PTEN at a glance

PTEN mutations occur frequently in a variety of human cancers, including endometrial carcinoma, glioblastoma multiforme, skin and prostate cancers (Chalhoub and Baker, 2009). They include a mutational spectrum that is scattered along the entire gene and predominantly represent monoallelic mutations. A majority of PTEN mutations cause truncations of the protein. In addition to cancer-associated somatic mutations, PTEN is often mutated in tumor-prone germline diseases.

Genetic studies in mice have unambiguously demonstrated that PTEN is a potent tumor suppressor in almost all organs. Tissue-specific Pten-knockout mice develop tumors in numerous organs (Knobbe et al., 2008; Suzuki et al., 2008). Intriguingly, subtle changes in PTEN gene expression, mRNA regulation and protein synthesis are sufficient to have a substantial impact on tumorigenesis, a phenomenon that is referred to as quasi-insufficiency (Carracedo et al., 2011). Furthermore, the notion of PTEN haploinsufficiency seems to apply to human cancer because monoallelic mutation of PTEN without loss or mutation of the second allele is prevalent in breast and prostate cancers (Salmena et al., 2008).

PTEN is involved in regulating many cellular processes. Many of these functions can be attributed to its lipid phosphatase activity. For example, PTEN regulates cell proliferation and apoptosis. These effects are mediated by suppressing AKT (also known as protein kinase B) activation. Because AKT activation requires PtdIns(3,4,5)P₃, dephosphorylation of this lipid second messenger by PTEN results in inhibition of AKT and subsequent alteration of the function of AKT substrates, such as forkhead box protein O (FOXO), the E3 ubiquitin ligase MDM2, and BCL2 antagonist of cell death (BAD) (Tamguney and Stokoe, 2007). PTEN has a crucial role in regulating the self-renewal and differentiation of human embryonic stem cells and hematopoietic stem cells as well as the timing of follicle activation through the regulation of oocyte growth (Reddy et al., 2008). PTEN also regulates the chemotaxis of neutrophils (Heit et al., 2008), and all these functions require its lipid phosphatase activity.

In addition, PTEN can regulate various cellular events independently of its lipid phosphatase activity. For example, it can inhibit cell cycle progression by modulating the activity of the anaphase-promoting complex/cyclosome (APC/C) in the nucleus in a manner that is independent of PTEN enzymatic activity (Song et al., 2011). Furthermore, it can inhibit cell invasion and migration, probably by making use of its protein phosphatase activity (Tamura et al., 1999). PTEN can control the size of DNA-damaged cells by regulating the actin-remodeling process (Kim et al., 2011) through a mechanism that is likely to be independent of its lipid phosphatase activity. Deficiency of any of these functions can contribute to tumorigenesis.

As mentioned above, subtle changes in the regulation of gene expression and mRNA of PTEN have a substantial impact on the normal physiology of organisms. Therefore, both are delicately regulated through multiple mechanisms (see Poster).

The PTEN promoter is regulated by many transcription factors that operate at specific times and in different cell types. Transcription factors, such as early growth response protein (EGR1), tumor protein 53 (Tp53, also referred to as p53) and peroxisome proliferator-activated receptor gamma (PPARG) were reported to positively regulate PTEN expression (Martelli et al., 2011). PTEN transcription can also be repressed, for example, by the polycomb complex protein BMI1 during epithelial–mesenchymal transition (EMT) of nasopharyngeal epithelial cells, or through β-catenin/transcription factor (TCF)-mediated downregulation of EGR1 in ovarian cancer cells (Lau et al., 2011; Song et al., 2009). Intriguingly, in a tissue-specific manner, Notch signaling might lead to upregulation of PTEN by inhibiting one of the PTEN downregulators, the recombining binding protein suppressor of hairless (RBPJ, also known as and hereafter referred to as CBF-1) (Chappell et al., 2005; Whelan et al., 2007) or, conversely, to its downregulation by activating another PTEN downregulator, the transcription factor hairy and enhancer of split 1 (HES1) (Palomero et al., 2007). Furthermore, epigenetic silencing by hyper-methylation of the PTEN promoter occurs in some breast and colorectal cancers, and in gastric carcinoma.

PTEN mRNA has been reported to be susceptible to post-transcriptional regulation by a variety of microRNAs (miRNAs), including miRNA-21, miRNA-22 and miRNA-26a (Bar and Dikstein, 2010; Huse et al., 2009; Meng et al., 2007). These miRNAs can, thus, impact both normal biological and pathological functions of PTEN.

Additional complexity results from the regulation of PTEN expression through non-coding RNAs, such as the PTEN pseudogene PTENP1 mRNA. PTENP1 genetically resembles PTEN in its protein coding region. Owing to a mutation in its initiator codon PTEN1 mRNA, unlike PTENP mRNA, cannot be translated into a protein. The PTENP1 mRNA is generally subject to same miRNA-mediated regulation and, thus, can function as ‘decoy’ that sequesters miRNA-21. Interestingly, in sporadic colon cancer, PTENP1 undergoes a copy number loss that is concurrent with PTEN downregulation (Poliseno et al., 2010). Similarly, zinc finger E-box-binding homeobox 2 (ZEB2), which encodes another endogenous RNA, has been reported to serve as an miRNA decoy for the PTEN mRNA, and loss of ZEB2 mRNA contributes to melanomagenesis (Karreth et al., 2011). However, it is important to bear in mind that only 25% of cancer patients portray a correlation between the loss of PTEN protein and its mRNA level (Chen et al., 2011), which emphasizes the importance of PTEN regulation at the post-transcriptional and post-translational levels.

PTEN is subject to various post-translational modifications that regulate its enzymatic activity, interaction with other proteins and subcellular localization (see Poster).

The first phosphorylation sites mapped on PTEN were a cluster of serine and threonine residues in its C-terminal tail (Vazquez et al., 2000). Mutation of these residues to alanine leads to elevated membrane affinity, higher enzymatic activity and more-rapid degradation of PTEN. When these residues are phosphorylated, the C-terminal tail can interact with the N-terminal C2 and phosphatase domains (Odriozola et al., 2007; Rahdar et al., 2009), which suggests that phosphorylation of the C-terminal tail functions as an auto-inhibitory mechanism that controls both PTEN membrane recruitment and its lipid phosphatase activity.

Several kinases have been reported to phosphorylate PTEN. Casein kinase 2 (CK2) mainly phosphorylates Ser370 and Ser385 (Torres and Pulido, 2001), whereas glycogen synthase kinase 3 β (GSK3B) targets Ser362 and Thr366 (Al-Khouri et al., 2005). In contrast to the function of C-terminal tail phosphorylation, it seems that Thr366 phosphorylation can promote PTEN degradation (Maccario et al., 2007). Additionally, glioma tumor suppressor candidate region 2 (GLTSCR2, also known as PICT-1) has been shown to interact with PTEN, enhance its phosphorylation at Ser380 and stabilize the phosphatase (Okahara et al., 2006; Yim et al., 2007). Moreover, RhoA-associated kinase (ROCK) has been shown to phosphorylate PTEN at Ser229, Thr232, Thr319 and Thr321, which are all located in the C2 domain and promote membrane targeting of PTEN in chemoattractant-stimulated leukocytes (Li et al., 2005). Recently, a Src family tyrosine kinase, the fyn-related kinase (FRK, also known and hereafter referred to as RAK) has been reported to interact with PTEN and phosphorylate it on Tyr336, thereby protecting PTEN from proteasomal degradation mediated by neural precursor cell expressed, developmentally downregulated-4 (NEDD4) (Yim et al., 2009).

Similar to phosphorylation, acetylation can also regulate PTEN activity. The histone acetyltransferase K(lysine) acetyltransferase 2B (KAT2B, also known and hereafter referred to as PCAF) has been reported to interact with PTEN and promote PTEN acetylation on Lys125 and Lys128 in response to growth factor stimulation (Okumura et al., 2006). As these residues are within the catalytic pocket, PTEN acetylation by PCAF negatively regulates its enzymatic activity. PTEN is also acetylated on Lys402, which is located within the C-terminal PDZ-domain-binding motif Thr-Lys-Val sequence. This, potentially, affects the interaction between PTEN and PDZ-domain-containing proteins (Ikenoue et al., 2008). CREB-binding protein (CREBBP, also known and hereafter referred to as CBP) and the sirtuin SIRT1 have been identified as the main PTEN acetyltransferase and deacetylase, respectively.

Another mechanism that can potentially regulate the catalytic activity of PTEN is direct oxidation by reactive oxygen species (ROS). ROS can oxidize Cys124 in the active site, thereby forming an intramolecular disulfide bond with Cys71 (Lee et al., 2002). Oxidative inactivation of PTEN has been reported in studies describing the use of hydrogen peroxide or endogenous ROS production in macrophages (Kwon et al., 2004; Leslie et al., 2003). PTEN activity can also be indirectly inhibited by oxidation through modulation of PTEN-binding partners. Oxidation of the antioxidant PARK7 (also known as DJ-1) leads to binding to PTEN and the subsequent inhibition of the PTEN lipid phosphatase activity (Kim et al., 2009).

A few studies have demonstrated the importance of another redox mechanism, S-nitrosylation, in the regulation of PTEN. The level of S-nitrosylation on PTEN substantially increases in the early stages of Alzheimer disease, and this correlates with reduced PTEN protein levels and elevated AKT phosphorylation (Kwak et al., 2010a). Nitric oxide (NO) signaling induces PTEN S-nitrosylation, thereby inactivating the lipid phosphatase, downregulating its protein level through NEDD4-mediated degradation and leading to downstream activation of AKT. Another report has shown that PTEN is selectively S-nitrosylated on Cys83 by low concentrations of NO (Numajiri et al., 2011). Moreover, S-nitrosylated PTEN has been detected in the core and penumbra regions of ischemic mouse brains. S-nitrosylation and downregulation of PTEN activity in this region are likely to act as protective mechanisms through AKT activation.

Using a biochemical purification approach, NEDD4 was identified as an E3 ligase that ubiquitylates PTEN (Wang et al., 2007). NEDD4 physically interacts with PTEN and its overexpression leads to both mono- and polyubiquitylation of PTEN. Interestingly, monoubiquitylation of PTEN appears to be crucial for its nuclear import (Trotman et al., 2007). Consistent with the function of the PTEN C-terminal tail in regulating its stability, deletion of this region makes PTEN a stronger binding partner and better substrate for NEDD4 (Wang et al., 2008).

Two other E3 ligases have been reported to target PTEN, X-linked inhibitor of apoptosis protein (XIAP) (Van Themsche et al., 2009) and WWP2, a NEDD4-like protein family member (Maddika et al., 2011). Although the physiological relevance of these E3 ligases with regards to PTEN function has not yet been defined, it is possible that PTEN is regulated by multiple E3 ligases, depending on the context or special physiological circumstances.

However, in most experimental systems, PTEN appears to be a rather stable protein. Thus, what is the biological relevance of PTEN regulation through ubiquitin-mediated proteasomal degradation? It appears that NEDD4 regulates PTEN in a context-dependent manner to achieve specific functions. For example, NEDD4 is required for neuronal axonal branching in retinal ganglion cells (RGCs) and mainly functions through downregulating PTEN (Drinjakovic et al., 2010). Blocking NEDD4 function severely inhibits terminal branching in RGCs, whereas PTEN knockdown rescues the branching deficiency. Also, NEDD4-mediated PTEN ubiquitylation is essential in the regulataion of PI3K-AKT signaling for neuronal survival in response to Zn²⁺ (Kwak et al., 2010b). Furthermore, in cultured neuronal models, NO signaling not only induces PTEN S-nitrosylation but also results in enhanced PTEN protein degradation through NEDD4-mediated ubiquitylation (Kwak et al., 2010a).

Interestingly, several cellular proteins have been reported to modulate the association between NEDD4 and PTEN, which might provide mechanistic insights into the context-dependent regulation of PTEN by NEDD4. In breast cancer cells, the tyrosine kinase RAK positively regulates PTEN stability by phosphorylating PTEN on Tyr336. This prevents PTEN from binding to NEDD4 and its subsequent degradation (Yim et al., 2009). Recently, the PY (Pro-Pro-x-Tyr)-motif-containing membrane proteins NEDD4-family-interacting proteins 1 and 2 (NDFIP 1 and 2, respectively), which are potent activators of NEDD4 family members, have been shown to promote NEDD4-mediated ubiquitylation and, thus, promote degradation and nuclear import of PTEN (Howitt et al., 2012; Mund and Pelham, 2010).

The enzymatic activity and the biological effects of PTEN are often regulated by its interaction with other proteins. The lipid phosphatase activity of PTEN can be stimulated by binding to p85, the regulatory subunit of PI3K (Chagpar et al., 2010). By contrast, the binding of shank-interacting protein-like 1 (SIPL1) or PtdIns(3,4,5)P₃–RAC-exchanger 2 (PREX2) to PTEN inhibits its lipid phosphatase activity, thereby promoting signaling through the PI3K pathway (Fine et al., 2009; He et al., 2010). Binding of the scaffold protein β-arrestin to PTEN controls the output of distinct PTEN signals during cell proliferation and migration; binding of β-arrestin to PTEN normally activates the lipid phosphatase activity and inhibits cell proliferation. However, following the activation of RhoA as a result of wounding, β-arrestin inhibits the lipid phosphatase-independent anti-migratory function of PTEN. In addition, the binding of several proteins to PTEN can regulate its stability (Lima-Fernandes et al., 2011). These proteins include membrane-associated guanylate kinase inverted 3 (MAGI), GLTSCR2, NDFIP1 and RAK, whose binding affects PTEN phosphorylation status, complex recruitment or NEDD4-mediated ubiquitylation.

The fact that the majority of cellular PTEN is found in the cytosol raises important questions, such as how PTEN is able to execute different cellular functions that often require its membranous lipid phosphatase activity. Accumulating evidence has suggested that a small subpopulation of cellular PTEN is sufficient for each specific biological function. A PTEN subpopulation can be defined by either specific post-translational modification(s) or its interaction with a particular associating factor. It should be emphasized that a functional subpopulation of PTEN might only represent a small portion of total cellular PTEN, thus it might evade evaluation when employing the approaches that are commonly used to monitor total cellular PTEN.

A subpopulation of PTEN is responsible for its lipid phosphatase activity at the plasma membrane. A study describing the use of single-molecule microscopy in living cells has shown that the interaction of PTEN with the plasma membrane is a dynamic process (Vazquez et al., 2006). PTEN binds to the membrane for only a few hundred milliseconds, and this time period is sufficient to hydrolyze PtdIns(3,4,5)P₃. The interaction of PTEN with the membrane requires the N-terminal lipid-binding motif. Membrane interaction is negatively regulated by C-terminal tail phosphorylation, which appears to constrain PTEN in a closed conformation and to limit its association with the membrane. Another study has demonstrated that phosphorylation of the C-terminal tail regulates an intra-molecular interaction between the phosphatase and C2 domains. This determines the percentage of PTEN that is associated with the membrane as well as the resulting PtdIns(3,4,5)P₃ levels (Rahdar et al., 2009). Interestingly, plasma membrane targeting of PTEN greatly enhances PTEN ubiquitylation through NEDD4, and both mono- and polyubiquitylation substantially inhibit PTEN phosphatase activity in vitro, even in the absence of proteasomal degradation (Maccario et al., 2010).

Interestingly, PTEN also localizes into the nucleus, and loss of PTEN nuclear localization has been shown to correlate with cancer progression in certain tissues (Song et al., 2008; Trotman et al., 2007). The nuclear subpopulation of PTEN is controlled by the dynamic regulation of its ubiquitylation, i.e. through NEDD4-mediated monoubiquitylation and deubiquitylation mediated through the network of ubiquitin carboxyl-terminal hydrolase 7 (USP7, also known as HAUSP) and PML (Song et al., 2008; Trotman et al., 2007).

The nuclear subpopulation of PTEN clearly has non-canonical functions (Lindsay et al., 2006). Earlier studies have shown that PTEN regulates p53 protein stability in both phosphatase-dependent and -independent manners (Li et al., 2006). A recent study has shown that genotoxic stress enhances the interaction between PTEN and the p53 family member p73 in the nucleus. This facilitates the binding of p73 to apoptotic promoters and induces the expression of p53-upregulated modulator of apoptosis (PUMA, also known as BBC3) and the apoptosis regulator BAX (Lehman et al., 2011). Nuclear PTEN has also been suggested to maintain chromosomal integrity through the physical interaction with centromere protein C 1 (CENPC1), which is an integral part of the kinetochore (Shen et al., 2007). Whether this function requires the lipid phosphatase activity of PTEN is under debate. More recently, nuclear PTEN has been found to interact with the anaphase-promoting complex/cyclosome (APC/C), an E3 ubiquitin ligase that promotes cell cycle progression by degrading components of the cell cycle machinery. Nuclear PTEN increases APC/C activity by enhancing the association between APC/C and its cofactor, CDH1 (also known as FZR1). This role of nuclear PTEN has been shown to be important for PTEN-mediated tumor suppression (Song et al., 2011).

The function of PTEN as a tumor suppressor has been well established over the past decade. Recent studies have revealed more unexpected – particularly lipid-phosphatase-independent – roles of PTEN in both cancer biology and normal physiology. PTEN activity is tightly regulated by various mechanisms on the transcriptional, translational and post-translational levels. Accumulating evidence suggests that PTEN can exert diverse physiological and pathological functions through its context-specific regulation. Furthermore, a highly regulated and dynamic subpopulation of cellular PTEN might be required specifically for a given contextual function of PTEN. In summary, to understand the molecular basis of contextual regulation of PTEN is crucial in order to understand the biological function of PTEN and, subsequently, to develop appropriate approaches to therapeutically target PTEN signaling.