Mullet, have you read anything about PGC-1 alpha or FGF21? They are hot topics at my university, and as I was sitting in a lecture on PGC, it struck me how similar the apparent aspects of it are to AP: increases muscular Glut4 expression, increases work capacity, decreases hyperglycemia and stunts gluconeogenesis.
PCG-1 is certainly interesting, but its universal coactivation of PPARå/∂/¥ means that that its phosphorylation in particular tissues - BAT/VAT/SAT, liver and skeletal muscle, in particular - is going to lead to somewhat paradoxical effects. For example, its coactivation of PPAR¥ is implicated in mitochondrial lipid accumulation via up-regulating glycerol-3-phosphate actyltransferase and other key lipogenic genes - whereas its similar coactivation of PPARå increases the transcription of genes involved in both skeletal muscle and hepatic oxidation of fatty acids.
As you say, Berberine and PGC-1 are similar in several senses - namely, in that they increase total mitochondrial density/capacity, increase ATP synthesis in response to physiological stimuli, and play key roles in substrate selection. However, they also differ in two key areas: their regulation of lipogenesis and their regulation of glucose oxidation.
As said previously, in certain cell conditions, PCG-1å/ß's coactivation of the PPARs has led to an imbalanced oxidation:accumulation/synthesis ratio for TAGs; leading, in some animal models, for hypotheses that PGC1 has prodiabetic effect. And, in respects to glucose oxidation, PGC1 has been shown to limit the oxidative phosphorylation of glucose by up-regulating PDK4 [pyruvate dehydrogenase kinase 4], a key negative regulator of glucose oxidation. Additively, these actions have led to the very paradoxical hypotheses I mentioned above - some claiming an anti-diabetic effect [via PPARå activation] and others a prodiabetic effect [via PPAR¥]. AMPk on the other hand, has been shown to limit
de novo lipogenesis and TAG biosynthesis in almost all tissue types, and it potently inhibits G3PAT; as well, it is a key regulator of mitochondrial oxidation of fatty acids primarily via regulation of the complimentary enzymes ACC [acetyl-CoA carboxylase] and MCD [malonyl-CoA decarboxylase]. In respects to glucose oxidation, phosphorylation of AMPk directly increases cytosolic HKII [hexokinase II activity] leading to higher rates of glucose consumption.
Now, where PGC-1 becomes troubling is in its dual nature to up-regulate lipogenesis and inhibit glucose catabolism in WAT tissues and skeletal muscle. This paradox is generally intuitive: its role in muscle-fiber phenotype variation is to convert TypeIIB [glycolytic] to TypeIIa/I [oxidative] - it achieves this by limiting total glucose phosphorylation capacity in skeletal muscle, sparing glycogen, and increasing total FA synthesis and accumulation. While this is amazing in fasted states - PGC-1 is going to upregulate FA synthesis and oxidation alike, leading to a positive FA, mitochondrial oxidative flux - it becomes counterproductive in the fed-state. While cell-surface GLUT4 is increased by PGC-1, total periphery --> mitochondria transport is decreased, and glycogen stores [particularly in myocytes] are already full due to limiting of HKII and this can lead to hyperglycemia [hence the paradox]. Where AMPk becomes superior to PGC-1, is that it is more diverse in its functionality. As a fully bi-directional enzyme, AMPk both increases and inhibits both lipogenesis and lipolysis; increases and inhibits the entire glycolytic pathway [gluconeogenesis, glyocgenolysis, glycogenesis, etc.]; and regulates catabolism and anabolism in a bi-directional fashion. This leads to a more favorable balance between glucose/FA oxidation and metabolism that avoids pro-diabetic/overtly pro-lipogenic consequences.