Multisite Phosphorylation of the Epidermal Growth Factor Receptor USE OF SITE-DIRECTED MUTAGENESIS TO EXAMINE THE ROLE OF SERINE/THREONINE PHOSPHORYLATION*

The major sites of serine and threonine phosphoryl- ation of the human epidermal growth factor (EGF) receptor observed in intact cells are ThrBs4, Thress, Ser’046, and Ser’04’. Phosphorylation of the EGF receptor is increased at these sites in cells treated with platelet-derived growth factor or phorbol ester. This increase in EGF receptor phosphorylation is associated with an inhibition of the high affinity binding of EGF to cell surface receptors and an inhibition of the receptor tyrosine protein kinase activity. In order to test the hypothesis that the phosphorylation of the EGF receptor is mechanistically related to the modulation of EGF receptor function, we replaced the major sites of serine and threonine phosphorylation with alanine residues.

The major sites of serine and threonine phosphorylation of the human epidermal growth factor (EGF) receptor observed in intact cells are ThrBs4, Thress, Ser'046, and Ser'04'. Phosphorylation of the EGF receptor is increased at these sites in cells treated with platelet-derived growth factor or phorbol ester. This increase in EGF receptor phosphorylation is associated with an inhibition of the high affinity binding of EGF to cell surface receptors and an inhibition of the receptor tyrosine protein kinase activity. In order to test the hypothesis that the phosphorylation of the EGF receptor is mechanistically related to the modulation of EGF receptor function, we replaced the major sites of serine and threonine phosphorylation with alanine residues. EGF receptors containing single point mutations or multiple mutations were expressed in Chinese hamster ovary cells.
Analysis of the regulation of the EGF receptor tyrosine protein kinase activity demonstrated that phorbol ester caused an inhibition of the tyrosine phosphorylation of wild-type receptors and receptors lacking ThrSBB, Serlo4', or Serlo4'. In contrast, the inhibition of EGF receptor tyrosine phosphorylation caused by phorbol ester was not observed for any of the mutated EGF receptors that lacked Thres4. These data are consistent with the hypothesis that the phosphorylation of the EGF receptor at Thres4 is required for the inhibition of the receptor tyrosine protein kinase activity caused by phorbol ester. Investigation of the apparent affinity of the EGF receptor demonstrated that treatment with phorbol ester caused an inhibition of the high affinity binding of 12'I-EGF to cells expressing wild-type EGF receptors and each of the mutated EGF receptors examined. We conclude that the regulation of the apparent affinity of the EGF receptor is independent of the major sites of serine and threonine phosphorylation of the EGF receptor.
The EGF' receptor is a 170-kDa transmembrane glycoprotein. The binding of EGF to the extracellular ligand binding domain of the receptor causes an increase in the tyrosine protein kinase activity of the receptor cytoplasmic domain * These studies were supported by National Institutes of Health Grants GM37845 and CA39240. The costs of oublication of this article were defrayed in part by the payment of-page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 USC. Section 1734 solely to indicate this fact.
(1). Incubation of cells with phorbol ester or with PDGF causes a decrease in the high affinity binding of lz51-EGF to cell surface receptors, a decrease in the receptor tyrosine protein kinase activity, and a marked increase in the state of phosphorylation of the EGF receptor at serine and threonine residues (reviewed by Schlessinger,Ref. 2). The apparent association of receptor phosphorylation with the modulation of receptor activity has been used as evidence for the hypothesis that receptor phosphorylation is mechanistically related to receptor regulation (3). One site of phosphorylation of the EGF receptor, Thr654, is a substrate for protein kinase C (4, 5). The phosphorylation of Thr654 has been observed after the treatment of WI-38 human fetal lung fibroblasts with phorbol ester (5) and with PDGF (6,7). In order to establish the role of Thr65" phosphorylation during protein kinase C regulation of high affinity EGF binding, the effect of site-directed mutagenesis of Thr654 has been investigated. Lin et al. (8) have reported that the inhibition of EGF binding caused by phorbol ester was not observed in cells expressing EGF receptors that lack Thr654. This result suggested that phosphorylation of Thr654 was required for the modulation of EGF binding caused by phorbol ester (8). However, subsequent studies demonstrated that the replacement of threonine-654 with either an alanine or a tyrosine residue caused no significant change in the regulation of the apparent affinity of the EGF receptor observed after treatment of cells with phorbol ester (9, 10) or with PDGF (11). Together, these data demonstrate that the mechanism of regulation of the high affinity binding of EGF to cell surface receptors is independent of the phosphorylation of the EGF receptor at Thr654 (9-11).
Protein kinase C causes an inhibition of the tyrosine protein kinase activity of the EGF receptor (12,13). To test the hypothesis that the phosphorylation of Thr654 is required for the inhibition of the receptor tyrosine protein kinase, we have examined the effect of mutagenesis of Thr654 (9). Phorbol ester treatment of cells expressing [Alaes4]EGF receptors did not cause an inhibition of EGF receptor tyrosine protein kinase activity (9). This lack of regulation of the [Ala654]EGF receptor suggests that the phosphorylation of Thr654 may be a critical event for the regulation by protein kinase C (9). Strong support for this hypothesis is provided by in vitro reconstitution experiments in which the phosphorylation of Thr654 by protein kinase C is associated with an inhibition of the receptor tyrosine protein kinase activity (9, 12-15). However, the contribution of Thr654 phosphorylation to the overall modulation of receptor function has not been resolved because evidence has been presented indicating that the EGF receptor tyrosine protein kinase activity can be regulated by mechanisms that are independent of Thr654 phosphorylation (10, 16).

Role of EGF Receptor Phosphorylation
Examination of the role of the phosphorylation of the EGF receptor at Thr654 demonstrates that phosphorylation at this site alone is not sufficient to account for the mechanism of transmodulation of the EGF receptor observed after treatment of cells with phorbol ester (9, 10) or with PDGF (11). Although Thr'je4 is the only site of phosphorylation on the EGF receptor catalyzed by protein kinase C in vitro (4,5,14), previous studies have demonstrated that protein kinase C activation in uivo results in the phosphorylation of the EGF receptor at several additional sites (5,(17)(18)(19). The observation that the phosphorylation state of a mutated [Ala654]EGF receptor was markedly increased after treatment with phorbol ester suggests that the phosphorylation of these serine and threonine residues may be related to the mechanism of EGF receptor transmodulation (9). The purpose of the study reported here was to investigate the role of the multiple sites of serine and threonine phosphorylation of the EGF receptor. In order for phosphorylation to serve a regulatory function, it is likely that a significant fraction of the EGF receptors will be phosphorylated at the sites that are critical for receptor modulation. Previous studies have demonstrated that the stoichiometry of the phosphorylation of the EGF receptor at many sites is extremely low (7). In addition, there are several sites with a measured stoichiometry of phosphorylation between 0.1 and 0.9 mol of phosphate per mol of receptor in pborbol ester and PDGF-treated cells (7). These major sites of phosphorylation represent a potential mechanism of regulation of EGF receptor function. To test this hypothesis, we examined the effect of mutagenesis of the major sites of EGF receptor phosphorylation. EGF receptors lacking individual phosphorylation sites as well as receptors lacking multiple phosphorylation sites were expressed in CHO cells. The properties of the wild-type and mutated EGF receptors were investigated in detail. We report that the mechanism of regulation of the apparent affinity of the EGF receptor is independent of the major sites of serine and threonine phosphorylation of the receptor. However, the data obtained strongly support the hypothesis that receptor phosphorylation is an important mechanism of regulation of the EGF receptor tyrosine protein kinase activity.  (20,21) and iodinated to a specific activity of 70-90 Ci/g as described (22). Restriction enzymes were from Boehringer. Porcine PDGF was obtained from Bioprocessing Ltd. Amethopterin and PMA were from Sigma. The monoclonal antiphosphotyrosine antibody (PY20) was obtained from ICN. Sequenase was obtained from United States Biochemical Corp.

Plasmid Construction
Oligonucleotide-directed mutagenesis of Thr654 (ACG), Thr6@ (ACA), Ser'046 (AGC), and Ser'047 (TCA) to alanine (GCA, GCG, or GCC) was carried out using 17-mer oligonucleotides according to Zoller and Smith (23) using methods described previously (24). Mutations were confirmed by sequencing using [35S]dATP, ddNTPs and Sequenase (25). The wild-type and mutated EGF receptor cDNAs were cloned as 4-kilobase XbaI-Hind111 fragments into the plasmid pX (11). This expression vector contains the murine dihydrofolate reductase gene as a selectable marker and allows the expression of the EGF receptor cDNA using the SV40 early promoter and polyadenylation signals and has been described previously (11).
Tissue Culture CHO cells expressing the human PDGF receptor (B-type) cDNA using a pZipNeoSV(X) vector (26) were obtained from Drs. L. Claes-son-Welsh and C.-H. Heldin (University of Uppsala, Sweden). The cells were maintained in Ham's F12 medium supplemented with 10% fetal bovine serum and 0.25 mg/ml G418 (Geneticin, Gibco). The cells were transfected with plasmids using the calcium phosphate technique. After 3 days, the cells were passaged and selected using modified Eagle's minimum a medium supplemented with 5% dialyzed fetal bovine serum, 0.5 pM amethopterin, and 0.25 mg/ml G418. Stable colonies were isolated using cloning rings and screened for the expression of EGF receptors by measuring the cell surface binding of "'1-EGF at 4 "C.
Analysis of lz51-EGF Binding CHO cells were seeded in 16-mm wells and grown to a density of 5 x lo4 cells/well. The cells were then incubated for 24 h in medium supplemented with 0.1% calf serum. The monolayers were then washed with 120 mM NaCl, 6 mM KCl, 1.2 mM CaCl*, 1 mM MgCl,, 25 mM HEPES (pH 7.4), 30 pM bovine serum albumin and incubated for 30 min at 37 "C in the same medium. The cells were treated without and with phorbol ester or PDGF at 37 "C and then rapidly cooled to 4 "C. The binding of ?-EGF to cell surface receptors was measured by incubation of the cells at 4 "C for 3 h as described (7). Nonspecific binding was estimated in incubations containing a 500fold excess of unlabeled ligand.

Autophosphorylation of the EGF Receptor in Zntact Cells
CHO cells were grown in 35-mm dishes and washed with serumfree medium. The cells were then incubated at 37 "C for 30 min in 1 ml of 120 mM NaCl. 6 mM KCl. 1.2 mM CaCL. 1 mM M&l,. 25 mM HEPES (pH 7.4), 36 pM bovine serum albumin. The cells were then incubated without or with EGF or phorbol ester for defined times, and the EGF receptors were isolated from the cells by immunoprecipitation and polyacrylamide gel electrophoresis as described (7). The state of tyrosine phosphorylation of the EGF receptors was investigated by a Western blot procedure (27) using a monoclonal antityrosine phosphate antibody (PYZO) and a iz51-goat anti-mouse Ig second antibody.

Analysis of EGF Receptor Tryptic 32P-Phosphopeptides
CHO cells were labeled with [32P]phosphate for 18 h by incubation with phosphate-free Dulbecco's modified Eagle's medium supplemented with 0.1% calf serum and 2 mCi/ml of [32P]phosphate. EGF receptors were isolated from the cells by immunoprecipitation of detergent extracts with a polyclonal anti-EGF receptor antibody as described (7). The immunoprecipitates (100 ~1) were reduced by heating at 60 "C for 15 min in the presence of 80 ~1 of 10% NaDodSOI, 14 mM dithiothreitol. After cooling, the EGF receptors were alkylated by adding 40 ~1 of 0.4 M iodoacetamide, 0.25 M Tris-HCl (pH 8.8) and incubation at room temperature for 15 min. Subsequently, 80 ~1 of 75% elvcerol. 25% 2-mercantoethanol was added, and the sample was heated to 66 "C for 15 min. After polyacrylamide gel electrophoresis, the gel slice containing the receptor was excised. The receptor was eluted with NaDodS04 and precipitated with trichloroacetic acid as described (28). The sample was then digested with 1 pg of tosylphenylalanyl chloromethyl ketone-treated trypsin in 100 mM Nethylmorpholine (pH 8.0). After 5 h, a second addition of trypsin was made, and the incubation was allowed to proceed for a further 19 h. Phosphopeptide mapping of the trypsin-digested EGF receptor was performed by reverse-phase HPLC using a Vydac Cls column (0.46 x 25 cm) equilibrated with 0.1% trifluoroacetic acid (7). Peptides were eluted with a linear gradient of acetonitrile (1% per minute) in 0.1% trifluoroacetic acid. Fractions were collected at 30-s intervals, and the [3ZP]phosphopeptides were detected by measuring the Cerenkov radiation associated with each fraction.

Phosphoamino
Acid Analysis Phosphoamino acid analysis was performed by partial acid hydrolysis (1 h at 110 "C in 6 M HCl) and thin layer electrophoresis by the method of Hunter and Sefton (29) as described (5).

Role of EGF Receptor Phosphorylation identified
and quantitated by a modification of the reverse-phase HPLC procedure described by Zimmerman et al. (30) using acetonitrile.
The radioactivity associated with the phenylthiohydantoins derived from the peptide that were released at each cycle was measured by liquid scintillation counting.

EGF Receptor Down-regulation
Two procedures were used to investigate the down-regulation of EGF receptors caused bv incubation of cells with EGF. Investigation of the C&l Surface Specific Binding of '25Z-EGF-CH0 cells were seeded in 16-mm wells and grown to a density of 5 x 10' cells/well.
The cells were washed with 120 mM NaCl, 6 mM KCl, 1.2 mM CaCI*, 1 mM MgC12, 25 mM HEPES (pH 7.4), 30 pM bovine serum albumin, 10 mM glucose and incubated for 30 min at 37 "C. The cells were treated without (control) or with 100 nM EGF at 37 "C for 0.5, 1, 2, 3, and 4 h. The cell monolayers were then washed three times at 37 "C over a period of 20 min and subsequently cooled to 4 "C. The specific binding of '*Y-EGF to cell surface receptors was then measured at 4 "C. CHO cells were seeded in 16-mm wells and grown to a density of 5 X lo4 cells/well. Cells were washed with 120 mM NaCl, 6 mM KCl, 1.2 mM CaCl*, 1 mM MgCl*, 25 mM HEPES (pH 7.4), 30 fiM bovine serum albumin and incubated for 30 min at 37 "C in the same medium. The cells were incubated with 1 nM iz51-EGF for 10, 20, 30, 40, 50, 60, and 90 min at 37 "C and rapidly cooled to 4 "C. The cells were washed without and with 50 mM NaCl, 150 mM glycine (pH 3.0) for 3 min at 4 "C. Intracellular EGF was estimated by determination of cell-associated '*'I-EGF following acid washina (31). Cell surface "'I-EGF was estimated by subtraction of the &&rated intracellular radioactivity from the measured total cell associated iz51-EGF. Nonspecific binding and accumulation of 'Y-EGF was determined in experiments using a 500-fold excess of EGF.

Analysis of r3H]Thymidine Incorporation by CHO Cells
CHO cells were seeded in 16-mm wells and grown to a density of 5 X 10' cells/well.
The cells were incubated for 46 h in medium supplemented with 0.1% calf serum. Growth factors were then added to the medium together with 5 &i/ml of [3H]thymidine. After a further 24 h of incubation, the incorporation of radioactivity into acid-insoluble material was measured as described (32).

Characterization of EGF Receptor
Phosphorylation Sites-Treatment of human tibroblasts with phorbol ester or with PDGF causes an increase in the phosphorylation state of the EGF receptor. Tryptic 32P-phosphopeptide mapping of the EGF receptor has demonstrated that the increase in phosphorylation occurs at multiple serine and threonine residues (5-7,17-19).
In order to identify the 32P-phosphopeptide that elutes at 34% acetonitrile, experiments were designed to examine the properties of this peptide. ( 476,129,81,61,69,54,62,46,51,38,38,27, and 26 cpm at cycles 1 through 15, respectively. These data indicate that the 32P-phosphopeptide contains a [32P]phosphoserine residue at position 3 from the amino terminus. Inspection of the cDNA sequence for the EGF receptor (34) indicated that there was only one predicted EGF receptor tryptic peptide that was consistent with the properties observed for the isolated 32Pphosphopeptide.
The predicted phosphorylation site was Ser1047. To test the hypothesis that Ser'047 is an EGF receptor phosphorylation site, we substituted this residue with alanine using site-directed mutagenesis and expressed the mutated receptor in CHO cells. It was expected that the mutated [A""] EGF receptor would be defective at the major site of serine phosphorylation.
However, 32P-phosphopeptide mapping of the mutated [A'047]EGF receptor indicated that the major site of serine phosphorylation, located on a tryptic 32P-phosphopeptide that eluted from the reverse-phase column at 34% acetonitrile, was still present (data not shown). This result suggested that Ser1047 may not be a receptor phosphorylation site. To account for this unexpected result, we tested the hypothesis that the EGF receptor may be phosphorylated at Ser 1o46 in cells expressing the [A'047]EGF receptor. We therefore substituted Ser'046 with an alanine residue to examine the effect of this mutation on the phosphorylation state of the EGF receptor. It was observed that the phosphorylation of the [A'046]EGF receptor was similar to the wild-type EGF receptor (data not shown). However, 32P-phosphopeptide maps of the EGF receptor in which both Ser1046 and Ser'047 were substituted with alanine residues did not contain the tryptic 32P-phosphopeptide that elutes at 34% acetonitrile (Fig. 1). These data indicate that both Ser'046 and Ser1047 can serve as phosphorylation sites on the EGF receptor. Evidence that the EGF receptor can be phosphorylated at Ser'046 and Ser'047 in A431 cells has recently been reported by Heisermann and Gill (19). Together, the data described above demonstrate that the major phosphorylation sites on the EGF receptor can be accounted for by Thr654, Thr669, Ser1046, and Ser1047. In order to examine the role of these phosphorylation sites for EGF receptor function, a series of mutant receptor cDNA constructs was prepared in which each of these sites was mutated by substitution with alanine. The wild-type and mutant EGF receptor cDNAs were then expressed in CHO cells together with the human B-type PDGF receptor (11) for 30 min. The EGF receptors were then isolated by immunoprecipitation and polyacrylamide gel electrophoresis. A marked increase in the phosphorylation state of the wildtype EGF receptor was observed after treatment of CHO cells with phorbol ester (Fig. 1A). An increase in EGF receptor phosphorylation was also observed after phorbol ester treatment of CHO cells expressing mutant receptors lacking either one or two phosphorylation sites. In contrast, phorbol ester caused only a small increase in phosphorylation of the EGF receptors when the four major phosphorylation sites (Thr'js4, Thrfi6' Ser'"46, and Serlo4') residues (Fig. 1A).
were substituted with alanine To further characterize the sites of phosphorylation of the wild-type and mutant EGF receptors, the phosphorylation state of the receptors was investigated by "'P-phosphopeptide mapping. The EGF receptors were digested with trypsin, and the ""P-phosphopeptides obtained were analyzed by reversephase chromatography (Fig. 1). It was observed that the maps obtained for all EGF receptor mutants in which Thr'jS4 was replaced by alanine lacked a "'P-phosphopeptide that eluted at 7% acetonitrile. Mutagenesis of Thr66g resulted in the loss of the "'P-phosphopeptide eluting at 27% acetonitrile. Furthermore, the "'P-phosphopeptide that elutes at 34% acetonitrile was not observed in maps of any of the EGF receptor mutants in which both Ser'046 and Ser'047 had been replaced with alanine residues (Fig. 1). No evidence for the phosphorylation of the mutant EGF receptors at sites that are not phosphorylated in the wild-type receptor was obtained (Fig.  1).
Examination of the [32P]phosphopeptide maps of the mutant EGF receptor lacking the four major sites of phosphorylation (Thres4, Thr@', Serlo4'j, and Ser'047) indicated that this receptor contained several minor sites of phosphorylation on serine residues. Treatment of the CHO cells with phorbol ester caused a small increase in the labeling of all of these sites with [32P]phosphate (Fig. 1). Two alternative hypotheses can account for the data obtained. First, it is possible that the 18-h incubation of the CHO cells with [32P]phosphate was not sufficient to allow for the complete equilibration of all of the phosphate pools in the cell. Consequently, it is possible that the treatment of CHO cells with phorbol ester altered the specific radioactivity of the [32P]phosphate labeling of the EGF receptor. A second hypothesis is that phorbol ester caused a small increase in the phosphorylation of several minor sites on the EGF receptor at a very low stoichiometry. As the increase in apparent phosphorylation observed was small and because the increase was found at all of the sites detected, further work will be required to distinguish between these hypotheses.

Regulation of "'I-EGF
Binding to Cell Surface Receptors by PDGF and Phorbol Ester-The effect of mutagenesis of the major phosphorylation sites of the EGF receptor on the expression of the receptor was examined by investigation of '*"I-EGF binding to cell surface receptors. Transfection of each of the cDNA constructs into CHO cells caused the expression of specific high affinity EGF binding sites at the cell surface. The binding isotherms were found to be curvilinear when plotted by the method of Scatchard (35). Analysis of the '?-EGF binding isotherms using the computer program LIGAND (36) showed that a two-site model provided a significantly better description of the experimental data than a one-site model (p > 0.9). The results of this analysis for the wild-type EGF receptor and the [A654A66gA1046A'047]EGF receptor (Fig. 2) are summarized in Table I.* Treatment of cells with phorbol ester has been reported to cause an acute decrease in the high affinity binding of lz51-EGF to cell surface receptors. This process ("transmodulation") has been proposed to be mechanistically related to EGF receptor phosphorylation (8). The effect of phorbol ester on the high affinity binding of '*'I-EGF to CHO cells expressing EGF receptors lacking the major sites of phosphorylation was therefore examined. Fig. 3 shows that phorbol ester caused a decrease in the specific cell surface binding of lz51-EGF (200 PM) to CHO cells expressing each of the mutated EGF receptors examined. Similar results were obtained in experiments using PDGF to cause EGF receptor transmodulation (data not shown). Investigation of the '251-EGF binding isotherm demonstrated the the effect of phorbol ester and PDGF was to cause a decrease in the high affinity binding of lz51-EGF (data not shown). Together, these data indicate that the phosphorylation of the EGF receptor at Thr654, Thr66g, Ser1046, and Ser'047 is not required for the transmodulation of the EGF receptor caused by phorbol ester and PDGF.
To investigate whether there was a qualitative difference between the transmodulation of the wild-type and mutant ' The analysis of the '*"I-EGF binding isotherm (Table I)   EGF receptors, a detailed study of PDGF and phorbol ester action was performed using CHO cells expressing the wildtype EGF receptor and the mutated [A654A669A1046A1047]EGF receptor. In initial experiments, the effect of phorbol ester and PDGF on the 1251-EGF binding isotherm was investigated (Fig. 2). Analysis of the binding isotherms using the computer program LIGAND CHO cells expressing wild-type or mutated EGF receptors were incubated without (open bars) or with (closed bars) 10 nM PMA for 30 min at 37 "C. The specific binding of 200 PM "'1-EGF to cell surface receptors was measured at 4 "C. The data are normalized (100%) to the specific binding of Y-EGF to cells treated without PMA. Similar results were obtained in three separate experiments for each CHO cell line.
affinity sites at the cell surface (Table I). The extent of the inhibition of iz51-EGF binding was observed to be greater for the wild-type receptor than for the mutated EGF receptor ( Fig. 2 and Table I).
The time course of PDGF and phorbol ester action was investigated on the transmodulation of the wild-type and the mutated [A654A66gA"'46A'047]EGF receptors. Similar results were obtained for both the wild-type and mutated EGF receptors (Fig. 4). Phorbol ester caused a rapid decrease in the binding of iz51-EGF to cell surface receptors which was sustained for at least 2 h (Fig. 4). In contrast to the sustained effect of phorbol ester, it was observed that the inhibition of "'1-EGF binding caused by PDGF was transient (Fig. 4). Maximal inhibition was observed after approximately 15 min of treatment with PDGF. EGF binding returned close to control levels after 1 and 2 h of treatment for the mutant and wild-type EGF receptors, respectively (Fig. 4). The effect of PDGF and phorbol ester concentration was examined on EGF receptor transmodulation. Fig. 5 shows that the dose-response of PDGF and phorbol ester to inhibit i'sI-EGF binding was similar for the wild-type and the [A654A66gA'046A1047]EGF receptors. At high concentrations of PDGF and phorbol ester it was observed that the maximum extent of inhibition of lz5 I-EGF binding was greater for the wild-type receptor than the mutated [A654A66gA1046A1047]EGF receptor (Fig. 5).
Regulation of EGF Receptor Endocytosis and Degradation-The endocytosis of EGF receptors was investigated by incubation of cells expressing wild-type and mutated EGF receptors with lz51-EGF at 37 "C. The accumulation of lz51-EGF into an intracellular compartment after receptor-mediated endocytosis was examined by washing the cell monolayers at 4 "C with 50 mM NaCl, 150 mM glycine (pH 3.0) to remove lz51-EGF bound to the cell surface. Rapid endocytosis of iz51-EGF was observed in experiments using cells expressing wildtype and mutated receptors (data not shown). No significant differences were found between wild-type EGF receptors and receptors mutated at Thr654, Thr6"', Ser'046, or Ser'047. Subsequent to endocytosis, the EGF receptor is degraded in lysosomes (2). The effect of EGF to cause down-regulation of the EGF receptor was investigated by two independent methods. First, the effect of EGF to cause a decrease in the expression of EGF binding sites at the cell surface was examined. Secondly, the effect of EGF on the expression of receptors by cells labeled with [35S]methionine was investigated, It was observed that EGF caused down-regulation of both wild-type and mutant receptors expressed in CHO cells (data not shown). We conclude that the substitution of Thrffi4, Thr6'jg, Ser1046, and Ser'047 with alanine residues caused no significant alterations to EGF receptor down-regulation.
Regulation of EGF Receptor Tyrosine Protein Kinase Activity by Protein Kinase C-The tyrosine protein kinase activity associated with the intracellular domain of the EGF receptor is stimulated by the binding of EGF to the extracellular domain of the receptor. One of the substrates for the tyrosine protein kinase activity is the receptor itself. Consequently, the addition of EGF to cells causes an increase in the autophosphorylation of the receptor on tyrosine residues. EGF also causes an increase in the phosphorylation state of the EGF receptor at serine and threonine residues (37). To examine the role of the phosphorylation of serine and threonine residues on the function of the receptor, the effect of the substitution of Thr654, Thr@', Ser'046, and Ser'047 with alanine on EGF receptor autophosphorylation was investigated. Fig.  6 shows that EGF stimulated the autophosphorylation of the receptor on tyrosine residues in cells expressing wild-type and all of the mutated EGF receptors examined.
Treatment of cultured cells with phorbol ester has been shown to inhibit the tyrosine protein kinase activity of the EGF receptor (12,13). To evaluate the role of the phosphoryl- Role of EGF Receptor Phosphorylation during Signal Transduction-The role of EGF receptor phosphorylation in the signalling pathway for cellular proliferation was investigated by measurement of the growth factor-stimulated incorporation of [3H]thymidine into DNA. Fig. 7 (38), the data presented here demonstrate that mitogenic signal transduction by the EGF receptor does not require Thr654, Thr'j6', Ser'046, or Ser1047 (Fig. 7). ation of the EGF receptor the effect of mutagenesis of Thr'jc4, In further experiments, the effect of PDGF to stimulate the Thr@', Ser'046, and Ser1o47 on the regulation of the autophos-growth of CHO cells was examined. It was observed that phorylation of the EGF receptor by phorbol ester was exam-PDGF increased the EGF-stimulated incorporation of [3H] ined. EGF stimulated tyrosine phosphorylation of the wild-thymidine by cells expressing both wild-type and mutated type receptor was observed to be inhibited following treatment EGF receptors (Fig. 7). Thus, the mechanism of PDGF action of cells with phorbol ester (Fig. 6). Similarly, phorbol ester to stimulate cellular proliferation does not require the phos-A B C caused a decrease in autophosphorylation of the mutated [A66g]-and [A1046A'047]EGF receptors (Fig. 6). These data indicate that the phosphorylation of Thr66g, Ser1046, and Ser1047 is not required for the inhibition of EGF receptor autophosphorylation caused by phorbol ester. In contrast, the phorbol ester-mediated regulation of receptor autophosphorylation was not observed in cells expressing the mutated [A654]EGF receptor. This result suggests that Thr654 is required for the regulation of receptor autophosphorylation caused by phorbol ester. Consistent with this hypothesis is the observation that all of the EGF receptors in which Thr654 was substituted with an alanine residue ( [A'j5*]-, [A654A66g] -, and [A654A66gA1046A1047] EGF receptors) were defective in the regulation of autophosphorylation caused by phorbol ester (Fig. 6). CHO cells expressing wildtype and mutant EGF receptors were incubated without or with 10 nM PMA for 30 min at 37 "C. The cells were then incubated without or with 100 nM EGF for 5 min at 37 "C. The EGF receptors were isolated by immunoprecipitation and polyacrylamide gel electrophoresis. Tyrosine phosphorylation of the receptors was determined by Western blot analysis using a monoclonal anti-tyrosine phosphate antibody and a ?-goat anti-mouse second antibody. The figure presents a representative autoradiograph of the Western blot analysis for each CHO cell line investigated. Similar results were obtained in three separate experiments for each cell line. phorylation of the EGF receptor at Thr654, Thr66g, SerlM6, and Ser1047.

DISCUSSION
Treatment of cultured cells with phorbol ester or with PDGF causes a rapid decrease in the high affinity binding of EGF to cell surface receptors and an inhibition of the tyrosine protein kinase activity of the EGF receptor. Associated with these changes in EGF receptor function is an increase in the extent of serine and threonine phosphorylation of the EGF receptor at several sites. The role of receptor phosphorylation is incompletely understood, but it has been proposed that the phosphorylation of the EGF receptor is mechanistically related to the process of EGF receptor transmodulation (reviewed by Schlessinger,Ref. 2). The purpose of the study reported here was to test this hypothesis. The approach that we have taken was to examine the effect of the mutagenesis of the major sites of phosphorylation on the EGF receptor.
Phosphopeptide mapping of the wild-type EGF receptor isolated from phorbol ester-treated cells demonstrates that there are three major tryptic phosphopeptides that elute from a reverse-phase Cl8 column at 7,27, and 34% acetonitrile (Fig.  1). These phosphopeptides have been characterized, and the phosphorylated residues have been identified as Thr654 (4,5), Thr669 (19,33), and Ser'047 ( Fig. 1 and Ref. 19), respectively. The mechanisms by which these sites are phosphorylated in phorbol ester-treated cells are distinct. Thr654 is a substrate for protein kinase C (4, 5), Thr669 is a substrate for a novel growth factor-stimulated protein kinase (33), and Ser1047 may be a substrate for a calmodulin-dependent protein kinase (19). The role of the phosphorylation of the EGF receptor at these major sites of phosphorylation was examined by replacement of these residues with alanine.
Phosphopeptide mapping of EGF receptors lacking Thr654 resulted in the loss of the phosphopeptide that eluted from the reverse-phase column at 7% acetonitrile, as expected (Fig.  1). Similarly, the mutagenesis of Thre9 caused the loss of the peptide eluting at 27% acetonitrile. However, this result was not expected because Heisermann and Gill (19) have reported that Ser671, which is located adjacent to Thr6@ in the phosphopeptide, is also a substrate for phosphorylation. The replacement of Thr669 with an alanine residue therefore appears to prevent the phosphorylation of Ser671. As Ser671 was not phosphorylated on the [A669]EGF receptor (Fig. l), the effect of site-directed mutagenesis of Ser671 was not investigated in the present study.3 In contrast to the results obtained for the mutagenesis of Thr654 and Thr669 (Fig. l), the substitution of Ser1047 with an alanine residue caused no significant alteration to the phosphopeptide map of the EGF receptor. This result could be accounted for by the phosphorylation of Ser'04'j in the [A'047]EGF receptor (19). To investigate this possibility, Ser'046 was replaced with an alanine residue. It was observed that the phosphopeptide map of the [Alo4'j]EGF receptor was similar to that of the wild-type EGF receptor. However, phosphopeptide maps of the [A1046A1047]EGF receptor lacked the phosphopeptide that elutes at 34% acetonitrile from the HPLC column (Fig. 1). These data indicate that both Ser1046 and Ser'047 are sites of EGF receptor phosphorylation (19). Substitution of all of the major sites of phosphorylation (Thr654, Thr669, Ser1046, and Ser'047) with alanine residues caused a marked decrease in the level of phosphorylation of the EGF receptor (Fig. 1). Phosphopeptide mapping demonstrated that the [A654A669A1046A1047]EGF receptor lacked all of the major sites of phosphorylation, but retained phosphorylation at several minor sites that were also observed in maps of the wild-type EGF receptor.
Inhibition of High Affinity EGF Binding by Protein Kinase C-EGF receptor transmodulation was examined by investigation of the high affinity binding of lz51-EGF to cell surface receptors. A decrease in the binding of lz51-EGF was observed when cells were incubated with phorbol ester or PDGF. Similar results were observed for cells expressing wild-type receptors and mutated EGF receptors (Fig. 3). Thus, phorbol ester and PDGF inhibit lz51-EGF binding to cells expressing mutated EGF receptors that lack the major sites of EGF receptor phosphorylation (Thr654, Thr669, Ser'046, and Ser1047). We conclude that the mechanism of action of PDGF and phorbol ester to cause a decrease in the apparent affinity of the EGF receptor is independent of receptor phosphorylation at Thr654, ThrM9, Serlo4'j, and Ser1047.
Further studies are required to identify the mechanism of EGF receptor transmodulation. Two general hypotheses can be presented to account for the mechanism of EGF receptor transmodulation. 1) Transmodulation of the EGF receptor may be caused by phosphorylation of the receptor. The data reported here demonstrate that the major sites of EGF receptor phosphorylation are not mechanistically related to transmodulation. Therefore, if receptor phosphorylation is relevant to transmodulation, the stoichiometry of phosphorylation at the potential regulatory sites is extremely low (Fig. 1). It is also possible that transmodulation is caused by the phosphorylation of a protein that interacts with the receptor. 2) Phosphorylation may not account for EGF receptor transmodulation. The decrease in the apparent affinity of the EGF receptor may therefore be accounted for by the interaction of the receptor with other cellular components. Rigorous tests of these hypotheses are required for future progress toward  (Fig. 1).
the understanding of the mechanism of EGF receptor transmodulation.

Inhibition of EGF Receptor Tyrosine Protein Kinase Activity
by Protein Kinase C-Treatment of cells with phorbol ester causes an inhibition of the EGF receptor tyrosine protein kinase activity (12,13). The role of EGF receptor phosphorylation during this process was examined by investigation of the effects of the substitution of the major receptor phosphorylation sites (Thr654, Thr66g, Serlo4'j, Ser'047) with alanine residues. It was observed that phorbol ester caused an inhibition of the autophosphorylation of EGF receptors lacking Thr@j' (Fig. 6). Similar results were observed for EGF receptors lacking Ser'046 and Ser1047 (Fig. 6). These data demonstrate that the phosphorylation of the EGF receptor at Thrmg, Ser'046, and Ser1047 is not required for the inhibition of the EGF receptor tyrosine protein kinase activity caused by phorbol ester. In contrast, no significant inhibition of receptor autophosphorylation was observed after substitution of Thr654 with alanine (Fig. 6). This result indicates that Thr654 is required for the inhibition of EGF receptor autophosphorylation.
The possible functional interaction between the phosphorylation of Thr654 and the phosphorylation of the EGF receptor at other sites was examined. The approach that we used was to compare the properties of receptors with single point mutations with receptors in which two or more phosphorylation sites were replaced with alanine residues. It was observed that phorbol ester did not inhibit the autophosphorylation of any of the mutated receptors that lack Thr654 (Fig.  6). In contrast, phorbol ester caused an inhibition of the autophosphorylation of the wild-type EGF receptor and all of the mutated receptors with Thr654 (Fig. 6). These data indicate a primary role for Thr654 compared with other receptor phosphorylation sites during regulation of the receptor tyrosine protein kinase activity by phorbol ester.
Evidence to support the hypothesis that phosphorylation of Thrm4 by protein kinase C accounts for the regulation of the receptor tyrosine protein kinase has been obtained from the investigation of the effects of PDGF on human fibroblasts. Comparison of different cell lines has demonstrated that there is a correlation between the ability of PDGF to cause phosphorylation of Thr654 and to cause an inhibition of the EGF receptor tyrosine protein kinase activity. Transmodulation of the EGF receptor caused by treatment of WI-38 human fetal lung fibroblasts with PDGF causes both an inhibition of the EGF receptor tyrosine protein kinase activity and an increase in the state of phosphorylation of Thr654 (6, 7). In contrast, treatment of FS4 human skin fibroblasts with PDGF does not increase Thr654 phosphorylation, and EGF receptor transmodulation occurs without an inhibition of the receptor tyrosine protein kinase activity4 (39). Thus, the ability of PDGF to cause an inhibition of the EGF receptor tyrosine protein kinase activity correlates with phosphorylation of the EGF receptor at Thr654 (6,7,39). More direct evidence that supports the hypothesis that Thr654 phosphorylation modulates the EGF receptor tyrosine protein kinase activity has been obtained from the results of in vitro reconstitution experiments in which the phosphorylation of Thr654 by protein kinase C is associated with an inhibition of the receptor tyrosine protein kinase activity (13-15). Together, these data strongly support the hypothesis that phosphorylation of Thr654 represents an important mechanism of regulation of the EGF receptor tyrosine protein kinase activity. However, it has recently been reported that the EGF receptor tyrosine protein kinase activity can be regulated by additional mechanisms. Treatment of cells with the non-phorbol ester tumor promoter thapsigargin causes a decrease in the tyrosine protein kinase activity of the EGF receptor without causing phosphorylation of the receptor at Thr654 (16). The regulation of the EGF receptor tyrosine protein kinase activity caused by thapsigargin therefore occurs by a mechanism that is independent of Thr654 phosphorylation. Further studies are warranted to identify regulatory mechanisms that are independent of Thr654 and to examine their physiological significance.
Role of EGF Receptor Phosphaylation during Mitogenic Signal Transduction-The role of EGF receptor phosphorylation was examined by comparison of the growth factorstimulated incorporation of [3H]thymidine into DNA by cells expressing wild-type EGF receptors with cells expressing mutated EGF receptors that lack all of the major sites of serine and threonine phosphorylation (Thr654, Thr%', Ser'046, and Ser'047). It was observed that EGF caused an increase in the incorporation of [3H]thymidine by cells expressing wild-type and mutated EGF receptors (Fig. 7). However, the mitogenic effects of EGF mediated by the mutated receptor lacking phosphorylation sites was observed to be significantly greater than that mediated by the wild-type receptor. The modest effect of EGF on stimulating [3H]thymidine incorporation by cells expressing wild-type EGF receptors is consistent with previous reports of functional desensitization of EGF receptor signalling (38,40). Investigation of the role of individual phosphorylation sites indicated that the marked increase in EGF receptor signalling could be accounted for by the mutagenesis of Thr 654 Livneh et al. (10) have reported that the . increased mitogenic responses mediated by the EGF receptor lacking Thr654 is caused by a defect in the desensitization of the EGF receptor. The data reported here are consistent with this hypothesis.
Conclusions-A summary of the conclusions about the role of EGF receptor phosphorylation that we have drawn from the results of this study is presented in Fig. 8 of the EGF receptor at Thr654 plays a critical role in the regulation of the tyrosine protein kinase activity of the EGF receptor. In contrast, the mechanism of regulation of the apparent affinity of the EGF receptor caused by PDGF and phorbol ester is independent of the phosphorylation of the EGF receptor at all of the major sites of serine and threonine phosphorylation.