Exercise: How Does It Promote Insulin Sensitivity?

Searching for the "Holy Grail"

The metabolic benefits of exercise are well known. Considerable data exist showing that a moderate daily exercise regimen can greatly improve insulin sensitivity. However, one problem with this apparently simple approach to treating insulin resistance is that most people find it difficult to maintain exercise regimens for long periods of time. A "holy grail" of physicians treating insulin resistance is to develop alternative treatment options that could target the key regulatory steps that couple exercise with its metabolic benefits. This would require a detailed understanding of the signaling pathways that regulate energy production, storage, and expenditure in skeletal muscle.

In recent years, adenosine monophosphate-activated kinase (AMPK) has emerged as a key regulator of energy-modulating pathways in skeletal muscle. The activity of AMPK is acutely increased by muscle contraction as well as other factors that increase the ratio of AMP to ATP in the cell. As such, it has been termed an "energy sensor." Recently, it has been shown that important energy-regulating hormones, such as leptin,^[1] as well as commonly used therapeutic agents, such as metformin^[2] and rosiglitazone, can modulate AMPK activity.

Activating AMPK has myriad consequences that would appear to shift the cell's metabolic activities away from substrate storage toward oxidative substrate disposal to create ATP; for example, the suppression of cholesterol, fatty acid, and triglyceride synthesis, and the activation of fatty acyl-CoA beta oxidation, glucose uptake, and glycogenolysis. AMPK is therefore an excellent target for therapies to overcome the clinical abnormalities that characterize the "metabolic syndrome."

Several experts in the field of AMPK biochemistry and exercise metabolism discussed the mechanisms and regulation of AMPK signaling pathways and the consequences of these effects on muscle energy metabolism and insulin sensitivity at the symposium entitled "The Role of AMP Kinase In Exercise on Substrate Metabolism," presented at the 62^nd Scientific Sessions of the American Diabetes Association.

Biochemical Aspects of AMPK Structure and Function

Bruce Kemp, PhD,^[3] University of Melbourne, Australia, reviewed key biochemical aspects of AMPK structure and function. AMPK has several isoforms derived from combinations of forms of its 3 subunits: alpha, beta, and gamma. The 2 chief isoforms in muscle are AMPK alpha1 and AMPK alpha2; the latter isoform is generally more likely to be regulated by exercise. AMPK's activity is increased by elevation in cellular AMP concentration, acute exercise, exercise training, and the activity of an upstream kinase termed AMPK kinase (AMPKK). The regulation of AMPKK has not been fully delineated; AMP appears to activate some but not all isoforms of AMPKK. Metformin appears to increase the activity of AMPK directly, while rosiglitazone does so by increasing the cellular concentration of its allosteric activator, AMP.

Human Exercise Studies of AMPK Activation

Erik Richter, MD,^[4] University of Copenhagen, Denmark, presented the results of detailed human exercise studies examining the relation of AMPK activation to downstream signaling events and physiologic consequences. These difficult studies measured the effects of exercise duration, exercise intensity, and muscle glycogen content using substrate balance measurements across the lower limbs, stable isotope infusions, and enzyme activities in skeletal muscle biopsies.

As background to these studies, it should be understood that a particularly well-studied pathway regulated by AMPK in vitro is that resulting from the phosphorylation of acetyl-CoA carboxylase (ACC) by AMPK. This action inhibits the enzymatic activity of ACC, causing the level of malonyl-CoA to fall; carnitine palmitoyl transferase I is then disinhibited, resulting in the movement of acylated fatty acids into mitochondria to undergo beta oxidation. Thus, one would expect this cascade to be activated during exercise in vivo.

Dr. Richter's group confirmed that short-duration exercise of varying intensities increased AMPK activity. Prolonged (3.5 hours) exercise of low intensity revealed a progressive increase in AMPK activity, but not that of ACC, which peaked after 1 hour and then declined. These findings suggested a rather surprising dissociation between AMPK activity and lipid oxidation, which is generally thought to be important to provide energy to muscle during prolonged exercise.

The investigators then examined the relative contributions of 3 AMPK-regulated functions that might operate to supply energy during exercise: glucose uptake, fatty acid oxidation, and inhibition of glycogen synthase activity. As to the first function, there did not appear to be a close temporal correlation in the exercising muscle between AMPK activity or ACC phosphorylation and glucose uptake. The situation with regard to exercise, AMPK activity, and lipid oxidation also appeared complex. With short, intense exercise, carbohydrate appeared to be the chief fuel for energy, yet malonyl-CoA content in the muscle biopsies declined rather than increased, as might be expected if carbohydrate oxidation were preferred to lipid oxidation. With prolonged, low-intensity exercise, there was a progressive temporal shift toward lipid oxidation for energy, yet there was not a corresponding progressive increase in ACC phosphorylation or decrease in malonyl-CoA content. Thus, it would appear that AMPK activation does not regulate lipid oxidation during exercise. However, an important caveat is that the other key regulator of malonyl-CoA concentration -- the activity of the enzyme malonyl-CoA decarboxylase -- was not measured. Finally, as to the exercise-associated decrease in glycogen synthase activity, this did occur in association with increased AMPK activity, but in normal exercising muscle this effect was masked by the more potent effect of the decline of glycogen concentration. A further layer of complexity was introduced into the regulatory system by the effect of chronic training. In well-trained muscle, the increments in AMPK activity were less than in untrained muscle for the same degree of exercise, probably because trained muscle has better maintenance of high-energy stores (eg, phosphocreatine) and pH during exercise.

AMPK Regulation of Glycogen Synthase

Laurie Goodyear, PhD,^[5] Joslin Diabetes Center, Boston, Massachusetts, further discussed the details of the mechanism whereby AMPK regulates glycogen synthase. It has been suggested that AMPK promotes glycogen breakdown by inhibiting glycogen synthase activity and activating glycogen phosphorylase. However, conflicting data emerge from in vitro and animal studies. Furthermore, although net glycogen breakdown occurs with exercise, there is actually an increase in both glycogen synthesis and glycogenolysis (ie, increased turnover).

Dr. Goodyear's group examined whether AMPK might actually increase glycogen synthase activity in the context of exercising muscle. As with the data presented by Dr. Richter in humans, the results were complex, suggesting an exercise-induced increase in glycogen synthase activity that is not necessarily associated with the increase in AMPK activity, or so associated only in specific muscle types (eg, those with red, oxidative fibers). Animal experiments utilizing different activating mutations of AMPK gamma-subunit isoforms suggest an inverse relationship between glycogen content and AMPK activity. However, the conditions of these experiments do not preclude the possibility of a metabolic feedback loop whereby exercise increases AMPK activity, which increases glucose transport, thereby enhancing glycogen synthesis. Increased muscle glycogen content might then feed back to inhibit AMPK activity.

AMPK Activation and Insulin Sensitivity

So the effects of AMPK during exercise are complex -- a fact of no surprise to those experienced in the regulation of metabolic pathways in health and disease. But what is the outcome of all this signaling? Does insulin sensitivity actually increase as a result of AMPK activation, and if so, how? These questions were addressed by Edward Kraegen, PhD,^[6] University of New South Wales, Sydney, Australia, who measured insulin sensitivity in rats treated with the compound AICAR, an activator of AMPK. AICAR treatment increased whole-body insulin sensitivity, and this was associated with increased insulin sensitivity in the liver as well as in selected skeletal muscle subtypes (eg, white fiber quadriceps). Unlike the findings reported in exercising humans, and consistent with the known effect of AMPK, muscle malonyl-CoA content decreased. Dr. Kraegen also cited a poster^[7] at the ADA meeting that described a significant loss of intra-abdominal fat as well as plasma and intramuscular lipid content as resulting from chronic AICAR treatment and a recent article^[8] demonstrating that AICAR reverses glucose-induced inhibition of the key insulin action kinase Akt, together with improved fatty acid oxidation. In sum, these results at least suggest that the pharmacologic activation of AMPK could improve insulin sensitivity in the liver as well as selected skeletal muscle subtypes, and that this action might be due to enhanced intramyocellular lipid oxidation and disposal.

While not all the details of AMPK's role in muscle metabolism are clear, the data presented point to exciting possibilities for this molecule as a target for pharmacologic treatment of insulin resistance in muscle. The fact that AMPK is already targeted by some of our current, effective insulin-sensitizing drugs (metformin, rosiglitazone) gives this notion added credence. We await further understanding of the mechanisms of AMPK signaling, as well as compounds that can modulate its actions.

References

1. Minokoshi Y, Kim YB, Peroni OD, et al. Leptin stimulates fatty-acid oxidation by activating AMP-activated protein kinase. Nature. 2002;415:339-343.
2. Zhang Z, Radzuik J. Inverse relationship between peripheral insulin removal and action: studies with metformin. Am J Physiol Endocrinol Metab. 2001;281:E1240-1248.
3. Kemp BE. AMP-kinase and AMPKK in exercise. Symposium: The role of AMP kinase in exercise on substrate metabolism. Program and abstracts of the 62nd Scientific Sessions of the American Diabetes Association. June 14-18, 2002; San Francisco, California. Diabetes, Volume 51, Supplement 2.
4. Richter E. AMP kinase regulation and action in human skeletal muscles. Symposium: The role of AMP kinase in exercise on substrate metabolism. Program and abstracts of the 62nd Scientific Sessions of the American Diabetes Association. June 14-18, 2002; San Francisco, California. Diabetes, Volume 51, Supplement 2.
5. Goodyear L. Effects of AMP kinase on glycogen synthesis. Symposium: The role of AMP kinase in exercise on substrate metabolism. Program and abstracts of the 62nd Scientific Sessions of the American Diabetes Association. June 14-18, 2002; San Francisco, California. Diabetes, Volume 51, Supplement 2.
6. Kraegen E. Effects of activation of AMP kinase on muscle insulin resistance. Symposium: The role of AMP kinase in exercise on substrate metabolism. Program and abstracts of the 62nd Scientific Sessions of the American Diabetes Association. June 14-18, 2002; San Francisco, California. Diabetes, Volume 51, Supplement 2.
7. Saha A, Kurowski T, Kaushik V, et al. Pharmacological activation of AMP-activated protein kinase: a target for the treatment of obesity. Program and Abstracts of the 62nd Scientific Sessions of the American Diabetes Association, June 14-18, 2002; San Francisco, California. Poster 1029. Diabetes, Volume 51, Supplement 2.
8. Ido Y, Carling D, Ruderman N. Hyperglycemia-induced apoptosis in human umbilical vein endothelial cells: inhibition by the AMP-activated protein kinase activation. Diabetes. 2002;51:159-167.