Phosphoprotein enriched in astrocytes 15 kDa (PEA-15), also known as PED-15, is a 15 kDa protein that is highly expressed in the nervous system with particularly high levels in astrocytes and neurons of the hippocampus. Chneiweiss and colleagues first characterized PEA-15 as a major substrate for Protein Kinase C (PKC) in astrocytes and later cloned the cDNA encoding the protein (Araujo et al. 1993; Estelles et al. 1996). The first indications of the functional significance of PEA-15 was identified when it was shown to be a phosphoprotein overexpressed in skeletal muscle in patients with type 2 diabetes (Condorelli et al. 1998). In this context, PEA-15 was shown to inhibit insulin-stimulated glucose transport, indicating it may be involved in the development of type 2 diabetes. PEA-15 was further found by expression cloning to inhibit H-Ras signaling to integrins (Ramos et al. 1998). This work provided an additional potential connection to cancer and the first suggestion that PEA-15 may modulate cell adhesion. This was quickly followed by the first two reports showing that PEA-15 can act as an antiapoptotic protein that blocks cell death by preventing Fas-Associated via death dominion protein (FADD) recruitment and activation of Caspase 8 in response to Fas or tumor necrosis factor receptor (TNFR) activation (Condorelli et al. 1999; Kitsberg et al. 1999). Thus, the initial papers already made it clear that PEA-15 is a multifunctional protein affecting many distinct cell signaling pathways. That has been borne out by more recent publications, indicating roles in inducing autophagy (Böck et al. 2010), regulating the DNA damage-induced cell cycle checkpoint (Nagarajan et al. 2014), and controlling hypoxia-induced apoptosis (Mergenthaler et al. 2012).
PEA-15 Structure and Binding Partners
PEA-15 binds several proteins known to regulate apoptosis, transcription, and proliferation and thereby affects these processes (Greig and Nixon 2014). Among identified interacting partners, some of the most extensively characterized are extracellular signal-regulated kinase 1 and 2 (ERK1/2) MAP Kinase, FADD, (Phospholipase D) PLD, ribosomal protein S6 kinase 2 (RSK2), and AKT. PEA-15 binding to ERK and FADD is regulated by phosphorylation at the two serines (Sulzmaier et al. 2012). This indicates that phosphorylation is one mechanism by which PEA-15 binding to its partners is determined. PEA-15 can also simultaneously bind ERK and the ERK substrate RSK2 and thereby act as a scaffold to target ERK to RSK2 (Vaidyanathan et al. 2007). PEA-15 binding to FADD acts predominantly to block FADD binding to and activation of Caspase 8 while binding to PLD appears to serve to activate that lipase (Fig. 1) (Fiory et al. 2009). Moreover, PEA-15 interaction with the JNK MAP kinase leads to JNK activation and induction of autophagy in glioma cells (Böck et al. 2010). In this way, the function of PEA-15 is determined by its binding partners.
PEA-15 in Glucose Transport
PEA-15 was identified by differential display to be overexpressed in the adipose and skeletal muscle tissues, as well as in skin fibroxblasts of type 2 diabetic patients and their first-degree relatives. Moreover, PEA-15 expression in skeletal muscle impairs glucose transporter type 4 (GLUT4) translocation and thereby inhibits insulin-stimulated glucose transport. This established early on that PEA-15 likely plays a significant role in type 2 diabetes (Condorelli et al. 1998). Subsequently, transgenic mice overexpressing PEA-15 were found to have some of the hallmarks of type 2 diabetes. These transgenic mice had impaired glucose tolerance, resistance to insulin action on glucose disposal, reduced insulin-stimulated glucose transport in fat and skeletal muscles, and impaired insulin effects on GLUT4 membrane translocation (Fiory et al. 2009). These effects may be caused in part by the ability of PEA-15 to directly bind and stabilize phospholipase D (PLD), thereby inducing PKC activity (Fiory et al. 2009). This hypothesis is further supported by the observation that inhibition of PEA-15 binding to PLD restores insulin-stimulated glucose uptake. Thus, targeting the PEA-15/PLD1 interaction may provide a new approach to improving sensitivity to insulin action in diabetics (Fiory et al. 2009).
PEA-15 in Apoptosis
PEA-15 regulates apoptosis by inhibiting the death receptor-activated extrinsic cascade. PEA-15 is expressed at particularly high levels in astrocytes and its N-terminal DED supports its interaction with other DED-containing proteins including FADD and possibly Caspase 8. This was an early indicator that PEA-15 may play an antiapoptotic role in these cells. Indeed, sensitivity to TNF-dependent apoptosis is significantly increased in astrocytes from PEA-15-null mice (Kitsberg et al. 1999). Moreover, this unusual sensitivity to TNF can be repaired by exogenous expression of PEA-15 in the knockout astrocytes. In separate experiments, PEA-15 was also reported to block the extrinsic apoptotic pathway in human mammary carcinoma cells (MCF-7). The likely mechanism is that by binding to both FADD and Caspase 8, PEA-15 inhibits their interaction and thereby prevents the formation of the death-inducing signaling complex (DISC). Thus, Caspase 8 and downstream caspases are not activated. This is the same mechanism suggested for the antiapoptotic effects of another DED family protein called cFLIP. The antiapoptotic effect of PEA-15 appears to extend to other cancer cell types as well. Human malignant glioma cells are often highly resistant to TNF-related apoptosis-inducing ligand (TRAIL)-induced apoptosis. Because of this, these gliomas cannot be treated with chemotherapies based on TRAIL activation. However, knockdown of PEA-15 in these cells restores their sensitivity to TRAIL (Fiory et al. 2009; Valmiki and Ramos 2009). These results suggest that reducing PEA-15 levels in some cancer cells may improve their response to death receptor-based therapies.
Phosphorylation at Ser116 stabilizes PEA-15 levels and promotes its interaction with FADD, while phosphorylation at Ser-104 abrogates PEA-15 binding to ERK. Moreover, only the dually phosphorylated PEA-15 is found in the death-inducing signaling complex (DISC) in TRAIL-resistant glioma cells, suggesting that dual phosphorylation is necessary for its protective role (Fiory et al. 2009). Indeed, treatment with inhibitors of PKC (known to phosphorylate PEA-15 at Ser104) abrogated the antiapoptotic function. Similarly, inhibitors of AKT/PKB block phosphorylation of PEA-15 at Ser116 and increase PEA-15 degradation leading to elevated TRAIL-mediated apoptosis. Thus, dual phosphorylation of PEA-15 is essential for it to block extrinsic apoptosis initiated by Fas, TNFR, or TRAIL (Fiory et al. 2009). PEA-15 levels are also regulated in apoptosis by the serine protease Omi/HtrA2. This pro-apoptotic mitochondrial protease is released in response to different cellular stresses and exerts pro-apoptotic function by promoting Caspase 3 activation through its protease activity. Omi/HtrA2 also promotes apoptosis by degrading PEA-15 and thus preventing PEA-15 antiapoptotic activity (Fiory et al. 2009). Hence, PEA-15 antiapoptotic function is regulated by phosphorylation, ubiquitination, and proteolytic degradation of the protein.
In addition to regulating extrinsic apoptotic cascades, PEA-15 also inhibits stress-induced apoptotic responses to oxidative agents, serum deprivation, and anisomycin treatment. In human kidney embryonic cells (HEK293), overexpression of PEA-15 decreased anisomycin- and H 2O 2-induced apoptosis and inhibited the phosphorylation of JNK1/2 and p38. Impaired activity of these stress kinases by PEA-15 correlated with inhibition of stress-induced Cdc-42, MKK4, and MKK6 activation (Fiory et al. 2009). Importantly, PEA-15 in combination with hexokinase II (HKII) (a HIF-1–regulated glycolytic enzyme) acts as a molecular switch that inhibits apoptosis after hypoxia. While HKII accelerates apoptosis in the absence of PEA15 and under glucose deprivation. In this way, a HKII/PEA-15 complex senses glucose availability and promotes cell survival during hypoxia but induces apoptosis under normoxia when glucose is not present (Mergenthaler et al. 2012).
PEA-15 is therefore well established as an important antiapoptotic protein that functions by multiple mechanisms to promote cell survival.
PEA-15 in the ERK Mitogen-Activated Protein Kinase (MAPK) Pathway
PEA-15 in Cancer Development and Progression
PEA-15 is reported to play diverse roles in many types of cancer including mammary, ovarian, skin, and glioma (Fiory et al. 2009; Ramos 2005). However, the effects of PEA-15 expression are not the same in each cancer, and indeed PEA-15 is reported to act in the manner of both a tumor promoter and a tumor suppressor. The functions of PEA-15 that make it particularly interesting in cancer research are its effects on apoptosis, chemoresistance, senescence, adhesion, and migration.
Cancer cells frequently become resistant to apoptosis-mediated cell death. PEA-15 can effectively block apoptosis induced by death receptors in many cellular contexts. In glioma, B cell carcinoma, as well as breast and small cell lung cancer, increased PEA-15 expression levels correlate with resistance to death receptor-induced apoptosis (Fiory et al. 2009). Similarly, PEA-15 inhibits apoptosis in a skin carcinoma mouse model by abrogating TPA-dependent Caspase 3 activation (Fiory et al. 2009). The ability of PEA-15 to prevent apoptosis suggests it could be a valuable target for the development of drugs to be used in combinatorial therapy to make cancer cells more prone to elimination. Thus, PEA-15 effects on apoptosis may partly explain cancer resistance to radiation- and chemotherapy-induced cell death.
Cancers also typically develop the ability to undergo increased, uncontrolled cell proliferation in the absence of apoptosis or senescence. PEA-15 binds to ERK, prevents ERK translocation into the nucleus, and thereby impairs ERK-dependent transcription and proliferation. This function of PEA-15 appears to be key in the ability of PEA-15 to limit proliferation of ovarian and breast cancers (Bartholomeusz et al. 2006). PEA-15 may also prevent proliferation of some cancers by inducing cellular senescence as described in fibroblasts. In both instances, PEA-15 levels may be controlled by the adenovirus E1A. In ovarian cancer E1A upregulates PEA-15, which prevents ERK nuclear signaling and thereby decreases cell proliferation (Bartholomeusz et al. 2006). Conversely, in mouse embryo fibroblasts, E1A downregulates PEA-15, promoting ERK nuclear localization and thereby preventing Ras-induced senescence. Thus, PEA-15 restriction of ERK to the cytoplasm is necessary in these cells for Ras to induce senescence. This provides an alternative method by which PEA-15 can prevent tumor cell proliferation even in the presence of an oncogene by promoting senescence (Fiory et al. 2009). In several cancers, PEA-15 is de-ubiquitinated and its expression stabilized by COPS5 in an ATM kinase-dependent manner (Nagarajan et al. 2014). In these cells, PEA15 expression oscillates throughout the cell cycle, and PEA15 loss accelerates cell cycle progression by activating CDK6 expression. Further supporting PEA-15 as a tumor suppressor, cells that lack PEA15 have a DNA damage-induced G2/M checkpoint defect due to increased CDC25C activity and higher (CDK1)/cyclin B activity. PEA-15 overexpression blocks oncogenic Ras transformation and PEA-15 is downregulated in colon, breast, and lung cancer (Nagarajan et al. 2014). Thus, PEA-15 can suppress tumorigenesis by negatively regulating tumor cell proliferation.
The initial discovery of PEA-15 as an inhibitor of Ras/MAPK signaling to integrins was the first indication of its potential role in cell adhesion, migration, and invasion. Since then PEA-15 has been reported to inhibit mammary tumor invasion (Glading et al. 2007), astrocytoma migration and invasion (Renault-Mihara et al. 2006), and neuroblastoma invasion (Gawecka et al. 2012) by differing mechanisms. Immunohistochemical microarray analysis shows that PEA-15 expression is correlated with decreased invasive behavior of breast carcinoma, and this effect seems to relate to the nuclear localization of activated ERK1/2. In these tumors, PEA-15 inhibits invasion by preventing ERK translocation to the nucleus. Indeed, PEA-15 mutants that cannot bind or sequester ERK are unable to inhibit invasion. Additionally, membrane-localized ERK1 that is unable to translocate to the nucleus also decreases invasion. These results reveal the significance of nuclear entry of ERK1/2 in tumor behavior and indicate that the PEA-15 inhibitory effect on cell invasion depends on its ability to bind ERK (Glading et al. 2007). Additionally, the use of organotypic culture analysis of highly invasive primary astrocytomas revealed that only tumor cells expressing low levels of PEA-15 migrated away from the originating explants. Similarly, PEA-15-null astrocytes demonstrated significantly elevated motility compared to wild-type control cells and this was reversed by transfection of PEA-15. Moreover, mouse embryo fibroblasts transfected with PEA-15 displayed reduced migration (Renault-Mihara et al. 2006). Pharmacological treatments in these experiments excluded participation of ERK1/2, PI3K/AKT, CamK II in this effect of PEA-15. Rather inhibition of astrocyte migration is dependent on PEA-15 downregulation of PKCδ (Fiory et al. 2009; Renault-Mihara et al. 2006). Finally, in neuroblastoma tumor cells, elevated PEA-15 expression impaired ERK activation of RSK2 and subsequent RSK2-driven migration (Gawecka et al. 2012). Independently of the mechanism, the loss of cell invasiveness strongly suggests a role for PEA-15 as a potential metastasis suppressor.
High PEA-15 expression correlates to better prognosis for several cancer types. Women with high PEA-15-expressing ovarian tumors survive longer than those with low PEA-15-expressing tumors. PEA-15 was thus proposed to be a potentially important prognostic marker in ovarian cancer (Fiory et al. 2009). Similarly, PEA-15 expression correlates with World Health Organization (WHO) grading criteria for astrocytic tumors. High PEA-15 levels correlate with lower grade tumors. Hence here too PEA-15 may prove to be a useful prognostic marker (Watanabe et al. 2010). Finally, PEA-15 expression levels inversely correlate with cell motility and invasiveness in astrocytomas, neuroblastomas, and mammary carcinomas (Gawecka et al. 2012; Glading et al. 2007; Renault-Mihara et al. 2006). Thus, PEA-15 expressions tend to reduce proliferation and invasion in some cancers and can thereby act as a tumor suppressor in these tumors. Whether PEA-15 acts to increase or decrease tumorigenesis therefore likely depends upon its phosphorylation status and the affected signaling pathways and microenvironment of the tumor.
PEA-15 in the Immune Response and Spatial Learning
Normal PEA-15 function has been investigated using PEA-15 null mice (Kitsberg et al. 1999). The PEA-15 null mice are born at Mendelian ratios, appear grossly normal, and did not evidence any abnormal brain size, neuron number, or structural defects. PEA-15 is highly expressed throughout the brain and regulates ERK, RSK2, and CREB function. As these proteins are important in cognitive function, the PEA-15 null mice were examined for nervous system function and cognitive defects. In these studies, the effects of loss of PEA-15 were investigated in a series of experiments designed to measure stress activity, as well as learning and sensory functions. PEA-15 knockout mice exhibit impaired spatial learning, while their fear conditioning, passive avoidance, egocentric navigation, and odor discrimination are normal. Additionally, PEA-15 knockout mice exhibit impaired forepaw strength. The knockout mice were generally normal with normal weight, pain sensitivity, and coordination and normal visual, auditory, and olfactory abilities (Ramos et al. 2009). These cognitive defects may be the result of abnormal ERK- and RSK2-dependent CREB transcription, but this remains to be tested.
PEA-15 null mice were noted by Dr. Chneiweiss to suffer enlarged spleens and lymph nodes when kept in uncontrolled antigen-rich environments. This largely went away when the mice were kept in controlled Hepa-filtered cages. Thus, a potential role for PEA-15 in the immune response was investigated. PEA-15 null mice have no significant defects in thymic or splenic lymphocyte cellularity or differentiation. However, activation of PEA-15 null T cells results in hyperproliferation in comparison to wild-type littermates. This increased proliferation results at least in part from increased ERK translocation into the nucleus and the resultant elevated activation of IL-2 transcription and secretion (Pastorino et al. 2010). In vitro studies conducted in Jurkat T cells confirm that PEA-15 negatively regulates T-cell receptor signaling restricting cell proliferation, ERK nuclear translocation, and IL-2 transcription (Pastorino et al. 2010). A second paper examining the PEA-15 null mice found similar effects of PEA-15 deletion on ERK localization, but different effects on production of cytokines such as Il-2 (reduced) and IFNγ (Kerbrat et al. 2015). In addition, they found that TCR-stimulated PEA-15 null CD4(+) T cells exhibited defective progression through the cell cycle associated with impaired expression of cyclin E and phospho Rb, two ERK1/2-dependent cell cycle proteins. These differences may be due to differences in the genetic background of the animals and intensity of TCR stimulation. Interestingly, there is no indication that apoptosis is altered in the PEA-15 null immune cells. Hence, PEA-15 is a novel player in T-cell homeostasis and thereby contributes to the regulation of the immune responses. Significantly, the effects of PEA-15 deletion on both cognitive function and immune response can be explained by changes in ERK MAP kinase signaling.
PEA-15 modulates signaling in extrinsic apoptosis, ERK MAP kinase, and PLD pathways. This leads to changes in cell death, autophagy, proliferation, and motility among others. Because of these effects, PEA-15 plays an important role in normal glucose metabolism, the immune response, and learning, while its abnormal expression is associated with type 2 diabetes and cancer. Therefore, it will be important to determine if manipulation of PEA-15 is a valid therapeutic approach in diabetes and cancer. However, the fact that PEA-15 expression can both enhance and impair cancer progression and influence immune function suggests this be pursued with caution. Before progressing, the context of any PEA-15 directed therapy must be well understood along with a better understanding of PEA-15 normal function. To this end, much work is still needed.
- Kerbrat S, Vingert B, Junier MP, Castellano F, Renault-Mihara F, Dos Reis TS, et al. Absence of the Adaptor Protein PEA-15 Is Associated with Altered Pattern of Th Cytokines Production by Activated CD4+ T Lymphocytes In Vitro, and Defective Red Blood Cell Alloimmune Response In Vivo. PLoS One. 2015;10:e0136885. doi:10.1371/journal.pone.0136885.PubMedPubMedCentralCrossRefGoogle Scholar
- Mergenthaler P, Kahl A, Kamitz A, van Laak V, Stohlmann K, Thomsen S, et al. Mitochondrial hexokinase II (HKII) and phosphoprotein enriched in astrocytes (PEA15) form a molecular switch governing cellular fate depending on the metabolic state. Proc Natl Acad Sci U S A. 2012;109:1518–23. doi:10.1073/pnas.1108225109.PubMedPubMedCentralCrossRefGoogle Scholar