Phosphorylation-dependent regulation cytosolic localization and oncogenic function of Skp2 by Akt/PKB

doi:10.1038/ncb1849

Nat Cell Biol. Author manuscript; available in PMC 2010 March 2.

Published in final edited form as:

Nat Cell Biol. 2009 April; 11(4): 420–432.

Published online 2009 March 8. doi: 10.1038/ncb1849

PMCID: PMC2830812

NIHMSID: NIHMS171642

Phosphorylation-dependent regulation cytosolic localization and oncogenic function of Skp2 by Akt/PKB

Hui-Kuan Lin,^2,^3,⁸ Guocan Wang,^1,² Zhenbang Chen,^1,^2,⁴ Julie Teruya-Feldstein,² Yan Liu,⁵ Chia-Hsin Chan,³ Wei-Lei Yang,³ Hediye Erdjument-Bromage,⁶ Keiichi I. Nakayama,⁷ Stephen Nimer,⁵ Paul Tempst,⁶ and Pier Paolo Pandolfi^1,^2,⁸

¹Cancer Genetics Program, Beth Israel Deaconess Cancer Center and Department of Medicine and Pathology, Beth Israel Deaconess Medical Center, Harvard Medical School, 330 Brookline Avenue, Boston, MA 02215, USA

²Cancer Biology and Genetics Program, Department of Pathology, Sloan-Kettering Institute, Memorial Sloan-Kettering Cancer Center, 1275 York Avenue, New York, NY, 10021, USA

³Department of Molecular and Cellular Oncology, The University of Texas M.D. Anderson Cancer Center, Houston, TX 77030, USA

⁵Molecular Pharmacology and Chemistry Program, Sloan-Kettering Institute, Memorial Sloan-Kettering Cancer Center, 1275 York Avenue, New York, NY, 10021, USA

⁶Molecular Biology Program, Sloan-Kettering Institute, Memorial Sloan-Kettering Cancer Center, 1275 York Avenue, New York, NY, 10021, USA

⁷Department of Molecular and Cellular Biology, Medical Institute of Bioregulation, Kyushu University, Fukuoka, Fukuoka 812-8582, Japan

⁸ Correspondence should be addressed to Pier Paolo Pandolfi: Email: ppandolf/at/bidmc.harvard.edu and Hui-Kuan Lin: Email: hklin/at/mdanderson.org

⁴Present address: Department of Cancer Biology, Meharry Medical College, 1005 Dr. D. B. Todd Jr. Blvd., Nashville, TN 37067-3599, USA.

The publisher's final edited version of this article is available at Nat Cell Biol

See other articles in PMC that cite the published article.

Abstract

Skp2 is an F-box protein that forms the SCF complex with Skp1 and Cullin-1 to constitute an E3 ligase for ubiquitylation. Ubiquitylation and degradation of the p27 is critical for Skp2-mediated cell cycle entry, and overexpression and cytosolic accumulation of Skp2 have been clearly associated with tumorigenesis although the functional significance of the latter has remained elusive. Here we show that the Akt/PKB interacts with and directly phosphorylates Skp2. We find that Skp2 phosphorylation by Akt triggers SCF complex formation and E3 ligase activity. Importantly, a phosphorylation-defective Skp2 mutant is drastically impaired in its ability to promote cell proliferation and tumorigenesis. Furthermore, we show that Akt-mediated phosphorylation triggers 14-3-3-β-dependent Skp2 relocalization to the cytosol, and we attribute a specific role to cytosolic Skp2 in the positive regulation of cell migration. Finally, we demonstrate that high levels of Akt activation correlate with Skp2 cytosolic accumulation in human cancer specimens. Our results therefore define a novel proto-oncogenic Akt/PKB-dependent signaling pathway.

The ubiquitin-proteasome system regulates the cell cycle through control of protein ubiquitylation and degradation¹^,². One of the key ubiquitin ligases (E3 ligase) in this process is the Skp1/Cul-1/F-box (SCF) complex, which consists of Skp1, Cullin-1 (Cul-1), RBX1, as well as an F-box protein, all required for its E3 ubiquitin ligase activity. Disruption of this complex severely ablates its enzymatic activity¹^,².

Skp2 (S-phase kinase associated protein-2) is a SCF F-box protein and is responsible for substrate recognition¹^,². It binds to p27 and targets it for ubiquitylation and degradation³^-⁵. Overexpression of Skp2 induces cell cycle entry, and the degradation of p27 is required for Skp2-mediated cell cycle progression⁶^,⁷. Skp2 deficiency displays elevated p27 protein levels and a profound impairment in proliferation accompanied by nuclear enlargement, polypoidy, and centrosome overduplication⁸^,⁹. Overexpression of Skp2 is frequently observed in human cancers of diverse histology, while in most human cancers reduced level of p27 represents an adverse prognostic marker¹^,². Skp2 cooperates with H-Ras^G12V to transform primary rodent fibroblasts¹⁰. Overexpression of Skp2 in the T-cell compartment cooperates with N-Ras to induce T cell lymphomas¹¹, while prostate specific expression of Skp2 leads to prostatic intraepithelial neoplasia (PIN)¹². These observations suggest that Skp2 overexpression may contribute to tumorigenesis.

Although substantial advances have been made in understanding the mechanisms that control its levels of expression, by contrast, the molecular mechanisms by which Skp2 activity within the SCF complex and its subcellular localization are regulated are currently unknown. This is of further relevance as, in human cancer, Skp2 is frequently found aberrantly localized in the cytosol. Here we demonstrate that phosphorylation of Skp2 by Akt/PKB constitutes a molecular switch that critically controls Skp2 SCF complex formation, localization and function.

Results

Akt/PKB interacts with and phosphorylates Skp2

Skp2 is phosphorylated during G1/S transition¹^,²^,¹³. Mitogens, such as epidermal growth factor (EGF), can also lead to Skp2 phosphorylation¹⁴. However, the functional relevance of this phosphorylation event is unclear and the kinases that execute it are still unknown. Since EGF can activate both phosphatidylinositol 3-kinase (PI3K)/Akt and mitogen activating protein kinase (MAPK) pathways, we speculated that Skp2 might be the phosphorylation target of one of these two pathways. We therefore tested whether Akt/PKB might be a Skp2 kinase. Skp2 was found to interact with Akt1 in reciprocal co-immunoprecipitation experiments (Fig. 1a-c). Interestingly, the interaction between endogenous Skp2 and Akt1 was detected in the presence of Insulin-like growth factor-1 (IGF-1) while the interaction was abolished by PI3K inhibitor LY294002 (LY), suggesting that Akt activity might favor the formation of the Akt/Skp2 complex (Fig. 1d). In support of this notion, we found that Akt1 kinase dead mutant (K179A) interacted with exogenous Skp2 much less effectively than the constitutive active Akt1 (data not shown). In glutathione S-transferase (GST)-pull down assays, Akt1 was able to interact with Skp2 directly (Fig. 1f).

Figure 1

Skp2 interacts with Akt

We next determined whether Skp2 was an in vitro substrate for Akt1. Skp2 was readily phosphorylated by recombinant active Akt1 (Fig. 2a). Skp2 phosphorylation by Akt1 was comparable to the phosphorylation of the TSC2 by Akt1, a well-known Akt substrate (Supplementary information Fig. S1b)¹⁵^-¹⁸. Using the Scansite program [http://scansite.mit.edu; ¹⁹] analysis, we found that Skp2 Ser (S) 72 lies within an Akt consensus site [(RXRXXS/T, where X is any amino acid)] identified at “medium stringency”, which is conserved from rat to human (Fig. 2b). To determine whether S72 is a site for Akt-mediated Skp2 phosphorylation, we mutated this residue from serine to alanine (S72A) and used this Skp2 mutant in in vitro kinase assays. Indeed, Akt-mediated phosphorylation of Skp2 S72A was markedly reduced (Fig. 2c, even though Skp2 S72A still interacted with Akt as efficiently as wild-type (wt) Skp2 (Fig. 1e). Similarly, in vivo phospho-labeling experiments performed in transfected cells with a constitutively active form of Akt1 (Mri-Akt) revealed phosphorylation of wt Skp2, but not of Skp2 S72A (Fig. 2d). We further confirmed that Akt1 induced Skp2, but not Skp2 S72A phosphorylation in vivo, by using a phospho-Akt substrate antibody (Fig. 2e).

Figure 2

Akt/PKB phosphorylates Skp2 at S72 in vitro and in vivo

In order to verify whether S72 is phosphorylated by Akt, we next performed mass spectrometry analysis. In vivo phosphorylated Skp2 (Skp2-P) and unphosphorylated Skp2 (Skp2-C) was isolated from 293T cells, digested with trypsin, and analyzed by MALDI-reTOF mass spectrometry (Figs. 2f and Supplementary information, Fig. S1a). Peptide patterns were then compared for differences. One m/z peak, at 1671.88 atomic mass units (amu), was observed in the spectra of “Skp2-P” that was absent from “Skp2-C.” The m/z value mapped to the predicted, monophosphorylated fragment of the Skp2 sequence (LKS⁷²KGS⁷⁵DKDFVIVR) with monoisotopic (¹²C) mass discrepancies of less than 40 ppm. Next, the same peptide was selectively retrieved by affinity metal chromatography²⁰ and reanalyzed by MALDI-TOF/TOF MS/MS sequencing. The presence of unique fragment ions confirmed the identity and the monophosphorylation state (characteristic loss of 98 amu) for the peptide and allowed us to narrow down the site of phosphorylation to either S72 or S75. In contrast, in similar experimental conditions, the phosphorylated peptide at 1671.88 amu was not identified when analyzing a Skp2 S72A mutant (designed as “S72A-P”), while the non-phosphorylated mutant peptide was detected, strongly suggesting that S72 is the site for Akt-mediated Skp2 phosphorylation in vivo. Similar results were also obtained when using Skp2 phosphorylated in vitro by recombinant Akt1 (data not shown and Supplementary information, Fig. S1c, d).

Lastly, to further demonstrate Akt-dependent Skp2 phosphorylation at S72 in vivo, we generated a phospho (S72)-Skp2 specific antibody. Once again, we found that Mri-Akt could phosphorylate Skp2, but not Skp2 S72A using this antibody (Fig. 2g). Importantly, in PC-3 prostate cancer cells, where Akt is constitutively active, endogenous Skp2 was phosphorylated at S72 (Fig. 2h), while LY drastically reduced its phosphorylation (Fig. 2h).

Since the p70 S6 kinase (S6K) is activated by Akt, we next determined whether S6K could mediate Akt-induced Skp2 S72 phosphorylation. Rapamycin, a well-known potent inhibitor of mTOR, abrogated S6K activity towards S6, while Skp2 S72 phosphorylation was unaffected (Fig. 2i), suggesting that S6K is not involved in Akt-dependent Skp2 S72 phosphorylation.

As aforementioned, EGF is known to induce Skp2 phosphorylation¹⁴. We therefore determined whether EGF would induce Skp2 phosphorylation at S72. EGF treatment induced Akt activation and S72 Skp2 phosphorylation, which was inhibited by Wortmannin (WN), but not by MEK1 inhibitor U0126, suggesting that EGF mediated Skp2 S72 phosphorylation occurs through an Akt-dependent pathway independently of MAPK-RSK activation (Fig. 2j). These results provide strong support for direct Skp2 phosphorylation at S72 by Akt/PKB both in vitro and in vivo.

S72 phosphorylation is critical for Skp2 biochemical and proto-oncogenic functions

We next determined whether Skp2 S72 phosphorylation modulates its E3 ligase activity toward p27. Wt Skp2 and the phosphomimetic Skp2 S72D mutant readily promoted endogenous and exogenous p27 ubiquitylation, whereas Skp2 S72A was profoundly impaired in this function (Fig. 3a and Supplementary information, Fig. S2a). Although p27 phosphorylation at T187 is proposed as an important event for the interaction of p27 with the Skp2 SCF complex¹, we found that, surprisingly, Skp2-mediated ubiquitylation of wt-p27 and p27-T187A was comparable (Supplementary information, Fig. S3a). Likewise, the interaction of Skp2 with wt-p27 and p27-T187A was also comparable (Supplementary information, Fig. S3b) consistent with the results from the Roberts group²¹ showing that the interaction of p27 with Skp2 regulated by p27 T187 phosphorylation is cell type-dependent.

Figure 3

Phosphorylation of Skp2 is required for Skp2 E3 ligase activity and function

Conversely, while wt Skp2 and Skp2 S72D promoted p27 degradation, Skp2 S72A markedly reduced this activity (Fig. 3b and Supplementary information, Fig. S2b). It should be noted that within this time frame there was no significant change in cell cycle distribution, suggesting that p27 degradation is not caused by varied cell cycle status (Supplementary information, Fig. S2c). The inability of Skp2 S72A in promoting p27 ubiquitylation and degradation was also not due to its inability to interact with p27 or a reduction of its expression level (Fig. 3b and Supplementary information, Fig. S2d). Furthermore, LY or WN suppressed Skp2-mediated p27 ubiquitylation (Fig. 3a). These results suggest Akt activity and Skp2 S72 phosphorylation promotes Skp2 E3 ligase activity.

Skp2 overexpression is known to promote cell proliferation by triggering p27 degradation⁶^,⁷. We therefore determined whether S72 phosphorylation regulates Skp2-mediated cell proliferation. Both Skp2 and the S72D mutant promoted cells in S-phase, while Skp2 S72A was markedly impaired in this activity in 293T cells (Fig. 3c). Similar results were also obtained in COS-1 cells (Supplementary information, Fig. S3c). LY or WN suppressed Skp2-mediated an increase in S-phase cells (Fig. 3c).

To study the role of Skp2 and Skp2 mutants in tumorigenesis, we generated LNCaP prostate cancer cells with stable overexpression of Skp2 or Skp2 mutants for tumorigenic potential. Overexpression of wt and Skp2 S72D but not Skp2 S72A markedly enhanced LNCaP proliferation and reduced p27 protein levels (Fig. 3d, e). While Skp2 and Skp2 S72D profoundly promoted tumorigenesis when compared to the mock control, Skp2 S72A completely lost this activity (Fig. 3f, g).

Skp2 phosphorylation regulates SCF complex formation and activity

Since the integrity of Skp2 SCF complex is critical for Skp2 E3 ligase activity, we next determined whether activation of the PI3K/Akt pathway could regulate the formation of the Skp2-SCF complex. Mri-Akt enhanced the interaction of exogenous Skp2 with endogenous Skp1 and Cul-1, whereas WN suppressed the formation of the Skp2/Skp1 complex (Fig. 4a). In this cell culture condition, we did not observe an appreciable change in S-phase by either Akt activation or inhibition (Supplementary information, Fig. S4), suggesting that the effect of Akt on Skp2 SCF complex formation is not due to changes in cell cycle distribution.

Figure 4

Akt-mediated Skp2 phosphorylation regulates SCF complex formation

Similarly, suppression of the PI3K/Akt pathway profoundly impaired endogenous Skp2-SCF complex formation, while we observed a slight increase in the formation of this complex upon Akt activation (Fig. 4b). In agreement with these findings, wt Skp2 and Skp2 S72D readily complexed with endogenous Skp1 and Cul-1, while Skp2 S72A significantly reduced its interaction with Skp1 and Cul-1 (Fig. 4c).

The Skp2 SCF complex isolated from cells transfected with wt Skp2 or its mutants was used for in vitro p27 ubiquitylation assay. Skp2-SCF complex readily promoted p27 ubiquitylation (Supplementary Fig. 5a, middle panel). Importantly, the Skp2 SCF complex failed to induce p27 ubiquitylation unless p27 was phosphorylated by the Cyclin A/CDK2 complex prior to this event (Supplementary Fig. 5a, right panel). Skp2 S72A ability to complex with Skp1 and Cul-1 and promote p27 ubiquitylation was markedly impaired when compared to wt Skp2 and Skp2 S72D mutant (Supplementary Fig. 5a). Furthermore, Mri-Akt significantly increased Skp2-SCF complex formation and p27 ubiquitylation, while LY reduced this effect (Supplementary Fig. 5b). We further demonstrated that the SCF complex contained phospho (S72) Skp2 (Fig. 4d, e), while alkaline phosphatase (CIP) induced Skp2 S72 dephosphorylation and disrupted the Skp2 and Skp1 interaction (Fig. 4f). In addition, the lysates with Skp1 depletion reduced Skp2 S72 phosphorylation and Cul-1 (Lane 2 versus lane 1) (Fig. 4e). GST pull down assays demonstrated that while Mri-Akt, which induced Skp2 S72 phosphorylation (Fig. 2d, g) profoundly increased Skp1 and Skp2 interaction in vitro, Skp2 S72A dramatically reduced its interaction with Skp1 (Fig. 4g). These results strongly suggest that Akt positively regulates Skp2-SCF complex formation, therefore enhancing it E3 ligase activity through Skp2 S72 phosphorylation.

Akt-mediated phosphorylation induces Skp2 cytosolic localization

Skp2 normally resides in the nucleus; however, cytosolic relocalization of Skp2 has been observed in several human cancers²²^-²⁵. However, the molecular mechanisms underlying this phenomenon and its biological consequence are currently unknown. Loss or mutations of the tumor suppressor PTEN is frequently observed in human cancers, resulting in marked Akt activation²⁶^,²⁷. Akt/PKB may represent a potential candidate for triggering Skp2 cytosolic relocalization given that phosphorylation of Akt substrates by Akt often results in change in protein localization²⁸^,²⁹. Immunofluorescence (IF) revealed that although Skp2 was localized in the nucleus, Mri-Akt induced Skp2 cytosolic relocalization in 293T cells (approximately 40% of cells). In contrast, this effect was markedly reduced with Skp2 S72A (Fig. 5a and Supplementary information, Fig. S6a). Importantly, Skp2 S72D relocalized to the cytosol even without Mri-Akt, but Mri-Akt did not further enhance the accumulation of S72D in the cytosol (Fig. 5a, Supplementary information, Fig. S6a and data not shown), suggesting Skp2 S72 phosphorylation regulates Skp2 cytosolic relocalization. We further demonstrated that Akt activation upon IGF-1 or Mri-Akt induced endogenous Skp2 cytosolic relocalization, as determined by IF and fractionation/Western blot analyses (Fig. 5b, c, f, and Supplementary information, Fig. S6b). Importantly, we found Skp2 S72 phosphorylation was predominantly detected in the cytosol, but not in the nucleus (Fig. 5d).

Fig. 5

Skp2 phosphorylation by Akt regulates Skp2 cytosolic localization

We next determined the kinetic of Skp2 S72 phosphorylation and cytosolic Skp2 localization upon IGF-1 stimulation. IGF-1 induced Akt activation and Skp2 S72 phosphorylation at 30 min, which lasted for several hours (Fig. 5e). Strikingly, WN markedly suppressed IGF-1-induced Skp2 S72 phosphorylation (Fig. 5e). Skp2 cytosolic accumulation already occurred at 1 hour after IGF-1 treatment, and was significantly blocked by WN (Fig. 5f). It should be noted that within this time frame there was no significant change in cell cycle distribution upon IGF-1 treatment (Supplementary information, Fig. S7), demonstrating that Skp2 S72 phosphorylation at S72 and cytosolic localization preceded cell cycle entry. We also obtained similar results in IMR90 cells. One hour after IGF treatment, up to 75-80% of cells displayed cytosolic Skp2 accumulation although the cell cycle distribution was minimally affected (Supplementary information, Fig. S8a, b, and d). Once again IGF treatment led to Akt activation accompanied by Skp2 S72 phosphorylation, which was abrogated by WN (Supplementary information, Fig. S8c). Our results therefore suggest that Akt induces Skp2 S72 phosphorylation in turn resulting in Skp2 cytosolic relocalization.

14-3-3β interacts with Skp2 and is required for Akt-mediated Skp2 cytosolic relocalization

Akt phosphorylation is known to trigger protein nuclear export. We therefore tested whether Skp2 S72 phosphorylation would prevent its nuclear localization. In fact, the Skp2 sequence PRKRLKS (where S is at residue 72) was identified as a putative nuclear localization sequence (NLS) (Supplementary information, Fig. S9a). We fused this putative NLS and its mutant sequences (S72A or S72D) to GFP and examined their ability to promote GFP nuclear import. We found that neither of them promoted GFP nuclear localization, while the NLS from SV40 large T-antigen fused to the GFP readily did (Supplementary information, Fig. S9a). These results suggested that the PRKRLKS sequence from Skp2 may not function sufficiently as an NLS and that the Akt-dependent Skp2 cytosolic accumulation may be due to enhanced nuclear export. In full agreement with this hypothesis leptomycin-B (LMB), which blocks protein nuclear export, significantly prevented Akt-mediated Skp2 cytosolic localization (Supplementary information, Fig. S9b). Although the PRKRLKS sequence from Skp2 is not functionally sufficient as a NLS in our assays, it is required for Skp2 nuclear localization as Gao et al. demonstrate that a Skp2 mutant devoid of this sequence primarily localizes in the cytoplasm (see Gao et al. submitted and accompanying paper).

The 14-3-3 protein controls a variety of biological processes through regulation of partner protein localization³⁰. It binds to phospho-serine/threonine containing motifs in a sequence-specific manner³⁰. As several Akt substrates interact with 14-3-3 upon Akt-mediated phosphorylation, which in turn results in a change of protein localization³¹^-³⁴, we therefore examined whether 14-3-3 mediates Skp2 cytosolic localization upon Akt activation. Co-immunoprecipitation experiments revealed that endogenous Skp2 interacted with endogenous 14-3-3β, but not with 14-3-3γ and 14-3-3θ (Fig. 6a), although we observed a weak interaction between exogenous Skp2 and 14-3-3γ (data not shown). Remarkably, suppression of the PI3K/Akt pathway by LY or PTEN significantly abolished this interaction (Fig. 6b). We further demonstrated that the interaction of Skp2 with 14-3-3β was enhanced by IGF-1 through the PI3K/Akt-dependent pathway (Fig. 6c). GST pull down assays revealed that 14-3-3β interacted with Skp2 directly and Akt profoundly enhanced this interaction (Fig. 6d).

Fig. 6

14-3-3β interacts with Skp2 and is essential for Akt-mediated Skp2 relocalization in the cytosol

While Skp2 and Skp2 S72D effectively interacted with 14-3-3β, a Skp2 S72A mutation compromised this interaction (Fig. 6e), suggesting that Skp2 S72 phosphorylation is critical for the interaction of Skp2 with 14-3-3β. We next determined whether 14-3-3β is required for Akt-mediated Skp2 cytosolic relocalization. 14-3-3β and 14-3-3γ siRNAs profoundly suppressed 14-3-3β and 14-3-3γ protein expression, respectively (Supplementary information, Fig. S10a). Notably, silencing 14-3-3β expression inhibited Akt-mediated Skp2 cytosolic relocalization (Fig. 6f and Supplementary information, Fig. S10b). Similarly, silencing 14-3-3β expression also reduced Skp2 S72D cytosolic localization (Supplementary information, Fig. S10c). As expected, silencing 14-3-3γ expression failed to affect Akt-mediated Skp2 cytosolic relocalization (Fig. 6f). These results suggest that 14-3-3β is required for Skp2 cytosolic relocalization induced by Akt.

Cytosolic Skp2 mediates cell migration

Skp2 overexpression is frequently observed in advanced human cancers. Skp2 may play a role in tumor invasion and metastasis. We examined whether Skp2 plays a role in cell migration, which underlies these steps in tumor progression. Overexpression of Skp2 and Skp2 S72D in MEFs markedly enhanced cell migration, whereas Skp2 S72A failed to promote cell migration (Fig. 7a). As Skp2 overexpression promotes cell migration, we determined whether Skp2 is required for cell migration. Strikingly, Skp2^-/- MEFs displayed a profound defect in cell migration (Fig. 7b, c). Similarly, in wound scratch assays, loss of Skp2 significantly delayed wound closure (Supplementary information, Fig. S11a). As advanced cancers often display cytosolic Skp2²²^-²⁵, we hypothesized that Skp2 relocalization to the cytosol would favor cell motility. To test this possibility, we created a Skp2-NES (nucleus export signal) construct where the NES was fused to the Skp2 C-terminal end. IF reveals that Skp2-NES was predominantly localized in the cytosol (Fig. 7d, upper panel). Remarkably, restoration of Skp2-NES rescued the migration defect in Skp2^-/- MEFs (Fig. 7c and Supplementary information, Fig. S11a). This function of Skp2-NES is likely independent from its ability to promote p27 degradation, as Skp2-NES did not form Skp2 SCF complex efficiently and failed to induce p27 down-regulation (Fig. 7d and Supplementary information, Fig. S11b).

Fig. 7

Cytosolic Skp2 positively regulates cell migration

Cytosolic Skp2 correlates with pAkt levels, PTEN loss, and metastasis in human cancers

As Akt induced Skp2 cytosolic relocalization, we next determined whether Skp2 localization correlates with pAkt and PTEN levels in cancer specimens. We immunoassayed Skp2, pAkt, and PTEN in colon and prostate tumor microarrays (TMAs). Of 84 cases of colonic adenocarcinoma, 16 cases showed cytoplasmic Skp2 staining in infiltrating carcinoma cells compared to normal colonic mucosa from the same tissue whole section or control tissue cores on the TMA, which showed scattered nuclear staining in the crypts and glands (Fig. 8a). Sequential sections showed high pAkt (Fig. 8a, c, and Supplementary information, Fig. S12a; p<0.003) and low PTEN levels (Fig. 8a, c, and Supplementary information, Fig. S12a; p<0.008) in these tumors. The intensity of cytoplasmic Skp2 staining in prostatic adenocarcinomas was lower than in colonic adenocarcinomas (Fig. 8b versus Fig. 8a). However, the correlation of cytoplasmic Skp2 and high pAKT was highly significant (Fig. 8b, c, and Supplementary information, Fig. S12b; p<0.001) and these cases also had low PTEN levels (Fig. 8b, c, and Supplementary information, Fig. S12b; p<0.008). In agreement with the role of cytosolic Skp2 in cell migration, we found that cytosolic Skp2 strongly correlates with lymph node (LN) metastasis in colon cancers (Fig. 8c and Supplementary information, Fig. S12a, p=0.03). These results demonstrate that cytosolic Skp2 strongly correlates with pAkt, PTEN loss, and metastasis in human cancers, further implying that Skp2 cytosolic relocalization may have a role in promoting tumor metastasis.

Fig. 8

Cytosolic Skp2 correlates with activation of the Akt kinase, PTEN loss, and metastasis in human cancer specimens

Evolutionary conservation of the Akt-Skp2 functional and biochemical cross-talk

S72 is conserved in many organisms from rat to primates (Fig. 2b). However, this residue is not conserved in the mouse Skp2 (mSkp2) raising the question on when and how this regulatory pathway has evolved. We therefore tested the consequences of Akt activation in murine cells. We found that both Akt-mediated mSkp2 phosphorylation and the functional consequences of this event still take place in murine cells. Although Akt did trigger mSkp2 phosphorylation in vivo and even in in vitro (Supplementary information, Fig. 13a-c), it phosphorylated mSkp2 in vitro much less efficiently than the human Skp2 protein (Supplementary information, Fig. 13a, right panel), demonstrating that mSkp2 is a poor direct Akt substrate. Importantly, however, mSkp2 SCF complex formation was also enhanced by Mri-Akt (Supplementary information, Fig. 13c). Akt-mediated mSkp2 cytosolic localization was also observed (Supplementary information, Fig. 13d). In agreement with these findings, in Pten^-/- MEFs, where Akt is constitutively active, mSkp2 was found to accumulate in the cytosol and to relocalize in the nucleus upon LY treatment (Supplementary information, Fig. 13e). These results therefore support the notion that the Akt-Skp2 pathway is evolutionarily conserved even though Skp2 may not be a direct target of Akt in murine cells.

Discussion

We show that Skp2 S72 phosphorylation by Akt is required for Skp2 E3 ligase activity and tumorigenesis. Skp2 S72 phosphorylation provides a molecular switch that orchestrates SCF complex formation. The proto-oncogenic PI3K/Akt pathway can therefore regulate Skp2 protein levels through the enhancement of Skp2 transcription³⁵ and stability (see the accompanying paper by Gao et al.), its subcellular localization, but, importantly, the function of the SCF complex through Skp2 phosphorylation. Collectively, these findings identify PI3K/Akt signaling as a major regulatory pathway of Skp2 function at multiple levels. Very recently, Skp2 was shown to be phosphorylated by Cdk2 at S64 whose phosphorylation also regulates Skp2 stability³⁶. However, we found that Skp2 phosphorylation at S64 by Cdk2 was not required for Skp2 SCF complex formation (Supplementary information, Fig. S14a).

It should be noted that varied Akt activity through the distinct culture conditions can affect the ability of Skp2 and Skp2 S72A to form a Skp2 SCF complex, as Skp2 S72A only exhibited a minor change in SCF complex formation compared to wt Skp2 in 0.1% FBS where Akt activity is low, while Skp2 S72A displayed an impairment in forming a SCF complex upon Mri-Akt activity compared to the wt Skp2 (Fig Supplementary information, Fig. S15a).

Additionally, it should be noticed that the scansite program also predicts Skp2 T21 as a possible additional Akt phosphorylation site at low stringency (data not shown). We indeed found that Akt-mediated phosphorylation was partially reduced towards a Skp2 T21A mutant in vitro. Moreover, the ability of Skp2 T21A mutant to form a SCF complex was also reduced (Supplementary information, Fig. S15b, c). These results suggest that T21 site on Skp2 may represent an additional site for the effect of Akt on Skp2 SCF complex formation, at least in vitro. However, since the stringency of this site as per scansite analysis is low, it remains to be determined if this site is at all relevant in vivo.

The crystal structure of Skp2 SCF complex has been resolved³⁷. However, the N-terminal Skp2 tail (aa 1 to aa 108 encompassing the S72 sites) was omitted in these structural studies, suggesting that the N-terminal Skp2 tail is not required for the formation of the SCF complex per se. By contrast, it may contain an inhibitory domain that, in the context of the full-length protein, hinders the Skp2 interaction interface for Skp1 binding. Akt-mediated phosphorylation of the N-terminal Skp2 tail at S72 may induce its conformational change in turn allowing the interaction of Skp2 with Skp1. In full support of this notion, a Skp2 mutant (aa 91-424) devoid of the N-terminal tail forms a SCF complex much more efficiently than the wt Skp2 in cytosol and nucleus (Fig. 4h, i).

Skp2 normally localizes in nucleus and aberrantly localizes into the cytosol during cancer progression through mechanisms unknown to date²²^-²⁵. We provide convincing evidence showing that cytosolic Skp2 localization is regulated by Akt through Skp2 S72 phosphorylation. Although Skp2 is also phosphorylated by Cdk2 at S64³⁶, Skp2 S64A did not affect Akt-mediated Skp2 S64A cytosolic localization (Supplementary information, Fig. S14b). Similarly, a Skp2 S64D mutation did not induce Skp2 S64D cytosolic localization, while Skp2 S72D did (Supplementary information, Fig. S14b), suggesting that Skp2 S64 phosphorylation does not regulate Skp2 cytosolic localization. Our results showing that Skp2 cytosolic localization strongly correlates with pAkt, PTEN loss, and cancer metastasis suggest that cytosolic Skp2 may play an important role in tumor metastasization. In full agreement with this notion, we find that Skp2 plays an essential and dose-dependent role in cell migration, while a cytosolic Skp2 mutant restores the migration defects in Skp2^-/- MEFs. In conclusion, we demonstrate that Akt-mediated Skp2 phosphorylation is critical for SCF complex formation, Skp2 cytosolic localization, and Skp2 biological functions thus providing the molecular basis and the biological relevance underlying the aberrant cytosolic localization of Skp2 in tumorigenesis.

Methods

Mice, cell culture, and reagents

Mouse embryonic fibroblasts (MEFs) from wild-type and Skp2^-/- mice were prepared as previously described⁸^,³⁸. 293T, PC-3, COS-1 (from ATCC), and IMR90 (from M. Pagano) were cultured in DMEM containing 10% Fetal Bovine Serum (FBS), while LNCaP prostate cancer cells were grown in RPMI containing 10% FBS. To clone Xp-Skp2, Skp2 was amplified by the polymerase chain reaction (PCR) using pcDNA3-Skp2 as a template and inserted into a pcDNA4/HisMax-TOPO vector (Invitrogen). To clone GFP-Skp2, Skp2 was amplified by PCR using pcDNA3-Skp2 as a template and inserted into pcDNA3.1/NT-GFP-TOPO vector (Invitrogen). pcDNA3-Skp2 S72A, pcDNA3-Skp2 S72D, Xp-Skp2 S64A, Xp-Skp2 S64D, Xp-Skp2 S72A, Xp-Skp2 S72D, and GFP-Skp2 S72A constructs were generated by the site-directed mutagenesis kit (Stratagene) according the manufacturer's standard procedure using pcDNA3-Skp2, Xp-Skp2, and GFP-Skp2 as templates, respectively. The Skp2-NES construct was generated by fusing the nuclear export signal (NES) (ELALKLAGLDINKTE) to the C-terminal end of Skp2. MSCV-PIG-Skp2-NES was generated by subcloning Skp2-NES into the Xho1 site of the MSCV-PIG vector. pBabe-Skp2, pBabe-Skp2 S72A, and pBabe-Skp2 S72D were generated by subcloning Skp2 and Skp2 S72A into the EcoRI site of the pBabe vector. GFP-PRKRLKS (from Skp2), GFP-PRKRLKA (from Skp2 S72A), GFP-PRKRLKD (from Skp2 S72D), or GFP-PKKKRKVD (from SV40 large T-antigen) constructs were generated by introducing the peptide encoding sequence in frame at the 3′ end of the GFP coding sequence in a pcDNA3.1/NT-GFP-TOPO vector (Invitrogen). (His)₆-ubiqutin, HA-p27, and pGEX-4X1-Skp1 constructs are gifts from Drs. D. Bohmann, M. Pagano, and P. Jackson, respectively. To clone Xp-mSkp2, the mouse Skp2 cDNA was amplified by RT-PCR from MEFs and was inserted into pcDNA4/HisMax-TOPO vector. Flad-Skp2 and Flag-ΔN-Skp2 (aa, 91-424) were from Dr. W. Wei. The p27-T187A construct was a gift form Dr. M.H. Lee. IGF-1, EGF, MG132, U0126, rapamycin and LY were obtained from Calbiochem. Wortmannin was purchased from Sigma. GST-Akt1 and recombinant active Akt1 were obtained from Cell Signaling.

Immunoprecipitation, immunoblotting and immunofluorescence

Immunoprecipitation (IP), immunoblotting (IB), and immunofluorescence (IF) were performed essentially as described with some modifications³⁸^,³⁹. For protein-protein interaction, cells were lysed by E1A lysis buffer (250 mM NaCl, 50 mM HEPES, [pH 7.5], 0.1% NP40, 5 mM EDTA, protease inhibitor cocktail (Roche). The following antibodies were used for IP, IB or IF: anti-Skp2 (IP: 1:250; IB: 1:1000; IF: 1:100, Zymed), anti-Skp2 (IB 1:1000, Santa Cruz), anti-phospho (S473)-Akt (IB: 1:1000, cell signaling), anti-Akt1 (IB: 1:1000, Cell Signaling), anti-phospho-Akt substrate antibody (IB: 1:1000, cell signaling), anti-Skp1 (IP: 1:200; IB: 1:1000, BD Transduction Lab), anti-14-3-3β (IP: 1:100; IB: 1:2000, Santa Cruz), anti-14-3-3γ (IB: 1:500, Santa Cruz), phospho (S72) Skp2 antibody (IB: 1:1000, Pocono Rabbit Farm and Laboratory Inc.), anti-Xpress (IP/IF: 1:500; IB: 1:5000, Invitrogen), anti-HA (IB/IF: 1:1000, Covance, Upstate), anti-α-tubulin (IB: 1:1000, Sigma), anti-β-actin (IB: 1:1000, Sigma), anti-GFP (IP: 1:500; IB: 1:1000, BD Clontech), anti-p27 (IP: 1:100; IB: 1:1000, BD Transduction Lab), and anti-E2F1 (IB: 1:500, Santa Cruz).

In vitro and in vivo phosphorylation assay

In vitro phosphorylation assay was performed as described previously⁴⁰. In brief, GFP-Skp2 or GFP-Skp2 S72A was immunoprecipitated with GFP antibody from 293T cells cultured in DMEM medium containing 0.5% FBS. Immune complexes were washed three times in RIPA lysis buffer 150 mM NaCl, 10 mM Tris (pH 7.5), 1% NP40, 0.5 % deoxycholate, 0.1% SDS, protease inhibitor cocktail (Roche), then washed twice in 1X kinase buffer (25 mM Tris-HCl (pH 7.5), 5 mM β-glycerophosphate, 2 mM dithiothreitol, 0.1 mM Na₃VO₄, 10 mM MgCl₂, 2 μM cold ATP), and incubated with 1μl recombinant active Akt1 kinase and 2 μCi [γ-³²P]-ATP in 50 μl total reaction buffer for 30 min at 30°C. Reactions were stopped by washing twice in kinase buffer and boiling in 2×SDS loading buffer. Proteins were resolved by 8% SDS-PAGE and transferred to the nitrocellulose membrane, and ³²P incorporation was detected by autoradiography. For in vivo labeling experiments, 293T cells cultured in DMEM containing 10% FBS were transfected with indicated plasmids for 24 h, and the medium was changed to phospho-free DMEM with 0.25 % dialyzed FBS containing 200 μCi/ml ortho-³²P for 4 h. Cells were lysed by RIPA buffer for IP and the Skp2 immunocomplex was subjected to 8% SDS-PAGE, followed by autoradiography.

Identification of phosphorylation sites by mass spectrometry

Gel-resolved proteins from in vitro and in vivo phosphorylation reactions were digested with trypsin, batch purified on a reversed-phase (RP) micro-tip, and an aliquot was analyzed by matrix-assisted laser desorption/ionization reflectron time-of-flight (MALDI-reTOF) mass spectrometry (MS; UltraFlex TOF/TOF; BRUKER Daltonics; Bremen, Germany) for peptide mass fingerprinting, as described ⁴⁰^,⁴¹. This served to confirm the identity of the proteins and to locate possible differences between the tryptic peptide maps of the phosphorylated and unphosphorylated forms. The remainder of the RP-eluted digest mixtures was then subjected to immobilized gallium (III) affinity chromatography for selective capture of phosphopeptides, followed by elution with phosphate buffer, desalting over an RP tip, and a second round of MALDI-reTOF MS²⁰. Peak m/z values were matched to the protein sequence, allowing for the likely presence of one or more phosphate groups. Mass spectrometric sequencing of the putative phosphopeptides was then carried by MALDI-TOF/TOF MS/MS analysis using the UltraFlex instrument in “LIFT” mode. Fragment ion spectra were inspected for the characteristic partial loss of 98 Da, indicating presence of phospho-serine or -threonine, and for the a″, b″, and y″ ions to compare with the computer-generated fragment ion series of the predicted tryptic peptides to locate the exact position of phosphoamino acids.

GST-pull down assays

For in vitro GST-Skp1 and Skp2 interaction, GST-Skp1 proteins purified from the bacterial lysates of BL21 competent cells transformed with pGEX-4X1-Skp1 using the Glutathione-Agarose beads according to the manufacturer's standard procedures. The GST-Skp1 proteins bound to glutathione Sepharose beads (Amersham Biosciences) were then incubated with the in vitro translated [³⁵S]-Skp2, which was preincubated with recombinant Akt kinase for 30 min at 30 °C, for 4 hours at 4°C in the interaction buffer (20 mM HEPES, PH 7.9, 150 mM KCl. 5mM EDTA, 0.5 mM DTT, 0.1% (v/v) Nonidet p-40, 0.1 % (w/v) BSA, 1mM PMSF, and 10% Glycerol), washed by the NETN buffer (20 mM Tris, pH 8.0, 100 mM NaCl, 6 mM MgCl₂, 1 mM EDTA, 0.5% Nonidet P-40, 1 mM DTT, 8% glycerol, 1 mM PMSF) 4 times, and subjected to 8% SDS-PAGE, followed by autoradiography. For in vitro Akt and Skp2 interaction, GST-Akt1 bound in agarose beads (Cell signaling) were incubated with in vitro translated [³⁵S] Skp2Skp2 at overnight in the interaction buffer, washed either by 1XPBS or NETN buffer 4 times, and subjected to 8% SDS-PAGE, followed by autoradiography.

Generation of phospho (S72)-Skp2 antibody

The Skp2 phospho (S72) peptide (PPRKRLKS^pKGSDKDF) was synthesized and injected into the rabbit, followed by affinity purification according to the manufacturer's standard procedures (Pocono Rabbit Farm and Laboratory Inc.). For the Western blot analysis, the Skp2 phospho (S72)-antibody was pre-incubated with the Skp2 non-phospho peptide (PPRKRLKSKGSDKDF) [5 (peptide): 1 (antibody); w/w] overnight at 4°C, prior to adding it to the Western membrane. The secondary antibody used for this phospho-antibody was Rabbit TrueBlot HRP (eBioscience).

Ubiquitylation assay in vitro and in vivo

In vitro p27 ubiquitylation assays was performed essentially as described⁴. In brief, the Skp2 SCF complex was immunoprecipitated from 293T cells and mixed with in vitro-translated [³⁵S]-p27 that was previously incubated with cyclin E/CDK2 for in vitro phosphorylation along with methylated ubiquitin and ubiquitin aldehyde for 60 min at 30 °C. The reaction was stopped with 2X SDS sample buffer and run on acrylamide gels. In vivo ubiquitylation assays were performed as described⁴². In brief, 293T cells were transfected with the indicated plasmids for 24 h, treated with vehicle, 20 μM LY, or 100 nM Wortmannin together with 20 μM MG132 for 6 h, and lysed by the denatured buffer (6M guanidine-HCl, 0.1M Na₂HPO₄/NaH₂PO₄, 10 mM imidazole). The cell extracts were then incubated with nickel beads for 3 h, washed, and subjected to the Western blot analysis.

RNA interference

293T cells were transfected with 14-3-3β or 14-3-3γ or nonspecific control siRNA using Oligofectamine reagent (Invitrogen) according to manufacturer's instructions. In brief, 100 nM final concentration of siRNA was used to transfect cells at 50 % confluency. Four hours after transfection, cells were recovered in full serum. Twenty-four hours later, cells were transfected with indicated plasmids together with siRNA using Lipofectamine2000 reagent in DMEM medium containing 0.5% before harvest. Cells were harvested 48 hr after siRNA transfection for IB and IF.

Cell growth and in vivo tumorigenesis assay

Brdu-incorporation assays were performed by using the In Situ Cell proliferation Kit, Fluos (Roche). LNCaP stable transfectant cells were generated as previously described⁴³. In brief, LNCaP cells were transfected with pcDNA3, pcDNA3-Skp2, or pcDNA3-Skp2 S72A and selected by G418 (300 μg/ml). The individual clone was picked and verified by Western blot analysis (data not shown). Two positive clones were pooled and used for Western blot analysis, cell growth, and the in vivo tumorigenesis assay (Fig.3). For cell growth assay, 2×10⁴ cells were seeded in 12 wells in triplicates, harvested, and stained with trypan blue at different days. Numbers of viable cells were directly counted under the microscope. For in vivo tumorigenesis assays, LNCaP stable cells (1×10⁶) mixed with matrigel (1:1) were subcutaneously injected into the left flank of 6-week-old athymic male nude mice (NCRNU-M, Taconic Farms Inc.). Tumor size was measured weekly using a caliper, and tumor volume was determined by using the standard formula: L×W²×0.52, where L is the longest diameter and W is the shortest diameter.

Cell migration and in vitro wound scratch assay

Cell migration assay and was done in 24 well transwell plate with 8-μm polyethylene terephalate membrane filters [Falcon cell culture insert (Becton-Dickinson)] separating the lower and upper culture chambers. In brief, MEFs were plated in the upper chamber at 5 × 10⁴ cells per well in serum-free DMEM medium. The bottom chamber contained DMEM with 10% FBS. Cells were allowed to migrate for 18 hours in a humidified chamber at 37°C with 5% CO2. After the incubation period, the filter was removed and non-migrant cells on the upper side of the filter were detached using a cotton swab. Filters were fixed with 4% formaldehyde for 15 min and cells located in the lower filter were stained with 0.1% crystal violet for 20 min and counted from three random fields. For the in vitro wound scratch assay, MEFs were grown to confluency and wounds were made using sterile pipette tips. Cells were washed in phosphate buffer saline (PBS) and replenished with DMEM containing 10% FBS. After overnight incubation, cells were fixed and photographed.

Immunohistochemistry

Paraffin immunohistochemistry was performed on tissue microarrays (TMAs) of prostate carcinoma (101 cases), colonic carcinoma (84 cases), and selected whole section slides of colonic carcinoma. Slides were deparaffinized and heated for antigen retrieval (steamer), followed by incubating with primary antibodies for overnight at 4°C. Biotinylated secondary and tertiary antibodies (avidin-biotin complex) were incubated and developed with a chromogen diaminobenzidine. Primary antibodies for Skp2 (2C8D9, 1:400, Zymed), PTEN (1:75, Cascade), and anti-phospho (S473) Akt (1:250, Cell Signaling) were used for this study. The statistic is analyzed by chi-square χ².

Supplementary Material

Figure S1. Identification of Skp2 phosphorylation by Akt in vivo and in vivo. (a) GFP-Skp2 was isolated form 293T cells transfected with GFP-Skp2 along with Mri-Akt and analyzed by mass spectrometry. (b) Flag-Skp2 or Flag-TSC2 was immunoprecipitated from 293T, incubated with recombinant active Akt for 30 min, and subjected to SDS-PAGE analysis. (c, d) GST-N-Skp2 or GST-N-S72A produced in bacteria were incubated with recombinant active Akt for 30 min and subjected to SDS-PAGE analysis (c) and mass spectrometry (d).

Figure S2. Skp2 S72 phosphorylation regulates p27 ubiquitylation and protein stability, but not Skp2-p27 interaction. (a) Skp2 and the Skp2 S72D mutant, but not Skp2 S72A, promote exogenous p27 ubiquitylation. 293T cells were transfected with the indicated plasmids, treated with 10 μM MG132 for 6 h and harvested for in vivo ubiquitylation assay. (b, c) Skp2 phosphorylation regulates p27 protein stability. 293T cells were transfected with the indicated plasmids for 24 h, treated with 20 μM cyclohexamide (CHX) at various time points, and harvested for Western blot analysis (b), and flow cytometry analysis (c). (d) Skp2 S72A retains the ability to interact with p27. 293T cells were transfected with the indicated plasmids and harvested for co-immunoprecipitation experiments and Western blot analysis.

Figure S3. Skp2 S72 phosphorylation regulates cells in S-phase. (a) p27 T187 phoshorylation is not required for Skp2-mediated p27 ubiquitylation. 293T cells were transfected with the indicated plasmids, treated with 10 μM MG132 for 6 h, and harvested for in vivo ubiquitylation assay (b). p27 T187 phosphorylation is not required for the interaction of p27 with Skp2. 293T cells were transfected with the indicated plasmids, treated with 10 μM MG132 for 6 h, and harvested for in vivo co-immunoprecdipitation assay. (c) Skp2 S72 phosphorylation affects Skp2-mediated an increase in S-phase cells. COS-1 cells were transfected with the indicated plasmids in a serum-starved condition (0.1% FBS) for 24 h, refreshed with 10% FBS for 16 h, incubated with 20 μM Brdu for 1 h, and harvested for quantification of Brdu-incorporation. 200-300 cells were scored and a representative result is shown from three independent experiments., **p<0.01, ***p<0.01 using Student's t-test, n=3. Scale bar, 50 μm.

Figure S4. Akt activity does not affect 293T cell cycle distribution upon serum-starvation. Cells transfected with the indicated plasmids were serum-starved for 1 day, treated with or without WN for 16 h in 0.1% FBS medium, and harvested for cell cycle analysis.

Figure S5 Skp2 S72 phosphoryaltion regulates Skp2 SCF complex formation and in vitor E3 ligase activity (a) The Skp2 S72A SCF mutant complex displays a reduced E3 ligase activity towards p27 in in vitro ubiquitylation assays. 293T cell were transfected with mock, XP-Skp2, XP-S72A, or XP-S72D. Cell lysates were immunoprecipitated with the XP antibody (left panel) and subjected to in vitro p27 ubiquitylation assays (right panel). (b) Akt activity regulates Skp2 SCF complex formation and Skp2 SCF E3 ligase activity towards p27 ubiquitylation in vitro. 293T cells were transfected with the indicated plasmids in the presence or absence of 20 μM LY for 6h, harvested for IP with XP antibody (right panel), and extracts subjected to in vitro p27 ubiquitylation assays (left panel).

Figure S6. Akt regulates Skp2 cytosolic relocalization. (a) S72 phosphorylation mediates Akt-induced Skp2 cytosolic localization. COS-1 cells were transfected with the indicated plasmids in the serum-free medium and harvested for IF analysis. (b) Akt promotes endogenous Skp2 cytosolic localization. Nuclear and cytosolic fractions from HeLa cells transfected with Mri-Akt in the serum-free medium were isolated and subjected to Western blot analysis. α-tubulin and E2F1 were used as cytosolic and nuclear markers, respectively. Scale bar, 10 μm.

Figure S7. Cell cycle distribution of 293T upon starvation and IGF-1 treatment. Cells were serum-starved for 2 days, treated with IGF-1 (100 ng/ml) in the presence or absence of WN for various times, and harvested for cell cycle analysis.

Figure S8. IGF-1 induces S72 Skp2 phosphorylation and Skp2 cytosolic localization in IMR90 cells. (a-c) Cells were serum starved (0.1% FBS) for 48 h, treated with 100 nM IGF with or without WN in the presence of 10% FBS, and harvested for flow cytometry (a), IF (b), and Western blot (c) analysis. (d) Quantitation of the experiments shown in (b). Three random fields were counted and results are presented as mean values ± s.d. from two independent experiments., **p<0.01, ***p<0.001 using Student's t-test, n=3. Scale bar, 50 μm.

Figure S9. PRKRLKS from Skp2 does not mediate nuclear import while Akt promotes Skp2 nuclear export. (a) The Skp2 sequence PRKRLKS (where S is at residue 72) was identified as a putative nuclear localization sequence (NLS) according to the(<http://cubic.bioc.columbia.edu/cgi/var/nair/loctree/query>http://cubic.bioc.columbia.edu/cgi/var/nair/loctree/query) software. NIH3T3 cells were transfected with the indicated plasmids for 2 days and harvested for IF analysis. We did not observe a preferential nuclear accumulation of GFP with any of the Skp2 derived sequence. By contrast, the SV40 NLS did cause GFP nuclear accumulation as predicted. (b) Akt promotes Skp2 nuclear export. 293T cells were transfected with the indicated plasmids in the serum-free medium, treated with leptomycin B (LMB, 10 ng/ml), and harvested for IF analysis. Scale bar, 20 μm.

Figure S10. Silencing 14-3-3β and 14-3-3γ expression by 14-3-3β and 14-3-3γ siRNA. (a) 293T cells were transfected with siRNAs as indicated and cells were harvested for Western blot analysis. Lamin A/C siRNA was used as unrelated siRNA control. (b, c) 293T cells were transfected with the 14-3-3β siRNAs and the indicated plasmids and harvested for IF analysis. Scale bar, 20 μm.

Figure S11. Cytosolic Skp2 restores cell migration defect in Skp2^-/- primary MEFs. (a) Skp2 is required for cell migration. Skp2 wt and Skp2^-/- primary MEFs were infected with the indicated viral constructs for 2 days, selected by 2 μg/ml puromycin for 3 days, and plated for in vitro wound healing assay. Results are presented as mean values ± s.d. from two independent experiments., ***p<0.001 using Student's t-test, n=3. Scale bar, 50 μm. (b) Skp2-NES displays a defect in Skp2 SCF complex formation. Total cell lysates from 293 T cells transfected with XP-Skp2 or XP-Skp2-NES were immunoprecipitated (IP) with XP antibody, washed, and followed by Western blot analysis.

Figure S12. Cytosolic Skp2 correlates with pAkt levels, PTEN loss, and tumor metastasis. (a) Correlation between Skp2 localization, pAkt levels, PTEN levels and metastasis in colon TMAs. (b) Correlation between Skp2 localization, pAkt levels, PTEN levels in prostate TMAs.

Figure S13. Akt regulates mouse Skp2 phosphorylation, cytosolic localization, and Skp2 SCF complex formation. (a) Akt induces mouse Skp2 and human Skp2 phosphorylation in vitro. Mouse Skp2 (mSkp2) immunopurified from 293T transfected with XP-mSkp2 using a XP antibody, incubated with Rec Akt for 30 min, and subjected to SDS-PAGE analysis (Left panel). XP-mSkp2 or XP-Skp2 isolated from 293T cells transfected with XP-mSkp2 or XP-Skp2 incubated with Rec Akt for 30 min, and subjected to SDS-PAGE analysis (Right panel). (b) Akt induces mSkp2 phosphorylation in vivo. 293T cells were transfected with XP-mSkp2 along with vector or Mri-Akt, and mSkp2 phosphorylation was determined by in vivo phospho-labeling. (c) Akt induces mouse Skp2 phosphorylation in vivo and positively regulates mouse Skp2 SCF complex formation. NIH3T3 cells were transfected with the indicated plasmids in the serum-free medium and harvested for co-immunoprecipitation experiments and the Western blot analysis. To detect in vivo mSkp2 phosphorylation, an anti-Ser/Thr antibody was used as primary antibody. (d) Akt induces cytosolic localization of mouse Skp2. NIH3T3 cells or 293T cells were transfected with the indicated plasmids in the serum-free medium for 2 days and harvested for IF analysis. (e) Immortalized wt or Pten^-/- MEFs were treated with or without LY (20 μM) for 3 h and harvested for fractionation and the Western blot analysis. Scale bar, 20 μm.

Figure S14. Mutation on Cdk2-mediated Skp2 phosphorylation site does not regulate Skp2 SCF complex formation and cytosolic localization. (A) Total cell lysates from 293 T cells transfected with XP-Skp2, XP-Skp2 S64A (XP-S64A), or XP-Skp2 S72A (XP-S72A) were immunoprecipitated with XP antibody, washed, and followed by Western blot analysis. (B) COS-1 cells were transfected with the indicated plasmids and harvested for IF analysis. Scale bar, 10 μm.

Figure S15. Skp2 T21 is a potential site for Akt-mediated Skp2 phosphorylation and Skp2 SCF complex formation. (a) 293T cells were transfected with the plasmids as indicated, serum starved with 0.1% FBS for 30h, and harvested for co-immunoprecipitation assay. (b) Akt phosphorylates Skp2 at T21 in vitro. XP-Skp2 and XP-mSkp2 T21A (XP-T21A), immunoprecipitated from 293T, were incubated with Rec Akt for 30 min and subjected to SDS-PAGE analysis. (c) T21 and S72 phosphorylation regulates Skp2 SCF complex formation. Total cell lysates from 293T cells transfected with Mock, XP-Skp2, XP-Skp2 T21A, or XP-S72A along with Mri-Akt in 0.1% FBS were immunoprecipitated with the XP antibody, followed by Western blot analysis.

Click here to view.^{(5.8M, pdf)}

Acknowledgments

We thank Drs D. Bohmann, P. Jackson, W. Wei, and M. Pagano, and M.H. Lee for reagents. We are also grateful to M. Asherov and I. Linkov in the Immunohistochemistry Pathology Core Laboratory, T. Matos for immunohistochemistry technical assistance, P. Bonner for data management, L. Lacomis for help with mass spectrometry, X.H. Zhu for technical advice, S. Clohessy for flow cytometry analysis. We also thank Drs. M.C. Hung and L. Cantley for insightful comments and suggestions and W. Wei for discussion and for sharing experimental results. Special thanks extended to B. Carver and L. DiSantis for editing and critical reading of the manuscript, as well as to all the members of the Pandolfi laboratory for comments and discussion. This work was supported by NIH grants RO1 CA-71692 and CA-74031 to P.P.P. and by M.D. Anderson Cancer Center Trust Scholar funds to H.K.L. The Microchemistry & Proteomics Core is supported by NIH grant P30 CA-08748.

Footnotes

Author Contributions: H.K.L. and P.P.P designed the experiments and wrote the manuscript; H.K.L., G.W. Z.C., Y.L., C.H.C. and W.L.Y. performed the experiments; J.T. performed the IHC and analyzed the data; K.I.N. provided the Skp2^-/- mice; S.N. provided the valuable suggestions; H.E. and P.T. performed the mass spectrometry analysis.

References

1. Bloom J, Pagano M. Deregulated degradation of the cdk inhibitor p27 and malignant transformation. Semin Cancer Biol. 2003;13:41–7. [PubMed]

2. Nakayama KI, Nakayama K. Regulation of the cell cycle by SCF-type ubiquitin ligases. Semin Cell Dev Biol. 2005;16:323–33. [PubMed]

3. Amati B, Vlach J. Kip1 meets SKP2: new links in cell-cycle control. Nat Cell Biol. 1999;1:E91–3. [PubMed]

4. Carrano AC, Eytan E, Hershko A, Pagano M. SKP2 is required for ubiquitin-mediated degradation of the CDK inhibitor p27. Nat Cell Biol. 1999;1:193–9. [PubMed]

5. Tsvetkov LM, Yeh KH, Lee SJ, Sun H, Zhang H. p27(Kip1) ubiquitination and degradation is regulated by the SCF(Skp2) complex through phosphorylated Thr187 in p27. Curr Biol. 1999;9:661–4. [PubMed]

6. Kossatz U, et al. Skp2-dependent degradation of p27kip1 is essential for cell cycle progression. Genes Dev. 2004;18:2602–7. [PubMed]

7. Sutterluty H, et al. p45SKP2 promotes p27Kip1 degradation and induces S phase in quiescent cells. Nat Cell Biol. 1999;1:207–14. [PubMed]

8. Nakayama K, et al. Targeted disruption of Skp2 results in accumulation of cyclin E and p27(Kip1), polyploidy and centrosome overduplication. Embo J. 2000;19:2069–81. [PubMed]

9. Nakayama K, et al. Skp2-mediated degradation of p27 regulates progression into mitosis. Dev Cell. 2004;6:661–72. [PubMed]

10. Gstaiger M, et al. Skp2 is oncogenic and overexpressed in human cancers. Proc Natl Acad Sci U S A. 2001;98:5043–8. [PubMed]

11. Latres E, et al. Role of the F-box protein Skp2 in lymphomagenesis. Proc Natl Acad Sci U S A. 2001;98:2515–20. [PubMed]

12. Shim EH, et al. Expression of the F-box protein SKP2 induces hyperplasia, dysplasia, and low-grade carcinoma in the mouse prostate. Cancer Res. 2003;63:1583–8. [PubMed]

13. Zhang H, Kobayashi R, Galaktionov K, Beach D. p19Skp1 and p45Skp2 are essential elements of the cyclin A-CDK2 S phase kinase. Cell. 1995;82:915–25. [PubMed]

14. Bilodeau M, et al. Skp2 induction and phosphorylation is associated with the late G1 phase of proliferating rat hepatocytes. FEBS Lett. 1999;452:247–53. [PubMed]

15. Dan HC, et al. Phosphatidylinositol 3-kinase/Akt pathway regulates tuberous sclerosis tumor suppressor complex by phosphorylation of tuberin. J Biol Chem. 2002;277:35364–70. [PubMed]

16. Inoki K, Li Y, Zhu T, Wu J, Guan KL. TSC2 is phosphorylated and inhibited by Akt and suppresses mTOR signalling. Nat Cell Biol. 2002;4:648–57. [PubMed]

17. Manning BD, Tee AR, Logsdon MN, Blenis J, Cantley LC. Identification of the tuberous sclerosis complex-2 tumor suppressor gene product tuberin as a target of the phosphoinositide 3-kinase/akt pathway. Mol Cell. 2002;10:151–62. [PubMed]

18. Potter CJ, Pedraza LG, Xu T. Akt regulates growth by directly phosphorylating Tsc2. Nat Cell Biol. 2002;4:658–65. [PubMed]

19. Yaffe MB, et al. A motif-based profile scanning approach for genome-wide prediction of signaling pathways. Nat Biotechnol. 2001;19:348–53. [PubMed]

20. Posewitz MC, Tempst P. Immobilized gallium(III) affinity chromatography of phosphopeptides. Anal Chem. 1999;71:2883–92. [PubMed]

21. Malek NP, et al. A mouse knock-in model exposes sequential proteolytic pathways that regulate p27Kip1 in G1 and S phase. Nature. 2001;413:323–7. [PubMed]

22. Drobnjak M, et al. Altered expression of p27 and Skp2 proteins in prostate cancer of African-American patients. Clin Cancer Res. 2003;9:2613–9. [PubMed]

23. Li Q, Murphy M, Ross J, Sheehan C, Carlson JA. Skp2 and p27kip1 expression in melanocytic nevi and melanoma: an inverse relationship. J Cutan Pathol. 2004;31:633–42. [PubMed]

24. Lim MS, et al. Expression of Skp2, a p27(Kip1) ubiquitin ligase, in malignant lymphoma: correlation with p27(Kip1) and proliferation index. Blood. 2002;100:2950–6. [PubMed]

25. Radke S, Pirkmaier A, Germain D. Differential expression of the F-box proteins Skp2 and Skp2B in breast cancer. Oncogene. 2005;24:3448–58. [PubMed]

26. Cantley LC, Neel BG. New insights into tumor suppression: PTEN suppresses tumor formation by restraining the phosphoinositide 3-kinase/AKT pathway. Proc Natl Acad Sci U S A. 1999;96:4240–5. [PubMed]

27. Di Cristofano A, Pandolfi PP. The multiple roles of PTEN in tumor suppression. Cell. 2000;100:387–90. [PubMed]

28. Brazil DP, Park J, Hemmings BA. PKB binding proteins. Getting in on the Akt. Cell. 2002;111:293–303. [PubMed]

29. Datta SR, Brunet A, Greenberg ME. Cellular survival: a play in three Akts. Genes Dev. 1999;13:2905–27. [PubMed]

30. Hermeking H. The 14-3-3 cancer connection. Nat Rev Cancer. 2003;3:931–43. [PubMed]

31. Brunet A, et al. Akt promotes cell survival by phosphorylating and inhibiting a Forkhead transcription factor. Cell. 1999;96:857–68. [PubMed]

32. Brunet A, et al. 14-3-3 transits to the nucleus and participates in dynamic nucleocytoplasmic transport. J Cell Biol. 2002;156:817–28. [PMC free article] [PubMed]

33. Fujita N, Sato S, Katayama K, Tsuruo T. Akt-dependent phosphorylation of p27Kip1 promotes binding to 14-3-3 and cytoplasmic localization. J Biol Chem. 2002;277:28706–13. [PubMed]

34. Sekimoto T, Fukumoto M, Yoneda Y. 14-3-3 suppresses the nuclear localization of threonine 157-phosphorylated p27(Kip1) Embo J. 2004;23:1934–42. [PubMed]

35. Mamillapalli R, et al. PTEN regulates the ubiquitin-dependent degradation of the CDK inhibitor p27(KIP1) through the ubiquitin E3 ligase SCF(SKP2) Curr Biol. 2001;11:263–7. [PubMed]

36. Rodier G, Coulombe P, Tanguay PL, Boutonnet C, Meloche S. Phosphorylation of Skp2 regulated by CDK2 and Cdc14B protects it from degradation by APC(Cdh1) in G1 phase. Embo J. 2008;27:679–91. [PubMed]

37. Zheng N, et al. Structure of the Cul1-Rbx1-Skp1-F boxSkp2 SCF ubiquitin ligase complex. Nature. 2002;416:703–9. [PubMed]

38. Lin HK, Bergmann S, Pandolfi PP. Cytoplasmic PML function in TGF-beta signalling. Nature. 2004;431:205–11. [PubMed]

39. Lin HK, Wang L, Hu YC, Altuwaijri S, Chang C. Phosphorylation-dependent ubiquitylation and degradation of androgen receptor by Akt require Mdm2 E3 ligase. Embo J. 2002;21:4037–48. [PubMed]

40. Ma L, Chen Z, Erdjument-Bromage H, Tempst P, Pandolfi PP. Phosphorylation and functional inactivation of TSC2 by Erk implications for tuberous sclerosis and cancer pathogenesis. Cell. 2005;121:179–93. [PubMed]

41. Sebastiaan Winkler G, et al. Isolation and mass spectrometry of transcription factor complexes. Methods. 2002;26:260–9. [PubMed]

42. Kim SY, Herbst A, Tworkowski KA, Salghetti SE, Tansey WP. Skp2 regulates Myc protein stability and activity. Mol Cell. 2003;11:1177–88. [PubMed]

43. Lin HK, Yeh S, Kang HY, Chang C. Akt suppresses androgen-induced apoptosis by phosphorylating and inhibiting androgen receptor. Proc Natl Acad Sci U S A. 2001;98:7200–5. [PubMed]