Fig 2: AMPK activation leads to PDE4B phosphorylation in intact hepatocytes.In a, wild-type (WT) or mutant recombinant mouse liver PDE4B was incubated for 1 h with non-radioactive ATP in the presence (+) or absence (-) of recombinant activated AMPK. Proteins (0.1 µg) were seperated by SDS–PAGE for immunoblotting with the indicated antibodies. In b and c, mouse hepatocytes from either WT (b) or both WT and AMPK a1-/-a2LS-/- mice (c) were serum-starved overnight and incubated for 1 h with the indicated concentrations of 991 or phenformin. The cells were collected and lysed for immunoblotting with the indicated antibodies, except for PDE4B, which was immunoprecipitated as described in the Methods section, before immunoblotting. In c, phosphorylation levels of AMPK and its targets ACC, Raptor and PDE4B were quantified by densitometry and expressed relative to the corresponding total protein levels or GAPDH before normalization as indicated. Representative immunoblots are shown and for blot quantification in c, the values are means±s.e.m. for n=3 (p-ACC, p-Raptor and p-AMPK) or n=4 (p-PDE4B) separate experiments. Statistical analysis was by a paired Student's t-test. *Indicates a significant difference (P<0.05).

Fig 3: How biguanides and compound 991 antagonize glucagon signalling.Unlike biguanides, treatment with 991 activates AMPK without increasing cellular AMP levels. Both biguanides and 991 activate the major PDE 4B isoenzyme in hepatocytes in an AMPK-dependent manner. Metformin and phenformin activate hepatic AMPK either directly or via a rise in AMP, which could compete with ATP to inhibit adenylate cyclase. Phosphorylation-induced activation of PDE4B by AMPK reduces glucagon-stimulated cAMP accumulation. As a consequence, PKA activation by glucagon and downstream signalling are decreased in hepatocytes incubated with 991, the effect being AMPK-dependent.

Fig 4: Phosphorylation-induced activation of mouse liver PDE4B.PDE4B was cloned from mouse hepatocyte cDNA. The recombinant protein was overexpressed in E. coli and purified. PDE protein was phosphorylated for 1 h with purified recombinant activated AMPK and/or purified PKA catalytic subunits and [?-32P] ATP, and analysed by SDS–PAGE followed by Coomassie blue staining and phosphorimaging for quantification (a,c). In b, PDE was phosphorylated for 1 h with recombinant activated AMPK and [?-32P]. Phosphorylation sites were identified by LC–MS/MS after trypsin digestion and radioactive peak separation by high-performance liquid chromatography (HPLC). The phosphorylation sites that were identified are underlined in the right hand panel. In d and e, recombinant PDE was phosphorylated as above but with non-radioactive ATP for PDE assay as indicated. In d, separate determinations of Vmax and KM were made by linear regression of double reciprocal (Lineweaver Burk) plots. In e, the basal PDE activities of the wild-type (WT), S118A, S125A and S304A mutant proteins were 1.97±0.25, 0.14±0.01, 1.59±0.15 and 0.32±0.09 mU per mg of protein, respectively. Values are means±s.e.m. for n=3 (c–e) separate experiments. Statistical analysis was by a paired Student's t-test. *Indicates a significant difference (P<0.05) compared with control incubations or between the indicated conditions.