Phospholipase C-ϵ Augments Epidermal Growth Factor-dependent Cell Growth by Inhibiting Epidermal Growth Factor Receptor Down-regulation*

The down-regulation of the epidermal growth factor (EGF) receptor is critical for the termination of EGF-dependent signaling, and the dysregulation of this process can lead to oncogenesis. In the present study, we suggest a novel mechanism for the regulation of EGF receptor down-regulation by phospholipase C-ϵ. The overexpression of PLC-ϵ led to an increase in receptor recycling and decreased the down-regulation of the EGF receptor in COS-7 cells. Adaptor protein complex 2 (AP2) was identified as a novel binding protein that associates with the PLC-ϵ RA2 domain independently of Ras. The interaction of PLC-ϵ with AP2 was responsible for the suppression of EGF receptor down-regulation, since a perturbation in this interaction abolished this effect. Enhanced EGF receptor stability by PLC-ϵ led to the potentiation of EGF-dependent growth in COS-7 cells. Finally, the knockdown of PLC-ϵ in mouse embryo fibroblast cells elicited a severe defect in EGF-dependent growth. Our results indicated that PLC-ϵ could promote EGF-dependent cell growth by suppressing receptor down-regulation.

Growth factor receptors control a wide variety of biological processes, including cell proliferation, differentiation, survival, and migration. The ligand-induced activation of receptor tyrosine kinases leads to the assembly of signaling protein complexes and the subsequent activation of downstream signaling pathways. The inappropriate activation and aberrant signaling of particular growth factor receptors have been associated with a number of diseases, including cancers. There are over 30 receptor tyrosine kinases tightly associated with cancer pathogenesis (1). Ligand-induced receptor down-regulation plays a key role in the tight regulation of the intensity and duration of receptor tyrosine kinase signaling to avoid aberrant cellular stimulation. Ligand-induced activation of receptors leads to the recruitment of the endocytic machinery and subsequently to endosomal sorting, which can result in lysosomal degradation of the receptor or in the recycling of the receptor back to the cell surface (2). The dysregulation of receptor tyrosine kinase down-regulation has recently been suggested as another potential cause of neoplastic growth. For example, alterations that uncouple receptor tyrosine kinases from c-Cbl-mediated ubiquitination and thereby from down-regulation are tightly associated with cancer pathogenesis (3). In addition, the aberrant expression or dysregulation of the machinery for the endocytosis or the lysosomal targeting of receptor tyrosine kinases is directly implicated in tumorigenesis. Hip1 interacts with AP2, 2 phosphoinositides, and clathrin, and it is thought to play a fundamental role in clathrin trafficking (4). The overexpression of Hip1 results in cell transformation and increased proliferation, which result from up-regulated EGF receptor levels (5). The Tsg101, a subunit of ESCRT-1, which is involved in lysosomal targeting, is deleted in some human cancers, and the functional inactivation of Tsg101 leads to the transformation of NIH3T3 cells (6,7).
AP2 plays an important role in the ligand-mediated downregulation of various receptors. It is a heterotetrameric complex (␣, ␤2, 2, and ␦2), and each subunit mediates various interactions. The ␣ and 2 subunits bind to phosphatidyl 4,5bisphosphate. The 2 subunit is also responsible for the sorting signal of surface cargo molecules, and the ␤2 subunit mediates AP2 interaction with clathrin (8 -12). AP2 utilizes these multiple interactions to help coat assembly events at the plasma membrane to ensure the highly selective internalization of surface receptors (13).
Phospholipase C (PLC) is activated by various growth factors or G protein-coupled receptor ligands and hydrolyzes phosphatidyl 4,5-bisphosphate into inositol 1,4,5-trisphosphate and diacylglycerol, which are implicated in calcium mobilization and protein kinase C activation, respectively (14). PLC-⑀ was identified to be a mammalian counterpart of PLC210, which was found to be a binding protein of LET-60, the nematode Ras (15). PLC-⑀ contains a CDC25 homology domain at its amino terminus and a pair of RA domains at the carboxyl terminus (16 -18). These structural features suggest that PLC-⑀ is involved in the signaling promoted by the Ras superfamily GTPases. In fact, the RA domain of PLC-⑀ specifically binds to the activated Ras, or Rap1A, which is known to be an important step in the activation and subcellular localization of PLC-⑀ (17,19).
In this study, we suggest a novel function of PLC-⑀ that is independent of its PLC activity. PLC-⑀ suppressed the downregulation of the EGF receptor through an RA domain-mediated interaction with AP2. The PLC-⑀-dependent up-regulation of the EGF receptor led to an increase in EGF-induced cell growth, which provided evidence that the novel mechanism was an involvement in the mitogenic role of PLC-⑀.

MATERIALS AND METHODS
Antibodies-Rabbit polyclonal antibody of PLC-⑀ was generated by immunizing CDC25 domain of PLC-⑀ obtained from Escherichia coli or obtained from Dr. Tohru Kataoka (Kobe University). Other antibodies used were mouse monoclonal anti-2 antibody (Transduction Laboratories, Lexington, KY), mouse monoclonal anti-␣-adaptin antibody (Transduction Laboratories, Lexington, KY), mouse monoclonal anti-FLAG antibody (Sigma), and rabbit polyclonal anti-EGF receptor antibody (Santa Cruz Biotechnology, Inc., Santa Cruz, CA). Rhodamine-conjugated goat anti-rabbit IgG and fluorescein isothiocyanate-conjugated goat anti-mouse IgG were purchased from Sigma. Horseradish peroxidase-conjugated goat anti-rabbit IgG and goat anti-mouse IgA, IgM, and IgG were from Kirkegaard & Perry Laboratories (Gaithersburg, MD).
Cell Culture-COS-7 cells were grown in DMEM containing 10% bovine calf serum (BCS), antibiotics, and glutamine, and OVCAR-3 cells and immortalized MEF cells were grown in DMEM containing 10% fetal bovine serum, antibiotics, and glutamine. Cells were grown to ϳ90% confluence for immunoprecipitation and Western blot experiments or 50% confluence for immunofluorescence experiments.
Yeast Two-hybrid Screening-PLC-⑀ RA domains (amino acids 1990 -2218) were cloned into the pLexA (BD Clontech) in frame with the LexA DNA-binding domain (referred to as pLexA-PLC-⑀). The yeast strain, EGY48, carrying the reporter gene was cotransformed with the bait plasmid, pLexA-PLC-⑀ and a human HeLa cDNA library fused to the VP16 activation domain. Transformation was carried out using the lithium acetate method (20). Leucine-positive colonies were identified by a filter-lifting assay for ␤-galactosidase activity. Library-derived DNA was prepared from candidate clones and analyzed by DNA sequencing.
EGF Receptor Internalization Assay-After 24 h of serum starvation, cells were incubated with 1 ng/ml of 125 I-labeled EGF for 5 min at 37°C. After incubation, cells were washed rapidly three times with cold DMEM. The cells were then incubated for 5 min with 0.2 M acetic acid (pH 2.8) containing 0.5 M NaCl at 4°C. The acid wash was combined with another short rinse in the same buffer, and used to determine the amount of surface-bound 125 I-EGF. The cells were lysed in 1 M NaOH to determine the intracellular radioactivity. Internalized 125 I-EGF was expressed as a percentage of the total (internalized and surface-bound).
EGF Receptor Recycling Assay-After 24 h of serum starvation, cells were incubated with 125 I-labeled EGF for 15 min at 37°C. Cells were rinsed with binding medium (1% bovine serum albumin in DMEM), and surface-bound EGF was removed by incubation with acidic solution (0.15 M NaCl, 0.1 M glycine, pH 3.0) for 5 min two times. Cells were rinsed again and chased with binding medium containing 100 ng/ml cold EGF for 15 min. The medium and cells were collected after the incubation. Cells were harvested by scraping with 1 M NaOH. Proteins in the medium were precipitated with 20% trichloroacetic acid, and cells (internalized EGF), medium pellet (recycled EGF), and medium supernatant (degraded EGF) were counted in a ␥-counter. Each data point was collected in duplicate. Recycled 125 I-EGF was expressed as a percentage of total (internalized, degraded, and recycled).
Analysis of EGF Trafficking-After 24 h of serum starvation, cells were treated with 0.5 g/ml rhodamine-conjugated EGF for the indicated times. To inhibit EGF receptor recycling, cells were pretreated with 100 M monensin for 30 min before the addition of 0.5 g/ml rhodamine-conjugated EGF for 30 min. After incubation, cells were washed, fixed, and stained for the indicated antibodies.
Thymidine Incorporation Assay-Cells were seeded in triplicate into 6-well plates at a density of 2 ϫ 10 5 cells/well and were transfected with PLC-⑀ siRNA or control scrambled siRNA. After 24 h, the cells were incubated with serum-free DMEM for 24 h to reach quiescence. The cells were then incubated in serum-free medium supplemented with 100 ng/ml EGF for 18 h prior to the addition of [ 3 H]thymidine for an additional 6 h. Thymidine incorporation was measured as previously reported (21).
Immunofluorescent Labeling and Confocal Microscopy-COS-7 cells expressing FLAG-tagged PLC-⑀ were cultured on poly-L-lysine slides. After incubation with rhodamine-EGF, cells were washed and fixed with 4% paraformaldehyde. After blocking with 1% horse serum in PBS, the cells were incubated with anti-FLAG antibody diluted in blocking solution for 3 h at room temperature. After extensive washing in PBS, the cells were incubated for an additional 1 h with a fluorescein isothiocyanate-conjugated secondary antibody. To visualize recycled EGF receptors in the plasma membrane, we expressed EGFR-RFP in the presence or absence of PLC-⑀ in COS-7 cells. After incubation for 12 h, the cells transfected with EGFR-RFP alone were mixed with the cells transfected with both EGF receptor and PLC-⑀. After serum starvation for 24 h, cells were either fixed or incubated with EGF for 15 min. After incubation of EGF, cells were either fixed or washed with cold PBS and acidic solution (0.15 M NaCl, 0.1 M glycine, pH 3.0) and then chased for another 15 min at 37°C to induce EGF receptor recycling. The fixed cells were stained for PLC-⑀ with anti-FLAG antibody. Cells were viewed with a Zeiss LSM 510 confocal microscope equipped with LSM 510 version 2.02 software, an argon (458and 488-nm) laser, and a helium/neon (543-and 633-nm) laser.
Plasmid Construction and Mutagenesis-FLAG-tagged mouse PLC-⑀ DNA was a generous gift from Dr. Tohru Kataoka (Kobe University). The lipase-inactive mutant of PLC-⑀ (referred to as PLC-⑀ LIM) was generated by substitution of His 1142 with Leu. PLC-⑀ ⌬CT was generated by PCR amplification of PLC-⑀ fragment containing amino acids . For the construction of the AP2 binding-deficient mutant, KLKE (residues 2214 -2217) sequence was changed into AAAA, or each amino acid in KLKE was changed into Ala by site-directed mutagenesis. Mouse 2 DNA was kindly provided by Dr. Jacques H Camonis (Institut Curie) and introduced into pcDNA-HA vector by PCR amplification. EGFP-2 was a generous gift from Dr. Margarita Martin (University of Barcelona). Previously, the 2 subunit tagged with EGFP (2-EGFP) was found to integrate into the endogenous AP2 complex by transient expression (22). The EGF receptor tagged with C-terminal RFP (EGFR-RFP) was previously described (23). For the inhibition of PLC-⑀-AP2 interaction, oligonucleotides corresponding to binding region peptide (GKFILKLKEQVQ; 2211-2220) were synthesized and subcloned into pFLAG-CMV2 vector (Sigma). To generate siRNA-resistant mouse PLC-⑀ cDNA, 5Ј-TAT TCC TAC AGC-3Ј was changed into 5Ј-TAC AGT TAT TCA-3Ј using site-directed mutagenesis.

RESULTS
PLC-⑀ Interacts with AP2-PLC-⑀ contains a CDC25 homology domain and RA domains, which are not found in other PLC isozymes (16 -18). These domains have been considered to be important regulatory modules for the specific function of PLC-⑀; thus, the elucidation of protein complexes involving these domains may lead to a better understanding of the roles of PLC-⑀. Given that the sequence conservation between RA domain family members is quite low, it has been suggested that the RA domains of PLC-⑀ may interact with other proteins rather than with members of the small GTPase superfamily (26). We sought to identify binding partners of the PLC-⑀ RA domain and performed a yeast two-hybrid analysis using bait containing the serial RA1 and RA2 domain of PLC-⑀. We screened ϳ5 ϫ 10 6 clones from a HeLa cell cDNA library. In total, five types of clones interacted specifically when tested for nutritional selection and ␤-galactosidase activity. Three of these clones were identified as the 2 subunit of AP2. The cDNA encoded amino acids 160 -385 of 2. AP2 is a heterotet-rameric complex that acts as an adaptor molecule during endocytosis (8).
We attempted to confirm the interaction between PLC-⑀ and AP2 using biochemical analysis (Fig. 1, A-C). As shown in Fig.  1A, immunoprecipitation of PLC-⑀ confirmed the interaction of endogenous AP2 with PLC-⑀ in P19 embryonal carcinoma cells. In addition, Fig. 1B demonstrated that deletion of the C terminus containing the RA1 and RA2 domain of PLC-⑀ abolished this interaction; however, a lipase-inactive mutant showed the same affinity for AP2 in COS-7 cells. We performed in vitro binding analysis using purified PLC-⑀ RA domain and 2 to examine whether interaction between PLC-⑀ and AP2 is direct or mediated by other proteins. As shown in Fig. 1C, GST pull-down analysis demonstrated that AP2 directly binds with the PLC-⑀ RA domain.
We compared the intracellular localization of endogenous PLC-⑀ and AP2 in cells in order to provide additional evidence . PLC-⑀ was immunoprecipitated with ␣-FLAG antibody, and the immune complex was subjected to immunoblotting with the indicated antibodies. C, the GST-RA domain (amino acids 1990 -2218) and His-2 were purified from E. coli. The GST-RA domain or GST was incubated with His-2, and bound 2 was detected with ␣-His antibody. Asterisks, nonspecific bands. D, COS-7 cells were transfected with FLAG-PLC-⑀ and EGFP-2. Cells were fixed and stained with ␣-FLAG antibody and then visualized with Rhodamine (b) for PLC-⑀ and GFP (a) for AP2, respectively, and the merged image is shown in c. The arrows indicate co-localization of PLC-⑀ and AP2.
for an association between PLC-⑀ and AP2 (Fig. 1D). In a previous study, immunostaining analysis with the ␣-PLC-⑀ antibody showed strong perinuclear staining and a less pronounced distribution throughout the cytoplasm in various cells (27). We used the 2 subunit tagged with EGFP (2-EGFP), which was found to integrate in the endogenous AP2 complex by transient expression (22). As shown in Fig. 1D, PLC-⑀ was observed around the perinuclear area and cytoplasm, which is consistent with the previous finding. 2 was dispersed throughout the cytoplasm as punctuate structures and also detected in the plasma membrane. Interestingly, PLC-⑀ was also detected at those punctuate structures around the perinuclear area and cell periphery. These results suggest that the association of the RA domain with the 2 subunit induces the formation of the PLC-⑀-AP2 complex in a PLC-⑀ lipase activity-independent manner.
RA2 Domain of PLC-⑀ Binds with AP2 in a Ras-independent Manner-The RA domain of PLC-⑀ is composed of the RA1 and RA2 domains, which are homologous and structurally related to one another. However, the critical charged residues in the interface of the RA2 domain that would face Ras are not conserved in the RA1 domain; thus, the RA1 domain is unable to interact with Ras family proteins (28). We first tried to elucidate the binding region of the RA domain in order to determine the functional role of the interaction between PLC-⑀ and AP2 (Fig. 2, A and B). As shown in Fig. 2B, GST pull-down analysis revealed that the RA2 domain alone is responsible for the interaction with the 2 subunit of AP2. We deleted either the N terminus or C terminus from the RA2 domain in an effort to further narrow down the binding region ( Fig. 2A). As shown in Fig. 2B, the deletion of the N terminus of the RA2 domain had no effect on its interaction with AP2; however, the deletion of the C terminus abolished the interaction between AP2 and the RA2 domain. In particular, the deletion of the last four amino acids in the RA2 domain (KLKE; 2214 -2217) was sufficient to abolish the interaction between AP2 and the RA2 domain. We then changed each amino acid or all four amino acids of KLKE into alanines and examined the interaction between PLC-⑀ and AP2. As shown in Fig. 2C, the E2217A and AAAA mutants did not co-immunoprecipitate with AP2 in the cells, which revealed that the glutamate residue of KLKE in the RA2 domain is essential to binding.
Recent structural analysis revealed that the ␤1, ␤2, and the ␣1-␣3 loop of the RA2 domain interact with switch I and switch II of Ras (28). The AP2 binding region is located at the end of ␤5. We examined whether Ras affects the interaction between AP2 and the RA2 domain (Fig. 2, D and E). As shown in Fig. 2D, we initially expressed constitutively active Ras (Ras V12) or a dominant negative mutant (Ras N17) of Ras in COS-7 cells and performed a pull-down analysis using GST-RA2 fusion proteins. The RA2 domain effectively associates with endogenous AP2 as well as with Ras, and co-incubation with Ras V12 did not change the association between the RA2 domain and AP2. Similarly, Fig. 2E shows that neither the constitutively active form nor the dominant negative form of Ras had an effect on the association between AP2 and PLC-⑀ in the cells. Overall, these results suggest that the RA2 domain of PLC-⑀ has a specific binding module that is utilized for the interaction with AP2 as well as that for Ras and that each interaction seems to be independent.
PLC-⑀ Promotes the Recycling of EGF Receptor-AP2 is an important adaptor molecule that is involved in the endocytosis of cell surface receptors. This led us to investigate whether PLC-⑀ can affect the endocytosis of cell surface receptors. Among various receptors, we focused on EGF receptor endocytosis, since EGF stimulation induces the translocation of PLC-⑀ into plasma membrane and subsequent hydrolysis of PIP 2 (17). To trace the endocytosis and trafficking of EGF receptors, COS-7 cells transfected with PLC-⑀ cDNA were incubated with rhodamine-EGF for various times at 37°C (Fig. 3A). The ␣-FLAG antibody and fluorescein isothiocyanate-conjugated secondary antibody were used to detect the expression of FLAG-tagged PLC-⑀. As shown in Fig. 3A (ac), the initial uptake of rhodamine-EGF was not altered by the overexpression of PLC-⑀. However, after a 30-min incubation at 37°C (Fig. 3A, g-i), the amount of intracellular rhodamine-EGF was significantly decreased in the PLC-⑀-transfected cells. We used monensin, a carboxylic ionophore that disrupts the recycling of internalized EGF receptors, to determine whether the reduced amount of EGF in PLC-⑀transfected cells was the result of enhanced EGF receptor recycling (Fig. 3A, j-l). Pretreatment with monensin restored the accumulation of rhodamine-EGF, which suggests the possibility that PLC-⑀ can promote the recycling of EGF receptors. We sought to measure the effects of PLC-⑀ on EGF receptor recycling. First, we examined whether PLC-⑀ affects the internalization of EGF receptors using 125 I-EGF (Fig. 3B). Consist- After serum starvation for 24 h, cells were incubated with rhodamine-labeled EGF (10 ng/ml) for the indicated times at 37°C (a-l). The recycling inhibitor monensin was pretreated 30 min before incubation with rhodamine-EGF (j-l). The cells were then fixed and stained for PLC-⑀ with anti-FLAG antibody. B, COS-7 cells were transfected with empty vector or PLC-⑀ plasmid. After 24 h of serum starvation, cells were incubated with 125 I-labeled EGF (1 ng/ml) for 5 min at 37°C. After incubation, the cells were washed rapidly three times with cold DMEM. The cells were then incubated for 5 min with 0.2 M acetic acid (pH 2.8) containing 0.5 M NaCl at 4°C. The acid wash was combined with another short rinse in the same buffer and used to determine the amount of surface-bound 125 I-EGF. The cells were lysed in 1 M NaOH to determine the intracellular radioactivity. Internalized 125 I-EGF was expressed as a percentage of total (internalized and surface-bound). C, COS-7 cells were transfected with empty vector or PLC-⑀ plasmid. After 24 h of serum starvation, cells were incubated with 125 I-labeled EGF (1 ng/ml) for 15 min at 37°C. The cells were then rinsed with binding medium, and surface-bound EGF was removed. The cells were then incubated with binding medium containing 100 ng/ml cold EGF for 15 min. Recycled 125 I-epidermal growth factor was expressed as a percentage of the total (internalized, degraded, and recycled). D, COS-7 cells were transfected with EGFR-RFP in the presence or absence of PLC-⑀ DNA. After incubation for 12 h, the cells transfected with EGFR-RFP alone were mixed with the cells transfected with both the EGFR-RFP and PLC-⑀. After serum starvation for 24 h, the cells were either fixed (a-c) or incubated with EGF for 15 min. After incubation of EGF, the cells were either fixed (d-f) or washed with cold PBS and chased for another 15 min at 37°C to induce EGF receptor recycling (g-i). The fixed cells were stained with anti-FLAG antibody to visualize PLC-⑀-transfected cells. The arrows indicate the cells transfected with PLC-⑀. The arrowheads indicate the EGF receptors recycled to the plasma membrane. E, OVCAR-3 cells were transfected with PLC-⑀ siRNA or control scrambled siRNA. Cell lysates were prepared and subjected to immunoblotting with the indicated antibodies. F, OVCAR-3 cells were transfected with PLC-⑀ siRNA or control scrambled siRNA. After serum starvation for 24 h, EGF receptor internalization was measured as described in B. G, OVCAR-3 cells were transfected with PLC-⑀ siRNA or control scrambled siRNA. After serum starvation for 24 h, EGF receptor recycling was measured as described in C. IB, immunoblot. ent with the results shown in Fig. 3A, EGF receptor internalization was not altered by PLC-⑀ overexpression. We then measured EGF receptor recycling by monitoring the trafficking of internalized 125 I-EGF (Fig. 3C). The results indicated that the extrinsic expression of PLC-⑀ can lead to an increase in EGF receptor recycling to the cell surface.
To further confirm the effects of PLC-⑀ on EGF receptor recycling, we expressed RFP-tagged EGF receptors with PLC-⑀ and tried to detect recycled EGF receptors on the plasma membrane (Fig. 3D). As shown in Fig. 3D (ac), the EGF receptor was primarily localized in the plasma membrane before stimulation, and the incubation of EGF for 15 min led to the almost complete internalization of the EGF receptors in both PLC-⑀transfected cells and nontransfected cells (Fig. 3D, d-f). After removing EGF from the media, the cells were chased for another 15 min to induce EGF receptor recycling (Fig. 3D, g-i). EGF receptors located in the plasma membrane were prominently detected in PLC-⑀transfected cells compared with nontransfected cells, which supports the positive role of PLC-⑀ in EGF receptor recycling.
We then examined whether endogenous PLC-⑀ is involved in EGF receptor recycling using the siRNA-mediated reduction of PLC-⑀ in cells (Fig. 3, E and G). Fig.  3E indicates that PLC-⑀ is specifically reduced by the PLC-⑀ siRNA without affecting the levels of other PLC isozyme, such as PLC-␥1 or PLC-␤3, in OVCAR-3 cells. As shown in Fig. 3, F and G, the internalization of EGF receptor was not altered by PLC-⑀ knockdown, but recycling was significantly decreased by PLC-⑀ knockdown, which demonstrates that endogenous PLC-⑀ is involved in the enhancement of EGF receptor recycling. Taken together, these results suggest that PLC-⑀ can promote EGF receptor recycling.
PLC-⑀ Suppresses EGF Receptor Degradation-Internalized receptors are delivered to early endosomes, which mature into multivesicular bodies, where EGF receptors undergo sorting to the plasma membrane for recycling or to the lysosome for destruction (2). We examined whether PLC-⑀ overexpression affected the degradation of EGF receptors (Fig. 4). Fig. 4A showed that following the addition of EGF to vector-transfected cells, total receptor levels as detected by Western blotting with an ␣-EGF receptor antibody were significantly reduced over a time period of 15-90 min. However, cells expressing PLC-⑀ showed a marked reduction in the rate of EGF receptor down-regulation. We examined whether PLC-⑀ affects the synthesis of EGF receptor after EGF stimulation, thus leading to the higher remaining EGF receptor levels in PLC-⑀-transfected cells compared with vectortransfected cells. As shown in Fig. 4B, the blockade of protein synthesis by cycloheximide did not abolish the PLC-⑀-dependent increase of remaining EGF receptor level after EGF stimulation, which suggests that PLC-⑀ is involved in the EGF receptor degradation pathway and not in the synthesis pathway. After serum starvation for 24 h, cells were incubated with EGF (10 ng/ml) for the indicated times at 37°C. The cells were then lysed and prepared for the immunoblotting (IB) with the indicated antibodies. The EGF receptor level after EGF stimulation was quantified and expressed as a percentage of the EGF receptor level of unstimulated control cells. B, COS-7 cells were transfected with empty vector or PLC-⑀ DNA. After 24 h of serum starvation, cells were pretreated with either cycloheximide (CHX) or Me 2 SO for 30 min and then incubated with EGF (10 ng/ml) for 30 min at 37°C. The cells were then lysed and prepared for the immunoblotting with the indicated antibodies. The EGF receptor level after EGF stimulation for 30 min was quantified and expressed as a percentage of the EGF receptor level of unstimulated control cells. Results are shown as the means Ϯ S.D. C, OVCAR-3 cells were transfected with control scrambled siRNA or PLC-⑀ siRNA. After serum starvation, cells were treated with EGF (10 ng/ml) for the indicated times, and immunoblotting was performed using the indicated antibodies. The EGF receptor level after EGF stimulation was quantified and expressed as a percentage of the EGF receptor level in unstimulated control cells.
We then examined EGF receptor degradation in PLC-⑀ siRNA-transfected cells in order to investigate the role of endogenous PLC-⑀ (Fig. 4C). Consistent with results from the overexpression of PLC-⑀, EGF receptor degradation was accelerated by PLC-⑀ knockdown. These results demonstrate that PLC-⑀ plays a negative role in EGF receptor degradation.
The Interaction between PLC-⑀ and AP2 Is Responsible for the Decreased EGF Receptor Down-regulation-We investigated whether the PLC-⑀-dependent suppression of EGF receptor down-regulation was the result of the binding between AP2 and PLC-⑀ (Fig. 5). As shown in Fig. 5A, overexpression of wild type PLC-⑀ or a lipase-inactive mutant of PLC-⑀ led to increased EGF receptor recycling, but the expression of the PLC-⑀ E2217A mutant did not increase the recycling of the EGF receptor. Concomitantly, as shown in Fig. 5B, EGF receptor degradation was suppressed in cells transfected with wild type PLC-⑀ and the lipase-inactive mutant in comparison with cells transfected with a vector or with E2217A mutant. These results are consistent with the findings of the recycling experiments and suggest that PLC-⑀ affects EGF receptor down-regulation by interacting with AP2 and that PLC activity is not required for this process.
We attempted to block the interaction between endogenous proteins in order to further confirm the importance of the AP2-PLC-⑀ interaction in the regulation of EGF receptor trafficking (Fig. 5, C and D). To block the binding between PLC-⑀ and AP2, we constructed a fusion construct encompassing 10 amino acids (2211-2220) in the AP2-binding region of the RA domain. Fig. 5C demonstrates that transfection of this construct in COS7 cells effectively inhibited the interaction between PLC-⑀ and AP2 (bottom). Furthermore, this construct suppressed EGF receptor recycling (top). In addition, as shown in Fig. 5D, EGF receptor degradation was facilitated by the expression of this construct. These results further support the conclusion that the PLC-⑀-mediated regulation of EGF receptor down-regulation is dependent on the interaction between PLC-⑀ and AP2.
We then investigated whether knockdown of AP2 affects PLC-⑀-dependent promotion of EGF receptor recycling. 2 siRNA effectively reduced AP2 level in MEF cells (Fig. 5E). As shown in Fig. 5F, knockdown of 2 subunit increased EGF receptor recycling. Furthermore, the PLC-⑀-dependent increase of EGF receptor recycling was abolished by AP2 knockdown.TheresultsconfirmedthatAP2isinvolvedinPLC-⑀dependent regulation of EGF receptor recycling.
PLC-⑀ Promotes EGF-dependent Thymidine Incorporation by Up-regulating the EGF Receptor-The enhanced recycling of EGF receptors can reportedly lead to increased EGF-dependent proliferation (29,30). We investigated whether PLC-⑀ overexpression promotes EGF-dependent thymidine incorporation by increasing EGF receptor recycling (Fig. 6). As shown in Fig.  6A, the expression of wild type PLC-⑀ or a lipase-inactive mutant of PLC-⑀ significantly increased the EGF-dependent thymidine incorporation of COS-7 cells; however, the E2217A mutant had no effect. These results suggest that PLC-⑀ is capable of promoting EGF-dependent cell growth by an RA domainmediated interaction with AP2.
We then examined whether PLC-⑀ knockdown led to the suppression of EGF-dependent thymidine incorporation. Fig.  6B indicates that PLC-⑀ knockdown in MEF cells led to a sig-  (n ϭ 3). E, MEF cells were transfected with the indicated siRNAs. After incubation for 36 h, the cells were lysed and prepared for the immunoblotting with the indicated antibodies. F, MEF cells were transfected with indicated siRNAs. After serum starvation, cells were incubated with 125 I-labeled EGF (1 ng/ml) for 15 min at 37°C. The cells were rinsed with binding medium, and surface-bound EGF was removed. Then cells were then incubated with binding medium containing 100 ng/ml cold EGF for 15 min. Recycled 125 I-EGF was expressed as a percentage of the total (internalized, degraded, and recycled). WT, wild type; IB, immunoblot. nificant inhibition of EGF-induced thymidine incorporation, which is recovered by the add-back of PLC-⑀ construct, which is resistant to the PLC-⑀ siRNA. In all, these results corroborate the contribution of PLC-⑀ in growth factor-dependent cell growth.

DISCUSSION
Thus far, the RA domain of PLC-⑀ was considered to be a binding module for Ras family G-proteins, and it was thought to play an important role in the activation of PLC-⑀ (17,19). However, our yeast two-hybrid analysis revealed that this domain interacts with various binding proteins as well as with Ras family proteins. Among the novel binding partners of the RA2 domain of PLC-⑀, AP2 plays a key role in the PLC-⑀-dependent regulation of the EGF receptor. Although both AP2 and Ras utilized the RA2 domain as a binding region for PLC-⑀, the binding of AP2 with the RA2 domain was not affected by Ras, and vice versa. The reported NMR and x-ray crystallography analyses indicate that the AP2 binding region of RA2 corresponds to the end of the ␤5 sheet, which is separate from the interaction surface for switch I or switch II of Ras (28). In addition, EGF receptor down-regulation was suppressed in cells transfected with the Ras-binding deficient mutant of PLC-⑀ (K2151E/K2153E) (data not shown). We speculate that the RA2 domain of PLC-⑀ can mediate novel binding in a manner independent of Ras family proteins and that it may be involved in the novel function of PLC-⑀.
There have been several reports on the role of PLC-⑀ in cell growth. PLC-⑀ was shown to induce platelet-derived growth factor-dependent cell growth in BaF3 cells expressing a platelet-derived growth factor receptor mutant that was incapable of activating PLC-␥ (31). Carcinogen-induced skin tumor formation was also reduced in PLC-⑀ knock-out mice, which suggests a mitogenic function of PLC-⑀ (32). In contrast, SW480 cells derived from colorectal tumors microinjected with PLC-⑀ DNA showed a reduction in cell proliferation and an increase in cell death (27). Our results support the mitogenic function of PLC-⑀, since the expression of PLC-⑀ elicits EGF-dependent cell growth, and a reduction in PLC-⑀ expression in several cancer cells led to a reduction in cell growth. However, the mechanism underlying PLC-⑀-dependent cell growth elucidated in this study is different from those of previous reports, suggesting that small G-protein-dependent PLC-⑀ activity is critical for cell growth or tumor formation (31,32). In the present study, we showed that PLC-⑀ increased EGF receptor recycling, which enhanced the receptor-mediated cell growth. The lipase activity of PLC-⑀ was dispensable for this process. Recent studies have established that the enhanced recycling of surface receptors effectively leads to augmented ligand-induced proliferation by continuously replenishing the surface receptor levels necessary for prolonged signaling (29,30).
Stimulation of cells with growth factors induces the activation of both PLC-␥1 and PLC-⑀, but the isozyme-specific functions remain elusive (33). Our analysis revealed that both PLC-␥1 and PLC-⑀ are involved in different stages of EGF receptor down-regulation. We previously determined that PLC-␥1 is a guanine nucleotide exchange factor for dynamin-1 and that it is important for the internalization of the EGF receptor (34). PLC-⑀ altered the endosomal trafficking of EGF receptors and enhanced receptor recycling. PLC-⑀ effectively inhibits EGF receptor down-regulation in PLC-␥1 knock-out cells, which implies that PLC-⑀ regulates EGF receptor trafficking independent of PLC-␥1 (Supplemental Fig. 1). Both PLC-␥1 and PLC-⑀ seem to promote mitogenic signaling via different underlying mechanisms. Although PLC-␥1 activates endosomal Erk and downstream signaling, PLC-⑀ blocks the termination of EGF receptor signaling, leading to cell growth.
Previous reports have suggested that AP2 can be detected in early endosomes and that the EGF receptor-AP2 interaction may persist after the internalization of the receptor (35). Beck et al. (36) suggested the possible role of AP2 in endosome formation through their in vitro vesicle aggregation experiments. . Effect of PLC-⑀ on EGF-dependent cell growth. A, COS-7 cells were transfected with the indicated plasmids. After 24 h, cells were incubated with serum-free DMEM for 24 h to reach quiescence. Cells were incubated in serum-free medium supplemented with 100 ng/ml EGF for 18 h prior to the addition of 3 H-labeled thymidine for an additional 6 h. B, MEF cells were transfected with PLC-⑀ siRNA in the presence or absence of the PLC-⑀ construct, which is resistant to siRNA-mediated knockdown. After serum starvation for 24 h, cells were incubated in serum-free medium supplemented with 100 ng/ml EGF for 18 h prior to the addition of 3 H-labeled thymidine for an additional 6 h. Thymidine incorporation was measured as described under "Materials and Methods." IB, immunoblot; wt, wild type. Furthermore, the L1010A/L1011A EGF receptor mutant, which is incapable of associating with AP2, showed a reduction in EGF-induced degradation, although its internalization rate was unchanged (37). Also, our analysis revealed that knockdown of endogenous AP2 increased the EGF receptor recycling in MEF cells (Fig. 5F). Taken together, these results suggest that AP2 can promote the endosomal trafficking of the EGF receptor to degradative pathways. PLC-⑀ seems to associate with the AP2 on endocytic vesicles and thereby negatively regulate the function of AP2 in EGF receptor trafficking to degradative pathways. The interaction between PLC-⑀ and AP2 is induced by EGF stimulation (Supplemental Fig. 2), and this suggests that PLC-⑀ is recruited into AP2 engaged in EGF receptor trafficking. The fact that the 2 subunit of AP2 is responsible for the interaction with EGF receptor suggests that PLC-⑀ may inhibit the EGF receptor-AP2 interaction. The detailed mechanism underlying the PLC-⑀-dependent regulation of EGF receptor trafficking remains to be revealed.
The results of our study demonstrate that PLC-⑀ promotes growth factor-dependent cell growth. They also provide evidence for a novel function of PLC-⑀ in the up-regulation of growth factor receptors.