Siah Proteins Induce the Epidermal Growth Factor-dependent Degradation of Phospholipase Cϵ^*

Under the control of cell surface receptors, phosphoinositide-specific phospholipase C (PLC)² isozymes hydrolyze phosphatidylinositol 4,5-bisphosphate to generate two intracellular products, inositol 1,4,5-trisphosphate and diacylglycerol, which are implicated in calcium mobilization and protein kinase C activation, respectively. So far, 14 PLC isoforms have been cloned in mammals. Based on their functional and structural characteristics, they have been grouped into five classes: PLCβ (β1-β4), PLCδ (δ1-δ4), PLCγ (γ1 and γ2), PLCϵ, PLCζ, and PLCη (η1 and η2) (1-3).

PLCϵ plays a role in the interplay between PLC and small GTPases. Various G-proteins directly activate PLCϵ. For example, RhoA was found to stimulate PLCϵ activity by interacting with a 65-amino acid insert within the catalytic core of PLCϵ (4). Lysophosphatidic acid and thrombin stimulates PLCϵ by activating Gα₁₂ and/or Gα₁₃ and downstream RhoA (5). Moreover, Ras-family GTPases activate PLCϵ by binding with the RA domain of PLCϵ. Adrenaline and prostaglandin E₁ have been reported to activate PLCϵ by triggering adenylyl cyclase-coupled receptors, and Rap2B, which is regulated by Epac (a guanine-nucleotide-exchange factor regulated by cAMP), has been found to associate with the RA domain of PLCϵ during PLCϵ activation (6, 7). In addition, EGF treatment was found to induce an association between the PLCϵ RA domain and activated Ras, and this resulted in the recruitment of PLCϵ into the plasma membrane for activation (8).

Several physiological studies have indicated that PLCϵ is involved in development and cell growth. PLCϵ knockout mice had cardiac dysfunction resulting from defective heart development or were susceptible to hypertrophy in response to chronic cardiac stress (9, 10). Recently, mutations of PLCϵ in individuals with severe nephrotic syndrome were identified, and PLCϵ knockdown in zebra fish led to a loss of the filtration barrier maintained by glomerular podocytes, which in combination demonstrate the importance of PLCϵ in kidney development and function (11). Moreover, ablation of PLCϵ activity in mice led to reduced carcinogen-induced skin tumor formation, providing evidence that PLCϵ plays a positive role in tumor cell growth (12). Furthermore, the overexpression of PLCϵ in BaF3 cells induced platelet-derived growth factor-dependent cell growth (13).

Siah proteins are homologues of Drosophila SINA, which is a ring-finger protein involved in R7 cell development in the eye (14). Three murine (Siah1a, Siah1b, and Siah2) and two human (SIAH1 and SIAH2) homologues have been identified. The mammalian Siah proteins are highly homologous; Siah1a and Siah1b are 98% identical, whereas Siah1 proteins and Siah2 protein diverge significantly only at their N termini (15, 16). Siah proteins are RING finger proteins with E3 ligase activity and have been implicated in the ubiquitination and proteasome-dependent degradation of various substrate molecules. Substrates of Siah proteins are quite diverse and include transcriptional regulators (17-19), membrane receptors (20, 21), a microtubule-associated motor protein (22), and other proteins. In particular, the involvement of Siah proteins in cell growth regulation has been suggested in many reports. Siah1 expression is induced by tumor suppressor p53 in mammals and the overexpression of Siah1 inhibits cell proliferation and promotes apoptosis (23-25). Moreover, Siah-induced β-catenin degradation is important for the negative regulation of cell proliferation (26, 27), and the Siah-induced degradation of Kid is important for mitosis and contributes to cell growth arrest (28). Furthermore, mutations of Siah proteins in several cancers have been reported (29). These reports imply that Siah has tumor suppressor functions in some experimental settings.

To identify novel PLCϵ regulatory proteins, we performed yeast two-hybrid analysis using the RA domain of PLCϵ, and Siah1 and Siah2 were identified as PLCϵ-binding proteins. Here, we demonstrate that Siah proteins induce the proteasomal degradation of PLCϵ after EGF stimulation. Src-dependent phosphorylation was found to be required for the EGF-dependent degradation of PLCϵ. Moreover, the Siah-dependent degradation of PLCϵ was found to act as an important negative regulator of PLCϵ-dependent cell growth.

MATERIALS AND METHODS

Antibodies—Rabbit polyclonal antibody of PLCϵ was obtained from Dr. Tohru Kataoka (Kobe University, Japan). Mouse monoclonal antibody of Siah1 was described previously (30). Other antibodies used were: goat polyclonal anti-Siah2 antibody and goat polyclonal anti-AH receptor antibody (Santa Cruz Biotechnology, Santa Cruz, CA), mouse monoclonal anti-FLAG antibody (Sigma), mouse monoclonal anti-Myc antibody (Invitrogen), mouse monoclonal anti-HA antibody (Sigma, St. Louis, MO), 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, antibiotics, and glutamine. Wild-type and Siah-deficient MEF cells were grown in DMEM containing 10% fetal bovine serum, 0.02 mm β-mercaptoethanol, antibiotics, and glutamine as previously described (30). Cells were grown to ∼90% confluence for immunoprecipitation and Western blot experiments.

Yeast Two-hybrid Screening—PLCϵ 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-PLCϵ). The yeast strain, EGY48, carrying a reporter gene was cotransformed with the bait plasmid, pLexA-PLCϵ, and a human HeLa cDNA library fused to the VP16 activation domain. Transformation was carried out using the lithium acetate method (31). Leucine-positive colonies were identified by a filter-lifting assay for β-galactosidase activity. Library-derived DNA was prepared from candidate clones and analyzed by DNA sequencing.

Immunoprecipitation—Cells were lysed with TGH buffer (1% Triton X-100, 10% glycerol, 50 mm NaCl, 50 mm HEPES, pH 7.3, 1 mm EGTA, 1 mm sodium orthovanadate, 10 mm sodium fluoride, 1 mm phenylmethysulfonyl fluoride, 10 mg/ml leupeptin, 10 mg/ml aprotinin). Lysates were then centrifuged at 14,000 × g for 10 min at 4 °C. Super-natants were incubated with anti-HA, anti-FLAG for 3 h and then washed with TGH buffer three times. Immunoprecipitates were subjected to SDS-PAGE and Western blotting.

PLCϵ Phosphorylation Analysis—COS-7 cell-transfected Siah proteins were serum-starved for 20 h and then incubated with 1 mCi of [³²P]orthophosphate in 2 ml of phosphate-free DMEM for 4 h at 37 °C. Cells were treated with 100 ng/ml EGF for the indicated times. Cell lysates were prepared, and Siah proteins were immunoprecipitated. The immunocomplexes were subjected to SDS-PAGE and autoradiography. The amounts of immunoprecipitated Siah proteins were measured by immunoblotting and used for the quantitation of relative Siah phosphorylation.

Cell Growth Assay—MEF cells were seeded in triplicate into 6-well plates at a density of 2 × 10⁵ cells per well and were transfected with PLCϵ siRNA or control scrambled siRNA. After 24 h, cells were incubated with serum-free DMEM for 24 h to reach quiescence. The cells were incubated in serum-free medium supplemented with 100 ng/ml EGF for 18 h prior to addition of ³H-labeled thymidine for additional 6 h. Thymidine incorporation was measured as previously reported (32).

Plasmid Construction and Mutagenesis—FLAG-tagged mouse PLCϵ DNA is a generous gift from Dr. Tohru Kataoka (Kobe University, Japan). For the construction of Siah binding-deficient mutant, evolutionally conserved VLK (2173-2175) sequence was changed into AAA (called PLCϵ 3A) or Lys-2186 residue was changed into alanine (called PLCϵ K2186A) by site-directed mutagenesis. PLCϵ K2151E/K2153E is a Ras binding-deficient mutant. Siah1 wild-type and Y100F/Y126F mutant DNA was kindly provided by Dr. Zhiheng Xu (Columbia University) and introduced into pcDNA-HA vector by PCR amplification. To construct phosphorylation-deficient mutant of Siah1, Tyr-47, Tyr-199, and Tyr-223 were changed into phenylalanines (called Siah1 TM). For expression of Siah1 in knockout MEF cells, wild-type Siah1 or Siah1 TM were PCR-amplified and cloned into a lentivirus-derived C-FUW vector (33).

Production of Lentivirus-harboring Siah1—Lentivirus infection was performed as previously described (34).

PLCϵ Knockdown in MEF Cells—Synthetic siRNA against PLCϵ and control scrambled siRNA was introduced into MEF cells using LipofectAMINE (Invitrogen). PLCϵ siRNA (GCCAAATATTCCTACAGCA) and control scrambled siRNA (ACTGTCACAAGTACCTACA) have been previously described (35).

RESULTS

Interaction between PLCϵ and Siah Proteins—We sought to identify binding partners of the PLCϵ RA domain and performed a yeast two-hybrid analysis using bait containing the serial RA1 and RA2 domain of PLCϵ. Our yeast two-hybrid analysis revealed that the RA domain of PLCϵ interacts with various proteins other than small GTPases. Siah1 and Siah2 were identified as novel PLCϵ RA domain-binding proteins. The positive clones obtained from a HeLa cell cDNA library contained the substrate binding domain of the Siah proteins (Siah1, 177-282; Siah2, 217-324). We examined whether PLCϵ interacts with Siah proteins in cells by using co-immunoprecipitation analysis. Both Siah1 and Siah2 were found to associate with PLCϵ in COS-7 cells (Fig. 1A), but Siah2 did not interact with PLC-γ1 or PLC-β1 (Fig. 1B). These results indicate that Siah proteins interact with PLCϵ by specifically recognizing the RA domain, which is present only in the PLCϵ isozyme.

We then investigated whether endogenous PLCϵ and Siah2 can form a complex in MEF cells. As shown in Fig. 1C, Siah2 was immunoprecipitated with α-Siah2 antibody and the immunocomplex contained PLCϵ, which indicates that PLCϵ-Siah2 complex exists under physiological conditions.

The RA2 Domain Contains Siah Binding Regions Distinct from the Ras Binding Region—The RA domains of PLCϵ are composed of RA1 and RA2 domains. Both domains have similar ubiquitin-like folds, but only the RA2 domain can associate with activated Ras (36). In the present study, we explored the binding region of the RA domains in detail. Glutathione S-transferase-pulldown analysis revealed that the RA2 domain is responsible for the interaction with Siah2, and further deletion analysis enabled us to narrow this down to several amino acids (Fig. 2A). We mutated several evolutionally conserved amino acids and examined their ability to interact with Siah2. In particular, the mutation of Lys-2186 into Ala led to the 85% inhibition of the interaction between PLCϵ and Siah2, whereas mutation of VLK (2173-2175) into AAA had little effect on the interaction with Siah2. In addition, the Ras binding-deficient mutant of PLCϵ (K2151E/K2153E) had almost the same affinity for Siah2 as did the wild type (Fig. 2B). According to previous structural analysis of the PLCϵ RA domain (36), the Siah binding region of PLCϵ corresponds to the loop between β3 and β4 of the RA2 domain, which is removed from the Ras binding region (β1, β2, and the α1-β3 loop of RA2). In fact, the K2186A mutant showed no difference in binding affinity for the constitutively active H-RasV12 mutant (Fig. 2C). Taken together, these results support the possibility that the RA2 domain has a specific Siah binding region that is quite distinct from the Ras binding region.

EGF Stimulation Induced the Interaction between Siah2 and PLCϵ—EGF stimulation evokes PLCϵ activity through Ras-dependent translocation to the plasma membrane (8). In the present study, we investigated whether EGF treatment affects the interaction between Siah and PLCϵ. Interestingly, EGF treatment induced an association between PLCϵ and Siah2 in COS-7 cells (Fig. 3A). The interaction was initially detected at 5 min after EGF stimulation and was most prominent at 30 min after stimulation. We then investigated whether Ras-dependent activation of PLCϵ is required for the EGF-dependent association between Siah2 and PLCϵ. To this end, PLCϵ activation was blocked by the expression of a dominant negative mutant, RasN17. The expression of the RasN17 mutant abolished EGF-dependent Siah2 binding with PLCϵ (Fig. 3B), which indicates that PLCϵ activation is required for the association with Siah2. Our findings demonstrate that Siah2 associates with activated PLCϵ after EGF stimulation.

EGF Stimulation Induced the Siah-dependent Degradation of PLCϵ—Siah proteins function as E3 ubiquitin ligases and mediate the ubiquitination of various substrate molecules by direct interaction (17-22). Thus, we investigated whether the EGF-dependent association between Siah proteins and PLCϵ can lead to the ubiquitination of PLCϵ. We found that PLCϵ was efficiently ubiquitinated in COS-7 cells upon EGF stimulation (Fig. 4A). Many Siah substrates are known to be subjected to degradation by the ubiquitin-proteasome pathway (17-22), and in the present study, the expression of Siah2 also led to the EGF-dependent degradation of PLCϵ, which was blocked by co-incubation with MG132 (a proteasomal inhibitor), indicating that Siah-mediated ubiquitination leads to the proteasomal degradation of PLCϵ (Fig. 4B). The PLCϵ K2186A mutant was not ubiquitinated by EGF stimulation (Fig. 4C). Concomitantly, the PLCϵ K2186A mutant was not efficiently degraded by EGF stimulation (Fig. 4D), which demonstrates that the interaction between PLCϵ and Siah2 is required for the EGF-induced PLCϵ degradation. These results suggest that PLCϵ is subjected to Siah-dependent ubiquitination and degradation upon EGF stimulation.

The EGF-dependent Activation of Src Leads to the Phosphorylation of Siah Proteins—The growth factor-dependent ubiquitination of substrate molecules by Siah proteins has not been reported to date. We attempted to identify the upstream mediator responsible for the activation of Siah proteins upon EGF stimulation. It was recently reported that Siah1 is phosphorylated on tyrosine residues by a JNK-dependent pathway after camptothecin treatment, and that this tyrosine phosphorylation plays an important role in the stability and function of Siah1 (37). Thus, we examined whether EGF treatment could induce the tyrosine phosphorylation of Siah proteins. Interestingly, both Siah1 and Siah2 were found to be phosphorylated on tyrosine residues after EGF treatment in COS-7 cells (Fig. 5A). We tested several pharmacological inhibitors to identify potential upstream kinases responsible for the EGF-induced phosphorylation of Siah proteins. Src tyrosine kinase inhibitor PP2 blocked the EGF-induced phosphorylation of Siah2 (Fig. 5B), whereas the EGF-dependent phosphorylation of Siah2 was unaffected by SP600125 (a JNK inhibitor), which suggested a novel phosphorylation-dependent means of Siah protein regulation. There are five conserved tyrosine residues in Siah1 and Siah2, and Tyr-100 and Tyr-126 of Siah1 have been reported to be phosphorylated by camptothecin treatment. Thus, we investigated whether Tyr-100 and Tyr-126 of Siah1 are also phosphorylated by EGF stimulation. Mutation of both residues to phenylalanines had no effect on the EGF-dependent phosphorylation of Siah1 (Fig. 5C). Thus, we substituted Tyr-47, Tyr-199, and Tyr-223 for phenylalanine and re-examined EGF-induced phosphorylation. Individual mutations of these tyrosine residues resulted in a partial reduction of the tyrosine phosphorylation of Siah1, but the mutation of all three tyrosine residues completely blocked this phosphorylation (Fig. 5C). These results imply that Siah is phosphorylated at multiple tyrosine residues via a Src-dependent pathway after EGF stimulation.

Siah Phosphorylation Is Required for the EGF-induced Degradation of PLCϵ—To elucidate the role of the Src-dependent tyrosine phosphorylation of Siah proteins, we pretreated cells with Src tyrosine kinase inhibitor PP2 before EGF stimulation and examined the ubiquitination of PLCϵ in these cells. Src inhibition led to the suppression of the EGF-dependent ubiquitination of PLCϵ, whereas pretreatment of SP600125 had no effect on PLCϵ ubiquitination (Fig. 6A). To further confirm the role of Siah phosphorylation, we utilized a phosphorylation-deficient mutant of Siah1 (Siah1 TM) that had phenylalanine substitutions at Tyr-47, Tyr-199, and Tyr-223. Siah1 TM did not effectively induce the ubiquitination of PLCϵ after EGF treatment, whereas wild-type and the Siah1 Y100F/Y126F mutant efficiently ubiquitinated PLCϵ (Fig. 6B). Concomitantly, the EGF-dependent degradation of PLCϵ was impaired in the cells transfected with Siah1 TM (Fig. 6C), as compared with the cells transfected with wild-type Siah1. Taken together, these results show that the Src-dependent phosphorylation of Siah is required for the EGF-induced ubiquitination and degradation of PLCϵ.

Siah Is Required for the EGF-dependent Degradation of PLCϵ in MEFs—We attempted to determine whether the EGF-dependent degradation of PLCϵ is impaired in the absence of the Siah proteins. To this end, we utilized Siah1a/Siah2 double knockout MEF cells. PLCϵ was subjected to EGF-dependent degradation in MEF cells derived from wild-type mice (Fig. 7A, panel a), but the degradation of PLCϵ was blocked in the absence of Siah1a and Siah2 (Fig. 7A, panel b). To clarify the involvement of Siah, we expressed wild-type Siah1 or Siah1 TM in knockout cells by lentiviral infection (Fig. 7B). Add-back of wild-type Siah1 restored the EGF-dependent degradation of PLCϵ (Fig. 7A, panel c), however, the add-back of Siah1 TM did not (Fig. 7A, panel d). These results confirm the role of endogenous Siah proteins in the EGF-dependent degradation of PLCϵ.

Siah Suppressed PLCϵ-induced Cell Growth—Various reports suggest that PLCϵ can promote cell growth (12, 13), and our previous results revealed that EGF-dependent cell growth was enhanced by PLCϵ. Thus, we speculated that the Siah-dependent degradation of PLCϵ can contribute to the negative regulation of EGF-dependent cell growth. To test this hypothesis, we first examined whether Siah1 can inhibit the EGF-dependent cell growth potentiated by PLCϵ expression. As shown in Fig. 8A, PLCϵ expression enhanced EGF-dependent cell growth in HEK293 cells and co-expression of Siah1 abolished the PLCϵ-dependent enhancement of cell growth. We then measured the EGF-dependent cell growth in Siah1a/Siah2-double knockout cells and in Siah1-add-back cells. The add-back of Siah1 in double knockout cells suppressed EGF-dependent cell growth (Fig. 8C). We then reduced PLCϵ in Siah1a/Siah1-double knockout cells and in Siah1-add-back cells to investigate whether Siah-dependent growth inhibition is attributable to PLCϵ degradation (Fig. 8B). PLCϵ reduction in double knockout cells led to EGF-dependent growth inhibition, and notably, the suppression of Siah-dependent cell growth was abolished by PLCϵ knockdown (Fig. 8C). Taken together, these results suggest that Siah inhibits EGF-dependent cell growth by removing PLCϵ.

DISCUSSION

Although the activation mechanisms and physiological functions of PLCϵ have been largely revealed, little is known about the negative regulation of PLCϵ. The present study provides evidences that growth factor-activated PLCϵ is subjected to proteasomal degradation. Siah proteins interact with the RA2 domain of PLCϵ and promote PLCϵ ubiquitination in the process. Our findings demonstrate that the RA2 domain of PLCϵ plays a role in the inactivation of PLCϵ as well as in the activation of PLCϵ after growth factor stimulation. The RA2 domain of PLCϵ binds with activated Ras or Rap, and this binding is critical for the growth factor-dependent translocation to the membrane and the activation of PLCϵ (8, 36). It is interesting to note that the blockade of the Ras-dependent activation of PLCϵ led to the inhibition of the EGF-dependent interaction between Siah2 and PLCϵ, which implies that PLCϵ activation may be a prerequisite for the EGF-induced ubiquitination and degradation of PLCϵ. In addition, the EGF-dependent interaction between Siah2 and PLCϵ is most prominent at 30 min after EGF stimulation (Fig. 3A), whereas the EGF-induced translocation of PLCϵ into the membrane for activation reportedly begins at 5 min after stimulation and ends at 40 min after stimulation (8), which indicates that the Siah-induced ubiquitination of PLCϵ is induced after the activation of PLCϵ. Thus, we speculate that activated PLCϵ is subjected to Siah-dependent ubiquitination and degradation to terminate PLCϵ downstream signaling. This is the first study to explore the molecular mechanism underlying the negative regulation of PLCϵ.

Several studies have reported that PLCϵ promotes cell growth. For example, carcinogen-induced tumor formation was inhibited in PLCϵ-deficient mice (12), and the overexpression of PLCϵ in BaF3 cells potentiated platelet-derived growth factor-dependent cell growth (13). Consistent with previous reports, we found that PLCϵ knockdown in MEFs led to the suppression of EGF-dependent cell growth. We conclude that it is necessary to control the level of PLCϵ in cells to prevent aberrant cell growth. Many reports have implicated Siah proteins in cell growth control. Siah is a transcriptional target of p53 and contributes to p53-induced apoptosis and tumor suppression (23-25). Moreover, β-catenin degradation by genotoxic stress was mediated by Siah1, which leads to the suppression of cancer cell growth (26, 27). Siah1 was also found to induce growth arrest by inhibiting cytokinesis via the degradation of kinesin (Kid) (28). The involvement of Siah proteins in the regulation of growth factor-dependent cell growth was newly revealed in the present study. PLCϵ was found to be a novel substrate of Siah proteins, and it was degraded by Siah proteins after EGF stimulation. We demonstrated that Siah proteins contribute to the negative regulation of growth factor-induced MEF cell growth by mediating PLCϵ degradation. The fact that PLCϵ reduction abolished Siah-dependent growth inhibition (Fig. 7) suggests that PLCϵ is a major substrate of Siah proteins, which needs to be degraded to regulate cell growth in MEF cells.

The present study suggests a novel regulatory mechanism for Siah-mediated substrate degradation. We found that the Src-dependent phosphorylation of Siah is required for the EGF-induced degradation of PLCϵ. Whether Src directly phosphorylates Siah proteins or activates other protein kinases remains unclear. Previous reports have shown that Siah proteins are phosphorylated in different environments. For example, Siah2 was phosphorylated on Thr-24 and Ser-29 by p38 MAPK under hypoxia, and this was found to be important for the degradation of PHD-3 (38). Greene et al. (37) reported that Siah1 is phosphorylated on Tyr-100 and Tyr-126 upon camptothecin treatment via the activation of the JNK pathway. This phosphorylation increased the stability of Siah1 and its association with the adaptor protein POSH. Our analysis revealed that Siah1 is phosphorylated on Tyr-47, Tyr-199, and Tyr-223 via Src activation. We speculate that the EGF-induced phosphorylation of Siah proteins contributes to the binding of proteins with PLCϵ and the subsequent degradation of PLCϵ, because pretreatment with the Src tyrosine kinase inhibitor PP2 reduced the EGF-induced interaction between PLCϵ and Siah proteins as well as the ubiquitination of PLCϵ (data not shown). Thus, Siah phosphorylation appears to be an important regulatory mode that mediates various interactions with substrate molecules or adaptor proteins in different cellular contexts.

In summary, we identified Siah1 and Siah2 as negative regulators of PLCϵ. PLCϵ is subjected to EGF-dependent degradation via Siah-induced ubiquitination and proteasomal degradation processes. Siah is phosphorylated by EGF stimulation, and this phosphorylation is required for PLCϵ degradation. Physiologically, the Siah-induced degradation of PLCϵ contributes to the negative regulation of the EGF-dependent cell growth potentiated by PLCϵ.