Glycogen synthase has been shown to be phosphorylated in vitro by five protein kinases, cyclic‐AMP‐dependent protein kinase (sites la, 1b and 2). phosphorylase kinase (site 2), glycogen synthase kinase 3 (sites 3a,3b and 3c), glycogen synthase kinase 4 (site 2) and glycogen synthase kinase 5 (site 5) [Cohen,P. et al. (1982)Eur. J. Biochem. 124, 21–35; Picton, C. et al. (1982) Eur. J. Biocliem. 124, 37–45]. All seven serine residues have now been found to be phosphorylated in vivo.Glycogen synthase purified from rabbits that had been injected with l‐propranolol contained 0.33 mol of phosphate/mol subunit in site la, 0.38 in site Ib, 0.39 in site 2, 1.27 in sites 3a+3b+3c and 0.65 in site 5, totalling 3.0 mol/mol subunit. The activity ratio (∓ glucose‐6 P) for these preparations was 0.21 and their Ka for glucose‐6‐P was 1.4mM.Glycogen synthase purified from animals that had been injected with I‐adrenaline contained 0.59 mol phosphate/mol subunit in site la, 0.65 in site lb, 1.03 in site 2, 2.43 in sites 3a + 3b + 3c and 0.65 in site 5, totalling 5.3 mol/mol subunit. The activity ratio of these preparations was 0.04 and their Ka for glucose‐6‐P was 6.0mM.The results show that cyclic‐AMP‐dependent protein kinase, glycogen synthase kinase 3 and glycogen synthase kinase 5 act as glycogen synthase kinases in vivo. Glycogen synthase kinase 4 may be largely responsible for the phosphorylation of site 2 in propranolol‐treated animals, and cyclic‐AMP‐dependent protein kinase and phos‐ phorylase kinase for the increased phosphorylation of site 2 in response to adrenaline.The activity of glycogen synthase in vivo in propranolol‐treated animals is largely determined by the state of phosphorylation of sites 3a, 3b and 3c, while the inactivation in response to adrenaline is due to increased phosphorylation at site 2 as well as at sites 3a, 3b and 3c. The finding that the phosphorylation of sites 3a, 3b and 3c is increased from 1.27 mol/mol subunit to 2.43 mol/mol subunit by adrenaline was unexpected, since this enzyme is not regulated by cyclic AMP. Possible explanations for this result are discussed.Phosphofructokinase was isolated as a byproduct of the purification of glycogen synthase. The phosphate content of this enzyme did not differ significantly between l−propranolol‐treated (0.15 mol/mol subunit) and adrenaline‐treated (0.18 mol/mol subunit) animals. These results indicate that muscle phosphofructokinase is not a substrate for cyclic‐AMP‐dependent protein kinase in vivo.
Read full abstract