Wednesday, July 27, 2016

Why Liver is Likely More Important than Muscle for Pre-Diabetes: Part 2

Muscle is argued to be the primary defect in insulin resistance.  The main evidence for this comes from the presence of impaired insulin signalling and glucose metabolism in people with insulin resistance and type 2 diabetes, and their first degree relatives [1].  Whether you think the contribution of muscle to postprandial glucose uptake is closer to 75% or 30% (see earlier post), it’s a decent contribution.  However, this line of research is limited as similar studies haven’t been done in the liver, since liver biopsies are far more invasive than muscle or fat biopsies (and unfortunately my PhD won’t resolve this)

One of the difficulties with mechanistic research in disease models is that you take a snapshot in time and find several pathologies are present.  The trick then is to figure out what are the possible causes of disease as opposed to the consequences of disease or just unrelated effects.  A method of doing this is to investigate what’s happening at the earliest time points in the progression of disease, rather than drawing conclusions from biomarkers in poorly controlled diabetes

My honours project followed on from a time course study in mice that found the following [1]:

·         Glucose intolerance was present when first measured after 3 days on a high fat diet (HFD)
o   Whole body insulin resistance was present when first measured at 1 week with the euglycemic-hyperinsulinemic clamp (higher glucose infusion rate)
§  This was due to liver insulin resistance  (this wasn’t associated with a defect in Akt activation or inflammation, but with a relatively mild increase in lipid accumulation)
§  Adipose tissue was also insulin resistant, but muscle wasn’t yet
·         Muscle insulin resistance was present when next measured at 3 weeks
o   This was associated with no further deterioration in glucose tolerance, but worsened whole body insulin resistance and increased fasting insulin
·         Glucose tolerance or tissue insulin resistance didn’t worsen even several weeks after the initial defect

This is consistent with other studies in rodents and humans:

·         Liver, but not muscle insulin resistance was present after mice [2] and rats [3] were put on a HFD at 3 days when first measured with the clamp (and associated with liver ER stress [2]).  Muscle insulin resistance was present when next measured at 3 weeks (which was associated with muscle lipid accumulation) [3]
·         Short-term overfeeding (3-7 days) in healthy humans causes hepatic insulin resistance [4] [5] [6] [7] [8], whereas muscle insulin resistance was not present at this point [4] [6]
·         Short term energy restriction almost normalises fasting glucose after 2-7 days, before significant weight loss, and this is associated with improvements in liver insulin sensitivity, but not muscle insulin sensitivity, which improves several weeks later [9] [10]

Since pre-diabetes and insulin resistance develop very quickly in mice on a HFD, one of the hypotheses of my honours project was that the pre-diabetes and insulin resistance would be normalised in a similar timeframe.  This was one of the novel aspects of my honours project, as earlier diet reversal studies in rodents haven’t looked at early points, but instead after 3, 4 and 16 weeks respectively [11] [12] [13].  We found pre-diabetes and insulin resistance was normalised in 7 days after switching mice from a high fat diet to the standard laboratory low fat chow diet.  We didn’t directly measure changes in insulin sensitivity in the liver and muscle, but some evidence suggests it was likely due to changes in liver glucose metabolism rather than muscle:

·         The HFD group had a higher change from baseline in endogenous glucose at 15 minutes, suggesting elevated glucose production between 0-15 minutes.  This was completely normalised in the HFD→CHOW group (figure 3B)
·         The HFD group had higher exogenous glucose at 30 and 60 minutes but not at 120 minutes, indicating a defect in early glucose disposal (15-60 minutes).  As the liver gets first access to glucose, this could suggest a defect primarily in liver glucose uptake that was completely normalised in the HFD→CHOW group (figure 3C and 3E)
·         The HFD group showed evidence of elevated futile glucose cycling in the liver, indicating impaired glucose metabolism, and this was completely normalised in the HFD→CHOW group (figure 3G)
·         While tissue lipid accumulation isn’t always associated with insulin resistance, the HFD→CHOW group normalised 60% of their liver triglycerides (figure 5A) and 32% of their muscle triglycerides (data not shown)

The takeaway from this is not so much that muscle isn’t important, just that it seems that changes in muscle insulin resistance generally occur more slowly and can be sufficiently compensated for by the beta cells or other organs**.  Metabolic changes happen quickly.  Whatever the mechanisms behind the rapid changes in glucose control are they have to be capable of changing quickly.  Lipid accumulation might be too slow and obesity is definitely too slow.  And these mechanisms are more likely to originate in the liver rather than muscle.

If this stuff interests you I strongly recommending reading this review

* Muscle insulin receptor knockout (MIRKO) mice have normal glucose levels and insulin levels, but have elevated plasma triglycerides and free fatty acids [14].  This is in contrast to liver insulin receptor knockout (LIRKO) mice that prior to liver failure have severe insulin resistance and hyperglycemia, but reduced or normal circulating fatty acids and triglycerides [15] [16]

** Also, in figure 1B bellow, note the absence of impaired glucose tolerance and insulin resistance people with duchenne muscular dystrophy, and the profound insulin resistance but normal glucose tolerance of those who were wheelchair bound (WC) 

“With such a severe reduction in muscle mass, glucose intolerance rather than insulin resistance would be the expected consequence. Therefore, the loss of muscle mass in this group is probably unrelated to their insulin resistance. Instead, the inactivity itself, which accompanies the loss in muscle tissue, is probably a major factor in the development of the insulin resistance.” [17]

Sunday, July 24, 2016

Why Liver is Likely More Important than Muscle for Pre-Diabetes: Part 1

My honours project recently got published.  I recommend reading the paper [1], but if that is too technical you can read my summary of it here.  A reasonable question you might have is why we focussed on the liver and not muscle, because after all, isn’t muscle supposed to be the main organ involved in glucose control?

The contribution of muscle to whole body glucose uptake is often over-exaggerated.  People bring up the fact that muscle is responsible for ~70-80% of insulin stimulated glucose uptake and use this to claim that muscle is responsible for ~70-80% of glucose uptake in the postprandial state (following a meal or an oral glucose tolerance test (OGTT))

The ~70-80% figure is correct, but the problem is that is that it comes from studies using a technique called the ‘euglycemic-hyperinsulinemic clamp’.  The euglycemic-hyperinsulinemic clamp is the gold standard to measure insulin resistance** and involves infusing glucose and insulin into the blood to maintain euglycemia (normal glucose levels ¬5 mmol/l) and hyperinsulinemia (high insulin levels), often higher than what occur following an oral glucose tolerance test (see below [2] [3])


This method clearly doesn’t accurately represent what happens following a meal or OGTT, where muscle is responsible for ~30% of glucose uptake* [4].  The conditions of the clamp favour muscle glucose uptake over other key organs such as the liver for a few of reasons:

·         The main glucose transporter in muscle (GLUT4) has a higher affinity for glucose than the main glucose transport in the liver (GLUT2) [5] and the muscle version of hexokinase (the enzyme catalysing the first step of glycolysis) has a higher affinity for glucose than liver glucokinase [6].  These differences mean muscle takes up more glucose during euglycemic conditions than the liver [6]
·         When you eat a meal or do an OGTT the liver has the benefit of glucose and insulin levels being ¬2-3x higher in the portal vein compared to systemic circulation.  This enhances liver glucose uptake and proportionally decreases glucose uptake by other tissues [4, 7, 8], but doesn’t occur when glucose and insulin are infused directly into the bloodstream

So it’s somewhat correct to say that muscle is responsible for ~70-80% of insulin stimulated glucose uptake, but this shouldn’t be extrapolated to what happens following a meal or OGTT where both insulin and glucose transiently increase 

* In the fasted state the brain takes up most of the glucose and muscle is responsible for ~18% of glucose uptake (see below) [9]