only support 140mV I'm guessing that both voltages are below this value.

Here PCarn, the triangles, shows significant ROS generation at low delta psi with a modest rise to high delta psi, the latter being a little under 140mV. Glutamate shows a delta psi responsive rate of ROS generation.

In this final graph, ignoring the O2 consumption, we can look at the end derived delta psi values from three substrates:

The (expected) delta psi for 18microM PCarn is at 140mV (this is the max which PCarn at any concentration will support in this model), for saturating glutamate with blocked complex II it's 170mV and for succinate (driving through complex II) it's somewhere around 190mV.

What I am driving at is that palmitate, whilst generating ROS, does this without the requirement for, and not as a result of, high delta psi. For glutamate the higher the delta psi, the higher the ROS.

There is no suggestion that PCarn at 18microM is uncoupling yet delta psi is relatively low. What interests me is whether this is being achieved through the inhibition of complex I. In the last post I talked about the role of glutathiolation of proteins as a consequence of localised ROS generation and the role which that glutathiolation might have in the avoidance of situations where unbridled supply of metabolic substrate would lead to very high delta psi and an unacceptable rise in ROS generation. Recall Nick Lane's concept that a cell will sacrifice ATP production from ox-phos rather than generate excess ROS.

How much glutathione was available and what state complex I's glutathiolation might be were not in the remit of this study. So a lot of unknowns.

These are isolated mitochondria. There are, in real life, many controls to the supply of calories to mitochondria, as near by as the carnitine palmitoyltransferases, further away at the cell surface (GLUT4 and CD36) and as far away as adipocytes and hepatocytes. 

Again I have further ideas about ROS and it has to be made very clear that I am making huge extrapolations from isolated mitochondria towards what might happen in an intact cell with insulin exposure, insulin receptors, phosphatases, a controllable supply of Ca2+, superoxide dismutase, catalase etc etc, all in the correct locations to do what is impossible for an isolated mitochondrion to do.

So caution. But ultimately ROS from fatty acid oxidation generates ROS without the need for a high delta psi. Primarily NADH supplying substrates produce ROS in association with (and probably caused by) high delta psi.


Sunday, February 05, 2023

Warning: Politics

Over the years I have only very rarely touched upon politics, outside of the corruption intrinsic to certain types of scientific publication. This post is more political. You know where the back-button is if you need it!

This is the settlement section of my own personal sewage treatment works, better known as a septic tank, about which I know far more than I would prefer. Certain components of household waste float to the top, much as some politicians do in political parties. I can definitely see Boris Johnson up near 12 o'clock, Richboy Sunak close by on his right, Nadhim Zahawi over towards 3 o'clock and I recently had to have the sh*t pumped out to get rid of the smell from Matt Hancock (so he's gone, thank goodness). Liz Truss has sunk.

I'm largely apolitical. If I had to characterise myself it would be mildly left of centre. I had never heard of Prof Sunetra Gupta before I heard her on the RCPath webinar talking about covid and epidemiology in early 2020. That was well before the days of the Great Barrington Declaration.

She is very comfortable to describe herself as a staunch socialist. This is her talking about Jeremy Corbyn. For those outside the UK he was the very left wing Labour Party leader rendered unelectable by the UK main stream media. This is what Prof Gupta had to say:

Ultimately her point is that Corbyn would not have made a morbidly obese type 2 diabetic bus driver work through the whole of the pandemic while some zero risk middle management bloke was paid to stay at home hiding behind his sofa and working on Zoom to carry on "running" the country. One of my neighbours, a renal transplant recipient with insulin resistance from immunosuppressive meds, had to get a job as a Tesco delivery driver because he lost his pre-covid job as a direct result of the first lockdown. Go figure.

I have to state that I have almost never, ever even considered voting for the Labour Party in any election. I don't do the whole "Looney Left" thing. But I have done so just once.

That was in 2019 and it was only because Corbyn was the party leader.

Not because I approve of all of his policies and life decisions (FFS, he's a vegetarian!) but, primarily, he is one of the few UK politicians who did not massively fiddle his parliamentary expenses claims in the 2008/9 scandal. He's not a crook. There might be a handful of other politicians around the world in this category but they're like hen's teeth. Recall my septic tank and what floats to the top.

So I have a great deal of respect for Jeremy Corbyn. Especially after the clear cut demonstration during the last three years of how the main stream media works with misinform to promote an agenda. Before covid I massively underestimated the influence of the MSM. How could anyone believe this crap? My bad.

So it was nice to see Liberty Places put out this tweet

Fact: would never go to Davos, never send arms to Ukraine, never ignore vaccine harms, never bow to the IMF, UN, NATO or EU technocrats. Many need to admit they got him wrong.

I don't think Liberty Places (not all of whose posts do I agree with) are exactly Socialists by any stretch of the imagination!

Corbyn was replaced by Starmer, who was interviewed by Emily Maitlis. There was about 30-40 minutes of drivel before the final quick fire questions, I listened to it all. This is the most telling few seconds, the rest is in the same vein:

The full interview video appears to have been removed from Youtube and I clipped the above few seconds from this guy's channel. Had to use Brave running under Linux to find it.

There is a reasonably accurate summary/partial transcript at The Conservative Woman. If you ever find the whole video then Starmer's hatred of Corbyn and his preference for the company of Piers Morgan (inset vomit emoji here) over Corbyn was emphatic and understandable, if morally reprehensible. Like the WEF/parliament clip, the Corbyn/Morgan clip is right at the end of the interview in the quickfire questions section.

This person, Starmer, is likely to be our next PM. As Lord Sumption discussed with Jay Bhattacharya, all those years ago in 2021, we might be witnessing the collapse of Western Democracy. That seems to be progressing pretty well.

The current divide is not between Left and Right, these are irrelevant. It's more individuals vs globalists.

Most people don't realise.


Edit: can't resist re publicising this from Bob Moran:

and this piece of ancient history too. Have to be fair to both sides: