Monoclonal antibodies to mouse epidermal growth factor (EGF) receptor were prepared by the immunization of rats with receptor glycoprotein purified from mouse liver by affinity chromatography on immobilized EGF. Purified mouse EGF receptor retained EGF-inducible autophosphorylating activity and was antigenic in rats and rabbits. The monoclonal antibodies cross react very poorly with human EGF receptor, while polyclonal rabbit antibodies immune precipitate human, rat and mouse EGF receptor equally well. The rabbit antibody blocks EGF binding to mouse fibroblast cells and, at 20-fold higher concentrations, stimulates uptake of tritiated thymidine into DNA. This indicates that antibodies bind at or close to the EGF-binding site and can mimic the effects of the growth factor. None of the monoclonals bind at the EGF site of the receptor.

Immunoprecipitation, immunoblotting, 125I-EGF cross linking, 125I-surface labelling, immunohisto-chemistry and autophosphorylation techniques were used to delineate the basis for the induction of EGF receptors when OC15 embryonal carcinoma (EC) cells differentiate into endodermal derivatives (END). EGF-stimulated autophosphorylation of a 170 × 103 Afr protein in solubilized OC15 EC cells is readily detectable, although intact EC cells do not bind or respond to EGF by all other tests. The results suggest that cryptic EGF receptors are present in EC stem cells, a finding with implications in development.

Epidermal growth factor (EGF) is a polypeptide mitogen which exerts its action through a receptor glycoprotein of Afr170000 which spans the cell membrane (reviewed by Adamson & Rees, 1981; Gregoriou & Rees, 1984; Herschman, 1985). The receptor possesses an endogenous tyrosine kinase activity which makes it a member of the src oncogene family (Yamamoto et al. 1983). Protein tyrosine kinases have been found ubiquitously in growth factor receptors and oncogenes, and are generally believed to mediate both normal mitogenic responses and transformation (Heldin & Westermark, 1984; Bradshaw, 1986). The v-erbB oncogene product (gp65-erbB) from avian erythroblastosis virus is a truncated form of the EGF receptor, having lost the EGF-binding domain but retaining kinase activity (Downward et al. 1984; Lin et al. 1984; Xu et al. 1984). It is commonly assumed that the unregulated activity of the v-erbB kinase accounts for its oncogenicity. However, the natural substrates for EGF receptor and the other tyrosine kinases are poorly defined (Hunter & Cooper, 1981; Brugge, 1986; Huang, 1986), and how the mitogenic signal passes to the nucleus to induce DNA replication is completely unknown.

The epidermal growth factor receptor has been studied most in A431 cells, a human vulval carcinoma cell line which produces two to three million copies of EGF receptor per cell. The A431 EGF receptor (Ullrich et al. 1984) protein consists of an external domain of 621 amino acids (Mr 110x103 when glycosylated) linked by a single transmembrane segment of 23 amino acids to the cytoplasmic domain of 524 amino acids (Mr-60 × 103). The cytoplasmic side of the molecule contains the kinase and three tyrosine autophosphorylation sites at the carboxy tail (Downward, Parker & Waterfield, 1984). These three sites comprise a Mr10 × 103 domain which can be cleaved from the receptor by the calcium-activated protease, calpain (Gates & King, 1982; Cassel & Glaser, 1982), leading to the common appearance of the EGF receptor as a Mr170/150 × lCr doublet (Cohen, Carpenter & King, 1980; Cohen, Fava & Sawyer, 1982; Linsley & Fox, 1980). The physiological significance (if any) of the calpain cleavage is not known and the role of autophosphorylation in the regulation of endogenous kinase activity is hotly debated (Bertics & Gill, 1985; Gullick, Downward & Waterfield, 1985; Downward, Waterfield & Parker, 1985; Gates & King, 1983).

We wish to study the structure and function of the EGF receptor in cells that respond mitogenically to EGF. In particular, our aim is to define the role, if any, for EGF receptors during murine development. Mouse embryonic and fetal tissues display 125I-EGF-binding activity on cell membranes (Adamson, Deller & Warshaw, 1981) and numbers and affinities of EGF receptors change during the gestation period (Adamson & Meek, 1984). At least some of these receptors in 13-14th day fetuses can respond to EGF by down regulation (Adamson & Warshaw, 1981). However, whether EGF receptors play a role in embryonic cell proliferation, in tissue maturation or in differentiated function in vivo remains unknown. An effective means of exploring the properties and functions of embryonic murine receptors is by the use of rabbit polyclonal antibodies and a range of rat monoclonal antibodies generated by immunization with affinitypurified mouse liver EGF receptor. To our knowledge, the production of monoclonal antibodies to mouse EGF receptors has not been described before. The polyclonal anti-EGF receptor blocks EGF binding to 3T3 cells and at higher concentrations mimics the mitogenic effect of EGF. Both the polyclonal and one monoclonal antibody (9D6) immunoprecipitate EGF receptor from mouse liver, 3T3 cells and the mouse embryonal carcinoma (EC) cell line OC15.

The induced differentiation of teratocarcinoma cell lines has proven to be a useful model for development and differentiation in vivo. Undifferentiated OC15 EC cells had been previously shown in our laboratory to bind little or no iodinated EGF, but after differentiation to endoderm-like cells (END, induced by retinoic acid) they bind and respond to EGF (Rees, Adamson & Graham, 1979; Adamson & Hogan, 1984). Using the antibodies that we have developed, we show here that undifferentiated OC15 EC cells synthesize EGF receptors with kinase activity, but these receptors are not expressed on the cell surface until after differentiation. The implications of this finding to embryonic development are discussed.

Purification of EGF receptor from mouse liver

Epidermal growth factor receptor was purified from adult mouse liver by affinity chromatography on EGF-Affigel. EGF (30 mg) was purified from male mouse submaxillary glands by the method of Savage & Cohen (1972). This was reacted with 3 · 6g of Affigel-10 (Bio-Rad Labs) by the anhydrous method as described by Cohen et al. (1982). EGF receptor was purified in a single step with the following modifications: 10 mM-EDTA was added to the homogenization buffer and receptor was eluted by recycling 2 ml of ethanolamine (2 · 5 mM, pH9-7) through the column 10 times to maximize the receptor concentration in the eluant. The yield of EGF receptor per column was approximately 100 fig per 15 g of fresh-frozen liver.

Generation of rabbit polyclonal antibody

The rabbit was initially injected (s.c.) with 100 μ g of purified mouse EGF receptor in complete Freund’s adjuvant and boosted 10 days later with 50 μ g. Serum was collected on the 10th day after the initial boost and after each subsequent boost at monthly intervals. Ig was partially purified by 40 % ammonium sulphate precipitation followed by exhaustive dialysis of the precipitate against phosphate-buffered saline (PBS). Samples of rabbit anti-mouse EGF receptor were stored at — 100°C. Protein was estimated by absorbance at 280 nm.

Generation of rat monoclonal antibodies

A female BDIX rat was injected i.p. with 20 μ g of purified EGF receptor in Freund’s complete adjuvant and then boosted twice at 10-day intervals with 10 μ g of EGF receptor. The spleen was harvested 3 days after the last boost and spleen cells fused with mouse myeloma cells (P3x63Ag8.541) in the presence of serum-free medium, 45 %, polyethylene glycol 4000 and DMSO as described by St Groth & Schneidegger (1980). Fused cells were selected in HAT medium.

Initial screening for EGF receptor-reactive hybridomas was performed by ELISA assay (Engvall, 1980). Positive hybridomas were subcloned after testing for their ability to react with purified mouse EGF receptor in ‘Western’ immunoblotting and to immunoprecipitate the receptor.

The three highly positive hybridomas (4D5, 5D4, and 9D6), which are described in this paper, were typed with an antibody-typing kit from Miles Scientific. All three were found to be producing IgG2a. Two nonreactive hybridomas (4B4, 5A2) which were used as negative controls were found to be IgMs.

For growth studies, monoclonal antibody 9D6 was partially purified by ammonium sulphate precipitation of serum-free medium conditioned by 9D6 cells, followed by exhaustive dialysis against PBS and sterile filtration.

Western immunoblotting

Purified EGF receptor or a mouse liver membrane preparation (Thom, Powell, Lloyd & Rees, 1977) was analysed on a 7 % SDS-PAGE gel (Laemmli, 1970) and the gel pressure-blotted onto nitrocellulose (Towbin, Staehelin & Gordon, 1979). The nitrocellulose was cut into strips corresponding to the gel lanes, then blocked with PBS containing 1 % BSA and 1 % either normal goat serum (for rabbit antibody) or normal rabbit serum (for monoclonals). Strips were washed extensively between each layer with NaCl(0 · 9 %)-Tween 20(0-05%). After washing, each strip was reacted with one hybridoma culture supernatant or dilution of rabbit serum for 48 – 72 h. Bound antibody was detected by biotin-linked anti-rabbit or anti-rat Ig and developed with a glucose oxidase ABC reaction kit (Vector Laboratories, CA).

OC15 embryonal carcinoma cell culture and differentiation

OC15 EC cells were cultured in alpha-modified MEM plus 10% FCS and split 1:8 every 2 days. To induce differentiation into END cells, OC15 cells were plated at 1 × 10s cells per 60 mm dish in the presence of l × 10− 6M-retinoic acid (RA; Sigma). Cells were grown for 3, 5, or 7 days in the continuous presence of RA. For cells collected on days 5 or 7, medium was renewed on days 3 and 5.

Immunohistochemical staining

In fluorescence tests, monolayers of cells grown on coverslips were washed in PBS and reacted unfixed as indicated in figure legends. Rabbit anti-EGF receptor Ig (200μg ml − 1) and undiluted monoclonal antibody supernatants were allowed to react with cells for 10 – 60 min before washing (20 min) and applying second antibody suitably diluted (Cappel Laboratories, PA). Immunofluorescence samples were viewed in a Nikon Optiphot microscope equipped with epifluorescence.

For greater sensitivity in testing mouse cells, the avidinbiotin-peroxidase procedure (Vector Laboratories, CA) was used. Cells were fixed in 3 · 7 % formaldehyde in PBS and then permeabilized with methanol. Peroxidase was located by a brown reaction product against cytoplasm counterstained with fast green, as described before (Adamson, Meek & Edwards, 1985).

Immunoprecipitation

Purified receptor, membranes, mouse liver homogenate or cultured cell homogenates were diluted in NET buffer (0 · 4M-NaCl, 5mM-EDTA, 0 · 05M-Tris, pH 8, 1 mM-PMSF, 1 % NP40, 0 · 02 % NaN3) and reacted with antibody over-night at 4°C. Protein A agarose beads (Sigma) were added to the mixture for 1 · 3 h to bind antigen-antibody complex and the supernatant discarded. For the rat monoclonal antibodies, it was necessary to precoat the beads with rabbit anti-rat Ig (Cappel), as the rat antibodies would not bind directly to protein A under the conditions we employed. After antibody-antigen was bound, the beads were washed extensively in RIPA buffer (0 · 15M-NaCl, 5mM-EDTA, 0 · l M-Tris, pH7 · 5, 0 · 1% SDS, 1% DOC, 1% NP40, 0 · 02 % NaN3) to remove nonspecifically bound proteins, then boiled in 2 × Laemmli buffer and analysed on 5 % or 7 · 5 % SDS-PAGE gel (Laemmli, 1970). Gels were stained, destained, dried and autoradiographed.

Receptor autophosphorylation

Purified receptor (3 μ g per sample) was autophosphorylated (Klein, Freidenberg, Cordera & Olefsky, 1985) in NET buffer containing 5mM-MnCl2, 12mM-MgCl2 and 5 – 10μCi of γ -32P-ATP (25 – 35 Ci mmol − 1, NEN). Reactions were carried out for 10 min at 0°C, with or without a 40 min pre-incubation with 2nM-EGF, and stopped by dilution in Laemmli buffer. Autophosphorylation of immunoprecipitated receptor (from mouse liver) was performed after receptor binding to protein A-agarose beads. The beads were reacted 10 min in the same solution as above then diluted and washed before boiling in Laemmli buffer.

For autophosphorylation of receptor in OC15 cells (Fig. 8), samples containing lysate from 106 cells (105 for A431) in 300 μl lysis buffer were aliquoted and incubated for lh on ice with or without 5μgml−1 EGF. Lysis buffer consisted of 0 · 4M-NaCl, 5mM-EDTA, 50mM-Tris-HCl, pH8 · 0, ImM-PMSF, and 0 · 6% Brij 35. The phosphorylation reaction was started by the addition of 100 al of a solution containing 10 μCi γ -32P-ATP and MnCl2 and MgCl2 to make final concentrations of these ions of 5 mM and 12 mM, respectively.

The ATP reaction was carried out for exactly 4 min, then stopped by the addition of 400 μ l of the buffer described above but with 1 % NP40 instead of Brij 35 and also containing 2mM-cold ATP, 20 mM-inorganic pyrophosphate, 200mM-sodium fluoride, and 0 · 01 mM-sodium metavanadate. Antibody was added immediately and incubated overnight at 4°C. Processing then continued as described above.

Iodinated EGF binding to receptor

Receptor grade EGF (purchased from Collaborative Research) was iodinated by the immobilized lactoperoxidase: glucose oxidase method (Enzymobeads, Bio-Rad) using the manufacturer’s directions. Iodinated EGF (the TCA-precipitable peak) was separated from free iodine by a Sephadex G-50 column.

The ability of anti-EGF receptor antibodies to block iodinated EGF binding was tested on monolayers of NIH 3T3 cells in 24-well plates (Costar). Confluent cell layers were washed with ice-cold binding buffer (DMEM plus 15mM-HEPES and Imgml−1 BSA, pH 7 · 4), then incubated 2h at 0°C in binding buffer containing [125I]-EGF (2 × 106ctsmin − 1) in the presence or absence of antibodies or 100-fold excess cold EGF. After incubation, the wells were washed twice with ice-cold-binding buffer, the cell monolayer dissolved in 0 · 5M-NaOH and counted in a gamma counter.

Covalent cross linking of iodinated EGF to receptors

Covalently cross-linked 12SI-EGF was used as an affinity label for the EGF receptor in some immunoprecipitation experiments. OC15 EC or END cells (as described below) were allowed to bind 125I-EGF (106ctsmin − 1 per plate) as described for the 3T3 cells, in the presence or absence of 2μg cold EGF. After 2h of binding, the cells were washed twice with binding buffer and 5 ml of fresh binding buffer was added to each 60 mm dish. To this was added 100pl of DMSO containing disuccinimidyl suberate (DSS, Pierce) to give a final concentration of cross linker of 0 · 65 mM (Massagué & Like, 1985). The cross-linking reaction proceeded for 15 min at 0°C, after which the monolayers were washed with a solution containing 0 · 25M-sucrose, 10 mn-Tris, ImM-EDTA, 0 · 1 mM-phenyl-methy) sulphonyl fluoride and 5 μ ml−1 leupeptin and scraped off in the presence of this buffer. The collected cells were centrifuged and resuspended in 0 · 5 ml of the above buffer containing 1 % v/v Triton X-100 and allowed to lyse for 30min on ice. Aliquots of this lysate were immunoprecipitated and run on a gel as described.

To cross link iodinated EGF to liver membranes, the same method was employed. Binding, cross linking and lysis took place in an Eppendorf tube, in 1 ml volume with the amounts of additives proportionally reduced. Washes were performed by centrifugation of the membrane pellet and aspiration of the supernatant.

Surface iodination of live cells

Cells were grown to confluence in 60 or 100 mm dishes and approximately 3 – 6 × 106 cells of each type were iodinated in situ on ice with 1 mCi of [125I]sodium iodide and 0 · 2 mg ml−1 of lactoperoxidase according to the method of LeBien, Boue, Bradley & Kersey (1982). Uncross-linked iodine was removed by ice-cold PBS washes and the cells were lysed on the plate by addition of NET buffer.

Metabolic labelling of OC15 EC and END cells

OC15 EC and END cells grown as described above were preincubated for 2h in DMEM containing 10% dialysed feta) calf serum and lacking methionine, after which 100 μ Ci ml−1 of [35S]methionine (NEN) was added, and the cells were incubated for 1 h. After PBS washing, the cells were lysed in the buffer described by Oshima (1981). Aliquots of the lysate containing 107 TCA-precipitable ctsmin−1 were precleared on protein A-agarose beads in the absence of antibody and then immunoprecipitated as described.

Uptake of [3H]thymidine into DNA

NIH 3T3 cells were seeded in 24-we)l plates at 5 × 104 cells per well and grown for 4 days in DMEM + 10 % FCS. They were then serum-deprived for 8h by replacing the medium with DMEM + 0-5 % FCS. At the end of this period, the medium was replaced with serum-free defined medium containing transferrin (10pg ml−1), insulin (10 μ g ml−1) and linoleic acid-albumin (Ipgml−1) as the only proteins. Purified rabbit or monoclonal antibodies were added at the indicated concentrations, ±10 μ g ml−1 EGF (culture grade, Collaborative Research). Positive control wells received fresh DMEM +10% FCS or lOngml−1 EGF in serum-free medium. After 16-20h in the presence of these additives, 0 · 5pCi of tritiated thymidine (20Cimmol−1; NEN) was added to each well and the cells incubated an additional 4 h. The medium was then removed and the cells washed with ice-cold PBS. Ice-cold 10% TCA was added for 15 min, followed by two more PBS washes. The precipitated cellular components were dissolved by the addition of 0 · 5 ml 0 · 5 M-NaOH per well. After neutralization with 50 μ l of 6M-HC1, the contents of the wells were transferred to scintillation vials and counted in Beta Phase (West Chem).

Properties of the mouse EGF receptor

Despite its relative scarcity in mouse liver (approximately 6 × 105 copies per ceil as opposed to 2 · 3 × 106 in A431 cells; Cohen et al. 1982), EGF receptor is nevertheless relatively easy to purify in milligram amounts using an EGF affinity column. The limiting factor for this protocol is the requirement for milligram quantities of pure EGF, but once the EGF-Affigel has been made, it can be used again and again with little loss of activity. The mouse liver EGF receptor, as eluted directly from the column appears as a doublet of Aμ170 and 150 × 103 (Fig. 1A). The purified receptor retains strong autophosphorylating activity, as shown in Fig. IB, which is a 10-min film exposure of 3 pg of autophosphorylated receptor. Phosphorylation occurs primarily on the intact (Mr170 × 103) receptor, but a small amount of Mr 150 × 10a protein phosphorylation is observed if the gel is exposed longer. The receptor retains at least some ability to bind EGF (a property which is readily lost during purification; Cohen et al. 1982), since both autophosphorylation and phosphorylation of exogenous substrates (Fig. 1C) are enhanced by preincubation with EGF. This purified and partially active receptor was used as an immunogen in a rat and a rabbit with the aim of producing a range of specific antibodies.

Fig. 1.

Electrophoresis and autoradiography of EGF receptor purified using affinity chromatography on EGF-Affigel. (A) Coomassie blue staining of M, markers (left) and EGF-Affigel eluate (right) after electrophoresis on a 7% acrylamide gel. (B,C) Autoradiography of EGF-Affigel eluate after phosphorylation with γ -32p-ATP, preincubated 40min, room temperature, with or without 100 nm-EGF (Klein et al. 1985). (C) Phos-phorylation of casein (Sigma C-4765) by purified EGF receptor, 15% acrylamide gel.

Fig. 1.

Electrophoresis and autoradiography of EGF receptor purified using affinity chromatography on EGF-Affigel. (A) Coomassie blue staining of M, markers (left) and EGF-Affigel eluate (right) after electrophoresis on a 7% acrylamide gel. (B,C) Autoradiography of EGF-Affigel eluate after phosphorylation with γ -32p-ATP, preincubated 40min, room temperature, with or without 100 nm-EGF (Klein et al. 1985). (C) Phos-phorylation of casein (Sigma C-4765) by purified EGF receptor, 15% acrylamide gel.

Antibody production and selection

Monoclonal-antibody-generating hybridomas were produced from the fusion of rat spleen lymphocytes with mouse myeloma cells and initially screened by reaction with the EGF-Affigel purified mouse liver receptor in ELISA. Three highly positive hybridomas, designated 4D5,5D4 and 9D6, were subcloned and their properties further tested.

Nitrocellulose blots (‘Westerns’) of purified mouse EGF receptor fractionated on a 7% acrylamide gel showed that the rabbit antibody and monoclonals 4D5 and 9D6 reacted specifically with a MR 170 – 150 × 103 protein (Fig. 2). The rabbit antibody was also capable of absorbing to the receptor in a nitrocellulose-blotted preparation of liver membrane proteins (Fig. 2C) and also recognizes only one additional band of Mr78× 103.

Fig. 2.

Western immunoblotting using antibodies to mouse EGF receptor. (A-C) Rabbit sera against a blot of purified receptor (A,B) or liver membrane preparation (C). (A) Preimmune serum. (B,C) Rabbit anti-EGF receptor serum, 1:1000 dilution. (D-F) Rat monoclonal antibodies; hybridoma culture media against a blot of purified receptor. (D) 4B4 (ELISA-negative). (E) 4D5. (F) 9D6. Methodology is described in the text.

Fig. 2.

Western immunoblotting using antibodies to mouse EGF receptor. (A-C) Rabbit sera against a blot of purified receptor (A,B) or liver membrane preparation (C). (A) Preimmune serum. (B,C) Rabbit anti-EGF receptor serum, 1:1000 dilution. (D-F) Rat monoclonal antibodies; hybridoma culture media against a blot of purified receptor. (D) 4B4 (ELISA-negative). (E) 4D5. (F) 9D6. Methodology is described in the text.

Immunoprecipitation

A number of ELISA-positive monoclonals were screened for their ability to precipitate the purified, autophosphorylated receptor. Three monoclonals, 4D5, 5D4 and 9D6, and the rabbit antibody were found to immunoprecipitate especially well. Results of immunoprecipitation of EGF receptor from whole adult mouse liver by the rabbit polyclonal antibody and monoclonals 4D5, 5D4 and 9D6 are shown in Fig. 3. For this experiment, the receptor was labelled by autophosphorylation with γ -32P-ATP while bound to the agarose beads. The rabbit antibody (lanes 3 and 4) and monoclonal 9D6 (lanes 11 and 12) proved to be the most effective for immunoprecipitation. In the right lane of each pair, autophosphorylation was carried out after preincubation with EGF. Unlike the results with the free receptor, EGF did not enhance autophosphorylation when the receptor was complexed with antibodies on protein A-agarose beads. Furthermore, in the case of the rabbit antibody, the presence of EGF appeared to decrease the intensity of the phosphorylation activity (Fig. 3, lane 4). In most experiments using this methodology, Mr 170 and 150 × 103 forms were radiolabelled almost equally.

Fig. 3.

Mouse liver EGF receptor immunoprecipitated by polyclonal and monoclonal antibodies. (A) Lane 1, purified EGF receptor, γ-32P-ATP autophosphorylated. Lanes 2–12, whole adult mouse liver, detergent-solubilized, immunoprecipitated on protein A-agarose with antibodies to EGF receptor and labelled with γ-32P-ATP as described in Materials and methods. Immunoprecipitation with (2) rabbit preimmune serum, (3,4) rabbit anti-EGF receptor Ig, labelled in the absence (3) and presence (4) of 100 μg EGF, (5) monoclonal control, no antibody, (6) negative monoclonal, (7,8) monoclonal antibody 4D5 in the absence (7) and presence (8) of EGF, (9,10) monoclonal 5D4 in the absence (9) and presence (10) of EGF, (11,12) monoclonal 9D6 in the absence (11) and presence (12) of EGF. (B) Mouse liver membranes with boundJI-EGF cross linked with DSS, homogenized and immunoprecipitated as described in Materials and methods. Immunoprecipitated with (13) rabbit anti-EGF receptor, (14) negative monoclonal antibody, (15) monoclonal 4D5 and (16) monoclonal 9D6.

Fig. 3.

Mouse liver EGF receptor immunoprecipitated by polyclonal and monoclonal antibodies. (A) Lane 1, purified EGF receptor, γ-32P-ATP autophosphorylated. Lanes 2–12, whole adult mouse liver, detergent-solubilized, immunoprecipitated on protein A-agarose with antibodies to EGF receptor and labelled with γ-32P-ATP as described in Materials and methods. Immunoprecipitation with (2) rabbit preimmune serum, (3,4) rabbit anti-EGF receptor Ig, labelled in the absence (3) and presence (4) of 100 μg EGF, (5) monoclonal control, no antibody, (6) negative monoclonal, (7,8) monoclonal antibody 4D5 in the absence (7) and presence (8) of EGF, (9,10) monoclonal 5D4 in the absence (9) and presence (10) of EGF, (11,12) monoclonal 9D6 in the absence (11) and presence (12) of EGF. (B) Mouse liver membranes with boundJI-EGF cross linked with DSS, homogenized and immunoprecipitated as described in Materials and methods. Immunoprecipitated with (13) rabbit anti-EGF receptor, (14) negative monoclonal antibody, (15) monoclonal 4D5 and (16) monoclonal 9D6.

In section B of Fig. 3, liver membranes were immunoprecipitated after iodinated EGF was bound to receptors and cross linked with disuccinimidyl suberate (DSS). As shown in Fig. 3B, lane 13, no band is visible when the rabbit antibody is used to precipitate 125I-EGF-labelled receptor. This result, taken with the effect of EGF seen in lane 4, first led us to suspect that the rabbit antibody binds to the receptor at or close to the EGF-binding site. Monoclonals 4D5 (lane 15) and 9D6 (lane 16), however, precipitate the 125l-EGF-labelled receptor. It should also be noted that while both the Mr170 and 150 × 103 forms of the protein can be phosphorylated, only the 170 × 103 form appears to bind EGF.

Biological effects of antibodies

We assayed the ability of antibodies to stimulate [3H]thymidine uptake and to block EGF-induced [3H]thymidine uptake in quiescent 3T3 cells. Results are shown in Figs 4, 5. The rabbit antibody has proven especially interesting since it has a biphasic effect on cells. At a relatively low concentration (13μgml−1 of partially purified antibody protein), EGF binding to 3T3 cell monolayers is more than 50% blocked (Fig. 4, inset) and this is accompanied by a more than 50% reduction in EGF-induced pHJthymidine uptake by these cells (Fig. 4). At 26 μ g antibody protein per ml, the EGF effect is even further blocked, but the antibody itself slightly stimulates DNA synthesis. At 130μgml−1 protein, the effect of the antibody is quite evident and the stimulation of thymidine uptake in the presence of EGF and rabbit antibody is additive. At 520 to 1300μg ml−1 antibody protein, the stimulatory effect is maximal. The same degree of stimulation occurs with or without EGF since EGF binding is completely blocked at these higher concentrations. The effect seems to be specific, as it cannot be mimicked by the same concentration of a similarly prepared rabbit antibody to mouse alphafetoprotein, although anti-AFP at very high concentrations has a slight stimulatory effect. EGF binding is also blocked by rabbit antibody in PSA5E differentiated teratocarcinoma cells and in mouse liver membranes with very similar kinetics (results not shown).

Fig. 4.

Incorporation of [3H]thymidine into DNA by quiescent 3T3 cells after treatment with rabbit anti-EGF receptor in the presence and absence of EGF. Antibody ± EGF was added to cells in serum-free medium as described in Materials and methods. Clear bars, incorporation of [3H]thymidine into cells cultured in the presence of the given concentration of antibody protein; stippled bars, the same but with 10 μ g ml−1 EGF added. Each bar represents quadruplicate measurements and this experiment was repeated with similar results. SFM, serum-free medium; FCS, 10% fetal calf serum in DME. Inset: blocking of the binding of iodinated EGF to quiescent 3T3 cells in the presence of rabbit anti-EGF receptor. Each point is the average of two or three experiments.

Fig. 4.

Incorporation of [3H]thymidine into DNA by quiescent 3T3 cells after treatment with rabbit anti-EGF receptor in the presence and absence of EGF. Antibody ± EGF was added to cells in serum-free medium as described in Materials and methods. Clear bars, incorporation of [3H]thymidine into cells cultured in the presence of the given concentration of antibody protein; stippled bars, the same but with 10 μ g ml−1 EGF added. Each bar represents quadruplicate measurements and this experiment was repeated with similar results. SFM, serum-free medium; FCS, 10% fetal calf serum in DME. Inset: blocking of the binding of iodinated EGF to quiescent 3T3 cells in the presence of rabbit anti-EGF receptor. Each point is the average of two or three experiments.

Fig. 5.

The effect on [3H]thymidine incorporation into DNA in quiescent 3T3 cells of treatment with monoclonal anti-EGF receptor 9D6 in the presence and absence of EGF. Each bar represents quadruplicate measurements and this experiment was repeated with similar results. SFM, serum-free medium; FCS, 10 % fetal calf serum in DME. Antibody with EGF (stippled bars) or without EGF (clear bars) was added to serum-free medium.

Fig. 5.

The effect on [3H]thymidine incorporation into DNA in quiescent 3T3 cells of treatment with monoclonal anti-EGF receptor 9D6 in the presence and absence of EGF. Each bar represents quadruplicate measurements and this experiment was repeated with similar results. SFM, serum-free medium; FCS, 10 % fetal calf serum in DME. Antibody with EGF (stippled bars) or without EGF (clear bars) was added to serum-free medium.

The effects of monoclonal antibody 9D6 on tritiated thymidine uptake in quiescent 3T3 cells is shown in Fig. 5. Unlike the rabbit antibody, 9D6 does not appear to block EGF-induced [3H]thymidine uptake. This is consistent with the immunoprecipitation data which indicate that 9D6 does not bind to receptor at the EGF-binding site. 9D6 does not appear to block 125I-EGF binding in binding competition experiments (not shown). Interestingly, 9D6 does stimulate [3H]thymidine uptake at the highest antibody concentrations tested (Fig. 5) compared to control serum-free medium not containing antibodies.

Immunohistochemical reactions

Rabbit anti-EGF receptor stains the cell margins of mouse and human cells known to have receptors.

Dramatic indirect immunofluorescence can be observed on living A431 cells because of the very large number of receptors. Fig. 6A shows antibody binding to A431 ceils at 4° C. If the cells are warmed to 20°C, receptor clustering is observed (Fig. 6B). For demonstrating EGF receptors on mouse cells, the more sensitive avidin-biotin-peroxidase method on formalin-fixed tissues gives better results. Fig. 6C shows that differentiated OC15 END cell margins are clearly stained, as are intracellular vesicles. OC15 EC cells stain weakly in the cytoplasm which is sparse in these cells (data not shown). Rat monoclonal antibodies give poor immunohistochemical reaction with either mouse or human cells.

Fig. 6.

Indirect immunofluorescence and immunoperoxidase staining of rabbit anti-EGF receptor on live A431 cells and fixed OC15 END cells. Cells were labelled with the primary antibody at 4 °C (A) or 25°C (B) for 15min, washed with PBS, and labelled with rhodamine-linked goat anti-rabbit for 15 min at 4°C. (C) OC15 END cells (5 days after induction with retinoic acid) stained by the avidin-biotin-peroxidase procedure with rabbit antibodies to EGF receptor at 15μgml*. (D) END cells similarly stained using a preimmune Ig preparation at the same concentration. Bar, 50 μ M.

Fig. 6.

Indirect immunofluorescence and immunoperoxidase staining of rabbit anti-EGF receptor on live A431 cells and fixed OC15 END cells. Cells were labelled with the primary antibody at 4 °C (A) or 25°C (B) for 15min, washed with PBS, and labelled with rhodamine-linked goat anti-rabbit for 15 min at 4°C. (C) OC15 END cells (5 days after induction with retinoic acid) stained by the avidin-biotin-peroxidase procedure with rabbit antibodies to EGF receptor at 15μgml*. (D) END cells similarly stained using a preimmune Ig preparation at the same concentration. Bar, 50 μ M.

Analysis of EGF receptors on differentiating embryonal carcinoma cells

(1) By 125l-EGF cross linking

Since rabbit antibodies do not react well with ligand-occupied receptors, we used immunoprecipitation with rat monoclonal 9D6 to detect the expression of EGF receptor during OC15 EC cell differentiation. Iodinated EGF was allowed to bind to monolayers of either EC cells or END cells (differentiated by plating at low density and growth in RA for 5 days), and then covalently cross linked to its receptor by DSS. The cells were lysed and equal amounts of cell lysate immunoprecipitated as described and run on a 7·5 % agarose gel. The results are shown in Fig. 7A. A single radiolabelled band is seen at Mr50×103 for END cells labelled with 125I-EGF (Fig. 7A, lane 4), but not if the cells are labelled in the presence of 100-fold excess cold EGF (lane 3). No band is visible in EC lysate from an equivalent number of cells (lanes 1 and 2).

Fig. 7.

Detection of EGF receptors expressed on the surface of OC15 EC and OC15 END cells by immunoprecipitation. (A) Iodinated EGF was covalently cross linked to receptor and immunoprecipitated with monoclonal antibody 9D6. l25I-EGF bound to 2×106 OC15 EC cells in the presence (1) and absence (2) of 500-fold excess unlabelled EGF; 12SI-EGF bound to 2×106 OC15 END cells (after 5 days of differentiation) in the presence (3) and absence (4) of 500-fold excess unlabelled EGF. (B) Surface iodination of cells in vivo was followed by immunoprecipitation with rabbit antibody to EGF receptor. Lane 5, 106 OC15 EC cells; lane 6, 106 OC15 END after 3 days of differentiation; lane 7; 106 OC15 END cells after 5 days of differentiation; lane 8, 5×104 A431 cells.

Fig. 7.

Detection of EGF receptors expressed on the surface of OC15 EC and OC15 END cells by immunoprecipitation. (A) Iodinated EGF was covalently cross linked to receptor and immunoprecipitated with monoclonal antibody 9D6. l25I-EGF bound to 2×106 OC15 EC cells in the presence (1) and absence (2) of 500-fold excess unlabelled EGF; 12SI-EGF bound to 2×106 OC15 END cells (after 5 days of differentiation) in the presence (3) and absence (4) of 500-fold excess unlabelled EGF. (B) Surface iodination of cells in vivo was followed by immunoprecipitation with rabbit antibody to EGF receptor. Lane 5, 106 OC15 EC cells; lane 6, 106 OC15 END after 3 days of differentiation; lane 7; 106 OC15 END cells after 5 days of differentiation; lane 8, 5×104 A431 cells.

Fig. 8.

Immunoprecipitation analyses of autophosphorylated and metabolically labelled 0C15 EC and END EGF receptor. (A) Lysates from 106 cells (105 cells for A431 cells; lanes 1 and 2) were allowed to react with γ-32P-ATP with or without pretreatment with EGF before immunoprecipitation overnight with rabbit antibody as described in Materials and methods. Lanes 3 and 4, mouse 3T3 fibroblasts; lanes 5 and 6, OC15 EC cells; lanes 7 and 8, 3-day OC15 END cells; lanes 9 and 10, 5-day OC15 END; lanes 11 and 12, 7-day OC15 END; lanes 13 and 14, 7-day OC15 END cells seeded at one-fifth the density. M, M, markers. Gel contained 7·5% polyacrylamide. Autoradiograph exposed for 3 days. (B) Lanes 1 and 2, 10* OC15 EC cells treated with trypsin before solubilization as described in Materials and methods. Lanes 3 and 4, without trypsin treatment. EGF was present in samples in even-numbered lanes. It was necessary to immunoprecipitate by overnight exposure when using rabbit antibodies reacting with EGF receptors which had been pretreated with EGF. Gel contained 5% polyacrylamide. Autoradiograph exposed for 5 days.(C) Immunoprecipitation of EGF receptor from metabolically labelled cells. Cells were labelled for lh with [35S]methionine in met-free medium. Lysates containing 10’ctsmin−1 of 35S-labelled protein were immunoprecipitated with rabbit anti-EGF receptor and analysed on a 5% polyacrylamide gel. Lanes 1·3, OC15 EC cells; lanes 4·6, OC15 END cells differentiated 3 days; lanes 7–9, OC15 END cells differentiated 5 days. Lanes 1, 4 and 7 were immunoprecipitated with preimmune rabbit Ig; lysates in lanes 2, 5 and 8 were immunoprecipitated for 90 min with rabbit anti-EGF receptor antibodies; lysates in lanes 3, 6 and 9 were left overnight in contact with antibody.

Fig. 8.

Immunoprecipitation analyses of autophosphorylated and metabolically labelled 0C15 EC and END EGF receptor. (A) Lysates from 106 cells (105 cells for A431 cells; lanes 1 and 2) were allowed to react with γ-32P-ATP with or without pretreatment with EGF before immunoprecipitation overnight with rabbit antibody as described in Materials and methods. Lanes 3 and 4, mouse 3T3 fibroblasts; lanes 5 and 6, OC15 EC cells; lanes 7 and 8, 3-day OC15 END cells; lanes 9 and 10, 5-day OC15 END; lanes 11 and 12, 7-day OC15 END; lanes 13 and 14, 7-day OC15 END cells seeded at one-fifth the density. M, M, markers. Gel contained 7·5% polyacrylamide. Autoradiograph exposed for 3 days. (B) Lanes 1 and 2, 10* OC15 EC cells treated with trypsin before solubilization as described in Materials and methods. Lanes 3 and 4, without trypsin treatment. EGF was present in samples in even-numbered lanes. It was necessary to immunoprecipitate by overnight exposure when using rabbit antibodies reacting with EGF receptors which had been pretreated with EGF. Gel contained 5% polyacrylamide. Autoradiograph exposed for 5 days.(C) Immunoprecipitation of EGF receptor from metabolically labelled cells. Cells were labelled for lh with [35S]methionine in met-free medium. Lysates containing 10’ctsmin−1 of 35S-labelled protein were immunoprecipitated with rabbit anti-EGF receptor and analysed on a 5% polyacrylamide gel. Lanes 1·3, OC15 EC cells; lanes 4·6, OC15 END cells differentiated 3 days; lanes 7–9, OC15 END cells differentiated 5 days. Lanes 1, 4 and 7 were immunoprecipitated with preimmune rabbit Ig; lysates in lanes 2, 5 and 8 were immunoprecipitated for 90 min with rabbit anti-EGF receptor antibodies; lysates in lanes 3, 6 and 9 were left overnight in contact with antibody.

(2) By surface radioiodination

Rabbit antibodies reacted with cell surface iodinated EGF receptors provided confirmation that OC15 EC cells (Fig. 7B, lane 5) have little or no detectable surface receptors, while OC15 END gradually acquire receptors over 5 days (Fig. 7B, lanes 6 and 7). Lane 8 shows EGF receptor immunoprecipitated from 5 × 104 A431 cells (as opposed to 106 OC15 cells in other lanes). The A431 band is five times as heavy as the 5-day END band by densitometric scanning, indicating that differentiated OC15 END cells contain about 1/100 of the number of receptors on A431 cells. This gives approximately 30000 EGF receptors per END cell, in good agreement with our previous binding assay data (Rees et al. 1979).

(3) By autophosphorylation of EGF receptor in detergent lysates

This method ensures that both intracellular and surface EGF receptor protein kinase are detected. To accurately measure EGF receptor kinase autophosphorylation in cells such as OC15 END, which have very low receptor numbers, it is necessary to optimize conditions. Brij 35 detergent at 0 · 6 % is sufficient to solubilize most of the cellular proteins without affecting kinase activity. The addition of phosphatase inhibitors allowed the reaction to be carried out in the whole cell lysates prior to immunoprecipitation, pre-serving EGF stimulation effects. Fig. 8 shows that the kinase is highly responsive to the stimulatory activity of EGF (even numbered lanes). As expected, A431 human cells display an amount of phosphorylating activity that overwhelms the conditions needed to detect protein kinase activity from tenfold more mouse cells (Fig. 8A, lanes 1, 2). As OC15 END cells differentiate over 3, 5 and 7 days, a steady increase in receptor autophosphorylation activity (lanes 8 – 14) parallels the rise in surface-iodinated receptor seen in Fig. 7. Fig. 8, lane 14 shows that OC15 cells seeded most sparsely, achieve significantly greater EGF-autophosphorylating activity 7 days of differentiation when compared to cells seeded at fivefold higher density (lane 12). Presumably, this is explained by a greater degree of differentiation and fewer remaining EC cells in these cultures. Surprisingly, there is a readily detectable phosphorylated Mr170 × 103 species in OC15 EC cells (especially when stimulated with EGF, lane 6) at a level not very much different from that seen in the mouse 3T3 fibroblast line (lanes 3, 4) and 3d OC15 END cells (lanes 7, 8). Identical results were obtained using monoclonal antibody, 9D6, to immunoprecipitate the labelled receptors (data not shown).

(4) By tryptic digestion of surface receptors

In order to test whether the EGF receptors in EC cells are exposed at the surface, the experiment was repeated to include cells that were treated with 0 · 25 % trypsin for 15 min at room temperature before neutralizing the trypsin with calf serum and trypsin inhibitor. Fig. 8B shows that trypsin had very little effect on the autophosphorylating activity of receptors stimulated with EGF after solubilization. Trypsin had a slight, but reproducible, stimulatory effect on the level of phosphorylation of receptors not stimulated with EGF (Fig. 8B, lane 1, a faint Mr170 × 103 band is visible in the original autoradiograph). Densitometry and averaging the relative intensities of 32P-labelled receptors in lanes 2 and 4 showed there was no significant difference between them. We concluded, therefore, that the EGF receptors in teratocarcinoma stem cells are not exposed at the cell surface, a result that confirmed EGF-binding studies and cell surface iodination experiments. Trypsin incubation of OC15 END cells almost completely destroyed the EGF-stimulated phosphorylation of the Mr 170×103protein however (data not shown).

(5) By metabolic labelling

To establish that EGF receptors are being synthesized in OC15 EC as well as END cells, immunoprecipitated, [35S]methionine-labelled proteins were analysed (Fig. 8C). By this method, EC cells (lane 3) appeared to synthesize almost as much EGF receptor as 3-day END cells (lanes 5, 6) but proportionally more was in the lowermost of three bands (approx. Mr 160×103), while END cells contained more of the upper band of a close doublet (Mr 170×103) Control lysates treated with a nonreactive rabbit Ig preparation did not precipitate any band in this region of the gel (lanes 1, 4 and 7). Radioactive protein at the top of the gel is fibronectin which is adsorbed by protein A, together with other coprecipitated matrix proteins. OC15 EC cells synthesize more fibronectin and type IV collagen than END cells as can be seen by bands at greater than Mr200× 103. The lower Mr bands in lane 3 may be breakdown products of EGF receptor which would indicate that intracellular receptors are more rapidly turned over. Identical results were obtained with 9D6 monoclonal antibody but bands were weaker (data not shown).

Epidermal growth factor receptor purified from mouse liver appears as a Coomassie-blue-stained doublet of Mt 170 and 150 ×103 on acrylamide gels. The two forms are present in roughly equal amounts and increasing the amount of protease inhibitors and calcium chelators does not eliminate the Mr-170×103 form, as this treatment does for the A431 cell receptor (Yeaton, Lipari & Fox, 1983). Proteolytic cleavage of the intracellular C-terminus of the A431 receptor removes all three of the tyrosines (residues 1068, 1148 and 1174) which are the autophosphorylation sites (Gullick et al. 1985). In our hands, purified but unimmunoprecipitated mouse fiver EGF receptor appears to autophosphorylate the Mr170×103 band almost exclusively. However, autophosphorylation is equally strong on both bands if the reaction is carried out on protein A-agarose beads after immunoprecipitation (Fig. 3). This observation could rule out the assumption that all the autophosphorylation sites have been removed from the mouse Mr,450 ×103 receptor form. Alternatively, it is possible that binding and immobilization of receptor-antibody complex on agarose beads permits phosphorylation of another, less-frequent normal site or even an aberrant site. A similar finding in A431 cells was reported by Gates & King (1985). Relevant to these speculations, we have also found that 125I-EGF covalently cross linked to the receptor produces a labelled protein only at Mr170–180×103 This could be explained if EGF protected the receptor from further degradation even in the intracellular domain.

Inoculation of a rabbit with purified mouse EGF receptor has produced an antibody that reacts very specifically and avidly with EGF receptor even with-out antibody affinity purification. The rabbit antibody immunoprecipitates well and reacts strongly in immunohistochemical reactions with both mouse and human EGF receptor (Fig. 6). The polyclonal antibody binds to EGF receptor at or close to the EGF-binding site since it can completely block EGF binding. At higher concentrations, binding of the rabbit antibody can mimic EGF by stimulating DNA synthesis. The latter property could be nonspecific, since the antibody is not a completely pure preparation. However, it is not likely to be due to the presence of rabbit serum growth factors since these were removed during ammonium sulphate precipitation and dialysis. As a control, we did a parallel series of experiments with an anti-AFP antibody preparation and there was no stimulation of thymidine uptake into DNA except weakly (5 %) at the highest concentration used. The rabbit antibody can induce capping of the EGF receptor on A431 cells and it is widely believed (Gullick et al. 1985; Schreiber et al. 1983; Downward et al. 1985) that this is an essential step for the propagation of the mitogenic signal. Therefore, it seems most likely that the ability of rabbit antibody to induce [3H]thymidine uptake in 3T3 cells is related to its ability to induce clustering of the receptor. The fact that the blocking effect occurs at a 20-fold lower antibody concentration than the growth stimulation effect may allow us to use this antibody to produce either effect differentially in systems we wish to study.

We have produced several anti-EGF receptor monoclonals from the fusion of one spleen. Antibodies designated 4D5, 3B4, 5D4 and 9D6 immunoprecipitate EGF receptor from mouse fiver. Monoclonal 9D6 immunoprecipitates mouse EGF receptor protein but only weakly reacts with the human receptor (about 1 % compared to the polyclonal antibody). We are currently employing 9D6 to detect EGF receptor expression in mouse fetal tissues. Monoclonal 9D6 stimulated [3H]thymidine uptake in quiescent 3T3 cells at the highest concentrations tested, though it does not appear to bind at the EGF-binding site. It cannot be ruled out that this is a nonspecific effect due to some other nondialysable factor produced by the hybridoma cells. However, it is also possible that the effect is produced by antibody binding to another site on the EGF receptor, possibly by promoting cross linking of receptors or by inducing a conformational change. Das, Knowles, Biswas & Bishayee (1984) have also described a mouse monoclonal antibody that enhances kinase activity, although it does not bind at the EGF-binding site.

EGF receptor expression in differentiating embryonal carcinoma cells

Our earlier observations showed that EGF-binding activity is undetectably low in embryonal carcinoma cells (Rees et al. 1979) except for F9 cells (Adamson & Hogan, 1984). OC15 EC cells differentiate into endoderm-like cells that bind 125I-EGF with normal kinetics at 30000 sites per cell. These new cell surface EGF receptors respond to EGF by increased cellular proliferation (Rees et al. 1979). Using antibodies prepared to mouse EGF receptors, we were able to confirm that OC15 END cells express a Af,470x1o3 EGF receptor on cell membranes and that this receptor autophosphorylates in response to EGF (Figs 7A,B, 8A).

OC15 EC cells do not display a Mr170×103 cell surface protein that can bind 125I-EGF (Fig. 7A) or that can be immunoprecipitated by anti-EGF receptor antibodies from cell surface iodinated EC cells (Fig. 7B). However, several results combined strongly suggest that they do contain EGF receptors: (1) immunoperoxidase reactions show weak though reproducible intracellular staining in OC15 cells (not shown); (2) [35S]methionine-labelled immunoprecipitates contain three bands specifically immunoprecipitated by anti-receptor antibodies, which appear to correspond to Mr160 and 170×103 forms. In END cells, a doublet at Mr170×103 predominates (Fig. 8C); (3) Trypsin-treated or untreated OC15 EC cells solubilized in detergent produce a 32P-phosphorylated product at Mr170×103 in response to EGF that is specifically immunoprecipitated with anti-receptor antibodies (Fig. 8A,B). The Mr 170 ×103 band in EC cells is identical in migration to the cell surface species in END cells (Fig. 8A) and therefore appears to be a mature glycosylated form, since EGF receptors do not bind EGF unless they are glycosylated (Slieker & Lane, 1985; Soderquist & Carpenter, 1986). However, the OC15 EC cell receptor may have lower affinity for EGF, since increasing the EGF concentration in the solubilized extracts enhances the autophosphorylating activity (maximized in Fig. 8B, but not in Fig. 8A). Adult rat liver also contains an intracellular pool of low affinity EGF receptors with kinase activity in addition to high-affinity cell surface receptors (Dunn, Connolly & Hubbard, 1985). We do not yet know the status and origin of the M,60 ×103 EGF receptor band present in EC cells. It could be a precursor of the final Mr 170×103 form since both forms are metabolically labelled equally in 1h. The Mr160×103 form could also be a product of normal degradation after down regulation or degradation during processing. Pulsechase studies will be needed to determine this and these together with peptide mapping of the lower band are currently underway.

The presence of a Mr170 ×103 apparently mature form of EGF receptor that is entirely intracellular is an exciting finding that tempts speculation about possible developmental roles. Is it a latent or cryptic form awaiting the signal to be expressed as stem cells differentiate into various kinds of epithelial and mesodermal cells? Does a similar phenomenon occur in early embryonic ectoderm cells? Are the receptors an intracellular form produced after down regulation from binding autocrine EGF or TGFα? Alternatively, does the receptor/kinase perform a developmentally specific intracellular function that is totally independent of ligand binding?

This work was supported by Public Health Service grant CA 28427 and P30 CA 30199 from the National Cancer Institute.

Adamson
,
E. D.
,
Deller
,
M. J.
&
Warshaw
,
J.
(
1981
).
Functional EGF receptors on embryonic cells
.
Nature, Lond
.
291
,
656
659
.
Adamson
,
E. D.
&
Hogan
,
B. L. M.
(
1984
).
Expression of EGF receptor and transferrin by F9 and PC13 teratocarcinoma cells
.
Differentiation
27
,
152
157
.
Adamson
,
E. D.
&
Meek
,
J.
(
1984
).
The ontogeny of epidermal growth factor receptors during mouse development
.
Devi Biol
.
103
,
62
71
.
Adamson
,
E. D.
,
Meek
,
J.
&
Edwards
,
S. A.
(
1985
).
Product of the cellular oncogene, c-fos, observed in mouse and human tissues using an antibody to a synthetic peptide
.
EMBO J
.
4
,
941
947
.
Adamson
,
E. D.
&
Rees
,
A. R.
(
1981
).
Epidermal growth factor receptors
.
Mol. cell. Biochem
.
34
,
129
152
.
Adamson
,
E. D.
&
Warshaw
,
J.
(
1981
).
“Down-regulation” of epidermal growth factor receptors in mouse embryos
.
Devi Biol
.
90
,
430
434
.
Bertics
,
P. J.
&
Gill
,
G. N.
(
1985
).
Self-phosphorylation enhances the protein-tyrosine kinase activity of the epidermal growth factor receptor
.
J. biol. Chem
.
260
,
14 641
14 647
.
Bradshaw
,
T. K.
(
1986
).
Cell transformation: the role of oncogenes and growth factors
.
Mutagenesis
1
,
91
97
.
Brugge
,
J. S.
(
1986
).
The p35/p36 substrates of protein-tyrosine kinases as inhibitors of phospholipase A2
.
Cell
46
,
149
150
.
Cassel
,
D.
&
Glaser
,
L.
(
1982
).
Proteolytic cleavage of epidermal growth factor receptor
.
J. biol. Chem
.
257
,
9845
9848
.
Cohen
,
S.
,
Carpenter
,
G.
&
King
,
L.
, Jr
(
1980
).
Epidermal growth factor-receptor-protein kinase interactions
.
J. biol. Chem
.
255
,
4834
4842
.
Cohen
,
S.
,
Fava
,
R. A.
&
Sawyer
,
S. T.
(
1982
).
Purification and characterization of epidermal growth factor receptor/protein kinase from normal mouse liver
.
Proc. natn. Acad. Sci. U.S.A
.
79
,
6237
6241
.
Das
,
M.
,
Knowles
,
B. B.
,
Biswas
,
R.
&
Bishayee
,
S.
(
1984
)
Receptor modulating properties of an antibody directed against the epidermal growth factor receptor
.
Eur. J. Biochem
.
141
,
429
434
.
Downward
,
J.
,
Parker
,
P.
&
Waterfield
,
M. D.
(
1984
).
Autophosphorylation sites on the epidermal growth factor receptor
.
Nature, Lond
.
311
,
483
485
.
Downward
,
J.
,
Waterfield
,
M. D.
&
Parker
,
P. J.
(
1985
)
Autophosphorylation and protein kinase C phosphorylation of the epidermal growth factor receptor
.
J. biol. Chem
.
260
,
14538
14546
.
Downward
,
J.
,
Yarden
,
Y.
,
Mayes
,
E.
,
Scrace
,
G.
,
Totty
,
N.
,
Stockwell
,
P.
,
Ullrich
,
A.
,
Schlessinger
,
J.
&
Waterfield
,
M. D.
(
1984
).
Close similarity of epidermal growth factor receptor and v-erbB oncogene protein sequences
.
Nature, Lond
.
307
,
521
527
.
Dunn
,
W. A.
,
Connolly
,
T. P.
&
Hubbard
,
A. L.
(
1986
)
Receptor-mediated endocytosis of epidermal growth factor by rat hepatocytes: receptor pathway
.
J. Cell Biol
.
102
,
24
36
.
Engvall
,
E.
(
1980
).
Enzyme immunoassay ELISA and EMIT
.
Meth. Enzymol
.
70
,
419
439
.
Gates
,
R. E.
&
King
,
L. E.
, Jr
(
1982
).
Calcium facilitates endogenous proteolysis of the EGF receptor-kinase
.
Mol. cell. Endocrinol
.
27
,
263
276
.
Gates
,
R. E.
&
King
,
L. E.
(
1985
).
Different forms of the epidermal growth factor receptor kinase have different autophosphorylation sites
.
Biochemistry
24
,
5209
5215
.
Gregoriou
,
M.
&
Rees
,
A. R.
(
1984
).
Properties of a monoclonal antibody to epidermal growth factor with implications for the mechanism of action
.
EMBO J
.
3
,
929
937
.
Gullick
,
W. J.
,
Downward
,
J.
&
Waterfield
,
M. D.
(
1985
).
Antibodies to the autophosphorylation sites of the epidermal growth factor receptor protein-tyrosine kinase as probes of structure and function
.
EMBO J
.
4
,
2869
2877
.
Heldin
,
C.
&
Westermark
,
B.
(
1984
).
Growth factors: mechanism of action and relation to oncogenes
.
Cell
37
,
9
20
.
Herschman
,
H. R.
(
1985
).
The EGF receptor
, pp.
169
200
.
In Control of Animal Cell Proliferation, vol. I
(ed.
A. L.
Boynton
&
H. L.
Leffert
).
New York
:
Academic Press
.
Huang
,
K.
,
Wallner
,
B. P.
,
Mattaliano
,
R. J.
,
Tizard
,
R.
,
Burne
,
C.
,
Frey
,
A.
,
Hession
,
C.
,
Mcgray
,
P.
,
Sinclair
,
L. K.
,
Chow
,
E. P.
,
Browning
,
J. L.
,
Ramachandran
,
K. L.
,
Tang
,
J.
,
Smart
,
J. E.
&
Pepinsky
,
R. B.
(
1986
).
Two human 35 Kd inhibitors of phospholipase A2 are related to substrates of pp60v src and of the epidermal growth factor receptor/kinase
.
Cell
46
,
191
199
.
Hunter
,
T.
&
Cooper
,
J. A.
(
1981
).
Epidermal growth factor induces rapid tyrosine phosphorylation of proteins in A431 human tumor cells
.
Cell
24
,
741
752
.
Klein
,
H. H.
,
Freidenberg
,
G. R.
,
Cordera
,
R.
&
Olefsky
,
J. M.
(
1985
).
Substrate specificities of insulin and epidermal growth factor receptor kinases
.
Biophys. Biochem. Res. Comm
.
127
,
254
263
.
Laemmli
,
U. K.
(
1970
).
Cleavage of structural proteins during the assembly of the head of bacteriophage T4
.
Nature, Lond
.
277
,
680
685
.
Lebien
,
T. W.
,
Boue
,
D. R.
,
Bradley
,
J. G.
&
Kersey
,
J. H.
(
1982
).
Antibody affinity may influence antigenic modulation of the common acute lymphoblastic leukemia antigen in vitro
.
J. Immun
.
129
,
2287
2292
.
Lin
,
C. R.
,
Chen
,
W. S.
,
Kruiger
,
W.
,
Stolarsky
,
L. S.
,
Weber
,
W.
,
Evans
,
R. M.
,
Verma
,
I. M.
,
Gill
,
G. N.
&
Rosenfeld
,
M. G.
(
1984
).
Expression cloning of human EGF receptor complementary DNA: gene amplification and three related messenger RNA products in A431 cells
.
Science
224
,
843
848
.
Linsley
,
P. S.
&
Fox
,
C. F.
(
1980
).
Controlled proteolysis of EGF receptors: evidence for transmembrane distribution of the EGF binding and phosphate acceptor sites
.
J. Supramol. Struct
.
14
,
461
471
.
Massagué
,
J.
&
Like
,
B.
(
1985
).
Cellular receptors for type p transforming growth factor: Ligand binding and affinity labeling in human and rodent cell lines
.
J. biol. Chem
.
260
,
2636
2645
.
Oshima
,
R. G.
(
1981
).
Developmental expression of murine extraembryonic endodermal cytoskeletal proteins
.
J. biol. Chem
.
257
,
3414
3421
.
Rees
,
R. A.
,
Adamson
,
E. D.
&
Graham
,
C. F.
(
1979
).
Epidermal growth factor receptors increase during the differentiation of embryonal carcinoma cells
.
Nature, Lond
.
281
,
309
311
.
De St Groth
,
S. F.
&
Schneidegger
,
D.
(
1980
).
Production of monoclonal antibodies: Strategy and tactics
.
J. Immunological Methods
35
,
1
21
.
Savage
,
C. R.
, Jr
&
Cohen
,
S.
(
1972
).
Epidermal growth factor and a new derivative
.
J. biol. Chem
.
247
,
7609
7611
.
Schreiber
,
A. B.
,
Libermann
,
T. A.
,
Lax
,
I.
,
Yarden
,
Y.
&
Schlessinger
,
J.
(
1983
).
Biological role of epidermal growth factor receptor clustering: Investigation with monoclonal anti-receptor antibodies
.
J. biol. Chem
.
258
,
846
853
.
Slieker
,
L. J.
&
Lane
,
M. D.
(
1985
).
Post-translational processing of the epidermal growth factor receptor
.
J. biol. Chem
.
260
,
687
690
.
Smart
,
J. E.
,
Opperman
,
H.
,
Czernilofsky
,
A. P.
,
Purchio
,
A. F.
,
Erickson
,
R. L.
&
Bishop
,
J. M.
(
1984
).
Characterization of sites for tyrosine phosphorylation in the transforming protein of Rous sarcoma virus and its normal cellular homologue (pp60vjnr)
.
Proc. natn. Acad. Sci. U.S.A
.
78
,
6013
6017
.
Soderquist
,
A. M.
&
Carpenter
,
G.
(
1986
).
Biosynthesis and metabolic degradation of receptors for epidermal growth factor
.
J. memb. Biol
.
90
,
97
105
.
Thom
,
D.
,
Powell
,
A. J.
,
Lloyd
,
C. W.
&
Rees
,
D. A.
(
1977
).
Rapid isolation of plasma membranes in high yield from cultured fibroblasts
.
Biochem. J
.
168
,
187
194
.
Towbin
,
H.
,
Staehelin
,
T.
&
Gordon
,
J.
(
1979
).
Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: Procedure and some applications
.
Proc. natn. Acad. Sci. U.S.A
.
76
,
4350
4354
.
Ullrich
,
A.
,
Coussens
,
L.
,
Hayflick
,
J. S.
,
Dull
,
T. J.
,
Gray
,
A.
,
Tam
,
A. W.
,
Lee
,
J.
,
Yarden
,
Y.
,
Libermann
,
T. A.
,
Schlessinger
,
J.
,
Downward
,
J.
,
Mayes
,
E. L. V.
,
Whittle
,
N.
,
Waterfield
,
M. D.
&
Seeburg
,
P. H.
(
1984
).
Human epidermal growth factor receptor cDNA sequence and aberrant expression of the amplified gene in A431 epidermoid carcinoma cells
.
Nature, Lond
.
309
,
418
425
.
Xu
,
Y.
,
Ishii
,
S.
,
Clark
,
A. J. L.
,
Sullivan
,
M.
,
Wilson
,
R. K.
,
Ma
,
D. P.
,
Roe
,
B. A.
,
Merlino
,
G. T.
&
Pastan
,
I.
(
1984
).
Human epidermal growth factor receptor cDNA is homologous to a variety of RNAs overproduced in A431 carcinoma cells
.
Nature, Lond
.
309
,
806
810
.
Yamamoto
,
T.
,
Nishida
,
T.
,
Miyajima
,
N.
,
Kawai
,
S.
,
Ooi
,
T.
&
Toyoshima
,
K.
(
1983
).
The erbB gene of avian erythroblastosis virus is a member of the src family
.
Cell
35
,
71
78
.
Yeaton
,
R. W.
,
Lipari
,
M. T.
&
Fox
,
C. F.
(
1983
).
Calcium-mediated degradation of epidermal growth factor receptor in dislodged A431 cells and membrane preparations
.
J. biol. Chem
.
258
,
9254
9261
.