Centaurin-α1 Is a Phosphatidylinositol 3-Kinase-dependent Activator of ERK1/2 Mitogen-activated Protein Kinases*

Centaurin-α1 is known to be a phosphatidylinositol 3,4,5-triphosphate (PIP3)-binding protein that has two pleckstrin homology domains and a putative ADP ribosylation factor GTPase-activating protein domain. However, the physiological function of centaurin-α1 is still not understood. Here we have shown that transient expression of centaurin-α1 in COS-7 cells results in specific activation of ERK, and the activation is inhibited by co-expression of a dominant negative form of Ras. We have also found that a mutant form of centaurin-α1 that is unable to bind PIP3 fails to induce ERK activation and that a phosphatidylinositol 3-kinase inhibitor LY294002 inhibits centaurin-α1-dependent ERK activation. Furthermore, transient knockdown of centaurin-α1 by small interfering RNAs results in reduced ERK activation after epidermal growth factor stimulation in T-REx 293 cells. These results suggest that centaurin-α1 contributes to ERK activation in growth factor signaling, linking the PI3K pathway to the ERK mitogen-activated protein kinase pathway through its ability to interact with PIP3.

sequence and the transcription initiation site in the H1 promoter (15,16).
Cell Cultures and Transfection-COS-7 cells, 293-EBNA cells (Invitrogen), and T-REx 293 cells (Invitrogen) were cultured in Dulbecco's modified Eagle's medium (DMEM) containing 5 or 10% fetal calf serum. The cells were maintained in 5% CO 2 at 37°C. The cells were seeded on a 6 wells/cell culture dish at a density of 1.5 ϫ 10 5 cells/well. Transfection of these cells was performed using Lipofectamine Plus reagent (Invitrogen) or FuGENE 6 reagent (Roche Applied Science) according to the manufacturer's protocols. The total amount of DNA was adjusted using pSR␣-HA, pSR␣-Myc, or pcDNA3.1(ϩ) vector.
EGF Stimulation-3 h after transfection using Lipofectamine Plus reagent, the medium was exchanged to DMEM containing 10% fetal calf serum, and the cells were incubated for 18 h. Thereafter, the cells were cultured in low serum medium (DMEM supplemented with 0.35% glucose and 1% heat-inactivated horse serum) for 20 h. The cells were stimulated with 30 nM EGF for 2 min at 37°C.
Kinase Assays and Immunoblotting-21 h after transfection, COS-7 cells were lysed and immune complex kinase assay was performed as previously described (17). The following proteins were used as substrates: myelin basic protein for ERK2 and ERK5, glutathione S-transferase-c-Jun-  for c-Jun N-terminal kinase, and glutathione S-transferase-ATF2-  for p38. After SDS-PAGE, radioactivity was analyzed using an image analyzer (Bio-Rad). The samples were subjected to immunoblotting with anti-HA, anti-GFP, or anti-Myc antibodies. Immunoblotting was also performed as previously described (17).
Cell Staining-Cells were fixed in 3.7% formaldehyde for 10 min. The coverslips were washed three times with phosphate-buffered saline for 5 min. Fluorescence images were observed using a Confocal Microscope (Bio-Rad).
Luciferase Assays-293-EBNA cells were seeded into 5% fetal calf serum-containing DMEM in a 96-well cell culture plate to a final density of 1 ϫ 10 4 cells/well and cultured for 24 h. Thereafter, the cells were transfected with 50 ng of pcDNA3.1-human centaurin-␣ 1 , 60 ng of pFR-Luc (Stratagene), 0.25 ng of pFA2-Elk-1 (Stratagene), and 5 ng of siRNA expression plasmid or an empty vector in each well using FuGENE6. Twenty-four h after transfection, the reporter activity was measured using PicaGene LT2.0 (Toyo Ink) according to the manufacturer's instructions.
Establishment of Cell Lines Expressing siRNA for Centaurin-␣ 1 -The siRNA expression plasmids were introduced into T-REx 293 cells using FuGENE 6. Twenty-four h after transfection, the medium was exchanged to DMEM containing 1 g/ml puromycin and 5% fetal calf serum. Seven days later, the puromycin-resistant cells were collected and used as cell lines in which siRNAs of human centaurin-␣ 1 were induced by doxycycline.
Quantitative Analysis of Centaurin-␣ 1 mRNA-The quantitative real-time PCR reaction was performed with an ABI PRISM 7000 sequence detection system according to the manufacturer's instructions. The sequence of dual fluorophore-labeled probe for human cen-FIGURE 1. Centaurin-␣ 1 activates ERK specifically in a Ras-dependent manner. A, COS-7 cells were co-transfected with HA epitope-tagged or GFP-fused centaurin-␣ 1 and each of four kinds of Myc-tagged MAP kinases, ERK2, ERK5, c-Jun N-terminal kinase (JNK), and p38. The activities of these MAP kinases were analyzed by in vitro kinase assays after immunoprecipitation with an anti-Myc antibody. Total lysates were also analyzed by immunoblotting with anti-HA, anti-GFP, and anti-Myc antibodies. B, COS-7 cells were co-transfected with Myc-tagged centaurin-␣ 1 , Myc-tagged ERK2, and each of four kinds of Myc-tagged dominant negative forms of low molecular weight G-proteins, Ras, Cdc42, Rho, and Rac. 21 h after transfection, the phosphorylation of ERK was detected by immunoblotting with an anti-phosphorylated ERK antibody. Total lysates were also immunoblotted with an anti-Myc antibody. As a positive control, a MEK inhibitor U0126 was added in the culture medium of COS-7 cells co-transfected with centaurin-␣ 1 and ERK.

RESULTS
Centaurin-␣ 1 Activates ERK Specifically-In the previous report, it has been shown that overexpression of human centaurin-␣ 1 can activate ERK in mammalian cells (9). To examine whether this activation is specific for the ERK MAP kinase cascade among the four major MAP kinase cascades, HA epitope-tagged or GFP-fused centaurin-␣ 1 was coexpressed with four kinds of Myc-tagged MAP kinases, ERK2, ERK5, c-Jun N-terminal kinase, or p38 in COS-7 cells, and the activity of these MAP kinases was measured by in vitro kinase assays. As a result, only ERK2 showed the enhanced kinase activity when co-expressed with centaurin-␣ 1 (Fig. 1A). This result indicates that centaurin-␣ 1 is a specific activator of ERK among four MAP kinases.
Ras Is Involved in the Activation of ERK by Centaurin-␣ 1 -It is known that Ras, a low molecular weight GTP-binding protein, activates the Raf-MEK-ERK kinase pathway. To examine whether Ras is involved in the activation of ERK by centaurin-␣ 1 , we co-transfected COS-7 cells with centaurin-␣ 1 and each of four kinds of dominant negative forms of low molecular weight G-proteins Ras, Cdc42, Rho, and Rac and measured the activity of ERK MAP kinase. As a result, centaurin-␣ 1 failed to activate ERK when co-expressed with a dominant negative form of Ras (Fig. 1B). On the other hand, dominant negative forms of other low molecular weight G-proteins did not affect the activation of ERK by centaurin-␣ 1 (Fig. 1B). A MEK inhibitor U0126 also inhibited the activation of ERK by centaurin-␣ 1 . These results indicate that the activation of ERK by the overexpression of centaurin-␣ 1 in COS-7 cells is mediated by Ras.
PH Domains in Centaurin-␣ 1 Are Important for ERK Activation-Centaurin-␣ 1 has two PH domains in its N-terminal region. It has been reported that the PH domains are necessary for the ability of centaurin-␣ 1 to bind PIP 3 (13). Moreover, the RC1RC2 mutant form of centaurin-␣ 1 , in which Arg-149 and Arg-273 in the PH domains are mutated to Cys, is no longer recruited to the plasma membrane after EGF stimulation (18). To evaluate the role of the PH domains in the activation of ERK by centaurin-␣ 1 , we constructed the RC1RC2 mutant and examined its ability to activate ERK. Overexpression of the RC1RC2 mutant resulted in much weaker activation of ERK than that of wild-type centaurin-␣ 1 in COS-7 cells ( Fig. 2A). To investigate the localization of centaurin-␣ 1 in the cell and to study the role of the PH domain for the translocation of centaurin-␣ 1 by EGF stimulation, COS-7 cells were transfected with a GFP-fused form of wildtype or RC1RC2 mutant centaurin-␣ 1 . Before and after EGF stimulation, we determined their subcellular localization using the laser scanning confocal microscope. Before the stimulation, both wildtype and the RC1RC2 mutant centaurin-␣ 1 largely localized to the nucleus and a small part of wild-type centaurin-␣ 1 localized to the plasma membrane (Fig. 2B). After the stimulation, although wildtype centaurin-␣ 1 was strongly recruited to the plasma membrane, the RC1RC2 mutant did not change their localization (Fig. 2B). These results suggest that the recruitment of centaurin-␣ 1 to the plasma membrane, which is regulated by its PH domains, is important for the activation of ERK in COS-7 cells.
LY294002 Inhibits Centaurin-␣ 1 -dependent Activation of ERK-As PIP 3 , which binds to the PH domain, is a product of PI3K, PI3K activity may be required for ERK activation by centaurin-␣ 1. We therefore examined the effect of a PI3K inhibitor, LY294002, on this centaurin-␣ 1 -dependent activation of ERK. COS-7 cells were co-transfected with centaurin-␣ 1 and ERK and treated with LY294002. Immunoblot analysis showed that LY294002 is able to inhibit the phosphorylation of ERK in a dose-dependent manner (Fig. 3A). This indicates that PI3K activity is required for the activation of ERK induced by centaurin-␣ 1 . To investigate the role of PI3K activity on the subcellular localization of centaurin-␣ 1 , COS-7 cells expressing GFP-fused centaurin-␣ 1 were treated with 20 M LY294002. Although a small part of GFP-fused centaurin-␣ 1 localized to the plasma membrane in the absence of LY294002, centaurin-␣ 1 failed to localize to the plasma membrane in the presence of LY294002 (Fig. 3B). Furthermore, the translocation of wild-type centaurin-␣ 1 to the plasma membrane by EGF stimulation was not detected in the presence of LY294002 (Fig. 3C). A MEK inhibitor U0126 did not inhibit the recruitment of centaurin-␣ 1 to the plasma membrane (Fig.  3C). These results indicate that the EGF-dependent recruitment of centaurin-␣ 1 requires PI3K activation but not ERK activation. We also demonstrated that exogenous PIP 3 is able to induce the translocation of centaurin-␣ 1 to the plasma membrane, whereas PI(4,5)P 2 does not induce such translocation (Fig. 3D). Without the histone carrier, phosphoinositides do not permeate the cell membrane (Fig. 3D). These results suggest that the PI3K-and PIP 3 -dependent recruitment of centaurin-␣ 1 to the plasma membrane is required for the activation of ERK induced by centaurin-␣ 1 .

Down-regulation of Centaurin-␣ 1 by siRNA Suppresses the Activation of ERK by EGF-
To study the function of endogenous centaurin-␣ 1 , we employed the siRNA approach. For the purpose of screening the sequences of siRNA that have an ability to knock down the expression of centaurin-␣ 1 mRNA, 293-EBNA cells were co-transfected with each of the siRNA expression plasmids, a centaurin-␣ 1 expression plasmid, pFR-Luc, and pFA2-Elk-1. As a result, we found that two sequences (si3 and si5) of 25 sequences we tested were able to suppress the reporter gene activity to less than half of that in the control experiment in which the empty vector was used instead of an siRNA expression vector (Fig. 4). The other 23 sequences including si4 had little or no effect on the reporter activity ( Fig. 4 and data not shown).
We tried to obtain 293-EBNA stable cell lines that constitutively express the double-stranded RNA having the sequence si3, si4, or si5. Although we were able to establish si4-expressing cell lines, we were unable to obtain si3-or si5-expressing cell lines. This may suggest that centaurin-␣ 1 plays an important role in cell survival or proliferation. Thereafter, we constructed the T-REx 293 cell lines in which the expression of short hairpin RNA (shRNA) of si3, si4, or si5 is induced by doxycycline. In T-REx 293 cells, the tetracycline repressor is constitutively expressed, and therefore shRNA is not expressed in the absence of doxycycline. To know the suppressive effect of siRNA on the expression of endogenous centaurin-␣ 1 mRNA, we examined the expression level of centaurin-␣ 1 mRNA in these cell lines after doxycycline treatment. As shown in Fig. 5A, si3 and si5 significantly attenuated the expression level of endogenous centaurin-␣ 1 mRNA. On the other hand, si4 had only a small effect on the mRNA level (Fig. 5A). Next, we examined the contribution of centaurin-␣ 1 in EGF signaling. The inducible siRNAexpressing cells were transfected with pFR-Luc, pFA2-Elk-1, and phRL-tk 24 h after doxycycline treatment. One day after the transfection, the cells were stimulated with EGF for 3 h, and luciferase activity was measured (Fig. 5B). The cells expressing shRNA si3 or si5 (but not those expressing si4) showed markedly decreased luciferase activity compared with the cells without induction of siRNA. This result indicates that centaurin-␣ 1 is required for the EGF-dependent activation of Elk-1 in T-REx 293 cells. To examine whether centaurin-␣ 1 is necessary for ERK activation by EGF, the phosphorylation state of ERK was analyzed after EGF stimulation with or without the pretreatment of cells with doxycycline. As shown in Fig. 5C, phosphorylation of ERK by EGF stimulation was decreased when the expression of centaurin-␣ 1 was reduced by siRNA si3 or si5. In contrast, we could not detect the decrease of phosphorylated ERK in cells in which si4 was expressed (Fig.   5C). This result indicates that centaurin-␣ 1 contributes to the phosphorylation of ERK in the EGF signaling pathway.

DISCUSSION
It is well known that, after growth factor stimulation, both the PI3K pathway and the ERK MAP kinase pathway are activated (19,20).   Ͻ 0.001). B, each cell line was pretreated with or without doxycycline (500 ng/ml) for 24 h and co-transfected with pFR-Luc, phRL-tk, and pFA2-Elk-1. 24 h after transfection, the cells were stimulated with EGF (0.1 ng/ml) for 3 h. Thereafter, the cells were lysed and luciferase activities were analyzed. Renilla luciferase levels were used as an internal control. Shown is the mean Ϯ S.E. from five experiments. C, each cell line was pretreated with or without doxycycline (500 ng/ml) for 48 h and stimulated with EGF (1 ng/ml) for 15 min. The cells were lysed, and the proteins were separated by SDS-PAGE and immunoblotted with anti-phosphorylated ERK and anti-ERK antibodies. The graphs show the ratio of the density of phosphorylated ERK to that of total ERK. Values are means Ϯ S.E. from five experiments (*, p Ͻ 0.05; **, p Ͻ 0.01).
Although these two pathways are activated independently, cross-talk between the PI3K and ERK pathways has been reported (21). However, the signaling molecules linking the PI3K pathway to the ERK MAP kinase cascade have been unclear. In this report, we have shown that centaurin-␣ 1 may act as an adaptor molecule that links the PI3K activation to the ERK MAP kinase cascade. Our results reported here suggest that, upon growth factor stimulation, PIP 3 , which is produced by activated PI3K, promotes the recruitment of centaurin-␣ 1 to the plasma membrane via the PH domains of centaurin-␣ 1 , and thereafter centaurin-␣ 1 activates Ras, resulting in the activation of the ERK MAP kinase cascade (see Fig. 6). Molecular mechanisms by which centaurin-␣ 1 activates Ras remain elucidated.
When centaurin-␣ 1 is overexpressed in COS-7 cells, a small part of wild-type centaurin-␣ 1 localized to the plasma membrane without EGF stimulation. This may be due to the unnaturally excess amount of centaurin-␣ 1 protein in the cells. Given that the RC1RC2 mutant form of centaurin-␣ 1 hardly localized to the plasma membrane and that LY294002 treatment decreased the localization of centaurin-␣ 1 to the plasma membrane, the basal activity of PI3K and the steady-state level of PIP 3 abundance may contribute to this localization.
In RNAi experiments, we used the HEK293 cell line to evaluate the role of centaurin-␣ 1 in EGF signaling, as we found that both centaurin-␣ 1 and the EGF receptor are endogenously expressed in this cell line (data not shown). The large scale functional genomics approaches identified centaurin-␣ 1 as a putative AP-1 regulatory oncogene and suggested that centaurin-␣ 1 could function to augment cell proliferation in a primary cell type and mediate cellular transformation (22). Our results demonstrating that centaurin-␣ 1 has an ability to activate the ERK MAP kinase cascade can well explain this oncogenic activity of centaurin-␣ 1 . In a more recent report (12), it was shown that centaurin-␣ 1 negatively regulates ARF6 activity through its function as an in vivo PIP 3 -dependent ARF6 GAP. In this case, centaurin-␣ 1 negatively regulates membrane trafficking and actin cytoskeleton architecture by inactivating ARF6. It has also been reported that centaurin-␣ 1 is a neuronal actinbinding protein and regulates the actin cytoskeleton (23). Given that one of the functions of centaurin-␣ 1 is to activate ERK MAP kinase, centaurin-␣ 1 may regulate the cell fate by promoting cell proliferation and inhibiting various cellular activities, including membrane trafficking and actin cytoskeleton organization.
In humans, centaurin-␣ 1 is strongly expressed in the brain, especially in neurons (24). In Alzheimer disease, the intraneuronal expression of centaurin-␣ 1 protein was shown to be elevated. It has also been reported that the immunostaining of centaurin-␣ 1 is concentrated in neuritic plaques, the neuropathological hallmark of Alzheimer disease (25,26). Centaurin-␣ 1 is for the most part in the nucleus, but during neuritogenesis centaurin-␣ 1 becomes localized more in the neurites and the cytoplasm. This may suggest the possibility that centaurin-␣ 1 recruited to the plasma membrane plays some role in neurite outgrowth (24). The activation of ERK by centaurin-␣ 1 may contribute to these phenomena. Interestingly, when we overexpressed centaurin-␣ 1 in COS-7 or PC12 cells, we often observed a number of neurite-like processes growing around the cell membrane. 4 Centaurin-␣ 1 is expressed ubiquitously and could regulate cell fate by acting as an adaptor molecule linking the PI3K pathway to the ERK MAP kinase cascade. The present functional characterization of centaurin-␣ 1 could provide new insights into the understanding of growth factor signaling in mammalian cells. . Centaurin-␣ 1 is involved in ERK activation by EGF stimulation. The EGF signal is transduced via two pathways; one involves centaurin-␣ 1 . Centaurin-␣ 1 is recruited to the plasma membrane in response to PI3K activation by EGF and activates Ras resulting in ERK activation.