Phosphorylation-Dependent Inhibition of Akt1

Abstract

Protein kinase B (Akt1) is a proto-oncogene that is overactive in most cancers. Akt1 activation requires phosphorylation at Thr308; phosphorylation at Ser473 further enhances catalytic activity. Akt1 activity is also regulated via interactions between the kinase domain and the N-terminal auto-inhibitory pleckstrin homology (PH) domain. As it was previously difficult to produce Akt1 in site-specific phosphorylated forms, the contribution of each activating phosphorylation site to auto-inhibition was unknown. Using a combination of genetic code expansion and in vivo enzymatic phosphorylation, we produced Akt1 variants containing programmed phosphorylation to probe the interplay between Akt1 phosphorylation status and the auto-inhibitory function of the PH domain. Deletion of the PH domain increased the enzyme activity for all three phosphorylated Akt1 variants. For the doubly phosphorylated enzyme, deletion of the PH domain relieved auto-inhibition by 295-fold. We next found that phosphorylation at Ser473 provided resistance to chemical inhibition by Akti-1/2 inhibitor VIII. The Akti-1/2 inhibitor was most effective against pAkt1^T308 and showed four-fold decreased potency with Akt1 variants phosphorylated at Ser473. The data highlight the need to design more potent Akt1 inhibitors that are effective against the doubly phosphorylated and most pathogenic form of Akt1.

Keywords: genetic code expansion; phosphoinositide dependent kinase 1; phosphoseryl-tRNA synthetase; protein kinase B; tRNASep.

Figure 1

Simplified schematic of protein kinase B (Akt1) activation via phosphorylation of sites Thr308 and Ser473. The transition from Akt1’s inactive state (PH-in) to its fully active state (ppAkt1^T308/S473) requires the release of pleckstrin homology (PH) domain–mediated auto-inhibition. This release occurs when Akt1’s PH domain interacts with PIP₃ (PH-out). In the PH-out conformation, Akt1 is more susceptible to phosphorylation at Thr308 and Ser473 by phosphoinositide dependent kinase 1 (PDK1) and mechanistic target of rapamycin complex 2 (mTORC2), respectively. Upon release from PIP₃, Akt1 distributes rapidly in the cytosol and translocates to the nucleus to phosphorylate >100 cellular proteins [2,17].

Figure 2

Production of Akt1 variants with programmed phosphorylation. To produce pAkt1^T308, PDK1 (Akt1’s natural upstream kinase) was co-expressed along with Akt1. To produce pAkt1^S473, the phosphoserine orthogonal translation system was used to genetically incorporate phosphoserine at position 473 in response to an amber (UAG) codon. The ppAkt1^T308,S473 variant was produced by combining both methods. WT: wild type.

Figure 3

Activity of full-length ppAkt1 and commercially available active Akt1. (A) Kinase activity assays over a 30 min time course show substantially reduced activity of commercial Akt1 (lot 1, green squares) compared to full-length ppAkt1 (blue diamonds) produced in E. coli. Commercial Akt1 lot 2 (red circles) showed highly variable but similar activity to full-length ppAkt1. (B) Tryptic peptides from commercial Akt1 and ppAkt1 were analyzed by parallel-reaction monitoring mass spectrometry (PRM-MS). The purchased active Akt1 (lot 1) showed a low-intensity peak for phosphorylation at Thr308 (green peak at a retention time of ~40 min) and high-intensity peak for non-phosphorylated Thr308 (cyan peak, retention time of ~37 min). (C) PRM-MS analysis of ppAkt1^T308,S473 showed a high-intensity peak for phosphorylation at Thr308 (green peak at a retention time of 40 min) and the non-phosphorylated Thr308 was undetectable.

Figure 4

Enzyme activity of ΔPH-Akt1 variants. (A) The activity of differentially phosphorylated ΔPH Akt1 variants with the GSK-3β substrate peptide was measured over a 30 min time course. Akt1 phosphorylated at 308 and 473 (ΔPH-ppAkt^S473,T308, blue diamonds) showed maximal activity compared to unphosphorylated ΔPH-Akt1 (gray circles) and singly phosphorylated Akt1 variants ΔPH-pAkt1^T308 (black diamonds) and ΔPH-pAkt1^S473 (pink triangles). (B) The basal activity of unphosphorylated ΔPH-Akt1 (gray circles) was compared to full-length unphosphorylated Akt1 (green triangles). All reported values represent the mean of triplicate experiments, with error bars indicating one standard deviation.

Figure 5

Impact of PH domain on the activity of differentially phosphorylated Akt1. (A) Apparent catalytic rates (k_app) and (B) normalized k_app values of full-length Akt1 variants (blue) and ΔPH-Akt1 variants (red) are shown. Error bars represent one standard deviation of triplicate measurements.

Figure 6

Structure of Akt1 in active and inhibitor bound forms. (A) Structure of ΔPH-pAkt1^Thr308 Ser473Asp (PDB 1O6K [35]) is shown in complex with ATP analog (ANP-PNP) and substrate peptide (purple). (B) Structure of the full-length Akt1 (unphosphorylated) is shown in complex with the Akti-1/2 inhibitor VIII (PDB 1O96 [36]) binding in the cleft between the kinase domain (green) and the N-terminal PH domain (red).

Figure 7

Chemical inhibition of full-length phosphorylated Akt1 variants. Inhibition of (A) pAkt1^S473, (B) pAkt1^T308, and (C) ppAkt1^T308,S473 with varying concentrations (0.001, 0.05, 0.5, 1, 5, 10 μM) of Akti-1/2 inhibitor VIII. The resulting IC₅₀ values are in Table 4.