Monday, 25 August 2014

Metformin mechanism of action

In the last post I linked to a study proposing that insulin mainly causes triglyceride accumulation by inhibiting fat oxidation. From 2006 there is another study detailing how metformin counters the ability of insulin to inhibit fat oxidation.

How does insulin stop fat burning? Theres a few ways this happens, and one of the ways is by increasing  malonyl-CoA. Insulin activates the enzyme acetyl-CoA carboxylase ( ACC for short ) that converts acetyl-CoA to malonyl-CoA,  (link)

malonyl-CoA in turn inhibits carnitine palmitoyltransferase (CPT) I, the enzyme that controls the transfer of long-chain fatty acids  into the mitochondria where they are oxidized (link). So essentially the enzyme ACC is the cellular switch between fat burning and fat making/storing. ( there are probably other cellular switch's aswell )

Controlling the enzyme ACC is therefore probably central to cellular energy dynamics.

If we turn off the ACC enzyme, we can stop making fat and start burning it how can we turn it off? Thats where metformin and AMPK enter the picture. AMPK directly shuts down ACC, by phosphorylation at several serine sites. AMPK responds to increased cellular levels of AMP, and we can activate AMPK by doing things like fasting, exercise, low-carbing, vinegar, acetate fermentation in gut from fibre,....... and probably most importantly, KEEPING INSULIN DOWN.

While none of this stuff is particularly new to seasoned researchers, what is "new-ish" is a 2013 study  that looked at changing the serine phosphorylation sites on the ACC enzyme to alanine. This now makes ACC highly resistant to inhibition by AMPK, allowing relentless lipogenesis and lowered fat oxidation. The result is that the rodents with these alanine knockin enzymes become very insulin resistant with fatty liver, and metformin DOESNT work on them.

What this shows is that the primary way metformin improves insulin sensitivity is by curbing the activity of the ACC enzyme.

There are a few corollaries here aswell....

- the rodents in the alanine knockin study didnt become obese even though they had reduced fat oxidation. This would *seem* to contradict the idea that insulin causes fat gain by lowering fat oxidation. However keep in mind that we are dealing with 2 studies that did completely things. In study 1 excess insulin was pumped into the rodents, and in study 2 the ACC enzyme is made mutant. It should go without saying these are certainly not equivalent. There are other ways insulin inhibits fat oxidation not to mention insulin does significantly more when it binds to adipose tissue other than just lowering fat oxidation/activating ACC.

- Increased basal malonyl-CoA levels have been found in muscle from obese and T2D subjects (link). They also found that the activity of the ACC enzyme was much higher in the basal state in these subjects. This brings me back to the idea of obesity/T2D being caused by excess insulin secretion, in particular fasting hyperinsulinemia. Alot of people seem to be under the impression that insulin secretion is primarily reactive to insulin resistance, and that insulin resistance appears FIRST and increased insulin secretion is compensatory. 

If metformin improves insulin sensitivity by curbing the ACC enzyme, and increased levels of malonyl-CoA and ACC activity are found in muscles of obese/T2D subjects, doesnt this suggest the insulin resistance is caused by the increased malonyl-CoA and ACC levels?

And what causes increased malonyl-CoA and ACC? ........ insulin!

Its *unproven*, but makes sense..... increased fasting insulin secretion -> increased fasting malonyl-CoA and ACC, -> resistant to insulin.

- sometime ago I made a post on carbohydrate-sensitive obesity. In that model we saw that sensitivity to carbohydrate induced obesity was predicted by having overly suppressed fat oxidation in the postprandial state. With respect to metformin stopping insulin from lowering fat oxidation, I would like to speculate that metformin may help prevent carbohydrates from making you fat by keeping fat oxidation elevated in the postprandial state. I would also speculate there are genetic differences in the ability of insulin to inhibit fat oxidation because the sensitivity of CPT1 and ACC can be influenced by genetics.

- Lastly, muscle IR is not the only defect in obesity/T2D, there is also hepatic IR and excessive glucose output that we think is primarily driven by hyperactivity of the FOXO1 transcription factor manufacturing the PEPCK enzyme. Read this comment here to learn how this ties in with all of the above.


  1. re PEPCK...
    Diabetes. 2008 Feb;57(2):306-14. Epub 2007 Oct 1.
    Metformin inhibits hepatic gluconeogenesis through AMP-activated protein kinase-dependent regulation of the orphan nuclear receptor SHP.