Epidermal Growth Factor Induces Ubiquitination of Eps15*

Epidermal growth factor (EGF) receptor pathway substrate clone 15 (Eps15) has been described as a 142-kDa EGF receptor substrate. It has been shown to bind to the EGF receptor, adaptor protein-2, and clathrin and is present at clathrin-coated pits and vesicles. Upon stimulation of cells with EGF or transforming growth factor α, Eps15 becomes rapidly and transiently phosphorylated on tyrosine residues. This phosphorylation coincides with an increase of 8 kDa in molecular mass. Here we show that this increase in molecular mass is not due to tyrosine phosphorylation. Instead, we found both by Western blotting and protein sequencing that this EGF-induced increase in molecular mass is the result of monoubiquitination. Eps15 ubiquitination but not tyrosine phosphorylation was inhibited under conditions that blocked EGF-induced internalization of the EGF receptor. Our results establish ubiquitination as a second form of EGF-stimulated covalent modification of Eps15.

Eps15 1 has been identified as a 142-kDa substrate of the EGF receptor (1). In quiescent cells Eps15 is associated to the EGF receptor, and upon EGF stimulation this association increases dramatically (2). In addition, Eps15 has been shown to bind to both adaptor protein-2 and clathrin (2,3). Subcellular fractionation and immunolocalization studies have shown that Eps15 is present in clathrin-coated pits and vesicles but not in early endosomes (2,4). Eps15 shares homology with the yeast proteins End3p and Pan1p. Both proteins contain multiple Eps15 homology domains, a motif proposed to mediate proteinprotein interaction, and have been implicated in the endocytosis of the ␣-factor and lipids, respectively (5,6).
Tyrosine phosphorylation of Eps15 is transient and occurs within 2 min of EGF stimulation (1). In addition, EGF stimulation results in the appearance on SDS-polyacrylamide gels of a slowly migrating band of Eps15 of approximately 150 kDa.
Tyrosine kinase activity of the EGF receptor was found to be required for this apparent increase in molecular mass of Eps15 (7). Expression of Eps15 cDNA in bacteria shows the presence of only the 142-kDa form, suggesting that Eps15 is undergoing an EGF-induced post-translational modification (1).
In this paper we investigated the nature of this post-translational modification of Eps15. We found that the appearance of the high molecular mass form of Eps15 is not due to EGFinduced hyperphosphorylation. Instead, we found that the 8-kDa increase in molecular mass was caused by monoubiquitination of Eps15. This indicates that EGF induces two different modes of post-translational modification of Eps15: tyrosine phosphorylation and ubiquitination.
Immunoprecipitations-Cells were grown in 60-, 100-, or 175-mm dishes (Nunc Life Technologies, Gaithersburg, MD) till 80% confluency. Cells were serum-starved in DMEM with 0% v/v FCS for 24 h before stimulation with 50 ng/ml EGF. Cells were lysed in RIPA buffer (20 mM Tris-HCl, pH 7.4, 150 mM NaCl, 0.5% Triton X-100, 0.1% SDS, 1 mM EDTA, 1 mM phenylmethylsulfonyl fluoride, 1 mM benzamidine, 100 mM NaF, and 1 mM Na 3 VO 4 ) at 4°C for 5 min, scraped from the dish, and centrifuged for 5 min at 12,000 ϫ g in an Eppendorf centrifuge. Total cell lysate samples were prepared by adding Laemmli sample buffer to the RIPA lysates. For immunoprecipitations, the RIPA lysates were incubated with 25 l of a 1:1 suspension of protein A-Sepharose for 1 h at 4°C and centrifuged. Supernatants were incubated with anti-Eps15 antibody (8) for 2 h at 4°C. Subsequently, protein A-Sepharose was added, and after a further incubation of 2 h, the immunoprecipitates were washed three times, once with RIPA-buffer, once with high salt buffer (20 mM Tris-HCl, pH 7.4, 0.5 M NaCl, and 1% Triton X-100), and finally once with low salt buffer (20 mM Tris-HCl, pH 7.4, 0.15 M NaCl, and 1% Triton X-100). For alkaline phosphatase treatment, immunoprecipitates were incubated in phosphatase buffer (50 mM Tris, pH 8.5, and 1 mM EDTA) with 50 units of alkaline phosphatase (Boehringer Mannheim, Mannheim, Germany) at 37°C for 30 min. Heat-inactivated alkaline phosphatase was prepared by incubating the alkaline phosphatase at 95°C for 15 min. The beads were boiled in 20 l of Laemmli sample buffer for 5 min, and proteins were separated by 8% SDSpolyacrylamide gel electrophoresis, Western blotted onto PVDF membrane (Immobilon-P, Millipore, Bedford MA, USA), and probed with rabbit polyclonal antibodies against Eps15, mouse monoclonal antiphosphotyrosine PY20 (Transduction laboratories, Lexington KY), or rabbit polyclonal anti-ubiquitin antibodies (antibody kindly provided by Dr. A. Ciechanover). Protein bands were visualized by Enhanced Chemiluminescence (Renaissance, DuPont NEN, Boston, MA) using peroxidase-conjugated goat-anti-rabbit or rabbit-anti-mouse immunoglobulins (Jackson ImmunoResearch, Pennsylvania, PA).
Edman Degradation-To determine the N-terminal amino acid sequence of Eps15, proteins of immunoprecipitates were separated on a 8% SDS-polyacrylamide gel and transferred to a PVDF membrane. The proteins were stained with Ponceau S (Sigma, St. Louis, MO), and the Eps15 bands were cut out of the membrane, washed thoroughly with distilled water, and subjected to Edman degradation (9, 10). High pressure liquid chromatography was used for analysis of degradation products.
Inhibition of Endocytosis-Inhibition of endocytosis in HER14 cells was performed by potassium depletion (11), by incubating the cells in hypertonic medium (12), by acidification of the cytosol (13), or by an incubation of the cells at 4°C (14).
For potassium depletion, cells were washed twice with depletion buffer (20 mM Hepes, pH 7.4, 0.14 M NaCl, 1 mM CaCl 2 , 1 mM MgCl 2 , and 1 g/l D-glucose). Subsequently, cells were incubated for 5 min with * This work was supported by Grant 17.182 (to S. v. D.) from the Life Sciences Foundation, which is subsidized by the Netherlands Organization for Scientific Research. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
§ To whom correspondence should be addressed. Internalization Analysis-Internalization assays were performed using a protocol modified from Haigler et al. (15). HER14 cells were grown on 6-well dishes to 80% confluency and incubated for 30 min with 1 ng/ml 125 I-EGF in DMEM containing 20 mM Hepes, pH 7.2, and 0.1% bovine serum albumin. Nonspecific EGF binding was measured in the presence of 500-fold molar excess of unlabeled EGF. Following this incubation, the medium was removed and the cells were washed twice with ice-cold phosphate-buffered saline. Surface-bound EGF was removed by successive incubation of the cells with 0.2 M sodium acetate, pH 3.5, containing 150 mM NaCl (acid wash) for 5 and 1 min, respectively. The acid wash solution represented the surface-bound radioactivity. Internalized ligand was determined by lysis of the cells by incubation with 1 N NaOH for 5 min at 37°C. The rate of endocytosis was expressed as the ratio of internal and surface-bound EGF.

RESULTS AND DISCUSSION
Eps15 Mobility Shift Is Not Due to Tyrosine Phosphorylation-To determine the nature of the EGF-induced increase in the molecular mass of Eps15, we investigated whether this increase is caused by tyrosine phosphorylation. HER14 fibroblasts expressing the human EGF receptor were stimulated with EGF, and Eps15 was immunoprecipitated from the cell lysates. One Eps15 immunoprecipitate was treated with alkaline phosphatase, whereas two controls were either left untreated or treated with heat-inactivated phosphatase. The proteins were separated on 8% SDS-polyacrylamide gels and blotted onto PVDF membrane. After detection of tyrosine phosphorylated Eps15 by anti-phosphotyrosine antibodies, two Eps15 bands of 142 and 150 kDa were visible in untreated cells. This indicates that both forms of Eps15 are phosphorylated (Fig. 1, Con). The two Eps15 bands of 142 and 150 kDa are each resolved into a tightly spaced doubled, and both forms of the doubled are tyrosyl phosphorylated. The reason for the slight difference in molecular mass is not known but could be due to differential splicing. Treatment of Eps15 with alkaline phosphatase resulted in the complete dephosphorylation of Eps15 (Fig. 1, AP), whereas treatment with heat-inactivated alkaline phosphatase did not change the phosphorylation state of Eps15 (Fig. 1, HI-AP). Reprobing the same blot with anti-Eps15 antibodies showed that irrespective of the phosphorylation state of Eps15, the 142-and 150-kDa forms were present (Fig. 1). These results demonstrate that the appearance of high molec-ular mass form of Eps15 is not the result of tyrosine phosphorylation.
Eps15 Mobility Shift Is Due to Monoubiquitination-The approximate increase of 8 kDa in the modified form of Eps15 stimulated us to investigate the possible monoubiquitination of Eps15. Ubiquitin is a highly conserved protein of about 8 kDa, which is abundant in eukaryotes. Ubiquitin is found free or covalently linked via its C terminus to NH 2 groups of one or more lysine residues of a variety of cytoplasmic, nuclear, and integral membrane proteins (16).
To investigate the monoubiquitination of Eps15, Eps15 was immunoprecipitated from HER14 cells that were either left unstimulated or stimulated with 50 ng/ml EGF. The protein samples were separated on 8% SDS-polyacrylamide gels, and the Western blot was probed with anti-Eps15 antibodies. A clear mobility shift was seen after EGF stimulation but not in unstimulated cells (Fig. 2). Subsequently, the Western blot was stripped and reprobed with anti-ubiquitin antibodies. In this case only the 150-kDa form of Eps15 was detected, demonstrating that Eps15 becomes monoubiquitinated upon EGF stimulation (Fig. 2). In addition to the appearance of the 150-kDa band, a slight staining of higher molecular mass Eps15 was detected upon EGF addition. This phenomenon was better visible upon longer exposures (data not shown). Eps15 of higher molecular mass was previously also found on Western blots containing immunoprecipitated Eps15 that were stained for phosphotyrosine residues (2). These observations suggest that Eps15 is not only monoubiquitinated but that a minority of Eps15 may also be multiubiquitinated.
The 150-kDa Form of Eps15 Contains Covalently Bound Ubiquitin-To obtain further proof for the ubiquitination of Eps15, the N-terminal sequences of the 142-and 150-kDa Eps15 isoforms were determined by Edman degradation. Because ubiquitin is conjugated via its C terminus to the target proteins, the N terminus of conjugated ubiquitin is still available for Edman degradation. Sequencing of the 142-kDa form of Eps15 did not result in any signal, most probably due to Nterminal blocking. Sequencing of the 150-kDa form of Eps15 resulted in a single protein sequence (Fig. 3). Comparison of these 10 amino acids with the published sequence of bovine ubiquitin revealed that the obtained amino acids are identical to the first amino acids of ubiquitin. Comparison of this sequence with sequences in the SWISS-PROT protein data base did not reveal a relevant match with any other protein than ubiquitin.
Based on both the Western blotting results and the N-terminal amino acid sequence, we conclude that Eps15 becomes ubiquitinated after stimulation of the cell with EGF. Because the increase in molecular mass of Eps15 is similar to the FIG. 1. Alkaline phosphatase treatment of Eps15 immunoprecipitates. HER14 cells were stimulated for 10 min with 50 ng/ml EGF and lysed in RIPA buffer. Eps15 was immunoprecipitated from the cell lysates, and the immunoprecipitates were either left untreated (Con) or treated with alkaline phosphatase (AP) or heat-inactivated alkaline phosphatase (HI-AP). Proteins were separated on an 8% SDS-polyacrylamide gel, and the Western blot was incubated with anti-phosphotyrosine antibodies (WB P-Tyr). Subsequently, the blot was stripped and incubated with anti-Eps15 antibodies (WB Eps15).

FIG. 2. Stimulation of HER14 cells with EGF induces monoubiquitination of Eps15
. HER14 cells were left unstimulated or stimulated for 10 min with 50 ng/ml EGF and lysed in RIPA buffer. Eps15 was immunoprecipitated from the cell lysates, and the proteins were separated on an 8% SDS-polyacrylamide gel. The Western blot was incubated with anti-Eps15 antibodies (WB Eps15). Subsequently, the blot was stripped and incubated with anti-ubiquitin antibodies (WB Ubiquitin). molecular mass of ubiquitin (8 kDa), we conclude that Eps15 becomes predominantly monoubiquitinated. Because the approximate ratio of the two Eps15 forms in EGF-stimulated cells was previously determined as 1:1, we estimate that about 50% of Eps15 becomes monoubiquitinated after stimulation of cells with EGF (2). Both forms of Eps15 become phosphorylated on tyrosine residues (Fig. 1), which indicates that ubiquitination of Eps15 is not required for its phosphorylation.
Monoubiquitination of proteins has not frequently been reported. Examples of monoubiquitination are the T cell antigen receptor (17), histone H2A (18), and cytochrome c (19). The yeast ␣-factor receptor has recently been shown to become either mono-or diubiquitinated (20). Multiubiquitination of proteins usually starts on one lysine residue (16). Subsequently, this ubiquitin becomes ubiquitinated, resulting in the formation of multiubiquitin chains. Examples of multiubiquitination include cytoplasmic and nuclear proteins but also integral membrane proteins such as receptors for EGF (21), growth hormone (22,23), platelet-derived growth factor (24), and the tumor necrosis factor (25). Protein ubiquitination has been implicated in many cellular processes (16). The most widely studied function of ubiquitination lies in the targeting of (multiubiquitinated) proteins for degradation to the 26 S proteasome. However, not all ubiquitinated proteins are degraded, suggesting additional functions for ubiquitination besides proteolysis. Treatment of cells with transcription inhibitors resulted in a reduced level of ubiquitinated histone H2B, suggesting a role for ubiquitination in chromatin organization (18). Recently, it has been suggested that ubiquitination plays a role in the activation of IB␣, a regulator of the transcription factor NFB (26). Interestingly, a new function for ubiquitination has been recently described for the ubiquitination of plasma membrane receptors. Ubiquitination of both the growth hormone receptor and the ␣-factor receptor in S. cerevisiae have been implicated in the endocytosis of these receptors (20,23,27).
The binding of Eps15 to both adaptor protein complex-2 and clathrin and the presence of Eps15 in clathrin-coated pits and vesicles of mammalian cells suggest a role for Eps15 in the endocytosis of the EGF receptor. To investigate the possible relationship between Eps15 ubiquitination and EGF receptor internalization, we examined the effect of blocking EGF receptor internalization on Eps15 ubiquitination. Internalization of EGF receptors was inhibited in four different ways: by incubation at low temperature, by depleting potassium from the cytosol, by a hypertonic shock of the cells, or by acidification of the cytosol. These methods inhibit different steps in endocytosis: hypertonic shock and incubation at low temperature prevent clustering of receptors (12), potassium depletion inhibits the assembly of coated pits (11), whereas acidification of the cytosol is suggested to inhibit pinching off of clathrin-coated pits from the plasma membrane (13). The effect of these conditions on EGF endocytosis was measured using 125 I-labeled EGF. In control cells EGF was rapidly internalized, whereas under all four endocytosis inhibiting conditions EGF internalization was inhibited for more than 80% (Fig. 4). Analysis of Eps15 phosphorylation revealed that in all cases Eps15 became phosphorylated on tyrosine residues (Fig. 5B). These results demonstrate that EGF receptor activity has not been affected by either of these treatments. Interestingly, it was recently reported by Vieira and co-workers (28) that inhibition of endocytosis using a dynamin mutant resulted in differences in EGF receptor substrate phosphorylation. This indicates that maximum Eps15 phosphorylation is already achieved before the EGF receptor internalization process is initiated. Detection of Eps15 by Western blotting showed that in control cells Eps15 became ubiquitinated after 10 min of EGF stimulation (Fig.  5A). However, when cells were incubated and stimulated at 4°C, ubiquitination of Eps15 was completely abolished (Fig.  5A, row 1). The same results were obtained when endocytosis was inhibited by alternative methods such as potassium depletion (Fig. 5A, row 2), acidification of the cytosol (Fig. 5A, row 3), and hypertonic shock (Fig. 5A, row 4). Together these data show that when endocytosis is inhibited Eps15 monoubiquitination is abolished, but tyrosine phosphorylation remains undisturbed.
An interesting question is the possible function of the monoubiquitination of Eps15. The absence of Eps15 ubiquitination under conditions that inhibit EGF receptor internalization indicates that either Eps15 ubiquitination is required for Eps15 endocytosis or EGF receptor endocytosis is a prerequi-FIG. 5. Inhibition of endocytosis prevents Eps15 monoubiquitination. HER14 cells were subjected to four different methods to inhibit endocytosis: low temperature, hypertonic shock, potassium depletion, or cytosol acidification. Subsequently the cells were left unstimulated or stimulated with 50 ng/ml EGF for 10 min. Proteins from the cell lysates were separated on 8% SDS-polyacrylamide gels, and the Western blots were incubated with anti-Eps15 antibodies (A). Immunoprecipitated (IP) Eps15 was separated in a similar manner, and the Western blots (WB) were incubated with anti-phosphotyrosine (␣-p-tyr) antibodies (B). The 142-kDa band of Eps15 and the ubiquitinated 150-kDa band of Eps15 (Eps15-Ub) are indicated.

FIG. 3. Obtained N-terminal amino acid sequence of the 150-kDa form of Eps15 as determined by Edman degradation.
HER14 cells were stimulated with EGF, and Eps15 was immunoprecipitated from all the lysates. Proteins were separated on an 8% SDS-polyacrylamide gel. The 150-kDa band of Eps15 was cut out of the PVDF membrane and subjected to Edman degradation. The obtained N-terminal sequence of the 150-kDa mouse Eps15 and the bovine ubiquitin sequence are shown.
FIG. 4. Inhibition of EGF receptor endocytosis. Inhibition of EGF receptor endocytosis in HER14 cells was done by four different methods as described under "Experimental Procedures." The rate of endocytosis in control cells (C) was set at 100%, and the endocytosis after treatment with low temperature, hypertonic shock, potassium depletion, or cytosol acidification (E) is presented as a percentage of the control.
site for Eps15 ubiquitination. The first possibility is an analogy to what has been reported for the growth hormone receptor in mammalian cells (23) and the ␣-factor receptor in yeast (20). In this case Eps15 ubiquitination could be involved in the early steps of endocytosis of the EGF receptor. However, inhibition of endocytosis of the ␣-factor receptor in yeast resulted in an increased ubiquitination of the receptor, which is in contrast to the results presented in this paper (20). Alternatively, our results may indicate that endocytosis is required for Eps15 ubiquitination. This would imply that Eps15 ubiquitination occurs exclusively at a post-surface endocytic transport step. We have shown previously that Eps15 localization is restricted to coated pits and coated vesicles and absent from early endosomes (2). The monoubiquitination of Eps15 could thus be involved in the targeting of coated pits to the early endosome or in the uncoating of the coated vesicle. Another possibility is that Eps15 ubiquitination could be involved in the targeting of Eps15 to the 26 S proteasome for its degradation, which is in fact the first described function of protein ubiquitination.