- Email: [email protected]
A radioimmunoassay for human epidermal growth factor receptor
ANALYTICAL
BIOCHEMISTRY
141,253-261
A Radioimmunoassay WILLIAM J. GULLICK,’
(1984)
for Human Epidermal
Growth
D. JULIAN H. DOWNWARD, AND MICHAEL D. WATERFIELD
JUDITH
Factor Receptor J. MARSDEN,
Department of’ Protein Chemistry, Imperial Cancer Research Fund Laboratories, LincolnS Inn Fields, London WC2A 3PX, England Received January 24, 1984 The development of a radioimmunoassay (RIA) for the human epidermal growth factor receptor solubilized with nonionic detergents which employs iodinated epidermal growth factor (“‘IEGF) as the specific ligand is described. A monoclonal antibody (Rl) that binds specifically to human EGF receptors [Waterheld, M. D., et al. (1982) J. Cell B&hem. 20, 149-1611 was used to separate solubilized receptors saturated with i’sI-EGF from free ligand by absorption to protein A-Sepharose, and the bound radioactivity was determined. The RIA was linear when increasing amounts of solubilized membrane protein were added and, when compared to the standard polyethylene glycol assay,was mom reproducible. In addition, the background nonspecific binding obtained in the presence of a hundred-fold excessof unlabeled EGF was lessin the RIA. Substitution of normal mouse serum for the monoclonal antibody gave very low nonspecific background ligand binding and avoided the use.of large amounts of unlabeled EGF in the assay. Two major classes of binding sites for EGF were observed in membrane preparations from the cervical carcinoma cell line A431 or from normal human placental tissue. These were present in ap proximately equal amounts, with apparent dissociation constants of 4 X lo-” and 4 X 1O-9 M. Upon solubilization with the nonionic detergent Triton X-100, only one class of EGF binding sites was detected in both cases, with a dissociation constant of 3 X IO-* M. The RIA can be used to monitor receptor purification and for quantitation of receptor number and affinity in various cell types. KEY WORDS: epidermal growth factor receptor; epidermal growth factor, radioimmunoassay: monoclonal antibodies; membrane proteins; glycoproteins.
The role of growth factors in the regulation of the growth of normal and transformed or tumor-derived cells has been the subject of intense study [for recent review see Ref. (l)]. Although a number of growth factors have been characterized, details of the structure and function of their receptors are much less complete. The receptors which have been best characterized are those for insulin (2-4), insulin-like growth factor I (4,5), insulin-like growth factor II (6,7), platelet-derived growth factor (8-lo), and epidermal growth factor [for reviews see (11,12)]. The receptors are all integral membrane proteins composed of one
(IGFIIR, PDGFR, EGFR)* or several (IR, IGFIR) polypeptides. EGF receptor (13) and insulin receptor ( 14) copurify with a protein kinase, as probably do the PDGF receptor (9) * Abbreviations: PDGF, Plateletderived growth factor; IGFIIR, Insulin-like growth factor II receptor; PDGFR, Platelet-derived growth factor receptor, EGFR, Epidermal growth factor receptor; IR, Insulin receptor; IGFIR, Insulin-like growth factor I receptor, EGF, Epidermal growth factor; mAb, Monoclonal antibody; PEG, polyethylene glycol; RIA, Radioimmunoassay; EGTA, ethylene glycol bis (&aminoethyl ether)-N,N’-tetraacetic acid; PMSF, phenylmethylsulfonyl fluoride; Hepes, 4-(2-hydroxyethylt 1-piperazineethanesulfonic acid; TFA, trifluoroacetic acid; EGFR 1, anti-human epidetmaI growth factor monoclonal antibody; PBS, phosphate-buffered saline; SDS, sodium dodecyl sulfate; Ig, immunoglobulin.
’ To whom correspondence should be addressed. 253
0003-2697184 $3.00 Copyright 0 1984 by Academic F’ress. Inc. All rights of reproduction in any form resewed
254
GULLICK
and IGFI receptor (l&16), whose activity is stimulated by binding of the respective growth factor ligands to the receptors, which themselves act as substrates for the protein kinase. Several cellular components also become phosphorylated in vivo ( 17), and other proteins or synthetic peptides can act as substrates in vitro (18-2 1). To further study the structure and function of the EGF receptor and its association with the protein kinase, this membrane protein must be purified from detergent lysates, and a prerequisite for such a purification is the ability to monitor the presence of the protein in a complex cell lysate with a reliable quantitative assay. Previously, detergent-solubilized EGF receptors have been quantitated by the binding of known specific activity iodinated EGF, and the receptor-ligand complexes generally separated from free ligand by polyethylene glycol precipitation (PEG) (22,23). With the availability of monoclonal antibodies (2429), specific immunoassays can be developed. We present here the characterization of a radioimmunoassay (RIA) for EGF receptors in solution using a mAb (24) and compare this method with the PEG assay and a gelfiltration assay. MATERIALS
AND
METHODS
A431 cells were grown at 37°C in Dulbecco’s modified Eagle’s medium containing 10% (v/v) newborn calf serum in an atmosphere of 10% C02-90% air. A431 cell membranes were prepared by the method of Thorn et al. (30). To obtain complete cell lysate, A43 1 cells were rinsed twice with PBS (10 mM NazHPG4, 1.7 mM KHZP04, 150 mM NaCl, 3 mM KCl, pH 7.2) containing 2 mM EGTA, and then lysed with 50 mM Tris-HCl buffer, pH 7.4, containing 1% (v/v) Triton X- 100, 150 mM NaCl, 25 mM benzamidine (Sigma), 5 mM EGTA, 0.1% (w/v) bovine serum albumin, and 1 mM phenylmethylsulfonyl fluoride (Sigma) (lysis buffer). The solution was then spun at 100,OOOg for 1 h at 4°C and the supematant was used for assays.
ET
AL.
Human placentas were obtained from a local hospital within 30 min of birth. Syncytiotrophoblastic plasma membrane vesicles were prepared from human placentas by a modification of the method of Smith et al. (3 1). Briefly, freshly delivered human term placentas were finely diced, the amnion and chorion being discarded. They were then washed in PBS with 2 IIIM EGTA, 2 mM benzamidine, 2 pg/ml soybean trypsin inhibitor, 2 &ml PMSF. The placental pieces were then stirred in 500 ml/placenta of buffer A (0.15 M NaCl, 20 II’IM Hepes, pH 7.4,2 IrIM EGTA, 2 mM benzamidine, 2 &ml trypsin inhibitor, 2 pg/ ml PMSF). After 1 h the mixture was filtered through gauze. The filtrate was centrifuged for 20 min at 4OOOg; the pellet from this spin was discarded, and the supematant recentrifuged at 100,OOOgfor 30 min. The pellet was washed, resuspended in buffer A, and stored at -70°C. Placental vesicles were solubilized by adding an equal volume of 50 mM Hepes buffer, pH 7.4, containing 2% (v/v) Triton X-100. The solution was then spun at 100,OOOg for 1 h at 4°C and the supematant was used for assays. EGF was purified from male mouse submaxillary glands by the method of Savage and Cohen (32) with the following modification. The final peak of protein to emerge from the Bio-Gel P- 10 sizing column was concentrated by passing it through a 5-ml syringe packed with silica coated with C 18 alkane chains (the contents of six SepPak cartridges, Waters). The column was washed with 0.1% (v/v) trifluoroacetic acid (TFA) in water containing 20% (v/v) acetonitrile (Rathbum, Scotland), and then the bound EGF was eluted with 0.1% TFA containing 60% acetonitrile. The organic solvent was evaporated under a stream of dry nitrogen, and the aqueous residue was lyophilized. The dried material was dissolved in 10 ml 20 mM ammonium acetate buffer, pH 5.6, and then applied to the ion-exchange column as in (32). The purified EGF had an amino acid composition identical to that predicted by its sequence (33); contained no trace of lysine, al-
RADIOIMMUNOASSAY
FOR EPIDERMAL
anine, or phenylalanine; and ran as a single, sharp, symmetrical peak on reverse-phase HPLC using a Dupont C8 column from which it eluted at 35% (v/v) acetonitrile in 0.1% TFA. EGF was iodinated by the soluble lactoperoxidase, glucose ox&se method. Lyophilized EGF (4.7 X lop9 mol) was dissolved in 20 ~1 0.5 M Na phosphate buffer, pH 7.4, to which was added 30 ~1 0.2 mg/ml lactoperox&se (&lb&hem), 12 ~17 units/ml glucose oxidase type V (Sigma), and 10 ~1 of carrierfree Naiz51 (Amersham). The reaction was started by adding 6 ~1 ~-glucose (90 mg/ml), and the reaction mixture was left at room temperature for 20 min. The solution was then applied to a column of Sephadex G25 coarse (0.7 X 19 cm) equilibrated in phosphate-buffered saline. The column was eluted and 0.45ml fractions were collected. The incorporation of label under these conditions was >90%, and the recovery of protein 60-70%. The specific activity of the lZSI-EGF at this stage was approximately 3 X 1017 cpm/mol. For the experiments here, however, the specific activity was reduced to approximately 6 X lO”j cpm/ mol by adding cold EGF to give a stock solution of 3.5 X 10e6 M. The specific activity of each preparation was determined exactly, and 1-pl standards were prepared, which were counted with each experiment to account for the isotopic decay. The monoclonal antibody R 1 was produced and purified as described by Waterfield et al. (24). The antibody was a mouse IgG2b which showed only heavy and light chains on SDSgel electrophoresis in the presence of reducing agents. The binding of 12’I-EGF to A-43 1 cell membranes or to placental vesicles was determined by the method of Carpenter et al. (34). Polyethylene glycol radioreceptor assays were performed as in (23). The gel-filtration radioreceptor assay was performed by incubating an aliquot of A431 cell lysate with a saturating concentration of 12’I-EGF (2 X 1O-7 M) in a final volume of 200 ~1 10 mM Na phosphate buffer, pH 7.4, containing 0.2% (v/v) Triton X-100 and 150 mM NaCl (Triton buffer). After 10 min in-
GROWTH
FACTOR
RECEPTOR
255
cubation at room temperature the solution was applied to a column of Sephadex G75 Fine (0.7 X 19 cm) equilibrated in Triton buffer. The column was eluted at a flow rate of 1 ml/min, and approximately 0.4-ml fractions were collected. RESULTS
In order to provide a quantitative measurement of EGF receptor concentration in detergent solution by RIA, several parameters were examined and optimized. Since the labeled l&and, lZSI-EGF, must be present in sufficiently high concentration to saturate all its specific binding sites, and because it has been reported that the affinity of the EGF receptor for EGF decreases upon solubilization of receptor containing membranes (23), we first determined the binding affinity of the EGF receptor in cell membranes prepared from human A431 epidermal carcinoma cells and from human placental tissue. Thus, membranes were incubated with increasing concentrations of 12’I-EGF, and the bound radioactivity was determined. As is shown in Fig. 1A, receptors were saturated at approximately 15 nM i2’I-EGF in each case. Scatchard analysis of the data from Fig. 1 gave a curvilinear plot. One interpretation of this is that there are at least two classes of binding sites for 12’I-EGF in A431 cell membranes (Fig. 1B) and placental membranes (Fig. 1C). The higher-affinity sites represented approximately 60% of the total apparent binding sites in each case. The measured dissociation constants for 12’I-EGF were 3.7 X 10-l’ and 3.1 X 10e9 M for the high- and low-affinity sites in A43 1 cell membranes and 4.7 X 10-l’ and 4.9 X 1Om9M for EGF receptors from placenta. We next solubilized the membranes with a final concentration of 1% (v/v) Triton-X-100, and centrifuged the solution to remove residual insoluble material. The binding affinity for lZSI-EGF of the solubilized receptors was then determined by RIA. The optimal concentration of mAb, the length of incubation, and the washing procedure are examined below. Figure 2A shows a binding curve for sol-
256
GULLICK
ET AL.
Concentration
of
lz5 I- EGF
(nM)
0.1
t
0
2
4
6 Bound
6 ‘=I-EGF
0
2
(cpm
x lo-*)
4
6
I
I 6
I
FIG. 1. The binding of increasing concentrations of “‘1-EGF to a fixed amount of cell membranes prepared from human A431 cells (0) or human placental tissue (O), and Scatchard analysis of the data from (A) for membrane preparations of A43 1 cells (B) and human placenta (C).
ubilized placental EGF receptors, and Fig. 2B that obtained with solubilized A431 cell receptors. Both receptor types were saturated at approximately 150 nM lz51-EGF, and Scatchard analysis gave an apparent single dissociation constant of 2.3 X lo-* M for placental receptors and 2.9 X lo-* M for A431 receptors, respectively (Figs. 2C and D). For subsequent RIA of solubilized receptors, we therefore used an ‘*%EGF concentration of 2 X lo-’ h4 to ensure lull occupancy of specihc binding sites. We next examined the concentration of mAb necessary to precipitate the EGF receptor. Using a saturating amount of ‘2SI-EGF, a fixed amount of EGF receptor, and increasing amounts of mAb, we obtained binding curves for A431 cell receptor and placental
receptor (Fig. 3). It should be emphasized that the assay has been designed to monitor purification of EGF receptor from whole-cell lysates as well as solubilized cell membranes. In order to minimize proteolysis of the receptor which can occur in cell lysates, we used an incubation time with mAb of only 1.5 h, and thus the reaction may not be at equilibrium at low concentrations of mAb. In subsequent RIA we therefore added 2 pg mAb per assay tube to give a final concentration of approximately 67 nM, which is in great excess and saturates its binding within the incubation period. The RIA was routinely Iierforrned as follows. The solution containing an unknown amount of EGF receptor ( 1O-40 ~1) was made to 200 ~1 with Triton buffer containing final
RADIOIMMUNOASSAY *b ;
FOR EPIDERMAL
GROWTH
FACTOR
RECEPTOR
257
A
5-
,--g0
1
2
-7 3
4
Concentration
of ‘*%
EGF
(nM
I 16*)
FIG. 2. The binding of increasing concentrations of “‘1-EGF to T&on X- 100~solubilized placental membranes (A) and A43 I membranes (B) measured in a radioimmunoassay. (0) Total ‘*‘I-EGF binding; (m) nonspecific ‘251-EGF binding in the presence of nonimmune mouse serum; (A, 0) specific ‘*‘I-EGF binding. Scatchard analysis of lZ51-EGF binding to the placental membranes (C) and A431 membranes (D).
concentrations of 2 X lo-’ M ‘*%EGF of known specific activity and 67 nM mAb. Assays were performed in triplicate or duplicate, depending on the amount of sample available. The solution was tumbled for 1 h at room temperature, and then 20 ~1 of a 1: 1 suspension of protein A-Sepharose in Triton buffer added. The mixture was tumbled for a further 30 min at room temperature, and then rapidly washed three times with l.O-ml aliquots of Triton buffer at 4°C by brief centrifugation and aspiration of the supematant. The washing procedure was shown to reduce the background nonspecific radioactivity to acceptable levels (Fig. 4). The specifically bound ‘*‘I-EGF did not appear to dissociate significantly dur-
ing the washing process (Fig. 4). The pellets were counted in a y counter, duplicate or trip licate values were averaged, background nonspecific values were subtracted, and the number of moles of EGF bound was calculated from the specific activity of ‘*%EGF. Since the number of EGF binding sites on each 170,000-Da EGF receptor polypeptide chain is not yet known, the concentration of receptor is expressed in moles of EGF bound per liter. Placental vesicles or A43 1 cell membranes were solubilized and RIAs performed using increasing amounts of membrane protein with saturating concentrations of 12%EGF and mAb. Background nonspecific binding was estimated either by adding a lOO-fold excess
258
GULLICK
Concentration
of antibody
(nM)
ET AL.
Protein
concentration
(mg/ml)
FIG. 3. Precipitation of EGF receptors from Triton XlOO-soiubilized A431 cell membranes (0) and placental membranes (0) labeled with a saturating concentration of r2’1-EGF by increasing concentrations of monoclonal antibody. Membrane
of cold EGF or by substituting normal mouse serum for the mAb. The nonspecifically bound material was less with the serum control and, since this avoided the use of large amounts of unlabeled EGF, we now estimate background binding by this method (Figs. 5A and B). Finally, we compared the standard PEG assay and a gel-filtration assay with the RIA described above. The objective of this was twofold: first, to establish that the mAb binds
Number
of
washes
4. Optimization of the washing procedure in the radioimmunoassay. Pellets from identical incubations were washed in parallel, and the total radioactivity obtained in the presence of monoclonal antibody (0) or with nonimmune mouse serum (0) was determined. The specifically bound radioactivity (A) was obtained by subtraction of the nonimmune serum background. FIG.
Protein
( p9 )
J Fraction
number
FIG. 5. RIA using increasing amounts of solubilized placental (A) or A431 cell membrane protein (B) and saturating concentrations of “~1-EGF and monoclonal antibody; total bound radioactivity in the presence of monoclonal antibody (O), nonimmune mouse serum (A, A), or a W-fold excessof unlabeled EGF (0). (C) PEG assayusing the same amounts ofA 1 cell protein present in (B); total radioactivity precipitated (a), total radioactivity bound in the presence of a IOO-fold excess of unlabeled EGF (0). (D) Gel filtration on a column of Sephadex G75 of 80 pg solubilized A431 cell membrane protein labeled with ‘251-EGF in the presence (- - -) or absence (-) of a lOO-fold excess of unlabeled EGF.
to all of the EGF receptors; and, second, to compare the ease and reproducibility of each method. Figure 5 shows the results of the BIA (Fig. 5B) using increasing amounts of A431 cell membrane extract, and results are compared with the PEG assay data (Fig. 5C). A sample containing the largest amount of membrane protein was also labeled with i2% EGF and run on a Sephadex G75 column to separate bound from free ligand (Fig. 5D). Since the PEG assay relies upon the preferential precipitation of all the high-molecularweight material from the solution (leaving the
RADIOIMMUNOASSAY
FOR
EPIDERMAL
GROWTH
FACTOR
RECEPTOR
259
their detergent-solubilized counterparts. One tissue is a human organ and the other a cell line derived from a cervical carcinoma. Interestingly, both intact cell membrane preparations showed two major classes of EGF receptors defined by their binding affinity for EGF. The ratio of high- to low-affinity binding sites is similar in both tissues. We have not examined here the small percentage of receptors reported by Kawamoto et al. (28) which possess very high affinities for EGF, although they may be of functional significance. The binding constants for EGF of the high- and low-affinity forms of receptors in each tissue are essentially indistinguishable. Since they are both of human origin this is not unreasonable, although the cell types are very different. The primary structures of the two receptor types are, however, very similar (41), and the mAb used here reacts with both receptor types. DISCUSSION The physical basis for the EGF receptor existing in two affinity states for EGF may be a A reproducible, accurate, and simple consequence of its local environment, such as method for measuring the concentration of EGF receptor in detergent solution will pro- the nature of its lipid neighbors or cytoskeletal vide an invaluable adjunct for many areas of interactions. Hillman and Schlessinger (38) receptor research. It will enable the quantihave reported that there are mobile and imtation of the proportion of EGF receptors mobile fractions of EGF receptors in cell present in a cell line or tissue that can be membranes, and that the former represents solubilized by detergents, and a measurement 50-70% of the total population at 23°C. of their binding affinity for EGF alter memUpon solubilization with Triton X- 100 the brane disruption. Second, an assay will be es- EGF receptors are converted to an even lowersential for monitoring receptor purification affinity state for EGF. In this case only one and reconstitution into liposomes or planar class of binding sites is present, having a Kd bilayers (35,36). of 2-3 X lo-’ M. There are several possible The only ligand available in a pure form explanations for this phenomenon, and the until recently that (by definition) bound spe- mechanism involved in this change of affinity cifically to EGF receptors was EGF. This molis not understood but it may involve disagecule can be iodinated on one or two tyrosine gregation of polypeptides from a functional residues without apparent effect on its binding multisubunit structure. It is clear, however, selectivity or affinity (37). There are two basic that a single EGF receptor molecule is suffisteps in a radioligand assay: First, the satu- cient to bind EGF in detergent solution (39). ration of specific binding sites with a ligand The practical consequence of the lowered of known specific activity; and then the sep- affinity of the solubilized EGF receptor is that aration of bound from free ligand while mainhigh concentrations of ligand (2 X IO-’ M) taining receptor occupancy. We, therefore, are required for saturation. This creates a determined the concentration of EGF nec- problem common to all assays employing laessary to saturate specific binding sites in two beled EGF, in that the higher the specific acpreparations of human cell membranes and tivity of the ligand the more sensitive is the
free ligand in the supematant), the only measure of nonspecific 12’1-EGF binding was to add a lOO-fold excess of cold EGF at each point. We also used this control for the gelfiltration assay, which separates bound from free l&and on the basis of effective molecular size. The RIA and PEG assays were performed using duplicate incubations for each amount of membrane protein, and it is apparent from Fig. 5 that the PEG assay is subject to more variability and higher nonspecific background values than is the RIA. The gel-filtration assay also suffered from high nonspecific 129EGF binding. The absolute amount of specifically bound EGF was highest for the RIA, next highest with the PEG assay, and rather less for the gel-filtration assay. Thus, the mAb recognizes all the specific EGF binding material present in the reaction mixture.
260
GULLICK
ET AL.
assay but the more radioactivity must be tolACKNOWLEDGMENTS erated. We use here ‘251-EGF with an activity We thank Dr. Peter Parker, Dr. Elaine Mayes, and Dr. of approximately 6 X lOi cpm mole-‘, which Paul Stroobant for useful discussions. We are grateful to means that a single reaction mixture of 200 the Queen Charlotte’s Maternity Hospital and to Dr. E. ~1 with an EGF concentration of 2 X lo-’ M Owen for help in obtaining placental tissue. contains 1.2 X lo6 cpm. Theoretically the speREFERENCES cific activity of the EGF could be increased by about one hundred times with a concom1. Guroff, G., Ed. (I 983) Growth and Maturation Facitant increase in sensitivity of the assay, but tors, Vol. 1, Wiley, New York. 2. Kasuga, M., Hedo, J. A., Yamada, K. M., and Kahn, the amount of radioactivity necessary would C. R. (1982) J. Biol. Chem. 257, 10392-10399. generally be considered unacceptable. The 3. Fujita-Yamaguchi, Y., Choi, S., Sakamoto, Y., and specific activity used here allows detection of Itakura, K. (1983) J. Biol. Chem. 258.5045-5049. EGF receptor concentrations as low as na- 4. Czech, M. P. (1982) Cell 31, 8-10. nomolar (a tube containing 200 ~1 of EGF 5. Blundell, T. L., Bedarkar, S., and Humbel, R. E. (1983) Fed. Proc. 42,2592-2597. receptor at 1Od9 M would have 1.2 X lo4 spe6. Perdue, J. F., Chart, J. K., Thibault, C., Radaj, P., cifically bound cpm). We are, however, curMills, B., and Daughaday, W. H. (1983) J. Biol. rently investigating alternative iodinatable liChem. 258, 7800-7811. gands such as mAbs, or fragments of mAbs, 7. August, G. P., Nissley, S. P., Kasuga, M., Lee, L., Greenstein, L., and Rechler, M. M. (1983) J. Biol. whose affinity of binding to the EGF receptor Chem. 258,9033-9036. is not diminished upon solubilization. These 8. Heldin, C-H., Ek, B., and Ronnstrand, L. (1983) Cell. reagents have the additional advantage that Biol. Int. Rep. 7, 543-544. molecules such as aTGFs (40), if present, 9. Pike, L. J., Bowen-Pope, D. F., Ross, R., and Krebs, would not compete for binding as they would E. G. (1983) J. Biol. Chem. 258,9383-9390. using ‘251-EGF. 10. Huang, J. S., Huang, S. S., Kennedy, B., and Deuel, T. F. (1982) J. Biol. Chem. 257, 8130-8136. The assay developed here was compared with the established PEG precipitation assay. 11. Carpenter, G. (1983) Mol. Cell Endocrinol. 31, l-19. 12. Schlessinger, J., Schreiber, A. B., Levi, A., Lax, I., The RIA was more reproducible and avoided Libermann, T., and Yarden, Y. (1983) CRC Crif. the use of large amounts of cold EGF as a Revs. Biochem. 14, 93-111. nonspecific binding control. The low back- 13. Cohen, S., Ushiro, H., Stoscheck, C., and Chinkers, M. (1982) J. Biol. Chem. 257, 1523-1531. ground binding of EGF in the RIA is probably a consequence of the assay possessing two lev- 14. Zick, Y., Kasuga, M., Kahn, C. R., and Roth, J. (1983) J. Biol. Chem. 258,75-80. els of specificity; the binding of EGF to the 15. Rubin, J. B., Shia, M. A., and Pilch, P. F. (1983) EGF receptor and the binding of the mAb to Nature (London) 305, 438-440. the EGF receptor. The total cpm bound in 16. Jacobs, S., Kull, F. C., Earp, H. S., Svoboda, M. E., Van Wyk, J. J., and Cuatrecasas, P. (1983) J. Biol. each assay using identical reaction mixtures Chem. 258,958 l-9584. was slightly higher for the RIA, probably be17. Cooper, J. A., Bowen-Pope, D. F., Raines, E., Ross, cause the washing procedure could be perR., and Hunter, T. (1982) Cell 31, 263-273. formed more rapidly, avoiding dissociation of 18. Cassel, D., Pike, L. J., Grant, G. A., Krebs, E. G., specifically bound ligand. and Glaser, L. (1983) J. Biol. Chem. 258,29452950. Finally, we have used the RIA to monitor 19. Baldwin, G. S., Kneael, J., and Monckton, J. M. ( 1983) EGF receptor purification by immunoaffinity Nature (London) 301, 435437. chromatography using mAb Rl coupled to 20. Baldwin, G. S., Stanley, I. J., and Nice, E. C. (1983) Sepharose (41). By using the same mAb in FEBS LETT. 153,257-261. 21. Stadtmauer, L. A., and Rozen, 0. M. (1983) J. Biol. solution to monitor EGF receptor concentraChem. 258,6682-6685. tion, any conditions which influenced the P. (1972) Proc. Nat/. Acad. Sci. USA 69, binding to the affinity column were detected 22. Cuatrecasas, 318-322. directly. Milligram quantities of EGF receptor 23. Carpenter, G. (1979) Life Sci. 24, 1691-1698. from A43 1 cells were obtained using this ap- 24. Waterfield, M. D., Mayea, E. L. V., Stroobant, P., Bennett, P. L. P., Young, S., Goodfellow, P. N., proach.
RADIOIMMUNOASSAY
FOR EPIDERMAL
Banting, G. S., and Ozanne, B. (1982) J. Cell. Biochem. 20, 149-161. 25. Watetield, M. D., Parker, P., and Mayes, E. (1983) Cell Biol. Int. Rep. 7, 535-536. 26. Schreiber, A, B., Lax, I., Yarden, Y., Eshhar, ‘Z., and Schlessinger, J. (198 1) Proc. Natl. Acad. Sci. USA n&,7535-7539. 27.
Schreiber, A. B., Libermann, T. A., Lax, I., Yarden, Y., and Schlessinger, J. (1983) J. Biol. Chem. 258, 846-853.
28.
29. 30. 31.
32.
Kawamoto, T., Sato, J. D., Le, A., Polikoff, J., Sate, G., and Mendelsohn, J. (1983) Proc. Natl. Acad. Sci. USA 80, 1337-1341. Richert, N. D., Willingham, M. C., and Pastan, I. (1983) J. Biol. Chem. 258, 8902-8907. Thorn, D., Powell, A. J., Lloyd, C. W., and Rees, D. A. (1977) Biochem. J. 168, 187-194. Smith, C. H., Nelson, D. M., King, B. F. Donohue, T. M., Ruzycki, S. M., and Kelley, L. K. ( 1977) Amer. .I. Obstet. Gynecol. 128, 190-196. Savage, C. R., and Cohen, S. (1972) J. Biol. Chem. 241, 7609-7611.
GROWTH
FACTOR
RECEPTOR
261
33. Savage, C. R., Inagami, T., and Cohen, S. (1972) J. Biol. Chem. 241,16 12-162 1. 34. Carpenter, G., King, L., and Cohen, S. (1979) J. Biol. Chem. 254,4884-489 1. 35. Anholt, R., Lindstrom, J., and Montal, M. (198 I) J. Biol. Chem. 256,4317-4387. 36. Nelson, N., Anholt, R., Lindstrom, J., and Montal, M. (1980) Proc. Natl. Acad. Sci. USA 77, 30573061. 37.
38. 39.
40.
41.
Burgess, A. W., Lloyd, C. J., and Nice, E. C. (1983) EMBO J. 2,2065-2069. Hillman, G., and Schlessinger, J. (1982) Biochemistrv 21, 1667-1672. Hoch, R. A., Nexo, E., and Hollenberg, M. D. ( 1980) J. Biol. Chem. 255, 10,737-10,743. Marquardt, H., Hunkapiller, M. W., Hood, L. E., Twardzik, D. R., DeLarco, J. E., Stephenson, J. R., and Todaro, G. J. (1983) Proc. Natl. dead. Sci. USA 80, 4684-4688. Downward, J., Yarden, Y.. Scrace, G., Totty, N., Stockwell, P., Ullrich, A., Schlessinger, J., and Waterfield, M. D. (1984) Nature (Londonj 307,521521.
BIOCHEMISTRY
141,253-261
A Radioimmunoassay WILLIAM J. GULLICK,’
(1984)
for Human Epidermal
Growth
D. JULIAN H. DOWNWARD, AND MICHAEL D. WATERFIELD
JUDITH
Factor Receptor J. MARSDEN,
Department of’ Protein Chemistry, Imperial Cancer Research Fund Laboratories, LincolnS Inn Fields, London WC2A 3PX, England Received January 24, 1984 The development of a radioimmunoassay (RIA) for the human epidermal growth factor receptor solubilized with nonionic detergents which employs iodinated epidermal growth factor (“‘IEGF) as the specific ligand is described. A monoclonal antibody (Rl) that binds specifically to human EGF receptors [Waterheld, M. D., et al. (1982) J. Cell B&hem. 20, 149-1611 was used to separate solubilized receptors saturated with i’sI-EGF from free ligand by absorption to protein A-Sepharose, and the bound radioactivity was determined. The RIA was linear when increasing amounts of solubilized membrane protein were added and, when compared to the standard polyethylene glycol assay,was mom reproducible. In addition, the background nonspecific binding obtained in the presence of a hundred-fold excessof unlabeled EGF was lessin the RIA. Substitution of normal mouse serum for the monoclonal antibody gave very low nonspecific background ligand binding and avoided the use.of large amounts of unlabeled EGF in the assay. Two major classes of binding sites for EGF were observed in membrane preparations from the cervical carcinoma cell line A431 or from normal human placental tissue. These were present in ap proximately equal amounts, with apparent dissociation constants of 4 X lo-” and 4 X 1O-9 M. Upon solubilization with the nonionic detergent Triton X-100, only one class of EGF binding sites was detected in both cases, with a dissociation constant of 3 X IO-* M. The RIA can be used to monitor receptor purification and for quantitation of receptor number and affinity in various cell types. KEY WORDS: epidermal growth factor receptor; epidermal growth factor, radioimmunoassay: monoclonal antibodies; membrane proteins; glycoproteins.
The role of growth factors in the regulation of the growth of normal and transformed or tumor-derived cells has been the subject of intense study [for recent review see Ref. (l)]. Although a number of growth factors have been characterized, details of the structure and function of their receptors are much less complete. The receptors which have been best characterized are those for insulin (2-4), insulin-like growth factor I (4,5), insulin-like growth factor II (6,7), platelet-derived growth factor (8-lo), and epidermal growth factor [for reviews see (11,12)]. The receptors are all integral membrane proteins composed of one
(IGFIIR, PDGFR, EGFR)* or several (IR, IGFIR) polypeptides. EGF receptor (13) and insulin receptor ( 14) copurify with a protein kinase, as probably do the PDGF receptor (9) * Abbreviations: PDGF, Plateletderived growth factor; IGFIIR, Insulin-like growth factor II receptor; PDGFR, Platelet-derived growth factor receptor, EGFR, Epidermal growth factor receptor; IR, Insulin receptor; IGFIR, Insulin-like growth factor I receptor, EGF, Epidermal growth factor; mAb, Monoclonal antibody; PEG, polyethylene glycol; RIA, Radioimmunoassay; EGTA, ethylene glycol bis (&aminoethyl ether)-N,N’-tetraacetic acid; PMSF, phenylmethylsulfonyl fluoride; Hepes, 4-(2-hydroxyethylt 1-piperazineethanesulfonic acid; TFA, trifluoroacetic acid; EGFR 1, anti-human epidetmaI growth factor monoclonal antibody; PBS, phosphate-buffered saline; SDS, sodium dodecyl sulfate; Ig, immunoglobulin.
’ To whom correspondence should be addressed. 253
0003-2697184 $3.00 Copyright 0 1984 by Academic F’ress. Inc. All rights of reproduction in any form resewed
254
GULLICK
and IGFI receptor (l&16), whose activity is stimulated by binding of the respective growth factor ligands to the receptors, which themselves act as substrates for the protein kinase. Several cellular components also become phosphorylated in vivo ( 17), and other proteins or synthetic peptides can act as substrates in vitro (18-2 1). To further study the structure and function of the EGF receptor and its association with the protein kinase, this membrane protein must be purified from detergent lysates, and a prerequisite for such a purification is the ability to monitor the presence of the protein in a complex cell lysate with a reliable quantitative assay. Previously, detergent-solubilized EGF receptors have been quantitated by the binding of known specific activity iodinated EGF, and the receptor-ligand complexes generally separated from free ligand by polyethylene glycol precipitation (PEG) (22,23). With the availability of monoclonal antibodies (2429), specific immunoassays can be developed. We present here the characterization of a radioimmunoassay (RIA) for EGF receptors in solution using a mAb (24) and compare this method with the PEG assay and a gelfiltration assay. MATERIALS
AND
METHODS
A431 cells were grown at 37°C in Dulbecco’s modified Eagle’s medium containing 10% (v/v) newborn calf serum in an atmosphere of 10% C02-90% air. A431 cell membranes were prepared by the method of Thorn et al. (30). To obtain complete cell lysate, A43 1 cells were rinsed twice with PBS (10 mM NazHPG4, 1.7 mM KHZP04, 150 mM NaCl, 3 mM KCl, pH 7.2) containing 2 mM EGTA, and then lysed with 50 mM Tris-HCl buffer, pH 7.4, containing 1% (v/v) Triton X- 100, 150 mM NaCl, 25 mM benzamidine (Sigma), 5 mM EGTA, 0.1% (w/v) bovine serum albumin, and 1 mM phenylmethylsulfonyl fluoride (Sigma) (lysis buffer). The solution was then spun at 100,OOOg for 1 h at 4°C and the supematant was used for assays.
ET
AL.
Human placentas were obtained from a local hospital within 30 min of birth. Syncytiotrophoblastic plasma membrane vesicles were prepared from human placentas by a modification of the method of Smith et al. (3 1). Briefly, freshly delivered human term placentas were finely diced, the amnion and chorion being discarded. They were then washed in PBS with 2 IIIM EGTA, 2 mM benzamidine, 2 pg/ml soybean trypsin inhibitor, 2 &ml PMSF. The placental pieces were then stirred in 500 ml/placenta of buffer A (0.15 M NaCl, 20 II’IM Hepes, pH 7.4,2 IrIM EGTA, 2 mM benzamidine, 2 &ml trypsin inhibitor, 2 pg/ ml PMSF). After 1 h the mixture was filtered through gauze. The filtrate was centrifuged for 20 min at 4OOOg; the pellet from this spin was discarded, and the supematant recentrifuged at 100,OOOgfor 30 min. The pellet was washed, resuspended in buffer A, and stored at -70°C. Placental vesicles were solubilized by adding an equal volume of 50 mM Hepes buffer, pH 7.4, containing 2% (v/v) Triton X-100. The solution was then spun at 100,OOOg for 1 h at 4°C and the supematant was used for assays. EGF was purified from male mouse submaxillary glands by the method of Savage and Cohen (32) with the following modification. The final peak of protein to emerge from the Bio-Gel P- 10 sizing column was concentrated by passing it through a 5-ml syringe packed with silica coated with C 18 alkane chains (the contents of six SepPak cartridges, Waters). The column was washed with 0.1% (v/v) trifluoroacetic acid (TFA) in water containing 20% (v/v) acetonitrile (Rathbum, Scotland), and then the bound EGF was eluted with 0.1% TFA containing 60% acetonitrile. The organic solvent was evaporated under a stream of dry nitrogen, and the aqueous residue was lyophilized. The dried material was dissolved in 10 ml 20 mM ammonium acetate buffer, pH 5.6, and then applied to the ion-exchange column as in (32). The purified EGF had an amino acid composition identical to that predicted by its sequence (33); contained no trace of lysine, al-
RADIOIMMUNOASSAY
FOR EPIDERMAL
anine, or phenylalanine; and ran as a single, sharp, symmetrical peak on reverse-phase HPLC using a Dupont C8 column from which it eluted at 35% (v/v) acetonitrile in 0.1% TFA. EGF was iodinated by the soluble lactoperoxidase, glucose ox&se method. Lyophilized EGF (4.7 X lop9 mol) was dissolved in 20 ~1 0.5 M Na phosphate buffer, pH 7.4, to which was added 30 ~1 0.2 mg/ml lactoperox&se (&lb&hem), 12 ~17 units/ml glucose oxidase type V (Sigma), and 10 ~1 of carrierfree Naiz51 (Amersham). The reaction was started by adding 6 ~1 ~-glucose (90 mg/ml), and the reaction mixture was left at room temperature for 20 min. The solution was then applied to a column of Sephadex G25 coarse (0.7 X 19 cm) equilibrated in phosphate-buffered saline. The column was eluted and 0.45ml fractions were collected. The incorporation of label under these conditions was >90%, and the recovery of protein 60-70%. The specific activity of the lZSI-EGF at this stage was approximately 3 X 1017 cpm/mol. For the experiments here, however, the specific activity was reduced to approximately 6 X lO”j cpm/ mol by adding cold EGF to give a stock solution of 3.5 X 10e6 M. The specific activity of each preparation was determined exactly, and 1-pl standards were prepared, which were counted with each experiment to account for the isotopic decay. The monoclonal antibody R 1 was produced and purified as described by Waterfield et al. (24). The antibody was a mouse IgG2b which showed only heavy and light chains on SDSgel electrophoresis in the presence of reducing agents. The binding of 12’I-EGF to A-43 1 cell membranes or to placental vesicles was determined by the method of Carpenter et al. (34). Polyethylene glycol radioreceptor assays were performed as in (23). The gel-filtration radioreceptor assay was performed by incubating an aliquot of A431 cell lysate with a saturating concentration of 12’I-EGF (2 X 1O-7 M) in a final volume of 200 ~1 10 mM Na phosphate buffer, pH 7.4, containing 0.2% (v/v) Triton X-100 and 150 mM NaCl (Triton buffer). After 10 min in-
GROWTH
FACTOR
RECEPTOR
255
cubation at room temperature the solution was applied to a column of Sephadex G75 Fine (0.7 X 19 cm) equilibrated in Triton buffer. The column was eluted at a flow rate of 1 ml/min, and approximately 0.4-ml fractions were collected. RESULTS
In order to provide a quantitative measurement of EGF receptor concentration in detergent solution by RIA, several parameters were examined and optimized. Since the labeled l&and, lZSI-EGF, must be present in sufficiently high concentration to saturate all its specific binding sites, and because it has been reported that the affinity of the EGF receptor for EGF decreases upon solubilization of receptor containing membranes (23), we first determined the binding affinity of the EGF receptor in cell membranes prepared from human A431 epidermal carcinoma cells and from human placental tissue. Thus, membranes were incubated with increasing concentrations of 12’I-EGF, and the bound radioactivity was determined. As is shown in Fig. 1A, receptors were saturated at approximately 15 nM i2’I-EGF in each case. Scatchard analysis of the data from Fig. 1 gave a curvilinear plot. One interpretation of this is that there are at least two classes of binding sites for 12’I-EGF in A431 cell membranes (Fig. 1B) and placental membranes (Fig. 1C). The higher-affinity sites represented approximately 60% of the total apparent binding sites in each case. The measured dissociation constants for 12’I-EGF were 3.7 X 10-l’ and 3.1 X 10e9 M for the high- and low-affinity sites in A43 1 cell membranes and 4.7 X 10-l’ and 4.9 X 1Om9M for EGF receptors from placenta. We next solubilized the membranes with a final concentration of 1% (v/v) Triton-X-100, and centrifuged the solution to remove residual insoluble material. The binding affinity for lZSI-EGF of the solubilized receptors was then determined by RIA. The optimal concentration of mAb, the length of incubation, and the washing procedure are examined below. Figure 2A shows a binding curve for sol-
256
GULLICK
ET AL.
Concentration
of
lz5 I- EGF
(nM)
0.1
t
0
2
4
6 Bound
6 ‘=I-EGF
0
2
(cpm
x lo-*)
4
6
I
I 6
I
FIG. 1. The binding of increasing concentrations of “‘1-EGF to a fixed amount of cell membranes prepared from human A431 cells (0) or human placental tissue (O), and Scatchard analysis of the data from (A) for membrane preparations of A43 1 cells (B) and human placenta (C).
ubilized placental EGF receptors, and Fig. 2B that obtained with solubilized A431 cell receptors. Both receptor types were saturated at approximately 150 nM lz51-EGF, and Scatchard analysis gave an apparent single dissociation constant of 2.3 X lo-* M for placental receptors and 2.9 X lo-* M for A431 receptors, respectively (Figs. 2C and D). For subsequent RIA of solubilized receptors, we therefore used an ‘*%EGF concentration of 2 X lo-’ h4 to ensure lull occupancy of specihc binding sites. We next examined the concentration of mAb necessary to precipitate the EGF receptor. Using a saturating amount of ‘2SI-EGF, a fixed amount of EGF receptor, and increasing amounts of mAb, we obtained binding curves for A431 cell receptor and placental
receptor (Fig. 3). It should be emphasized that the assay has been designed to monitor purification of EGF receptor from whole-cell lysates as well as solubilized cell membranes. In order to minimize proteolysis of the receptor which can occur in cell lysates, we used an incubation time with mAb of only 1.5 h, and thus the reaction may not be at equilibrium at low concentrations of mAb. In subsequent RIA we therefore added 2 pg mAb per assay tube to give a final concentration of approximately 67 nM, which is in great excess and saturates its binding within the incubation period. The RIA was routinely Iierforrned as follows. The solution containing an unknown amount of EGF receptor ( 1O-40 ~1) was made to 200 ~1 with Triton buffer containing final
RADIOIMMUNOASSAY *b ;
FOR EPIDERMAL
GROWTH
FACTOR
RECEPTOR
257
A
5-
,--g0
1
2
-7 3
4
Concentration
of ‘*%
EGF
(nM
I 16*)
FIG. 2. The binding of increasing concentrations of “‘1-EGF to T&on X- 100~solubilized placental membranes (A) and A43 I membranes (B) measured in a radioimmunoassay. (0) Total ‘*‘I-EGF binding; (m) nonspecific ‘251-EGF binding in the presence of nonimmune mouse serum; (A, 0) specific ‘*‘I-EGF binding. Scatchard analysis of lZ51-EGF binding to the placental membranes (C) and A431 membranes (D).
concentrations of 2 X lo-’ M ‘*%EGF of known specific activity and 67 nM mAb. Assays were performed in triplicate or duplicate, depending on the amount of sample available. The solution was tumbled for 1 h at room temperature, and then 20 ~1 of a 1: 1 suspension of protein A-Sepharose in Triton buffer added. The mixture was tumbled for a further 30 min at room temperature, and then rapidly washed three times with l.O-ml aliquots of Triton buffer at 4°C by brief centrifugation and aspiration of the supematant. The washing procedure was shown to reduce the background nonspecific radioactivity to acceptable levels (Fig. 4). The specifically bound ‘*‘I-EGF did not appear to dissociate significantly dur-
ing the washing process (Fig. 4). The pellets were counted in a y counter, duplicate or trip licate values were averaged, background nonspecific values were subtracted, and the number of moles of EGF bound was calculated from the specific activity of ‘*%EGF. Since the number of EGF binding sites on each 170,000-Da EGF receptor polypeptide chain is not yet known, the concentration of receptor is expressed in moles of EGF bound per liter. Placental vesicles or A43 1 cell membranes were solubilized and RIAs performed using increasing amounts of membrane protein with saturating concentrations of 12%EGF and mAb. Background nonspecific binding was estimated either by adding a lOO-fold excess
258
GULLICK
Concentration
of antibody
(nM)
ET AL.
Protein
concentration
(mg/ml)
FIG. 3. Precipitation of EGF receptors from Triton XlOO-soiubilized A431 cell membranes (0) and placental membranes (0) labeled with a saturating concentration of r2’1-EGF by increasing concentrations of monoclonal antibody. Membrane
of cold EGF or by substituting normal mouse serum for the mAb. The nonspecifically bound material was less with the serum control and, since this avoided the use of large amounts of unlabeled EGF, we now estimate background binding by this method (Figs. 5A and B). Finally, we compared the standard PEG assay and a gel-filtration assay with the RIA described above. The objective of this was twofold: first, to establish that the mAb binds
Number
of
washes
4. Optimization of the washing procedure in the radioimmunoassay. Pellets from identical incubations were washed in parallel, and the total radioactivity obtained in the presence of monoclonal antibody (0) or with nonimmune mouse serum (0) was determined. The specifically bound radioactivity (A) was obtained by subtraction of the nonimmune serum background. FIG.
Protein
( p9 )
J Fraction
number
FIG. 5. RIA using increasing amounts of solubilized placental (A) or A431 cell membrane protein (B) and saturating concentrations of “~1-EGF and monoclonal antibody; total bound radioactivity in the presence of monoclonal antibody (O), nonimmune mouse serum (A, A), or a W-fold excessof unlabeled EGF (0). (C) PEG assayusing the same amounts ofA 1 cell protein present in (B); total radioactivity precipitated (a), total radioactivity bound in the presence of a IOO-fold excess of unlabeled EGF (0). (D) Gel filtration on a column of Sephadex G75 of 80 pg solubilized A431 cell membrane protein labeled with ‘251-EGF in the presence (- - -) or absence (-) of a lOO-fold excess of unlabeled EGF.
to all of the EGF receptors; and, second, to compare the ease and reproducibility of each method. Figure 5 shows the results of the BIA (Fig. 5B) using increasing amounts of A431 cell membrane extract, and results are compared with the PEG assay data (Fig. 5C). A sample containing the largest amount of membrane protein was also labeled with i2% EGF and run on a Sephadex G75 column to separate bound from free ligand (Fig. 5D). Since the PEG assay relies upon the preferential precipitation of all the high-molecularweight material from the solution (leaving the
RADIOIMMUNOASSAY
FOR
EPIDERMAL
GROWTH
FACTOR
RECEPTOR
259
their detergent-solubilized counterparts. One tissue is a human organ and the other a cell line derived from a cervical carcinoma. Interestingly, both intact cell membrane preparations showed two major classes of EGF receptors defined by their binding affinity for EGF. The ratio of high- to low-affinity binding sites is similar in both tissues. We have not examined here the small percentage of receptors reported by Kawamoto et al. (28) which possess very high affinities for EGF, although they may be of functional significance. The binding constants for EGF of the high- and low-affinity forms of receptors in each tissue are essentially indistinguishable. Since they are both of human origin this is not unreasonable, although the cell types are very different. The primary structures of the two receptor types are, however, very similar (41), and the mAb used here reacts with both receptor types. DISCUSSION The physical basis for the EGF receptor existing in two affinity states for EGF may be a A reproducible, accurate, and simple consequence of its local environment, such as method for measuring the concentration of EGF receptor in detergent solution will pro- the nature of its lipid neighbors or cytoskeletal vide an invaluable adjunct for many areas of interactions. Hillman and Schlessinger (38) receptor research. It will enable the quantihave reported that there are mobile and imtation of the proportion of EGF receptors mobile fractions of EGF receptors in cell present in a cell line or tissue that can be membranes, and that the former represents solubilized by detergents, and a measurement 50-70% of the total population at 23°C. of their binding affinity for EGF alter memUpon solubilization with Triton X- 100 the brane disruption. Second, an assay will be es- EGF receptors are converted to an even lowersential for monitoring receptor purification affinity state for EGF. In this case only one and reconstitution into liposomes or planar class of binding sites is present, having a Kd bilayers (35,36). of 2-3 X lo-’ M. There are several possible The only ligand available in a pure form explanations for this phenomenon, and the until recently that (by definition) bound spe- mechanism involved in this change of affinity cifically to EGF receptors was EGF. This molis not understood but it may involve disagecule can be iodinated on one or two tyrosine gregation of polypeptides from a functional residues without apparent effect on its binding multisubunit structure. It is clear, however, selectivity or affinity (37). There are two basic that a single EGF receptor molecule is suffisteps in a radioligand assay: First, the satu- cient to bind EGF in detergent solution (39). ration of specific binding sites with a ligand The practical consequence of the lowered of known specific activity; and then the sep- affinity of the solubilized EGF receptor is that aration of bound from free ligand while mainhigh concentrations of ligand (2 X IO-’ M) taining receptor occupancy. We, therefore, are required for saturation. This creates a determined the concentration of EGF nec- problem common to all assays employing laessary to saturate specific binding sites in two beled EGF, in that the higher the specific acpreparations of human cell membranes and tivity of the ligand the more sensitive is the
free ligand in the supematant), the only measure of nonspecific 12’1-EGF binding was to add a lOO-fold excess of cold EGF at each point. We also used this control for the gelfiltration assay, which separates bound from free l&and on the basis of effective molecular size. The RIA and PEG assays were performed using duplicate incubations for each amount of membrane protein, and it is apparent from Fig. 5 that the PEG assay is subject to more variability and higher nonspecific background values than is the RIA. The gel-filtration assay also suffered from high nonspecific 129EGF binding. The absolute amount of specifically bound EGF was highest for the RIA, next highest with the PEG assay, and rather less for the gel-filtration assay. Thus, the mAb recognizes all the specific EGF binding material present in the reaction mixture.
260
GULLICK
ET AL.
assay but the more radioactivity must be tolACKNOWLEDGMENTS erated. We use here ‘251-EGF with an activity We thank Dr. Peter Parker, Dr. Elaine Mayes, and Dr. of approximately 6 X lOi cpm mole-‘, which Paul Stroobant for useful discussions. We are grateful to means that a single reaction mixture of 200 the Queen Charlotte’s Maternity Hospital and to Dr. E. ~1 with an EGF concentration of 2 X lo-’ M Owen for help in obtaining placental tissue. contains 1.2 X lo6 cpm. Theoretically the speREFERENCES cific activity of the EGF could be increased by about one hundred times with a concom1. Guroff, G., Ed. (I 983) Growth and Maturation Facitant increase in sensitivity of the assay, but tors, Vol. 1, Wiley, New York. 2. Kasuga, M., Hedo, J. A., Yamada, K. M., and Kahn, the amount of radioactivity necessary would C. R. (1982) J. Biol. Chem. 257, 10392-10399. generally be considered unacceptable. The 3. Fujita-Yamaguchi, Y., Choi, S., Sakamoto, Y., and specific activity used here allows detection of Itakura, K. (1983) J. Biol. Chem. 258.5045-5049. EGF receptor concentrations as low as na- 4. Czech, M. P. (1982) Cell 31, 8-10. nomolar (a tube containing 200 ~1 of EGF 5. Blundell, T. L., Bedarkar, S., and Humbel, R. E. (1983) Fed. Proc. 42,2592-2597. receptor at 1Od9 M would have 1.2 X lo4 spe6. Perdue, J. F., Chart, J. K., Thibault, C., Radaj, P., cifically bound cpm). We are, however, curMills, B., and Daughaday, W. H. (1983) J. Biol. rently investigating alternative iodinatable liChem. 258, 7800-7811. gands such as mAbs, or fragments of mAbs, 7. August, G. P., Nissley, S. P., Kasuga, M., Lee, L., Greenstein, L., and Rechler, M. M. (1983) J. Biol. whose affinity of binding to the EGF receptor Chem. 258,9033-9036. is not diminished upon solubilization. These 8. Heldin, C-H., Ek, B., and Ronnstrand, L. (1983) Cell. reagents have the additional advantage that Biol. Int. Rep. 7, 543-544. molecules such as aTGFs (40), if present, 9. Pike, L. J., Bowen-Pope, D. F., Ross, R., and Krebs, would not compete for binding as they would E. G. (1983) J. Biol. Chem. 258,9383-9390. using ‘251-EGF. 10. Huang, J. S., Huang, S. S., Kennedy, B., and Deuel, T. F. (1982) J. Biol. Chem. 257, 8130-8136. The assay developed here was compared with the established PEG precipitation assay. 11. Carpenter, G. (1983) Mol. Cell Endocrinol. 31, l-19. 12. Schlessinger, J., Schreiber, A. B., Levi, A., Lax, I., The RIA was more reproducible and avoided Libermann, T., and Yarden, Y. (1983) CRC Crif. the use of large amounts of cold EGF as a Revs. Biochem. 14, 93-111. nonspecific binding control. The low back- 13. Cohen, S., Ushiro, H., Stoscheck, C., and Chinkers, M. (1982) J. Biol. Chem. 257, 1523-1531. ground binding of EGF in the RIA is probably a consequence of the assay possessing two lev- 14. Zick, Y., Kasuga, M., Kahn, C. R., and Roth, J. (1983) J. Biol. Chem. 258,75-80. els of specificity; the binding of EGF to the 15. Rubin, J. B., Shia, M. A., and Pilch, P. F. (1983) EGF receptor and the binding of the mAb to Nature (London) 305, 438-440. the EGF receptor. The total cpm bound in 16. Jacobs, S., Kull, F. C., Earp, H. S., Svoboda, M. E., Van Wyk, J. J., and Cuatrecasas, P. (1983) J. Biol. each assay using identical reaction mixtures Chem. 258,958 l-9584. was slightly higher for the RIA, probably be17. Cooper, J. A., Bowen-Pope, D. F., Raines, E., Ross, cause the washing procedure could be perR., and Hunter, T. (1982) Cell 31, 263-273. formed more rapidly, avoiding dissociation of 18. Cassel, D., Pike, L. J., Grant, G. A., Krebs, E. G., specifically bound ligand. and Glaser, L. (1983) J. Biol. Chem. 258,29452950. Finally, we have used the RIA to monitor 19. Baldwin, G. S., Kneael, J., and Monckton, J. M. ( 1983) EGF receptor purification by immunoaffinity Nature (London) 301, 435437. chromatography using mAb Rl coupled to 20. Baldwin, G. S., Stanley, I. J., and Nice, E. C. (1983) Sepharose (41). By using the same mAb in FEBS LETT. 153,257-261. 21. Stadtmauer, L. A., and Rozen, 0. M. (1983) J. Biol. solution to monitor EGF receptor concentraChem. 258,6682-6685. tion, any conditions which influenced the P. (1972) Proc. Nat/. Acad. Sci. USA 69, binding to the affinity column were detected 22. Cuatrecasas, 318-322. directly. Milligram quantities of EGF receptor 23. Carpenter, G. (1979) Life Sci. 24, 1691-1698. from A43 1 cells were obtained using this ap- 24. Waterfield, M. D., Mayea, E. L. V., Stroobant, P., Bennett, P. L. P., Young, S., Goodfellow, P. N., proach.
RADIOIMMUNOASSAY
FOR EPIDERMAL
Banting, G. S., and Ozanne, B. (1982) J. Cell. Biochem. 20, 149-161. 25. Watetield, M. D., Parker, P., and Mayes, E. (1983) Cell Biol. Int. Rep. 7, 535-536. 26. Schreiber, A, B., Lax, I., Yarden, Y., Eshhar, ‘Z., and Schlessinger, J. (198 1) Proc. Natl. Acad. Sci. USA n&,7535-7539. 27.
Schreiber, A. B., Libermann, T. A., Lax, I., Yarden, Y., and Schlessinger, J. (1983) J. Biol. Chem. 258, 846-853.
28.
29. 30. 31.
32.
Kawamoto, T., Sato, J. D., Le, A., Polikoff, J., Sate, G., and Mendelsohn, J. (1983) Proc. Natl. Acad. Sci. USA 80, 1337-1341. Richert, N. D., Willingham, M. C., and Pastan, I. (1983) J. Biol. Chem. 258, 8902-8907. Thorn, D., Powell, A. J., Lloyd, C. W., and Rees, D. A. (1977) Biochem. J. 168, 187-194. Smith, C. H., Nelson, D. M., King, B. F. Donohue, T. M., Ruzycki, S. M., and Kelley, L. K. ( 1977) Amer. .I. Obstet. Gynecol. 128, 190-196. Savage, C. R., and Cohen, S. (1972) J. Biol. Chem. 241, 7609-7611.
GROWTH
FACTOR
RECEPTOR
261
33. Savage, C. R., Inagami, T., and Cohen, S. (1972) J. Biol. Chem. 241,16 12-162 1. 34. Carpenter, G., King, L., and Cohen, S. (1979) J. Biol. Chem. 254,4884-489 1. 35. Anholt, R., Lindstrom, J., and Montal, M. (198 I) J. Biol. Chem. 256,4317-4387. 36. Nelson, N., Anholt, R., Lindstrom, J., and Montal, M. (1980) Proc. Natl. Acad. Sci. USA 77, 30573061. 37.
38. 39.
40.
41.
Burgess, A. W., Lloyd, C. J., and Nice, E. C. (1983) EMBO J. 2,2065-2069. Hillman, G., and Schlessinger, J. (1982) Biochemistrv 21, 1667-1672. Hoch, R. A., Nexo, E., and Hollenberg, M. D. ( 1980) J. Biol. Chem. 255, 10,737-10,743. Marquardt, H., Hunkapiller, M. W., Hood, L. E., Twardzik, D. R., DeLarco, J. E., Stephenson, J. R., and Todaro, G. J. (1983) Proc. Natl. dead. Sci. USA 80, 4684-4688. Downward, J., Yarden, Y.. Scrace, G., Totty, N., Stockwell, P., Ullrich, A., Schlessinger, J., and Waterfield, M. D. (1984) Nature (Londonj 307,521521.