Phosphorylation of regulatory subunit of type I cyclic AMP‐dependent protein kinase: Biphasic effects of cyclic AMP in intact S49 mouse lymphoma cells

Intact S49 mouse lymphoma cells were used as a model system to study the effects of cyclic AMP (cAMP) and its analogs on the phosphorylation of regulatory (R) subunit of type I cAMP‐dependent protein kinase. Phosphorylation of R subunit was negligible in mutants deficient in adenylate cyclase; low levels of cAMP analogs, however, stimulated R subunit phosphorylation in these cells to rates comparable to those in wild‐type cells. In both wild‐type and adenylate cyclase‐deficient cells, R subunit phosphorylation was inhibited by a variety of N 6 ‐substituted derivatives of cAMP; C‐8‐substituted derivatives were generally poor inhibitors. Two derivatives that were inactive as kinase activators (N 6 ‐carbamoylmethyl‐5′‐AMP and 2′‐deoxy‐N 6 ‐monobutyryl‐cAMP) were also ineffective as inhibitors of R subunit phosphorylation. Preferential inhibition by N 6 ‐modified cAMP analogs could not be ascribed simply to selectivity for the more aminoterminal (site I) of the two cAMP‐binding sites in R subunit: Analog concentrations required for inhibition of R subunit phosphorylation were always higher than those required for activation of endogenous kinase; 8‐piperidino‐cAMP, a C‐8‐substituted derivative that is selective for cAMP‐binding site I, was relatively ineffective as an inhibitor; and, although thresholds for activation of endogenous kinase by site I‐selective analogs could be reduced markedly by coincubation with low levels of site II‐selective analogs, no such synergism was observed for the inhibitory effect. The uncoupling of cyclic nucleotide effects on R subunit phosphorylation from activation of endogenous protein kinase suggests that, in intact cells, activation of cAMP‐dependent protein kinase requires more than one and fewer than four molecules of cyclic nucleotide.