p85 regulatory subunit of PI3K mediates cAMP-PKA and estrogens biological effects on growth and survival

Cyclic adenosine 3050 monophosphate (cAMP) andprotein kinase A (PKA) cooperate with phosphatidylinositol 30 kinase (PI3K) signals in the control of growth and survival. To determine the molecular mechanism(s) involved, we identified and mutagenized a specific serine (residue 83) in p85aPI3K, which is phosphorylated in vivo and in vitro by PKA. Expression of p85aPI3K mutants (alanine or aspartic substitutions) significantly alteredthe biological responses of the cells to cAMP. cAMP protection from anoikis was reduced in cells expressing the alanine version p85aPI3K. These cells did not arrest in G1 in the presence of cAMP, whereas cells expressing the aspartic mutant p85D accumulatedin G1 even in the absence of cAMP. S phase was still efficiently inhibitedby cAMP in cells expressing both mutants. The binding of PI3K to Ras p21 was greatly reduced in cells expressing p85A in the presence or absence of cAMP. Conversely, expression of the aspartic mutant stimulatedrobust ly the binding of PI3K to p21 Ras in the presence of cAMP. Mutation in the Ser 83 inhibitedcAMP, but not PDGF stimulation of PI3K. Conversely, the p85D aspartic mutant amplifiedcAMP stimulation of PI3K activity. Phosphorylation of Ser 83 by cAMP–PKA in p85aPI3K was also necessary for estrogen signaling as expression of p85A or p85D mutants inhibitedor amplified, respectively, the binding of estrogen receptor to p85a andAKT phosphorylation induced by estrogens. The data presented indicate that: (1) phosphorylation of Ser 83 in p85aPI3K is critical for cAMP–PKA induced G1 arrest and survival in mouse 3T3 fibroblasts; (2) this site is necessary for amplification of estrogen signals by cAMP–PKA and relatedreceptors. Finally, these data suggest a general mechanism of PI3K regulation by cAMP, operating in various cell types and under different conditions.