Cell-cycle-regulated activation of Akt kinase by phosphorylation at its carboxyl terminus

Abstract

Akt, also known as protein kinase B, plays key roles in cell proliferation, survival and metabolism. Akt hyperactivation contributes to many pathophysiological conditions, including human cancers^1–3, and is closely associated with poor prognosis and chemo- or radio-therapeutic resistance⁴. Phosphorylation of Akt at S473 (ref. 5) and T308 (ref. 6) activates Akt. However, it remains unclear whether further mechanisms account for full Akt activation, and whether Akt hyperactivation is linked to misregulated cell cycle progression, another cancer hallmark⁷. Here we report that Akt activity fluctuates across the cell cycle, mirroring cyclin A expression. Mechanistically, phosphorylation of S477 and T479 at the Akt extreme carboxy terminus by cyclin-dependent kinase 2 (Cdk2)/cyclin A or mTORC2, under distinct physiological conditions, promotes Akt activation through facilitating, or functionally compensating for, S473 phosphorylation. Furthermore, deletion of the cyclin A2 allele in the mouse olfactory bulb leads to reduced S477/T479 phosphorylation and elevated cellular apoptosis. Notably, cyclin A2-deletion-induced cellular apoptosis in mouse embryonic stem cells is partly rescued by S477D/T479E-Akt1, supporting a physiological role for cyclin A2 in governing Akt activation. Together, the results of our study show Akt S477/T479 phosphorylation to be an essential layer of the Akt activation mechanism to regulate its physiological functions, thereby providing a new mechanistic link between aberrant cell cycle progression and Akt hyperactivation in cancer.

Using single live-cell imaging⁸, we found that Akt activation fluctuated across the cell cycle, inversely correlating with Cdt1 abundance⁹ (Fig. 1a and Supplementary Fig. 1a). Statistical analysis of immunostained HeLa cells further showed that Akt-pS473 has a similar periodic feature as geminin^9,10 (Fig. 1b). Notably, in several cancer cell lines (Fig. 1c, d and Supplementary Fig. 1b, c), Akt phosphorylation, but not total Akt abundance, fluctuated across the cell cycle. The periodic Akt phosphorylation mirrored the expression pattern of cyclin A2, the predominant mammalian cyclin A isoform¹¹, during cell cycle progression (Fig. 1c, d). Moreover, acute depletion of cyclin A2 or Cdk2, but not cyclin E, resulted in decreased Akt phosphorylation, with no significant impact on phosphorylation of Akt upstream kinases PDK1 and mTORC2 (Fig. 1d). This prompted us to evaluate whether Cdk2/cyclin A directly regulates Akt activation in a phosphorylation-dependent manner during the cell cycle¹².

Figure 1. Akt activity fluctuated during the cell cycle and mirrored the periodic cyclin A expression pattern.

a, Single live-cell imaging of the Akt activity reporter (green, EGFP–Akt AR (activity reporter)) and the cell cycle marker Cdt1 (red, mCherry–Cdt1) transiently expressed in non-synchronized HeLa cells. b, A representative heat map of cell-cycle-dependent Akt1-pS473 as a function of DNA content (x axis) and geminin expression as an indicator of cell cycle stages (y axis). Blue, low Akt-pS473; red, high Akt-pS473. Number of cells was more than 500. c, Akt phosphorylation fluctuated across the cell cycle. Immunoblot (IB) of whole-cell lysates (WCLs) derived from T98G cells synchronized by double thymidine block and released to normal cell cycle for the indicated periods. d, Depletion of cyclin A, but not cyclin E, led to reduced Akt phosphorylation across the cell cycle. Immunoblot of whole-cell lysates derived from HeLa cells synchronized by nocodazole and released for the indicated periods. Where indicated, short interfering RNA (siRNA) oligonucleotides were transfected into cells 24 h before synchronization.

In support of Akt as a Cdk2/cyclin A substrate, Akt isoforms interacted with cyclin A2 (Fig. 2a and Supplementary Fig. 2a). Furthermore, we identified four ‘RXL’ cyclin A-binding motifs¹³ in Akt1 (Fig.2b), all of which are evolutionarily conserved (Supplementary Fig. 2b). Mutation of R76CL or R273DL, and to a lesser extent, R200VL or R370TL (RXL to AXA) attenuated Akt1 interaction with cyclin A2 (Fig. 2c), and reduced Akt1 activity (Supplementary Fig. 2c). Consistently, depleting cyclin A2 or Cdk2 (Supplementary Fig. 2d–f) led to a significant reduction in Akt phosphorylation. More importantly, either acute treatment with Cdk2 inhibitors (Fig. 2d) or deletion of the cyclin A2 allele in cyclin A2^f/f primary mouse embryonic fibroblasts (MEFs) (Fig. 2e) led to a marked decrease in Akt phosphorylation without a significant perturbation of cell cycle progression (Fig. 2d and ref. 14), excluding a possible indirect cell cycle effect on Akt phosphorylation by inhibiting Cdk2/cyclin A.

Figure 2. Cdk2/cyclin A2 functioned as a physiological kinase phosphorylating Akt1 at both S477 and T479.

a, Akt1 interacted with cyclin A2 at endogenous levels. Immunoblot analysis of WCLs and anti-Akt1 immunoprecipitations (IP) derived from MCF7 or T47D cells. b, Illustration of four putative cyclin A binding motifs (RXL) in human Akt1. c, Deficiency in cyclin A binding led to attenuated Akt phosphorylation. Immunoblot of WCLs and human influenza hemagglutinin (HA)-immunoprecipitations derived from HeLa cells transfected with indicated HA-Akt1 constructs. d, Immunoblot of WCLs derived from HeLa cells treated with indicated Cdk2 inhibitors at various time points. Drug doses used were roscovitine (40 μM), CVT-131 (0.5 μM) and mimosine (50 μM). e, Immunoblot of WCLs derived from cyclin A2^f/f primary MEFs with or without adenoviral-Cre infection. f, Immunoblot of WCLs derived from indicated immortalized MEFs with or without Cre viral infection. g, Immunoblot of WCLs derived from cyclin A2^f/f mouse embryonic stem (ES) cells stably expressing Cre-ER treated with 2 μg ml⁻¹ tamoxifen (4-OHT) for the indicated days. h, Immunoblot of WCLs derived from HeLa cells stably expressing inducible pTRIPZ-cyclin-A2 treated with 500 ng ml⁻¹ doxycycline for the indicated periods. i, j, MDA-MB-231 cells depleted of Cdh1 by two independent shRNAs (i), or depleted of both Cdh1 and cyclin A2 (j), were injected into nude mice (n = 10 for each group) and monitored for tumorigenesis in vivo. *P < 0.05 (Student’s t-test). k, Relative percentages of 50 samples examined from patients with breast cancer bearing the indicated cyclin A2 and Akt1-pS477/pT479 status indexed by either low or high.

Notably, deletion of cyclin A2, but not cyclin A1 or cyclin E1/E2 alleles, caused a significant decrease of Akt phosphorylation (Fig. 2f, g), whereas conversely, ectopic expression of cyclin A2 (Fig. 2h and Supplementary Fig. 3a) resulted in elevated Akt phosphorylation coupled with enhanced in vitro anchorage-independent growth (Supplementary Fig. 3b, c). Moreover, depletion of Cdh1, the E3 ligase that controls cyclin A turnover¹⁵, resulted in increased cyclin A2 abundance and elevated Akt phosphorylation, leading to enhanced in vitro anchorage-independent growth (Supplementary Fig. 3d, e) and in vivo tumour formation (Fig. 2i and Supplementary Fig. 3f–h). More importantly, increased Akt phosphorylation and tumor-igenicity by depleting Cdh1 could be partly reversed by extra depletion of cyclin A2 (Fig. 2j and Supplementary Fig. 3i–l). Collectively, these results support Cdk2/cyclin A2 as a major physiological kinase that governs Akt phosphorylation and oncogenic functions.

Notably, Cdk2/cyclin A directly phosphorylated Akt1 in vitro on its carboxy (C)-terminal region (Supplementary Fig. 4a, b). Serial truncations showed Cdk2/cyclin A phosphorylation sites in the last four evolutionarily conserved residues and subsequent mutageneses pinpointed both S477 and T479 as Cdk2/cyclin A sites (Supplementary Fig. 4c–e). Similarly, Cdk2/cyclin A phosphorylated Akt2-S478 (Supplementary Fig. 4f). Interestingly, mutation of G478 to proline (G478P) to mimic the canonical Cdk2 ‘SP/TP’ phospho-motif^16,17, or to other bulky amino acids (L/W/R), did not significantly affect Cdk2/cyclin-A-mediated Akt phosphorylation (Supplementary Fig. 4g). Conversely, C-terminal addition of an α-helix¹⁸ (Supplementary Fig. 4g) or a green fluorescent protein (GFP) (Supplementary Fig. 4h) reduced Cdk2/cyclin-A-mediated phosphorylation of the engineered non-tail version of S477G478-, but not S477P478-Akt1. These data indicate that S477/T479 may belong to a new class of Cdk2/cyclin A phospho-motifs where relative structural flexibility at the Akt1 C terminus might override the requirement of an adjacent proline for Cdk2-mediated phosphorylation of canonical TP/SP sites^16,17 typically buried within defined structures.

Furthermore, mass spectrometry analyses confirmed Akt1 S477 and T479 phosphorylation¹⁹ (Supplementary Fig. 5a–d). To gain further mechanistic insights, we developed phospho-specific antibodies that recognize pS477/pT479-Akt1 (Supplementary Fig. 6a–f), pS477-Akt1 or pT479-Akt1 (Supplementary Fig. 7a–h) in vivo. Owing to the large consistency among three antibodies under our experimental conditions, we focused on examining the correlation of Akt1-pS477/pT479 and Akt activation in the remainder of the studies. As with pS473-Akt1, Akt1 tail phosphorylation fluctuated during the cell cycle (Fig. 1c, d and Supplementary Fig. 1b, c), and was subjected to regulation by Cdk2/cyclin A (Fig. 2d–h). Furthermore, a positive correlation between cyclin A2 expression and Akt1-pS477/pT479 was observed in breast cancer patient samples (Fig. 2k and Supplementary Fig. 8a, b). More importantly, similar to the reported phosphorylation of Akt-S473 by several upstream kinases dependent on upstream stimuli^5,20,21, Akt1-pS477/pT479 could be mediated by mTOR (Supplementary Figs 9a–c, 10a–g and 11a–c) or DNAPK (Supplementary Fig. 9d, e), in addition to Cdk2/cyclin A. In keeping with this notion, elevated Akt1-pS477/pT479 was detected in Pten heterozygous MEFs (Supplementary Fig. 10h). Interestingly, Akt1-pS477/pT479 was negatively regulated by PTEN in mice in a tissue-specific manner and largely with Akt-pS473 (Supplementary Fig. 10i).

Moreover, inactivation of mTORC2 by depleting Rictor led to a more dramatic reduction of Akt1-pS477/pT479 in response to insulin (Supplementary Fig. 11a) than under synchronized cell cycle conditions (Supplementary Fig. 12a). Conversely, depletion of Cdk2, but not Rictor (Supplementary Fig. 12a, b), resulted in a more robust reduction in Akt1-pS477/pT479 across the cell cycle. These findings suggest that Akt1-pS477/pT479 is possibly mediated by Cdk2/cyclin A, mTORC2 or DNAPK, under cell cycle progression, growth factor stimulation or DNA damaging conditions, respectively (Supplementary Fig. 13a, b). Notably, depletion of Cdk2, cyclin A2 or Rictor in human primary foreskin fibroblasts all resulted in reduced Akt1-pS477/pT479 (Supplementary Fig. 14a–c), highlighting both Cdk2/cyclin A and mTORC2 as upstream physiological kinases governing Akt1-pS477/pT479. We therefore next evaluated the contribution of Akt1-pS477/pT479 to Akt kinase activation under various cellular conditions.

In keeping with the notion that Akt1 tail phosphorylation is required to achieve full Akt1 activation, the phospho-deficient Akt1-S477A/T479A (Akt1-AA) mutant showed a dramatic suppression, whereas a phosphomimetic S477D/T479E (Akt1-DE) mutant showed enhanced Akt1 phosphorylation and activation (Fig. 3a–c and Supplementary Fig. 15a–c). However, caution needs to be taken in interpreting these results, as these mutations might not fully recapitulate the in vivo phosphorylation status. Strikingly, Akt1-AA was defective in pS473 even in the myristoylation-tagged constitutively active Akt1 (Fig. 3d), advocating a critical role for Akt1-pS477/pT479 in activating Akt (Supplementary Fig. 16a–c). Consistently, S477D/T479E effectively rescued Akt-pS473 and kinase activity of the cyclin A binding motif-defective mutant, R76A (Fig. 3e, f), as well as its anchorage-independent growth ability (Supplementary Fig. 16d). However, the rescue effect of DE on Akt1-R76A was not due to changes in its affinity with PIP3 (Supplementary Fig. 16e).

Figure 3. Akt1-S477/T479 phosphorylation triggered Akt1-S473 phosphorylation and enhanced Akt1 activation.

a, Akt1-S477D/T479E mutation led to elevated Akt-pS473. Immunoblot of WCLs and HA-immunoprecipitations derived from HeLa cells transfected with indicated Akt1 constructs. EV, empty vector. b, In vitro kinase assays of indicated affinity-purified Akt1 kinases with crosstide as a substrate. Experiments were performed in triplicate; data are shown as mean ± s.d. c, Akt1-S477D/T479E mutation resulted in enhanced Skp2 phosphorylation. Immunoblot of WCLs and Flag-immunoprecipitations derived from HeLa cells transfected with indicated Akt1 constructs together with Flag–Skp2. d, e, Immunoblot of WCLs and HA-immunoprecipitations derived from HeLa cells transfected with indicated Akt1 constructs. f, In vitro kinase assays of indicated affinity-purified Akt1 kinases with crosstide as a substrate. Experiments were performed in triplicate; data are shown as mean ± s.d. g, Relative 5-bromodeoxyuridine (BrdU) incorporation for quadruple knockout MEFs (cyclin E1^−/−/ E2^−/−/A1^−/−/A2^f/f) expressing indicated Akt1 constructs that were further infected with or without Cre to delete the cyclin A2 alleles. h, Akt1-S477D/T479E mutation led to sustained Akt phosphorylation across the cell cycle. Immunoblot of WCLs derived from HeLa cells transfected with indicated Akt1 constructs, synchronized by nocodazole and released for indicated periods. i, Akt1-depleted HeLa cells stably expressing Akt1-WT or -DE were subjected to cyclin A2 knockdown by lentiviral shRNA infections. The resulting cell lines were subcutaneously injected into nude mice (n = 10 for each group) and monitored for tumorigenesis. P values were calculated by Student’s t-test.

Consistent with S477/T479 being Cdk2/cyclin A sites, ectopic expression of Akt1-DE partly rescued cell cycle defects observed in quadruple knockout MEFs (cyclin E1^−/−/cyclin E2^−/−/cyclin A1^−/−/cyclin A2^f/f) after Cre infection (Fig. 3g and Supplementary Fig. 16f). Compared with the periodic phosphorylation of wild-type Akt1 (Akt1-WT), Akt1-AA was severely impaired, whereas Akt1-DE showed an elevated and constitutive S473/T308 phosphorylation across the cell cycle (Fig. 3h). More importantly, under several physiological conditions (Supplementary Fig. 17a–c), Akt-pS473/pT308 was severely compromised in Akt1-AA, but robustly elevated in Akt1-DE expressing cells. Consistently, depletion of cyclin A2 in Akt1-WT, but not Akt1-DE expressing cells, led to a significant reduction in Akt phosphorylation (Supplementary Fig. 17d) and decreased tumour formation in vivo (Fig. 3i and Supplementary Fig. 17e–g). Cumulatively, these data demonstrate that in response to several upstream signals, phosphorylation of Akt1-S477/T479 may govern the canonical Akt-pS473 to promote Akt activation. Thus, we next evaluated the precise molecular mechanism(s) linking Akt1-pS477/pT479 to pS473 and subsequent Akt activation.

As Akt2 crystal structures are reported in great detail^22,23, we next focused on understanding how Akt2-pS478 (equivalent to Akt1-pS477; Supplementary Figs 4d and 18a) modulates Akt2 kinase activity. Notably, Akt2-pS478 functioned synergistically with Akt2-pS474 (equivalent to Akt1-pS473) to allosterically activate Akt2 (Fig. 4a and Supplementary Fig. 18b). Mechanistically, Akt2-pS478 may create a new charge–charge interaction between S478D and R208 to stabilize Akt2 in its closed, active form (Supplementary Fig. 18c, d). Consistently, deletion of the Akt1-tail after amino acid 475 (termed 476Δ) led to significantly increased Akt phosphorylation (Supplementary Fig. 18e), suggesting that either phosphorylation or deletion of the tail region could lock Akt in its active conformation.

Figure 4. Akt tail phosphorylation triggered Akt activation to promote Akt oncogenic functions.

a, Stimulation of purified recombinant pΔPH-Akt2-ΔC kinase activity by indicated Akt2 C-terminal tail peptides. pΔPH-Akt2-ΔC, Akt2 residues 146–460, lacking the PH domain and C-terminal tail with T309 phosphorylated. Experiments were performed in triplicate; data are shown as mean ± s.d. No phos, no phosphorylation. b, mTOR in vitro kinase assays with degenerate peptide libraries as substrates. Experiments were performed in triplicate; data are shown as mean ± s.d. c, Akt1-S477D/T479E mutation led to enhanced binding with Sin1. Immunoblot of WCLs and HA- or Flag-immunoprecipitations derived from Akt1-depleted HeLa cells transfected with indicated Akt1 constructs together with Flag–Sin1. d, e, Soft agar assays using Akt1-depleted HeLa cells stably expressing WT-, AA- or DE-Akt1. f, g, Akt1-depleted HeLa cells stably expressing WT-, AA- or DE-Akt1 were injected subcutaneously into nude mice (n = 10 for each group) and monitored for tumorigenesis (f). Tumours were dissected and weighed (g). *P < 0.05 (Student’s t-test). h, Immunoblot analyses of the indicated mouse brain tissues derived from mice with the indicated cyclin A2 genetic status. i, Fluorescence-activated cell sorting analysis of cyclin A2^f/f mouse embryonic stem cells stably expressing Cre-ER in the presence of 2 μg ml⁻¹ tamoxifen for 4 days. j, Immunoblot analysis of WCLs derived from various indicated embryonic stem stable cell lines generated in i.

Intriguingly, in addition to stabilizing Akt, Akt tail phosphorylation may alter Akt kinase kinetics to accelerate the kinase reaction (Supplementary Fig. 18f). Furthermore, pS473 and pS477/pT479, both of which activate Akt, are intrinsically linked; as compared with Akt1-WT, Akt1-DE but not Akt1-AA was preferentially phosphorylated by mTOR in vitro (Supplementary Fig. 19a, b). This indicates that pS477/pT479 might prime Akt1 for mTORC2-mediated phosphorylation of S473 (refs 5, 24) (Fig. 4b). Mechanistically, enhanced phosphorylation of Akt1-DE at S473 by mTOR may be in part due to its increased interaction with Sin1 (Fig. 4c) and mTOR, but not due to changes in its association with the Akt inhibitor CTMP²⁵ or phosphatase PHLPP2 (ref. 26) (Supplementary Fig. 19c, d). However, phospho-mimetic mutation of S473 (S473D) (Supplementary Fig. 20a) or its mimicking peptide library (Supplementary Fig. 20b) led to reduced Akt1 tail phosphorylation, suggesting that the pS477/pT479 event may precede pS473.

Mutation of either S473 or S477/T479 to alanine significantly reduced Akt activity towards phosphorylating Skp2 or crosstides, whereas no significant additive effect was observed in S473A/S477A/T479A (Supplementary Fig. 21a, b), arguing for a possible functional redundancy between pS477/pT479 and pS473 for Akt catalytic activity. Consistently, Akt1-DE kinase activity was largely unchanged after depletion of Rictor (Supplementary Fig. 21c, d), whereas Akt1-DE partly rescued the deficient Akt kinase activity in Akt1-S473A (Supplementary Fig. 21e, f). Given the close proximity of S473 and S477/T479, these results indicate that in addition to promoting pS473, pS477/pT479 may also trigger Akt activation independently of, or partly compensate for, pS473.

Biologically, phosphorylation of the Akt1 tail activated Akt1, leading to elevated Skp2 or FOXO phosphorylation, which further promoted cell cycle progression (Supplementary Fig. 22a–c), or conferred resistance to the chemotherapeutic agents etoposide or camptothecin (Supplementary Fig. 22d, e), respectively. More importantly, compared with Akt1-WT, Akt1-DE-expressing cells showed growth advantage in both in vitro soft agar (Fig. 4d, e) and in vivo tumour formation (Fig. 4f, g and Supplementary Fig. 23a, b) assays, supporting a role for Akt1-pS477/pT479 in promoting Akt1 activation and signalling phenotypes associated with malignancy. Notably, Akt1-476Δ phenocopied Akt1-DE by showing elevated Akt phosphorylation (Supplementary Fig. 18e), probably through enhanced interaction with mTORC2 (Supplementary Fig. 24a), thereby promoting in vitro anchorage-independent growth and in vivo tumorigenesis (Supplementary Fig. 24 b–f).

To study the in vivo physiological significance of Cdk2/cyclin-A-mediated phosphorylation on Akt1-S477/T479 further, we generated brain-specific cyclin A2 knockout mice with nestin-Cre²⁷. Interestingly, a significant reduction in Akt1-pS477/pT479 was observed in cyclin A2^Δ/Δ olfactory bulbs coupled with elevated cleavage of caspase 3 (Fig. 4h), suggesting that Akt1-pS477/pT479 might be critical for cell survival in olfactory bulbs. Consistently, acute ablation of cyclin A2 in cyclin A2^f/f mouse embryonic stem cells led to induced cellular apoptosis (Fig. 4i and Supplementary Fig. 25a, d), providing a possible explanation for previous findings that deleting cyclin A2 abolished embryonic stem cell colony formation in vitro¹⁴. Furthermore, cyclin A2-deletion-induced elevation of cellular apoptosis in mouse embryonic stem cells could be partly rescued by expressing Akt1-DE, but not Akt1-AA (Fig. 4i, j and Supplementary Fig. 25), supporting the idea that cyclin A2 may govern cellular survival in vivo largely by promoting Akt activation.

To extend these findings and their clinical relevance to human pathophysiology, we observed a positive correlation between Akt1-pS477/pT479 and Akt1-pS473 in samples from patients with breast cancer and breast-cancer-derived cell lines (Supplementary Fig. 26a–e). Interestingly, high levels of pS477/pT479 occurred at a relatively higher rate than pS473 in an earlier breast cancer developmental stage (stage II) (Supplementary Fig. 26d), indicating that pS477/pT479 may serve as a better biomarker for early-stage breast cancer detection, although further investigations are warranted.

Taken together, our data unravel a new phosphorylation event on Akt1 at its extreme C-terminal residues, S477 and T479, to trigger Akt1 activation either through enhancing the association between Akt1 and mTORC2 to promote pS473, or by functionally compensating for pS473 to lock Akt1 in its active conformation. More importantly, our study directly couples Akt activity with cell cycle progression, two well-characterized hallmarks of human cancers^28–30. In this regard, our data suggest that cyclin A2 overexpression might exert its physiological functions in part by directly phosphorylating and activating Akt to trigger its pro-survival and oncogenic functions.

METHODS

Plasmids

pcDNA3-HA-Akt1/Akt2/Akt3 and pcDNA3-Myr-HA-Akt1 constructs were obtained from A. Toker and described previously³¹; pcDNA3-HA-Akt1-S477A, pcDNA3-HA-Akt1-T479A, pcDNA3-HA-Akt1-S477A/T479A, pcDNA3-HA-Akt1-S477D, pcDNA3-HA-Akt1-S479E, pcDNA3-HA-Akt1-S477D/T479E, pcDNA3-HA-Akt2-S478A, pcDNA3-HA-Akt2-S478D, pcDNA3-HA-Akt3-S474A, pcDNA3-HA-Akt3-S474D, pcDNA3-Myr-HA-Akt1-S477A/T479A, pcDNA3-Myr-HA-Akt1-S477D/T479E, pcDNA3-HA-Akt1-R76LAA, pcDNA3-HA-Akt1-R76LAA/DE, pcDNA3-HA-Akt1-R200LAA, pcDNA3-HA-Akt1R273LAA, pcDNA3-HA-Akt1-R370LAA and pcDNA3-HA-Akt1-4A constructs, as well as the shAkt1-resistant versions of Akt constructs, were generated using a QuikChange XL Site-Directed Mutagenesis Kit (Stratagene) according to the manufacturer’s instructions, with specific primer sequences available upon request. The various (glutathione S-transferase) GST–Akt1 plasmids used for in vitro kinase assays were constructed by sub-cloning corresponding PCR fragments into the pGEX-4T-1 vector by BamHI/EcoRI sites for amino (N)-terminal GST tag. pBabe-Myr-HA-Akt1-WT, S477A/T479A, S477D/T479E, R76A and R76A/DE retroviral vectors were generated by sub-cloning Akt1 from corresponding pcDNA3-Myr-HA-Akt1 vectors into the pBabe-HA-hygromycin vector or MSCV-hygromycin vector. Flag–Skp2 vector was constructed as described previously³¹. Cre adenoviruses were obtained from the laboratory of P.S. The phage-Cre construct was obtained from the laborator y of L. Glimcher. The Cre-ER construct was obtained from the laboratory of P. Pandolfi. The pTRIPZ-cyclin A2 was constructed by cloning the cyclin A2 allele into pTRIPZ vector by AgeI/ClaI sites. Akt activity reporter (Akt AR) was obtained from J. Zhang as described previously⁸. Cdt1 and geminin fucci reporters were obtained from A. Miyawaki as described previously⁹.

shRNAs and siRNAs

shRNA vectors to deplete endogenous Akt1 were described previously³¹. shRictor vectors were purchased from Addgene (1854). shRNA vectors to deplete endogenous Cdk2 were purchased from Open Biosystems (RHS4533-EG1017). shCdh1 and shPTEN vectors were described previously³². To generate the lentiviral shRNA constructs against human cyclin A2, the following sequences were cloned into the pLKO-puro vector (shCyclin A2 no. 1 sense: 5′-CCGGAAGGCAGCGCCCGTCCAACAACTCGAGTTGTTGGAGCGGCGCTGCCTTTTTTTG-3′; shCyclin A2 no. 1 anti-sense: 5′-AATTCAAAAAAAGGCAGCGCCCGTCC AACAACTCGAGTTGTTGGAGCGGCGCTGCCTT-3′; shCyclin A2 no. 2 sense: 5′-CCGGAACTACATTG ATAGGTTCCTGCTCGAGCAGGAACGTATCAATGTAGTTTTTTTG-3′; shCyclin A2 no. 2 anti-sense: 5′-AATTCAAAAAAACTACATTGATAGGTTCCTGCTCGAGCAGGAACGTATCAATGTAGTT-3′). Cyclin A2, Akt1, Cdk2 and cyclin E siRNA oligonucleotides and the siRNA transfection method have been described previously³¹.

Antibodies

Anti-Sin1 antibody for western blots was purchased from Millipore (07-2276). Anti-Akt1 antibody conjugated agarose beads for endogenous Akt1 immunoprecipitation were purchased from Cell Signaling Technology (3653S). Anti-mTOR antibody (2972), anti-pSer2481-mTOR antibody (2974), anti-Raptor antibody (2280), anti-Rictor antibody (9476), anti-phospho-Ser473-Akt antibody (4051), anti-phospho-Thr308-Akt antibody (2965), anti-phospho-Thr450-Akt antibody (9267), anti-Akt1 antibody (2938), anti-Akt total antibody (4691), anti-phospho-Thr389-S6K antibody (9205), anti-S6K antibody (9202), anti-phospho-Akt substrate (RxRxxpS/T) antibody (9614), anti-phospho-PDK1 antibody (3438), anti-phospho-TSC2-antibody (3617), anti-TSC2 antibody (3990), anti-phospho-T24/32-FOXO1/3a antibody (9464), anti-phospho-S253-FOXO 3a antibody (9466), anti-phospho-S256-FOXO1 antibody (9461), anti-FOXO1 antibody (2880), anti-FOXO3a antibody (2497), anti-phospho-pRAS40 (T246) antibody (2997), anti-pRAS40 antibody (2691), anti-Skp2 antibody (4313), anti-pS9-GSK3β antibody (9323), anti-GSK3β antibody (9315) and anti-Cdt1 antibody (8064) were purchased from Cell Signaling Technology. Anti-cyclin A2 (sc-751), anti-cyclin E (sc-247), anti-Cdk2, anti-cyclin B1 (sc-245) and polyclonal anti-HA antibody (sc-805) were purchased from Santa Cruz. Anti-geminin (ab12147) was purchased from Abcam. Anti-Tubulin antibody (T-5168), anti-Vinculin antibody (V-4505), polyclonal anti-Flag antibody (F-2425), monoclonal anti-Flag antibody (F-3165), anti-Flag agarose beads (A-2220), anti-HA agarose beads (A-2095), peroxidase-conjugated anti-mouse secondary antibody (A-4416) and peroxidase-conjugated anti-rabbit secondary antibody (A-4914) were purchased from Sigma. Monoclonal anti-HA antibody (MMS-101P) was purchased from Covance. Various anti-pS477-Akt1, anti-pT479-Akt1 and anti-pS477/pT479-Akt1 antibodies and the anti-pS72-Skp2 antibody were produced by Cell Signaling Technology.

Immunoblots and immunoprecipitation

Cells were lysed in EBC buffer (50 mM Tris pH 7.5, 120 mM NaCl, 0.5% NP-40) (for immunoprecipitation) supplemented with protease inhibitors (Complete Mini, Roche) and phosphatase inhibitors (phosphatase inhibitor cocktail set I and II, Calbiochem). The total protein concentrations of whole cell lysates were measured by a Beckman Coulter DU-800 spectrophotometer using the Bio-Rad protein assay reagent. The same amounts of whole cell lysates were resolved by SDS–PAGE and immunoblotted with indicated antibodies. For immunoprecipitation, 1000 μg lysates were incubated with the indicated antibody (1–2 μg) for 3–4 h at 4 °C followed by incubation for 1 h with Protein A Sepharose beads (GE Healthcare). Immunoprecipitants were washed five times with NETN buffer (20 mM Tris, pH 8.0, 100 mM NaCl, 1 mM EDTA, 0.5% NP-40) before being resolved by SDS–PAGE and immunoblotted with indicated antibodies.

Cell culture and cell viability assays

Cell culture and transfection procedures have been described previously^31,33. The WT and Akt1^−/−/Akt2^−/− MEFs were obtained from N. Hay as described previously³⁴. Retroviral shRNA virus packaging and subsequent infection of various cell lines were performed according to the protocol described previously³⁵. For cell viability assays, cells were plated at 10,000 per well in 96-well plates, and incubated with complete Dulbecco’s Modified Eagle Medium (DMEM) medium containing different concentrations of etoposide (Sigma, E1383), camptothecin (Sigma C156) and doxorubicin (Sigma, D1515) for 48 h. Assays were performed with a CellTiter-Glo Luminescent Cell Viability Assay Kit according to the manufacturer’s instructions (Promega).

Single-cell imaging

Single live-cell imaging procedures were done according to previously described methods³⁶ with some modifications as described below. Cells were plated on number 1.5 coverslips MatTek Dish (P35G-1.5-14-C). DMEM media (GIBCO) containing 25 mM HEPES, 25 mM glucose (except for the experiment of manipulating glucose concentration), 2 mM glutamine and 3% FBS (phenol red was excluded during image acquisition). Cells were imaged on a Nikon TE 2000E and a Nikon and Ti motorized inverted microscope with a 20×/0.75 or 60×/1.4 (oil immersion) numerical aperture objective lens. Dual emission ratio imaging was performed with a 436/10 filter, a diachronic mirror (Chroma 86002v1bs) and two emission filters (470/30 for cyan and 535/30 for yellow). All optical filters were obtained from Chroma Technologies. Images were acquired with a Hamamatsu ORCA-R2 cooled charge-coupled device (CCD) camera controlled with MetaMorph 7 software (Molecular Devices) and the Perfect Focus System for continuous maintenance of focus. Fluorescence images were background-corrected. For time-lapse experiments, images were collected every 30 min for 24 consecutive hours, with an exposure time of 50–100 ms and 2 × 2 binning, with illumination light shuttered between acquisitions. The ratios of yellow-to-cyan were then calculated at different time points and normalized by dividing all ratios by the emission ratio before stimulation, setting the basal emission ratio as 1.

Mapping the cell-cycle progression status in individual cells

The procedures for performing this experiment were as described previously³⁷. Briefly, to relate the elapsed time after cell division to cell-cycle phase and progression with the abundance or activities of the proteins of interest, we measured the distribution of DNA content in an asynchronously growing culture by flow cytometry using propidium iodide staining, as well as the gemenin abundance by a GFP–gemenin reporter as described previously⁹. HeLa cells expressing the mAG-hGem fucci reporter system⁹ were cultured in DMEM (Cellgro; DMEM 10-017-CV) with 10% FBS (Gibco; 26140) and 1% penicillin/streptomycin solution (Cellgro; 30-002-CI). Media were further supplemented with 3 μg ml⁻¹ blasticidin (InvivoGen; ant-bl-5b) to maintain selection of cells expressing mAG-hGem. For experiments, cells were plated on 24 mm × 60mm coverslips, no. 1.5 (VWR; 48393-252). Before plating, coverslips were sterilized by incubation for 20 min in 70% ethanol at room temperature and then dried in sterile conditions. Cells were typically plated at 1 × 10⁵ cells per millilitre in to 15 cm dishes that were pre-prepared with sterile coverslips as described above. Cells were fixed approximately 48 h after plating. The kinase activities of Akt were measured by anti-p-S473-Akt antibody. These distributions were fitted using a modification of the Dean–Jett model³⁸ to determine the number of cells in G1, S and G2 phases and were subsequently translated to the time spent in various cell-cycle phases using a previously published model³⁹.

Fixing and staining cells for immunofluorescence

To fix cells, coverslips were incubated for 10 min in 4% paraformaldehyde (Alfa Aesar, 30525-89-4) at room temperature. Cells were then permeabilized for 5 min in dry methanol at −20 °C and rehydrated in PBS. Cells were incubated with primary antibody (either Anti-phospho-Akt (S473) (Cell Signaling; 4060S) or Anti-Akt (Cell Signaling, 4060S)) overnight at 4 °C, followed by incubation with a fluorescent secondary antibody (Invitrogen, A-21429) for 1 h at room temperature. Antibody solutions were made in PBS with 2% bovine serum albumin (BSA). DNA was stained by incubating in 10 μM 4′,6-diamidino-2-phenylindole (DAPI) for 5 min. To label protein mass, fixed, permeabilized samples were incubated with 0.04 μg ml⁻¹ succinimidyl ester linked dye diluted in PBS (Alexa Fluor 647 carboxylic acid, succinimidyl ester; Invitrogen, A-20106). Following labelling procedures, cells were mounted on glass slides in ProLong Gold antifade (Life Technologies, P36930).

Microscopy

Slides prepared as described above were imaged with a Nikon Ti inverted fluorescence microscope with Perfect Focus controlled by Nikon Elements. We used the scan-slide function to image the full area of the slide at ×20 magnification. The microscope was surrounded by a custom enclosure to maintain constant temperature and atmosphere. The filter sets used were as follows: cyan fluorescent protein (CFP), 436/20 nm, 455 nm, 480/40 nm (excitation, beam splitter, emission filter); yellow fluorescent protein (YFP), 500/20 nm, 515 nm, 535/30 nm; and mCherry, 560/40 nm, 585 nm, 630/75 nm (Chroma). Images were acquired every 20 min in the phase and CFP channels and every 60 min in the YFP and mCherry channels. We acquired six z-sections with a step size of 0.75 μm in the YFP and mCherry channels. Image acquisition was controlled by MetaMorph software (Molecular Devices). This resulted in approximately 5,000–8,000 images per slide, leading to a total of about 100,000 cells. For larger cell counts, data from several slides was concatenated.

Image analysis

Image analysis used Matlab, with an algorithm written by R. Kafri, as described previously³⁷. Top-Hat transformation of images was used to remove background trends. Nuclei were identified by thresholding the DAPI image. Boundaries between adjacent, touching nuclei were identified by seed-based watershedding. Seeds were calculated as the regional maxima of the Gaussian smoothed image. Cell boundaries were then found by performing the same steps on the Alexa Fluor 647 scanning electron image using the nuclei as seeds. The integrated immunofluorescence intensity within the cell boundary was taken as the Akt or pS473-Akt amount.

Cdk2/cyclin A in vitro kinase assays

Cdk2/cyclin A in vitro kinase assay methods were adapted from those described previously^5,31. Briefly, 3 μg indicated GST–Akt1-Tail (268–480) fusion proteins were incubated with 50 ng commercially obtained recombinant active Cdk2/cyclin A proteins (NEB P6025), in the presence of 5 μCi [γ-³²P]ATP and 200 μM cold ATP in the NEB kinase reaction buffer for 30 min. The reaction was stopped by the addition of SDS containing lysis buffer and resolved by SDS–PAGE. Phosphorylation of GST–Akt1-Tail was detected by autoradiography.

For the cold Cdk2/cyclin A in vitro kinase assay, no [γ-³²P]ATP was added, and the other procedures remained the same. For the in vitro kinase assay on full length Akt1, HA-Akt1 was immunoprecipitated from 293T cells and subjected to phosphatase treatment for 30 min at 30 °C before adding into the kinase reaction.

Akt in vitro kinase assays

Various HA-Akt1 mutant proteins were HA-immunoprecipitated and stored in EBC buffer with 10% glycerol. Amounts of kinases affinity purified were determined by Coomassie staining. About 2 μg of each kinase was incubated with 50 μM crosstide (Millipore 12-331, based on the Akt phosphorylation site of GSK3), 200 μM ATP (with [γ-³²P]ATP) and reaction buffer (50 mM Tris pH 7.5, 1 mM MnCl₂, 2 mM DTT, 1 mM EGTA, 1 mg ml⁻¹ BSA) at 30 °C for 1 h. Aliquots of each reaction were spotted onto P81 phosphocellulose filters and washed extensively with 75 mM phosphoric acid. Filters were dried and radioactivity was determined by scintillation counting. Experiments were done in triplicate.

To test the ability of peptides based on the hydrophobic motif of Akt2 to activate the enzyme, the activity of pΔPH-Akt2-ΔC (Akt2 lacking the PH domain and C-terminal tail (residues 146–460) but phosphorylated on T309 by GST–PDK1, as described in ref. 23) was determined in the presence of various hydrophobic-motif peptides. pΔPH-Akt2-ΔC (50 nM) was combined with 50 mM substrate peptide (crosstide, based on the Akt phosphorylation site of GSK3), 500 μM ATP (with [γ-³²P]ATP), hydrophobic-motif peptides and buffer (50 mM Tris pH 7.5, 10 mM MgCl₂, 2 mM DTT, 1 mM EGTA, 1 mg ml⁻¹ BSA) at 30 °C for 1 h. Aliquots of each reaction were spotted onto P81 phosphocellulose filters and washed extensively with 75 mM phosphoric acid. Filters were dried and radioactivity was determined by scintillation counting. Experiments were done in triplicate.

mTOR in vitro kinase assay on peptide libraries

The reaction contained 100 ng mTOR (EMD/Calbiochem 475987, containing truncated mTOR kinase domain with amino acids 1,360–2,549), 50 mM biotinylated peptide mix and 100 μM cold ATP (with 0.8 μCi/ml⁻¹ [γ-³²P]ATP) in assay buffer (50 mM HEPES pH 7.5, 10 mM MnCl₂, 10 mM MgCl₂, 2 mM DTT, 0.5 mM EGTA). In each biotinylated peptide mix, one amino acid is fixed at the indicated position relative to a central serine/threonine residue. The remaining positions surrounding the central serine/threonine are degenerate (approximately equimolar mixtures of the 17 amino acids excluding cysteine, serine and threonine). The reactions were at 30 °C for approximately 7 h. Aliquots of each reaction were then spotted onto an avidin-coated membrane. The membrane was washed sequentially with 0.1% SDS in TBS, 1% H₃PO₄ and 2 M NaCl. After washing, the membrane was dried and exposed to a phospho-imager. The extent of incorporation of radiolabelled phosphate into each peptide was quantified using ImageQuant software.

Mass spectrometry analysis to detect HA-Akt-S477/T479 phosphorylation in vivo

The procedures of mass spectrometry analysis were performed as described previously^40,41 with minor modifications. Briefly, 293 cells were transiently transfected with the pcDNA3-HA-Akt1 plasmid and, 24 h after transfection, cells were treated with insulin for 30 min before collection. Whole-cell lysates were collected to perform HA immunoprecipitation. The HA immunoprecipitates were then resolved on SDS–PAGE and visualized by colloidal Coomassie blue. The band containing HA-Akt1 was excised and washed with 50% acetone. In-gel digestion of the protein was performed with trypsin and analysed by reversed-phase microcapillary/tandem mass spectrometry using a LTQ Orbitrap XL (Thermo Fisher Scientific) Hybrid Ion Trap-Orbitrap Mass Spectrometer. Tandem mass spectrometry centroid spectra collected by collision-induced dissociation in Top 6 data-dependent acquisition mode were searched against the concatenated target and decoy(reversed) Swiss-Prot protein database (version 2012_01) using the Sequest search engine with Proteomics Browser Software (W.S. Lane) with differential modifications for Ser/Thr/Tyr phosphorylation (+79.97) and differential modification of Met oxidation (+15.99, Msx). Phosphorylated and unphosphorylated peptide sequences were identified, and manual inspection and determination of the exact sites of phosphorylation were confirmed using FuzzyIons and GraphMod software (Proteomics Browser Software, W.S. Lane). False discovery rates of peptide hits (phosphorylated and non-phosphorylated) were estimated below 1.0% based on reversed database hits.

Mass spectrometry analysis to detect GST–Akt-S477/T479 phosphorylation in vitro

The mass spectrometry procedure was performed as described previously^40,41. Phosphorylated GST–Akt1 samples were prepared as described in the kinase assay section, with Cdk2/cyclin A or mTOR, respectively. Specifically, 1 μg of phosphorylated HA-Akt1 proteins were subjected to trypsin digestion before being analysed by mass spectrometry.

5-Bromodeoxyuridine labelling

5-Bromodeoxyuridine labelling assay was performed as described previously³². Experiments were repeated three times to generate the error bars.

Soft agar assay

The anchorage-independent cell growth assays were performed as described previously⁴². The solid medium consists of two layers. The bottom layer contains 0.8% noble agar and the top layer 0.4% agar. Briefly, 3 × 10⁵ cells were plated in the top layer. Complete DMEM medium (500 μl) was added every other day to keep the top layer moisture, and 3 weeks later the cells were stained with iodonitrotetrazolium chloride for colony visualization and counting. Three independent experiments were performed to generate the error bars.

Mouse xenograft assay

The tumorigenesis assay was as described previously⁴³. Each xenograft experiment was performed as described in the corresponding figure legends. For example, in Fig. 4f, g, briefly 2.5 × 10⁶ HeLa cells stably expressing HA-Akt1-WT, HA-Akt1-S477A/T479A or HA-Akt1-S477D/T479E (using empty vector as a negative control) were mixed with sterile 1× PBS (1:1) and injected into the flank of 10 male nude mice. Tumour size was measured weekly with a calliper, and the tumour volume was determined with the formula L × W² × 0.52, where L is the longest diameter and W is the shortest diameter. After 28 days, mice were killed and in vivo solid tumours were dissected, then tumour weights were measured and recorded.

Immunohistochemistry of tissue microarray (immunohistochemistry and immunostaining analyses)

Formalin-fixed and paraffin-embedded tissue microarrays of human breast tissues and breast cancer tissues were purchased from Imgenex (IMH-371). Immunohistochemical stainings for cyclin A2, Akt-pS473 and Akt-pS477/pT479 were performed as described previously⁴⁴. The cyclin A2 antibody for immunohistochemistry was purchased from Santa Cruz (sc-751), Akt-pS473 antibody for immunohistochemistry was purchased from Cell Signaling Technology (4060) and the Akt-pS477/pT479 antibody was generated by collaboration with the Cell Signaling Technology and further validated in this study.

Generation of cyclin A2^f/f nestin-Cre mice, and brain tissue analysis

Olfactory bulbs, cortices and cerebella from cyclin A2 ^flox/flox nestin-Cre mice⁴⁵, obtained by crossing cyclin A2 ^flox/+ mice with nestin-Cre mice, were collected on the postnatal day 2, lysed and analysed by immunoblots.

Generation of Akt1 expressing cyclin A1^−/−/cyclin A2^f/f/Cre-ER mouse embryonic stem cell lines

cyclin A1^−/−/cyclin A2^flox/flox mouse embryonic stem cells¹⁴ were electroporated with the pMSCV-CreER-puro expressing vector and then plated on a monolayer of arrested, puromycin-resistant feeder fibroblasts. After 2 days, cells were subjected to selection by 2 μg ml⁻¹ puromycin until resistant colonies became apparent (about 7 days). Single colonies were picked, expanded and tested for Cre-ER expression by immunoblot analysis. The resulting cells were infected with pBabe-hygro-HA-Akt1-DE or -AA retroviruses (using pBabe-hygroempty vector as a negative control) and selected in 1 μg ml⁻¹ puromycin in combination with 250 ng ml⁻¹ hygromycin for 3 days to eliminate non-infected cells and tested for Akt1 expression by HA immunoblot analysis.

Supplementary Material

01

Supplementary Figure 1. Akt activity fluctuated during the cell cycle progression.

a. Live cell imaging of HeLa cells expressing an Akt activity reporter (EGFP-Akt AR, in green color) as well as the cell cycle indicator, Cdt1 (in red color, mCherry Cdt1) that is expressed mostly in the G1 phase to demonstrate that the periodic Akt activation inversely correlated with Cdt1 expression across cell cycle. HeLa cells stably expressing the cell cycle indicator Cdt1 were transiently transfected with the Akt activity reporter. 36 hours post-transfection, cells were subjected to live cell imaging. Images were taken every 30 minutes for 24 hours.

b. Akt phosphorylation fluctuated during cell cycle transitions in U2OS cells synchronized by double thymidine. Immunoblot (IB) analysis of whole cell lysates (WCL) derived from U2OS cells that were synchronized by double thymidine block for 24 hours and then released back to the cell cycle for the indicated time periods.

c. Akt phosphorylation fluctuated during cell cycle transitions in HeLa cells synchronized by nocodazole. Immunoblot (IB) analysis of whole cell lysates (WCL) derived from HeLa cells that were synchronized by nocodazole (330 nM) for 24 hours and then released back to the cell cycle for the indicated time periods.

Supplementary Figure 2. Cdk2/Cyclin A regulated Akt1 phosphorylation at both the S477 and T479 sites.

a. All Akt isoforms interacted with Cyclin A2 in cells. Immunoblot (IB) analysis of whole cell lysates (WCL) and HA-immunoprecipitates (IP) derived from HeLa cells transfected with the various indicated HA-Akt1-mutants. 48 hours post-transfection, cells were harvested in EBC buffer for further biochemical analysis.

b. Schematic presentation of the four putative Cyclin A binding motifs (RXL) within Akt1, Akt2 and Akt3.

c. In vitro kinase assays to measure Akt kinase activity. Specifically, the indicated HA-Akt1 kinases were HA-immunoprecipitated from transfected 293T cells and thoroughly washed and resuspended in EBC buffer plus 10% glycerol. Kinase activities were determined as its ability to phosphorylate the crosstide as described in the materials and methods section. All kinase activities (cpm) were normalized as % of WT readings. Experiments were done in triplicates and the error bars represent mean ± SD.

d-e. Depletion of endogenous Cyclin A2 or Cdk2 resulted in reduced Akt phosphorylation. IB of WCLs derived from HeLa cells depleted of endogenous Cyclin A2 (d) or Cdk2 (e).

f. Cdk2^−/− MEFs were deficient in Akt phosphorylation in response to insulin or IGF-1. Cdk2^+/+ and Cdk2^−/− MEFs were cultured in FBS-free medium for 12 hours followed by insulin stimulation (100 nM) for the indicated time periods before harvesting for immunoblot analysis.

Supplementary Figure 3. Overexpression of Cyclin A2 or depletion of the upstream E3 ligase for Cyclin A2, Cdh1, led to increased Akt phosphorylaiton and subsequently enhanced cellular growth advantages.

a. Stable expression of Cyclin A2 resulted in elevated Akt phosphorylation in HeLa cells. HeLa cells were infected with pBabe-puro-HA-Cyclin A2 or pBabe-puro-EV and selected in 1 μg/ml puromycin for 3 days to eliminate non-infected cells. The resulting cells were subjected to IB analysis.

b-c. Induced expression of Cyclin A2 resulted in elevated Akt phosphorylation in HeLa cells. HeLa cells were infected with pTRIPZ-puro-HA-Cyclin A2 or pTRIPZ-puro-EV and selected in 1 μg/ml puromycin for 3 days to eliminate non-infected cells. 3×10⁴ of the resulting cells were inoculated in 0.4% top soft agar and cultured for 21 days (b) before quantitative analysis (c).

d. Depletion of endogenous Cdh1 resulted in elevated Akt phosphorylation in MDA-MB-231 cells. MDA-MB-231 cells were infected with two independent shCdh1 viruses and selected in 250 ng/ml hygromycin for 3 days to eliminate non-infected cells. The resulting cells were subjected to IB analysis.

e. 3×10⁴ of the resulting cells from (d) were inoculated in 0.4% top soft agar and cultured for 23 days before quantitative analysis.

f-h. 3×10⁶ of the endogenous Cdh1-depleted MDA-MB-231 cells from (d) were injected into nude mice (n=10 for each group) and monitored for tumor formation (f). Formed tumors were dissected (g) and weighed (h). As indicated p values were calculated by student’s t-test.

i. Depletion of endogenous Cyclin A2 could attenuate Akt phosphorylation caused by Cdh1 depletion in MDA-MB-231 cells. Endogenous Cdh1-depleted MDA-MB-231 cells were infected with shCyclin A2 viruses and selected in 1μg/ml puromycin in combination with 250 ng/ml hygromycin containing medium for 3 days to eliminate non-infected cells. The resulting cells were subjected to IB analysis.

j-l. 4×10⁶ of MDA-MB-231 cells depleted of both endogenous Cyclin A2 and Cdh1 from (i) or depleted of only Cdh1 in (d) were injected into nude mice (n=10 for each group) and monitored for tumor formation (j). Formed tumors were dissected (k) and weighed (l). As indicated p values were calculated by student’s t-test.

Supplementary Figure 4. Cdk2/Cyclin A phosphorylated Akt1 in vitro at both the S477 and T479 sites located in the extreme C-terminus of Akt1.

a. Schematic presentation of the various GST-Akt1 truncation mutants generated to pinpoint the in vitro Cdk2/Cyclin A-dependent phosphorylation sites in human Akt1.

b. In vitro Cdk2/Cyclin A kinase assays with the indicated series of GST-Akt1 C-terminal region truncations to narrow down the major region within the C-terminus of Akt1 that can be phosphorylated by Cdk2/Cyclin A in vitro.

c. In vitro Cdk2/Cyclin A kinase assays with indicated recombinant GST-Akt1 truncations. d. Schematic illustration of the evolutionary conservation of S477 and T479 sites in Akt1.

e-f. In vitro kinase assays depicting major Cdk2/Cyclin A phosphorylation sites on Akt1 (e) or Akt2 (f). g. In vitro Cdk2/Cyclin A kinase assays using indicated GST-Akt1-G478P mutated proteins to indicate that P478 adjacent to S477 is not critical for Cdk2/Cyclin A mediated phosphorylation on S477 site.

h. In vitro Cdk2/Cyclin A kinase assays with the indicated series of GST-Akt1 C-terminal region tagged with GFP to illustrate that unlike S477 and T479 that are located in the extreme C-terminus of Akt1, an adjacent proline residue at +1 position is critical for Cdk2/Cyclin A-mediated phosphorylation when these sites are buried in a defined structure.

Supplementary Figure 5. Mass spectrometry analysis identified Cdk2/Cyclin A-mediated phosphorylation of Akt1 at S477 and T479 both in vivo and in vitro.

a. Representative mass spectrometry spectrum to map the Akt1 S477 phosphorylation status in vivo upon insulin stimulation. The LC-MS/MS spectrum of the doubly phosphorylated triply charged peptide RPHFPQFpSYSApSGTAGR representing S473 and S477 in the Akt1 sequence. The neutral loss of phosphate confirms the phosphorylation status and sites are localized to S8 (S473 full length) and S12 (S477 full length) in the peptide based on the b- ion series (N-terminal fragments) starting at b₁₂ and the y- ion series starting at y₆ that contain a phosphate group. In addition, the ion starting at b₁₆ contains two phosphate groups and loss of phosphate, defining the double phosphorylation event. The ion at y₁₅ also contains two phosphate groups and shows phosphate loss.

b. Representative mass spectrometry spectrum to map the Akt1 T479 phosphorylation status in vivo upon insulin stimulation. The LC-MS/MS spectrum of the singly phosphorylated doubly charged peptide RPHFPQFSYSASGpTA representing T479 in the modified AKT1 sequence. The neutral loss of phosphate confirms the phosphorylation status and sites are localized to T14 (T479 full length) in the peptide based on the b- ion series (N-terminal fragments) starting at b₁₄ and the y- ion series starting at y₂ that contain a phosphate group. In addition, the b ions through b₁₃ contains no phosphate groups.

c. GraphMod analysis (PBS software) shows that the phosphorylation site mapped in (b) is confidently localized to the C-terminal Thr479 site on the in vivo peptide RPHFPQFSYSASGpTA based on six different scoring parameters from a Sequest database search.

d. Representative mass spectrometry spectrum to map the Akt1 S479 phosphorylation status by Cdk2/Cyclin A in vitro.

Supplementary Figure 6. Validation of the generated Akt1-pS477/pT479 specific antibody.

a. Validation of the generated Akt1-pS477/pT479 antibody by overexpressing the indicated Akt1-WT and Akt1-S477A/T479A constructs in cells. Immunoblot (IB) analysis of whole cell lysates (WCL) and HA-immunoprecipitates (IP) derived from HeLa cells transfected with the indicated various HAAkt1-mutants. 48 hours post-transfection, cells were harvested in EBC buffer for further biochemical analysis to examine the specificity of the generated anti-Akt1-pS477/pT479 antibody.

b. Validation of the generated Akt1-pS477/pT479 antibody by phosphatase treatment of cell lysates derived from 293T cells overexpressing Akt1-WT or Akt1-S477A/T479A constructs. Immunoblot (IB) analysis of whole cell lysates (WCL) derived from HeLa cells transfected with the Akt1-WT construct and treated with λ phosphatase to examine whether the generated various anti-phospho-Akt1-tail antibodies could detect phosphorylation-dependent signals.

c. Protein sequence illustration of synthesized Akt1 tail peptides.

d-e. Titration of the indicated Akt tail peptides with various phosphorylation status revealed that the generated Akt1-pS477/pT479 antibody specifically recognizes the pS477/pT479 epitope at this experimental condition.

f. Peptide blocking assays to demonstrate the phospho-epitope specificity of the generated Akt1-pS477/pT479 antibody to detect in vivo Akt1 phosphorylation. Immunoblot (IB) analysis of whole cell lysates (WCL) derived from 293T cells transfected with the indicated various HA-Akt1 constructs, using the Akt1-pS477/pT479 antibody, in the presence of the indicated Akt1 tail blocking peptides.

Supplementary Figure 7. Validation of the generated Akt1-pS477 and Akt1-pT479 specific antibodies.

a-b. Validation of the generated Akt1-pS477 (a) and Akt1-pT479 (b) antibodies by overexpressing Akt1-WT and Akt1-S477A or -T479A constructs in cells. Immunoblot (IB) analysis of whole cell lysates (WCL) and HA-immunoprecipitates (IP) derived from HeLa cells transfected with the indicated various HA-Akt1-mutants. 48 hours post-transfection, cells were harvested in EBC buffer for further biochemical analysis to examine the specificity of the anti-Akt1-pS477 or anti-Akt1-pT479 antibody.

c. Validation of the generated Akt1-pS477 and Akt1-pT479 antibodies by phosphatase treatment of cell lystates derived from 293T cells overexpressing Akt1-WT and Akt1-S477A or -T479A constructs. Immunoblot (IB) analysis of whole cell lysates (WCL) derived from HeLa cells transfected with the Akt1-WT construct and treated with λ phosphatase to examine whether the generated various anti-phospho-Akt1-tail antibodies could detect phosphorylation-dependent signals.

d. Protein sequence illustration of synthesized Akt1 tail peptides.

e-f. Titration of the indicated Akt tail peptides with various phosphorylation status revealed that the generated Akt1-pS477 or Akt1-pT479 antibody specifically recognizes the pS477 or pT479 epitope under this experimental condition.

g-h. Peptide blocking assays demonstrated the phospho-epitope specificity of the generated Akt1-pS477 and Akt1-pT479 antibodies to detect in vivo Akt1 phosphorylation signals. Immunoblot (IB) analysis of whole cell lysates (WCL) derived from 293T cells transfected with the indicated various HA-Akt1 constructs, using the Akt1-pS477 or Akt1-pT479 antibody, in the presence of the indicated Akt1 tail blocking peptides.

Supplementary Figure 8. A statistically significant positive correlation between Cyclin A2 expression and Akt1-pS477/pT479 was observed in 50 breast cancer patient clinic samples.

Serial sections of tissue arrays 50 patient breast specimens were subjected to immunohistochemistry with anti-Cyclin A2 and anti-Akt1-pS477/pT479 antibodies, respectively, and visualized by the DAB staining before imaging. Scale bar represents 50 μm. Representative images were shown in (a), and the summary of the IHC results was listed in (b). Scale bar represents 50 μm.

Supplementary Figure 9. mTOR or DNAPK participated in the phosphorylation of Akt1 at the S477 and T479 sites in vitro.

a. Representative mass spectrometry spectrum to map the Akt1-T479 phosphorylation status by mTOR in vitro.

b. Cdk2/Cyclin A and mTOR are capable of phosphorylating GST-Akt-tail fusion protein in vitro. In vitro kinase assays on GST-Akt1-tail (amino acids 409-480) by active recombinant Cdk2/Cyclin A or mTOR, respectively.

c. Cdk2/Cyclin A and mTOR are capable of phosphorylating full length HA-Akt1 in vitro. Immunoblot (IB) analysis of in vitro Cdk2/Cyclin A and mTOR kinase assays with the indicated HA-immunoprecipitated full length Akt1 proteins from 293T cells to demonstrate that both of the kinases are able to phosphorylate the Akt1 S477/T479 sites in vitro. Notably, only mTOR but not Cdk2/Cyclin A could phosphorylate Akt1 S473 and T450 sites, both of which are well-characterized mTOR sites.

d. DNAPK might also be involved in mediating Akt1-S477/T479 phosphorylation in DNA damaging conditions. Immunoblot (IB) analysis of whole cell lysates (WCL) derived from DNAPKcs^+/+ and DNAPKcs^−/− MEFs cultured in FBS-free medium for 12 hours followed by insulin stimulation (100 nM) or doxorubicin treatment (5 μM) for 30 min before harvesting.

e. IB analysis of WCLs derived from primary foreskin fibroblast cells treated with etoposide (5 μM) for 2 hours in the presence or absence of DNAPK inhibitor Nu7026 (25 nM).

Supplementary Figure 10. PI3K/mTOR regulated Akt1-S477/T479 phosphorylation in vivo.

a. Treatment with PI3K or mTOR inhibitors led to reduced Akt-pS477/pT479. Immunoblot (IB) analysis of whole cell lysates (WCL) derived from HeLa cells that were treated with the indicated inhibitor compounds (pp242: 10 μM, LY2940002: 10 μM, Akt VIII: 1 μM, rapamycin: 20 nM and S6K1-I: 10 μM) for 6 hours before harvesting.

b. Ablation of mTORC2 activity resulted in attenuated Akt-pS477/pT479. Immunoblot (IB) analysis of whole cell lysates (WCL) derived from Rictor ^+/+ and Rictor ^−/− MEFs, Sin1 ^+/+ and Sin1 ^−/− MEFs, as well as HeLa cells depleted of endogenous Rictor by infection with shRNA constructs targeting Rictor (using shGFP as a negative control).

c-e. Depletion of PTEN, which augments the PI3K/Akt signaling pathway, led to elevated AktpS477/pT479. Immunoblot (IB) analysis of whole cell lysates (WCL) derived from Pten ^+/+ and Pten ^−/− MEFs (c), as well as FF (foreskin fibroblasts) (d) or U2OS (e) cells depleted of endogenous PTEN by infection with shRNA constructs targeting PTEN (using shGFP as a negative control).

f-g. Paraffin blocks of WT-MEFs for validation of IHC conditions for the Akt1-pS477/pT479 antibody. WT MEFs were serum starved overnight and stimulated with 5 ng/ml insulin in the presence or absence of LY2940002 (1 μM). Then cells were subjected to IHC (f) or IB (g) analysis.

h. IHC analysis of Akt-pS473 and Akt1-pS477/pT479 status in both Pten-WT and Pten hyteozygous breast tissues. Scale bar represents 25 μm under 40x magnification.

i. IB analysis of WCLs derived from the indicated mouse tissues. 21-day old male mice were sacrificed and the indicated mouse tissues were freshly harvested. Where indicated, a.p.: anterior prostate and adr.: adrenal gland.

Supplementary Figure 11. mTORC2 phosphorylated Akt1 at S477/T479 sites upon insulin stimulation.

a. Ablation of Rictor led to dramatically reduced Akt S477/T479 phosphorylation. Immunoblot (IB) analysis of Rictor^+/+ and Rictor^−/− MEFs cultured in FBS-free medium for 12 hours followed by insulin stimulation (100 nM) for the indicated time periods before harvesting.

b. mTOR inhibition by pp242 in Rictor^+/+ MEFs led to dramatically reduced Akt S477/T479 phosphorylation. Immunoblot (IB) analysis of Rictor^+/+ MEFs cultured in FBS-free medium for 12 hours followed by insulin stimulation (100 nM), in the presence or absence of the mTOR inhibitor, pp242 (10 μM) for the indicated time periods before harvesting.

c. PI3K inhibition by either LY2940002 or mTOR inhibition by pp242 in Cdk2^+/+ MEFs led to dramatically reduced Akt S477/T479 phosphorylation. Immunoblot (IB) analysis of Cdk2^+/+ MEFs cultured in FBS-free medium for 12 hours followed by insulin stimulation (100 nM), in the presence of the PI3K inhibitor, LY2940002 (50 μM) or mTOR inhibitor, pp242 (10 μM) for the indicated time periods before harvesting.

Supplementary Figure 12. Cdk2/Cyclin A phosphorylated Akt1 S477/T479 sites during cell cycle transitions.

a. Depletion of endogenous Rictor did not noticeably affect Akt1 S477/T479 phosphorylation across the cell cycle. Immunoblot (IB) analysis of whole cell lysates (WCL) derived from HeLa cells depleted of endogenous Rictor via shRNA lentiviral infection (using shGFP as a negative control) that were synchronized by nocodazole (330 nM) for 24 hours and then released back to normal cell cycle for the indicated time periods before harvesting.

b. Depletion of endogenous Cdk2 resulted in dramatically reduced and delayed Akt1 S477/T479 phosphorylation across the cell cycle. Immunoblot (IB) analysis of whole cell lysates (WCL) derived from HeLa cells depleted of endogenous Cdk2 via shRNA lentiviral infection (using shGFP as a negative control) that were synchronized by nocodazole (330 nM) for 24 hours and then released back to normal cell cycle for the indicated time periods before harvesting.

Supplementary Figure 13. Schematic representation of a model indicating that in response to different upstream stimuli, different kinases are responsible for Akt tail phosphorylation.

a. Schematic illustration of various upstream signaling pathways that could activate Akt in part by phosphorylating Ser477 and Thr479 located in the extreme C-terminus of Akt. During cell cycle progression, Cdk2/Cyclin A is relatively more active to phosphorylate Akt1 at S477/T479; upon DNA damage, DNAPK is activated to trigger Akt1 tail phosphorylation; while under insulin or growth factor stimulation condition, mTORC2 may be the major active kinase to promote Akt1 activation via directly phosphorylating Akt1-S477/T479.

b. A table presenting critical Akt1 phosphorylation sites with corresponding identified upstream kinases. Notably, unlike mTOR that phosphorylates Akt1 T450, S473 and S477/T479, Cdk2/Cyclin A only mediates phosphorylation of S477/T479.

Supplementary Figure 14. Both Cdk2/Cyclin A and mTOR regulated Akt1-S477/T479 phosphorylation in primary foreskin fibroblast cells.

a. Treatment with Cdk2 inhibitors roscovitine or mimosine, PI3K inhibitor LY2940002 or mTOR inhibitor pp242 led to reduced Akt-pS477/pT479. Immunoblot (IB) analysis of whole cell lysates (WCL) derived from primary foreskin fibroblast cells that were treated with the indicated inhibitor compounds (pp242: 10 μM, LY2940002: 10 μM, roscovitine: 40 nM and mimosine: as indicated) for 18 hours before harvesting.

b-c. Depletion of Cdk2 (b), Cyclin A (b) or Rictor (c) led to elevated Akt-pS477/pT479. Immunoblot (IB) analysis of whole cell lysates (WCL) derived from primary foreskin fibroblast cells depleted of endogenous PTEN by infection with shRNA constructs targeting PTEN (using shGFP as a negative control).

Supplementary Figure 15. Akt1 tail phosphorylation facilitated Akt1 S473 phosphorylation to enhance Akt1 activation.

a-b. Akt1 tail phosphorylation facilitated Akt1 S473 phosphorylation. Immunoblot (IB) analysis of whole cell lysates (WCL) and HA-immunoprecipitates (IP) derived from HeLa cells transfected with the indicated Akt1 mutants.

c. In vitro kinase assay of indicated affinity-purified Akt1 kinases from 293T cells with recombinant GST-Skp2.

Supplementary Figure 16. Akt2 or Akt3 tail phosphorylation promoted Akt hydrophobic motif phosphorylation to modulate Akt activation, and phospho-mimetic DE mutant could partially rescue the loss of Cyclin A2.

a. Protein sequence alignment of the tail region of human Akt1, Akt2 and Akt3.

b-c. Akt2 or Akt3 tail phosphorylation promoted Akt hydrophobic motif phosphorylation to modulate Akt activation. Immunoblot (IB) analysis of whole cell lysates (WCL) and HA-immunoprecipitates (IP) derived from HeLa cells transfected with the indicated Akt2 or Akt3 constructs.

d. 3×10⁴ of HeLa cells expressing the indicated Akt mutants that were previously depleted of endogenous Akt1 were inoculated in 0.4% top soft agar and cultured for 24 days.

e. Akt1-R76A and R76A/DE mutants were not deficient in PIP3 interaction. The indicated Akt1 mutants were transfected into 293T cells and cells were harvested 48 hours post transfection. WCLs were obtained and subjected to pull down assays using PI(3,4,5)P3-conjugated agarose beads.

f. Quadruple knockout MEFs (QMEFs, Cyclin E1^−/−/E2^−/−/A1^−/−/A2^f/f) were infected with the indicated Akt mutant viruses and selected in 250 ng/ml hygromycin for 3 days to eliminate non-infected cells. Afterwards, generated stable cell lines were further infected with phage-Cre (EV as a negative control) and selected in 1 μg/ml puromycin for 3 days for eliminating non-infected cells. The cells. The resulting cells were subjected to BrdU labeling assay for evaluation of the cell cycle transition and cell proliferation status. Experiments were performed in triplicates and the error bars represent mean ± SD.

Supplementary Figure 17. Akt1 tail phosphorylation facilitated Akt1 S473 phosphorylation to enhance Akt1 activation.

a. Akt1 tail phosphorylation enhanced Akt S473 phosphorylation induced by insulin. Akt1-depleted HeLa cells transiently transfected with the indicated Akt1 constructs (with an empty vector as a negative control) were cultured in FBS-free medium for 12 hours followed by insulin stimulation (100 nM) for the indicated time periods before harvesting for immunoblot analysis.

b. Akt tail phosphorylation greatly promoted Akt1 S473 phosphorylation under different stimulation conditions. Akt1-depleted HeLa cells were infected with various retroviral constructs expressing Akt1-WT, AA or DE mutant (using an empty vector as a negative control) and selected with 1 μg/ml puromycin together with 250 ng/ml hygromycin for 72 hours to eliminate non-infected cells. The resulting cells were serum starved for 12 hours and treated with indicated stimuli (insulin: 100 nM; IGF-1: 100 ng/ml and PDGF: 100 ng/ml) for 30 minutes or (EGF: 100 ng/ml) 10 minutes before harvesting for immunoblot analysis.

c. The cell lines generated in (b) were serum starved for 12 hours and treated with indicated stimuli (insulin: 100 nM and doxorubicin: 5 μM for 30 minutes before harvesting for immunoblot analysis.

d. Endogenous Akt1 depleted HeLa cells were infected with the indicated pBabe-Myr-Akt1 viruses and selected in 250 ng/ml hygromycin to eliminate non-infected cells. Afterwards, each generated cell line was further infected with shCyclin A2 viruses and 3 days after selection in 1 μg/ml puromycin, the resulting cells were subjected to IB analysis.

e-g. 3×10⁶ of the generated cells in (d) were injected into nude mice (n=10 for each group) and monitored for tumor formation (e). Formed tumors were dissected (f) and weighed (g). As indicated, * represents p<0.05 calculated by student’s t-test.

Supplementary Figure 18. Structural illustration of the stabilization of closed Akt active conformation by Akt tail phosphorylation.

a. Protein sequence alignment of the carboxyl-terminal regions of Akt1 and Akt2 with PIFtide.

b. Akt2 tail (S478) phosphorylation together with S474 phosphorylation accelerated Akt2 kinase kinetics to enhance Akt2 activity towards phosphorylating crosstide. Dose response curve for the stimulation of pΔPH-Akt2-ΔC kinase activity by various Akt2 HM peptides.

c-d. A structural illustration of the stabilization of closed Akt active conformation by Akt tail phosphorylation. As shown in (c) (adapted from PDB 1OK6), when Akt2-S478 is not phosphorylated, no interaction between residues S473 and R208; when the WT-Akt2 tail was substituted with a PIFtide containing a S to D mutation at the position equals to Akt2-S478 (d), this S478D mutation creates a charge-charge interaction between D478 and R208 side chains to stablize the Akt2 in the active form, thus promoting Akt2 activation.

e. IB analysis of WCLs derived from 293T cells transfected with the indicated Akt1 constructs.

f. The indicated Akt mutants were transfected into HeLa cells and HA immunoprecipitation (IP) was performed to pull down various Akt mutants. Equal amount (~1 μg) of each Akt1 mutant IPs were used in the in vitro kinase assays with crosstides. Reactions were incubated for the indicated time periods before termination for measuring the ³²p-ATP incorporation.

Supplementary Figure 19. Akt1 tail phosphorylation promoted the interaction of Akt1 with mTORC2, to facilitate S473 phosphorylation.

a-b. mTOR prefers GST-Akt1-tail phosphorylated substrates for phosphorylation on S473. In vitro kinase assay (a) or immunoblot (IB) analysis (b) of the in vitro mTOR kinase assay products with the various indicated GST-Akt1 recombinant proteins.

c. Akt1-S477D/T479E mutant facilitated its binding with mTOR in cells. Immunoblot (IB) analysis of whole cell lysates (WCL) and HA-immunoprecipitates derived from Akt1-depleted HeLa cells that were transfected with the indicated Akt1 constructs and Flag-mTOR.

d. Akt1 tail phosphorylation did not significantly change Akt’s affinity with endogenous Akt1 inhibitors PHLPP2 and CTMP. Immunoblot (IB) analysis of whole cell lysates (WCL) and HA-immunoprecipitates (IP) derived from HeLa cells transfected with the indicated Akt1 constructs.

Supplementary Figure 20. Akt1 tail phosphorylation might not prefer a pre-phosphorylation occurred on S473.

a. Cdk2/Cyclin A in vitro kinase assays with various indicated GST-Akt1 mutants.

b. Quantification illustration of the mTOR in vitro kinase assays with the degenerate peptide libraries. The indicated amino acid was fixed in all peptide in the libraries with surrounding degenerate sequences. mTOR kinase assays were performed to test the effects of the fixed residue on the phosphorylation of a serine residue designated at 0 position. Experiments were performed in triplicates and the error bars represent mean ± SD.

Supplementary Figure 21. Akt1 tail phosphorylation may functionally compensate for S473 phosphorylation to activate Akt1.

a. Deficiency of Akt1-T308 phosphorylation but not Akt1-S473 phosphorylation could further attenuate phosphorylation of Skp2 in cells. Immunoblot (IB) analysis of whole cell lysates (WCL) and Flag-immunoprecipitates (IP) derived from Akt1-depleted HeLa cells that were transfected with the indicated Akt1 constructs together with Flag-Skp2.

b. In vitro kinase assay to measure Akt1 kinase activity. Specifically, indicated HA-Akt1 kinases were HA-immunoprecipitated from transfected 293T cells and thoroughly washed and resuspended in EBC plus 10% glycerol. Kinase activities were determined as its ability to phosphorylate crosstide as described in the materials and methods section. All kinase activities (cpm) were normalized as % of WT readings. Experiments were performed in triplicates and the error bars represent mean ± SD.

c. Depletion of Rictor led to attenuated Akt1-S473 phosphorylation in WT, but not S477D/T479E-Akt1 mutant. Immunoblot (IB) analysis of whole cell lysates (WCL) and HA-immunoprecipitates (IP) derived from HeLa cells depleted of Rictor transfected with the indicated Akt1 constructs (shGFP as a negative control).

d. Akt1-S477D/T479E mutant phosphorylated Skp2 to a similar level independent of Rictor. Immunoblot (IB) analysis of whole cell lysates (WCL) and Flag-immunoprecipitates (IP) derived from HeLa cells depleted of endogenous Rictor via the shRNA lentiviral infection (using shGFP as a negative control) that were transfected with the indicated Akt1 constructs with Flag-Skp2.

e. Akt1-S477D/T479E mutant partially compensated for the loss of Akt1 S473 phosphorylation towards phosphorylating Skp2 in cells. Immunoblot (IB) analysis of whole cell lysates (WCL) and Flag-immunoprecipitates derived from Akt1-depleted HeLa cells that were transfected with the indicated Akt1 constructs and Flag-Skp2.

f. In vitro kinase assay to measure Akt1 kinase activity. Specifically, indicated HA-Akt1 kinases were HA-immunoprecipitated from transfected 293T cells and thoroughly washed and resuspended in EBC plus 10% glycerol. Kinase activities were determined as its ability to phosphorylate crosstide as described in the materials and methods section. All kinase activities (cpm) were normalized as % of WT readings. Experiments were performed in triplicates and the error bars represent mean ± SD.

Supplementary Figure 22. Akt1 tail phosphorylation contributed to elevated oncogenic activity in multiple physiological settings.

a. Characterizing the various HeLa cell lines stably expressing the indicated Akt1 constructs. Akt1-depleted HeLa cells were infected with the various indicated retroviral constructs expressing Akt1-WT, AA or DE mutant (using an empty vector as a negative control) and selected with 1 μg/ml puromycin together with 250 ng/ml hygromycin for 72 hours to eliminate non-infected cells. The resulting cells were subjected to immunoblot analysis.

b-c. Akt1 tail phosphorylation enhanced S phase entry as evidenced by an increase in BrdU incorporation. Cell lines generated in (a) were subjected to BrdU labeling assay for evaluation of the cell cycle transition and cell proliferation status. Experiments were performed in triplicates and the error bars represent mean ± SD.

d-e. Akt1 tail phosphorylation conferred cellular resistance to various chemotherapeutic agents. The various HeLa cell lines generated in (a) were cultured in 10% FBS-containing medium with the indicated concentrations of etoposide (d) or CPT (e) for 48 hours before performing the cell viability assays. Data was shown as mean ± SD for three independent experiments. * indicates p < 0.05 (student’s t-test). CPT: Camptothecin.

Supplementary Figure 23. Akt1 tail phosphorylation promoted in vivo tumor growth in a xenograft mouse model.

Representative images of the xenograft experiments described in Fig. 4f-g illustrating the tumor sizes at 15 days post-injection (a), as well as the direct comparison of the dissected tumors between the four experimental groups (b).

Supplementary Figure 24. Akt1 tail deletion promoted in vitro anchorage-independent cell growth and in vivo tumor growth in a xenograft mouse model.

a. IB analysis of WCLs and Flag IPs derived from 293T cells transfected with the indicated constructs.

b. Endogenous Akt1 depleted HeLa cells were infected with pBabe-Hygro-Myr-Akt1-WT or -476Δ viruses and selected in 250 ng/ml hygromycin for 3 days to eliminate non-infected cells. 3×10⁴ of the resulting cells were inoculated in 0.4% top soft agar and cultured for 20 days before quantitative analysis.

c-f. The cell lines generated in (b) were injected into nude mice (n=10 for each group) and monitored for tumor formation (c-d). Formed tumors were dissected (e) and weighed (f). As indicated, * represents p<0.05 calculated by student’s t-test.

Supplementary Figure 25. Akt1 tail phosphorylation partially protected mouse ES cells from apoptotic stress caused by Cyclin A2 deletion.

Representative FACS analysis of the apoptotic levels in mouse ES cells stably expressing the indicated Akt mutants (EV, AA-Akt1 and DE-Akt1) upon 2 μg/ml tamoxifen treatment to deplete Cyclin A2 alleles for the indicated time periods. a, EV, without tamoxifen; b, DE-Akt1, without tamoxifen; c, AA-Akt1, without tamoxifen; d, EV-Akt1, 2 μg/ml tamoxifen treatment for 4 days; e, DE-Akt1, 2 μg/ml tamoxifen treatment for 4 days; f, AA-Akt1, 2 μg/ml tamoxifen treatment for 4 days.

Supplementary Figure 26. Akt1 tail phosphorylation largely correlated with S473 phosphorylation in breast cancer patient samples and breast cancer derived cell lines.

a-b. Akt1 tail phosphorylation status (pS477/pT479) positively correlated with Akt pS473 in normal human breast cancer tissues (a) and breast cancer patient breast tissues (b). Serial sections of tissue arrays of 30 normal breast specimens and 80 patient breast specimens were subjected to immunohistochemistry with anti-Akt1-pS473 and anti-Akt1-pS477/pT479 antibodies, respectively, and visualized by the DAB staining before imaging. Scale bar represents 25 μm.

c. Pie chart illustration of the relative distribution of the high levels of Akt1 S477/T479 and S473 phosphorylation status in all 80 human breast cancer patient samples. Notably, high levels of both Akt1 pS477/pT479 and pS473 were observed in 49% samples and 10% samples showed only high levels of pS473. Importantly, 19% samples beared only high levels of pS477/pT479, suggesting pS477/pT479 might depict a non-overlapping portion of breast cancers than pS473.

d. Bar graph illustration of the relative distribution of levels of Akt1 pS477/pT479 and pS473 at different breast cancer developmental stages in 80 human breast cancer patient samples examined.

e. Akt1 tail phosphorylation largely correlated with S473 phosphorylation in the panel of examined breast cancer derived cell lines. Immunoblot (IB) analysis of whole cell lysates (WCL) derived from a panel of breast cancer-derived cell lines.