Obesity impairs skeletal muscle AMPK signaling during exercise: role of AMPKα2 in the regulation of exercise capacity in vivo

OBJECTIVE: Skeletal muscle AMP-activated protein kinase (AMPK)α2 activity is impaired in obese, insulin resistant individuals during exercise. We determined whether this defect contributes to the metabolic dysregulation and reduced exercise capacity observed in the obese state. DESIGN: C57BL/6J wild-type (WT) mice and/or mice expressing a kinase dead AMPKα2 subunit in skeletal muscle (α2-KD) were fed chow or high fat (HF) diets from 3–16 weeks (wks) of age. At 15wks mice performed an exercise stress test to determine exercise capacity. In WT mice, muscle glucose uptake and skeletal muscle AMPKα2 activity was assessed in chronically catheterized mice (carotid artery/jugular vein) at 16wks. In a separate study, HF-fed WT and α2-KD mice performed 5wks of exercise training (from 15–20wks of age) to test whether AMPKα2 is necessary to restore work tolerance. RESULTS: HF-fed WT mice had reduced exercise tolerance during an exercise stress test, and an attenuation in muscle glucose uptake and AMPKα2 activity during a single bout of exercise (p<0.05 vs. chow). In chow-fed α2-KD mice running speed and time were impaired ~45% and ~55%, respectively (p<0.05 vs. WT chow); HF feeding further reduced running time ~25% (p<0.05 vs. α2-KD chow). In response to 5wks of exercise training, HF-fed WT and α2-KD mice increased maximum running speed ~35% (p<0.05 vs. pre-training) and maintained body weight at pre-training levels, whereas body weight increased in untrained HF WT and α2-KD mice. Exercise training restored running speed to levels seen in healthy, chow-fed mice. CONCLUSION: HF feeding impairs AMPKα2 activity in skeletal muscle during exercise in vivo. While this defect directly contributes to reduced exercise capacity, findings in HF-fed α2-KD mice show that AMPKα2-independent mechanisms are also involved. Importantly, α2-KD mice on a HF-fed diet adapt to regular exercise by increasing exercise tolerance, demonstrating that this adaptation is independent of skeletal muscle AMPKα2 activity.