Akt1 sequentially phosphorylates p27kip1 within a conserved but non-canonical region

p27kip1 (p27) is a multifunctional protein implicated in regulation of cell cycling, signal transduction, and adhesion. Its activity is controlled in part by Phosphatylinositol-3-Kinase (PI3K)/Akt1 signaling, and disruption of this regulatory connection has been identified in human breast cancers. The serine/threonine protein kinase Akt1 directly phosphorylates p27, so identifying the modified residue(s) is essential for understanding how it regulates p27 function. Various amino acids have been suggested as potential targets, but recent attention has focused on threonine 157 (T157) because it is located in a putative Akt1 consensus site. However, T157 is not evolutionarily conserved between mouse and human. We therefore re-evaluated Akt1 phosphorylation of p27 using purified proteins and in cells.

Here we show purified Akt1 phosphorylates human and mouse p27 equally well. Phospho-peptide mapping indicates Akt1 targets multiple sites conserved in both species, while phospho-amino acid analysis identifies the targeted residues as serine rather than threonine. P27 deletion mutants localized these sites to the N-terminus, which contains the major p27 phosphorylation site in cells (serine 10). P27 phosphorylated by Akt1 was detected by a phospho-S10 specific antibody, confirming this serine was targeted. Akt1 failed to phosphorylate p27S10A despite evidence of a second site from mapping experiments. This surprising result suggested S10 phosphorylation might be required for targeting the second site. We tested this idea by replacing S10 with threonine, which as expected led to the appearance of phospho-threonine. Phospho-serine was still present, however, confirming Akt1 sequentially targets multiple serines in this region. We took two approaches in an attempt to explain why different residues were previously implicated. A kinetic analysis revealed a putative Akt1 binding site in the C-terminus, which may explain why mutations in this region affect p27 phosphorylation. Furthermore, commercially available recombinant Akt1 preparations exhibit striking differences in substrate specificity and site selectivity. To confirm S10 is a relevant site, we first showed that full-length wild type Akt1 purified from mammalian cells phosphorylates both human and mouse p27 on S10. Finally, we found that in cultured cells under physiologically relevant conditions such as oxidative stress or growth factor deprivation, endogenous Akt1 causes p27 accumulation by phosphorylating S10.

Identifying where Akt1 phosphorylates p27 is essential for understanding its functional implications. We found that full-length wild type Akt1--whether purified, transiently overexpressed in cells, or activated in response to cellular stress--phosphorylates p27 at S10, a noncanonical but evolutionarily conserved site known to regulate p27 activity and stability. Using recombinant Akt1 recapitulating this specificity, we showed modification of p27S10 also leads to phosphorylation of an adjacent serine. These results integrate PI3K/Akt1 signaling in response to stress with p27 regulation through its major phosphorylation site in cells, and thus identify new avenues for understanding p27 deregulation in human cancers.