Synthesis, Turnover, and Down-regulation of Epidermal Growth Factor Receptors in Human A431 Epidermoid Carcinoma Cells and Skin Fibroblasts*

Epidermal growth factor (EGF) receptors extracted with Triton X-100 from human skin fibroblasts and A431 epidermoid carcinoma cells rapidly lose EGF-binding activity precipitable with polyethylene glycol. The presence of concanavalin A which can cross-link and, thereby, aggregate the receptors, allowed quanti- tative recovery of the lost EGF-binding activity. Scatchard analysis of EGF binding of Triton X-100-solubi-lized receptors showed that A431 cells and skin fibro- blasts possess approximately 1.5 X 10‘ and 7 X lo4 EGF-binding sites/cell, respectively, which exhibit similar affinities for the ligand. The heavy isotope density-shift method was employed to determine whether differ- ences in rates of receptor synthesis or decay account for the large difference in number of receptors/cell between the two cell types. After shifting cells to medium containing heavy (I5N, 13C, and ‘H) amino acids, light and heavy receptors, solubilized from total cellular membranes, were resolved by isopycnic banding on density gradients and then quantitated. It was demon-strated that A431 cells synthesize EGF receptors at a rate 12 times faster than skin fibroblasts and that the half-life for receptor decay of A431 cells is somewhat longer (tlIz = 16 h) than that ( t l / ~ = 9 h) of fibroblasts. Down-regulation of cell surface and total cellular EGF-binding capacity in A431 cells occurs with a tl/z of 2-3 h and

Epidermal growth factor (EGF) receptors extracted with Triton X-100 from human skin fibroblasts and A431 epidermoid carcinoma cells rapidly lose EGFbinding activity precipitable with polyethylene glycol. The presence of concanavalin A which can cross-link and, thereby, aggregate the receptors, allowed quantitative recovery of the lost EGF-binding activity. Scatchard analysis of EGF binding of Triton X-100-solubilized receptors showed that A431 cells and skin fibroblasts possess approximately 1.5 X 10' and 7 X lo4 EGFbinding sites/cell, respectively, which exhibit similar affinities for the ligand. The heavy isotope density-shift method was employed to determine whether differences in rates of receptor synthesis or decay account for the large difference in number of receptors/cell between the two cell types. After shifting cells to medium containing heavy (I5N, 13C, and 'H) amino acids, light and heavy receptors, solubilized from total cellular membranes, were resolved by isopycnic banding on density gradients and then quantitated. It was demonstrated that A431 cells synthesize EGF receptors at a rate 12 times faster than skin fibroblasts and that the half-life for receptor decay of A431 cells is somewhat longer (tlIz = 16 h) than that ( t l /~ = 9 h) of fibroblasts.
Down-regulation of cell surface and total cellular EGF-binding capacity in A431 cells occurs with a t l / z of 2-3 h and results in a 70-83% decrease in receptor level in 12 h. Scatchard analysis revealed that these changes in EGF binding were due to an alteration of receptor number and not EGF-binding affinity. Rates of EGF receptor synthesis and inactivation/decay were determined by the heavy isotope density-shift method. No change in the rate of receptor synthesis occurred as a consequence of EGF receptor down-regulation. Downregulation, however, caused a decrease in receptor half-life from 16 to 4.5 h. These results indicate that EGF-dependent regulation of EGF receptor level in A431 cells involves an alteration of the rate of receptor inactivation.
The binding of epidermal growth factor to its specific receptors on the plasma membrane of target cells triggers an array of biological responses including both short term and long term effects. The short term effects which are initiated within minutes after binding include ruffling of the plasma membrane * This work was supported by Research Grants from the American Cancer Society (BC0387) and the National Institutes of Health (AM-14574). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. + Postdoctoral Fellow of the American Cancer Society.
(1) and activation of several transport systems (2-4). The long term effects of EGF,' however, require continuous occupancy of the receptor by the ligand for 12-16 h (5); these effects include induction of DNA, RNA, and protein synthesis and mitogenesis (6).
Since the response to EGF is determined by the number of receptors occupied, the magnitude of the effect is likely to depend both upon the concentration of the ligand and the number of fLnctiona1 receptors the cell possesses. There is now compelling evidence that cells have the capacity to regulate their responsiveness to peptide hormones by modulating the levels of specific receptors. Ligand-induced receptor downregulation is a well-documented mechanism for the modulation of the sensitivity of a cell to homologous hormone (7). Ligand-induced "down-regulation'' of receptors has been demonstrated in culture for many peptide hormones (7), including EGF (8-10). In vivo studies have shown that the decreased cellular sensitivity to certain hormones, notably insulin, in various disease states is associated with a reduced level of functional cell surface hormone receptors (11).
Unambiguous approaches have not been used to investigate the mechaqism(s) by which the level of cellular EGF receptors is regulated, for example, to determine whether changes in receptor level are due to altered rates of receptor synthesis or degradation. The heavy isotope "density-shift'' technique has been employed sucessfully to determine whether altered synthesis or turnover of insulin receptors is responsible for the changes in receptor level brought about by preadipocyte differentiation (12,13), insulin-induced down-regulation (14-16), and glucocorticoid-induced up-regulation (16). This approach permits the identification of newly synthesized receptor after shifting cells to medium containing amino acids that are >95% enriched in nitrogen-15, carbon-13, and deuterium (I'N, '"(2, 'H). The incorporation of heavy amino acids into receptor protein increases the density of newly synthesized receptor. Thus, "new heavy" and "old light" receptor, solubilized in detergent, can be resolved by isopycnic density-gradient centrifugation. The relative positions and amounts of heavy and light receptor in the gradient are then determined. By foliowing the amounts of heavy and light receptor present at increasing times of exposure of the cells to heavy amino acids, the rates of new heavy receptor synthesis and old light receptor decay can be calculated.
In the present investigation, the heavy isotope density-shift method was used to investigate the control of EGF receptor level in human cells. Human A431 epidermoid carcinoma cells and skin fibroblasts, which possess about 1.5 X lo6 and 7 x ' The abbreviations used are: EGF, epidermal growth factor; Con A, concanavalin A; Hepes, 4-(2-hydroxyethyl)-l-piperazineethanesulfonic acid; PBS, phosphate-buffered saline.

11489
Synthesis, Turnover, and Down-regulation of EGF Receptors lo4 EGF-binding sites/cell, respectively, were employed. Two questions were addressed: 1) Do A431 cells maintain a higher steady state receptor level than fibroblasts because of a higher rate of receptor synthesis or because of a lower rate of receptor inactivation/decay? and 2) Is EGF-induced "down-regulation" of EGF receptor level the result of an altered rate of receptor synthesis or receptor inactivation/decay?

EXPERIMENTAL PROCEDURES"'
Gel I CUI t u r e ~r r v l o~~l y been SubJected t o &he r a p l d EGF debindlng pmtocol (below).

RESULTS
lz5I-EGF Binding and Assay of E G F Receptors Solubilized from Human A431 Cells and Skin Fibroblasts-EGF receptors are quantitatively extracted by 1% (v/v) Triton X-I00 from total cellular membranes isolated from human A431 epidermoid carcinoma cells and fibroblasts. At least 90% of the '"I-EGF-binding activity originally present in the membranes is recovered in the 230,000 X g supernatant, leaving less than 5% in the pellet (results not shown). Soluble EGFbinding activity, measured by the method of Carpenter (22), decayed rapidly at 4 "C (about 90% in 24 h) and somewhat more slowly at -20 "C (about 85% in 2 days). Moreover, when subjected to equilibrium centrifugation in CsCl density gradients (initial CsCl concentration of 2.5 M ) , a procedure to isopycnically band the EGF receptor, all 12'I-EGF binding activity was lost. Since this procedure was needed to resolve and quantitate light and heavy receptors after labeling with heavy ("N, "C, *H) amino acids, a modification of the "' 1-EGF-binding assay was sought to circumvent this problem.
It was discovered that the inclusion of Con A in the EGFbinding assay mixture restored "'I-EGF-binding activity lost during storage or isopycnic banding of solubilized EGF receptor. The concentration dependence of Iz5I-EGF binding by soluble EGF receptor from human A431 epidermoid carcinoma cells that had been stored at -20 "C is shown in Fig. 1. Since maximal EGF binding occurred at the level of 100 pg of Con A/assay, this level was used in all subsequent studies.
The possibility was considered that tetravalency of Con A is required for its stimulatory effect in the binding assay perhaps by promoting aggregation of soluble EGF-receptor complexes through cross-linking of their oligosaccharide moieties. It is evident that tetravalency is required for the Con A effect since succinylated Con A, which can only bind mannosyl groups bivalently, is relatively ineffective in the EGF-binding assay (Fig. 1). The inclusion of 200 mM a-methyl mannoside, an excellent ligand for the lectin, prevents the Con A effect in the EGF-binding assay (results not shown).
An additional experiment supported this view. Soluble re-6 0 r I 1 Lectin Added.pg

FIG. 1. Effect of concanavalin A and succinylated concanavalin A on "'I-EGF binding to and precipitation with solubi-
lized receptor in the polyethylene glycol receptor precipitation assay. Triton X-100-solubilized extracts of total cellular membranes from 0.25 plate (6 cm) of A431 cells were incubated with 2 nM Iz5I-EGF in the absence and presence of the indicated levels of Con A or succinylated Con A under the conditions of the soluble EGF receptorbinding assay. The "'I-EGF-receptor complex was then precipitated with polyethylene glycol. See "Experimental Procedure" for details. ceptor from A431 cells was incubated with l2'1-EGF either in the presence or absence of 100 pg of Con A; '"I-EGF-receptor complexes were precipitated by polyethylene glycol only when Con A was present (Table I). When the supernatant from the sample without Con A (after polyethylene glycol addition and centrifugation) was subsequently treated with Con A, most of the '"I-EGF-receptor complexes were precipitated. Taken together these results indicate that Con A cross-links and aggregates EGF-receptor complexes, and thereby facilitates their precipitation by polyethylene glycol.
The dependence upon Con A of the precipitation of "'I-EGF-receptor complexes occurred over a wide range of EGF receptor concentrations in both human A431 and skin fibroblast extracts (Fig. 2, A and B ). Since the curves are biphasic, all subsequent '"I-EGF-binding assays were conducted in the linear range, i.e. below 60 pg of soluble membrane protein from A431 cells ( Fig. 2A) and below 20 pg of protein from fibroblasts ( Fig. 2B). It is not understood why the receptor concentration dependence is biphasic, but it is possible that an inhibitor or competitor of EGF binding is present in the solubilized membrane preparations.
The characteristics of specific '"I-EGF binding to Triton X-100-solubilized receptors from human A431 cells and skin TABLE I Effect of concanavalin A on theprecipitability of '''I -EGFreceptor complexes by polyethylene glycol Solubilized receptor from A431 cells (60 pg of protein) was incubated with 1 nM '"I-EGF for 45 min at 23 "C either in the presence (A) or absence (B) of 100 pg of concanavalin A. The binding assay mixture was then "precipitated' with polyethylene glycol (PEG) as described under "Experimental Procedures" and "'I-EGF binding determined. In the case (C) where Con A was not present during the initial incubation with l2'1-EGF, the lectin was added to the PEGcontaining supernatant after centrifugation and the resultant mixture was further incubated for 10 min at 23 "C and then recentrifuged. All results are corrected for nonspecific '2sII-EGF binding. fibroblasts are compared in Fig. 3, A and B, respectively. The binding isotherms for the receptors of both cell types are similar, and the Scatchard plots are curvilinear in both cases. From the limiting slopes in the high affinity region of the respective Scatchard plots, the apparent equilibrium binding constants are estimated to be hn = 1.5-2.5 x 10"' M. The total number of solubilized EGF-binding sites, determined by extrapolation to the x axis of the Scatchard plots, were 1.5 X loti and 7 X lo4 sites/cell for A431 cells and skin fibroblasts, respectively. The number of cell surface EGF-binding sites/ cell (results not shown) represents 85-95% of the total number of sites/cell estimated from total detergent-sollbilized receptor and is in good agreement with the number Veported by others (23-25) for surface sites in these cell types. It appears, therefore, that the intracellular pool of active EGF receptors constitutes a small fraction, 5-15%, of the total number of receptors/cell.
Isopycnic Banding of Triton X-100-solubilized EGF Receptors in CsCl Density Gradients-To determine the rates of synthesis and decay of EGF receptors by the heavy isotope density-shift method, it was necessary to establish conditions for the isopycnic banding and quantitation of the receptor in CsCl gradients. The procedure employed was similar to that developed by us for the insulin receptor (12). Briefly, EGF receptors were quantitatively extracted (390%) from total cellular membranes with 1% Triton X-100. The solubilized receptors were then mixed with CsCl and centrifuged at 4 "C for 18 h to achieve isopycnic equilibrium. The CsCl gradients were then fractionated and the EGF receptor located and quantitated by determining the "'I-EGF-binding capacity of each fraction. Fig. 4, A and B (insets), shows the banding profiles in CsCl density gradients for EGF receptors from A431 cells and skin fibroblasts, respectively. The recovery of binding activity in the gradients was essentially quantitative (390%) when Con A was included in the binding assays. In each case a receptor peak was obtained which banded at a density corresponding to a refractive index, $?, of 1.3620 (vertical dashed line L in Fig. 4). As will be shown later "heavy" receptor from cells cultured in medium containing amino acids labeled with 15N, "(2, and 'H bands at a position data are shown in the insets. '""IEGF-binding assays were conducted as described in the legend to Fig. 2.

Synthesis, Turnover, and Down-regulation of EGF Receptors
in the gradient corresponding to a refractive index of 1.3675 (uertical dashed line H in Fig. 4). Of interest is the fact that both light and heavy EGF receptors banded a t positions of lower density than the corresponding light and heavy forms of the insulin receptor ($' = 1.3635 and 1.3690, respectively; Refs. 12, 14, and 16). Also, in addition to the main peak, a shoulder of greater density constituting 4 0 % of the total activity was consistently observed in gradients of A431 cell extracts.
An empirical relationship exists, as for the insulin receptor (13, 14), between peak height and peak area for the Iz5I-EGFbinding activity. This allows quantitation of the relative amounts of light and heavy EGF receptors in CsCl density gradients. Fig. 4, A and B, shows that for EGF-binding activity of A431 cells and fibroblasts, peak area and height are linearly related; thus, peak area is determined by multiplying peak height by the slope, i.e. 7.1 and 6.5 for human A431 cells and fibroblasts, respectively. For gradients with both heavy and light receptor, peak height for the heavy receptor is used directly to determine peak area since there is minimal contribution of light EGF-binding activity to heavy EGF-binding activity at the peak point for heavy EGF receptor (Fig. 4, A  and B, insets). Light EGF-binding activity is determined by subtracting heavy peak area from the total area under both peaks. This approach has been used successfully to estimate the relative amounts of heavy and light insulin receptors from several cell types on CsCl density gradients (13,14,16).

Heavy Isotope Density-Shift Analysis of the Synthesis and Decay of EGF Receptors in Human A431 Cells and Skin
Fibroblasts-To distinguish between newly synthesized and previously synthesized EGF receptors, confluent cell monolayers were shifted from medium containing light ("N, '%, 'H) amino acids to medium containing heavy ("N, '"C, 'H) amino acids. At various time intervals after the shift to heavy medium (3, 6, 9, 12, and 16 or 20 h) heavy and light EGF receptors were solubilized from total cellular membranes with Triton X-100, resolved on CsCl density gradients, and quantitated. Typical profiles of heavy and light receptors from both cell types 3,9, and 16 or 20 h after the shift are shown in Fig. 5 (A-F). The size of the light receptor peak decreased with time as the size of the heavy receptor peak increased concomitantly.
After correction for nonspecific "'I-EGF binding, the relative amounts of heavy and light EGF receptors were calculated from the peak areas as described above. These results, plotted in the form of progress curves, illustrate the rates of formation of newly synthesized heavy receptor and the decay of previously-synthesized receptors for human A431 cells (Fig.  SA) and skin fibroblasts (Fig. 6 B ) .
Replots of the light receptor decay data of Fig. 6, A and B, in semilogarithmic form show that the inactivation of EGF receptors in A431 cells and fibroblasts is a fiist order process (Fig. 6C). The half-lives for receptor decay differ somewhat, 16 h for A431 cells and 9 h for skin fibroblasts. The kinetic constants for the rates of synthesis and decay of EGF receptors in the two cell types are summarized in Table 11. While the rates of receptor inactivation differ by less than 2-fold, the rate of receptor synthesis by A431 cells is more than 12-fold faster than by fibroblasts. These results were quite reproducible. The mean half-life for receptor decay was 18.4 * 2.2 h (S.E.) for five experiments with A431 cells, and 10.0 k 1.2 h for two experiments with fibroblasts. Together the differences in the rates of both processes, i.e. receptor synthesis and decay, account for the more than 20-fold higher level of EGF receptors in A431 cells than in skin fibroblasts in the steady state determined by Scatchard analysis (Fig. 3, A and B ) .    Fig. 10C for B.

'Estimated from the slopes of light receptor inactivation in Fig.
' Calculated from ks/ku = total number of sites/cell at steady state. ks was estimated from the limiting initial slope of heavy receptor synthesis.
A431 Cells-Exposure of A431 epidermoid carcinoma cells to GGF causes a reduction, i.e. down-regulation, in the level of cell surface EGF receptors. In the presence of 50 nM EGF, cell surface EGF-binding capacity is reduced to 30% of its original level in 12 h, the tl,' for the process being 2-3 h (Fig. 7 A ) . The rate of loss of receptor extractable with Triton X-100 from total cellular membranes ( i e . internal as well as cell surface membranes) follows similar kinetics, dropping to 30% of control cells in 12 h (Fig. 7 A ) . Down-regulation is concentration-

Synthesis, Turnover, and
Down-regulation of EGF Receptors dependent, half-maximal down-regulation in 12 h occurring at about 1 X IO-* M EGF (Fig. 7 B ) .
To determine whether this modulation of EGF-binding capacity was the result of a change in absolute number of receptors, affinity of the receptor for EGF, or both, the EGFbinding activities of total soluble receptors from control and down-regulated cells were compared. Cells were incubated for 12 h either in the absence or in the presence of 10 or 50 nM EGF to induce down-regulation. The EGF-binding isotherms for receptors extracted with Triton X-100 from total cellular membranes are compared in Scatchard plots in Fig. 8. It is evident that the loss of EGF-binding capacity during downregulation is due to a reduction in the number of receptors with little or no change in the affinity of the receptors for the ligand. Receptor levels for A431 cells down-regulated with 10 and 50 nM EGF, estimated from the Scatchard plots (Fig. a), were 40 and 17%, respectively, of untreated controls. were calculated as described in the text using the data shown and described in the legend to Fig. 5. C shows semilogarithmic plots for the decay of light receptor in both cell types using the results from A and B. The synthetic and degradative rate constants for the two cell types determined from these results are summarized in Table 11.

20
To assess the effect of EGF-induced down-regulation on the rates of synthesis and turnover of EGF receptors, the heavy isotope density-shift method was employed. After reaching confluence, A431 cells were exposed to medium containing 50 nM EGF for 12 h to induce the down-regulated state. The cells were then shifted to medium containing heavy (15N, 13C, 'H) amino acids with or without 50 nM EGF. At various times (0, 3, 6, 9, 12, and 18 h) after the density shift, receptor was extracted from total cellular membranes and banded isopycnically in CsCl density gradients, and light and heavy receptor were quantitated as described above. Typical banding profiles for light and heavy EGF receptors from control and EGF Kinetics and EGF concentration dependence of EGF-induced down-regulation of the "'1-EGFbinding capacity of A431 cells. In A , cell monolayers were exposed to 50 nM EGF for the time indicated and were then subjected to the rapid acidic wash procedure to remove any bound EGF. lZ5I-EGF binding to cell surface or total cellular Triton X-100-solubilized receptor was determined as described under "Experimental Procedure." The experiment shown was repeated three times with similar results. In B, cells were exposed to the indicated concentrations of EGF for 12 h after which "'1-EGF binding to total cellular Triton X-100-sohbilized receptor was determined as described above. The amounts of heavy and light EGF receptor at each time point were calculated as described under "Experimental Procedures" and the text using the data shown and described in the legend to Fig. 9. C shows semilogarithmic plots for the decay of light receptor in control and down-regulated cells using the results from A and B. The synthetic and degradative rate constants for control and down-regulated cells are summarized in Table 11.
down-regulated cells before and at 18 h after the shift to "heavy" medium are shown in Fig. 9.
Comparison of the relative amounts of light receptor (represented by the peak areas in Fig. 9, A and B a t 0 h), just before the density shift, shows that chronic exposure to EGF caused the expected down-regulation of total cellular EGF receptor level. Eighteen hours after the density shift, however, the size of the light receptor peak, relative to the heavy receptor peak, was markedly lower in EGF down-regulated cells than in control cells (Fig. 9, D and C ) suggesting that an increased rate of receptor decay had occurred. From the integrated areas of the light and heavy receptor peaks at all sampling times after the density shift, progress curves for the decay of light receptor and the formation of newly synthesized heavy receptor were generated (Fig. 10). These kinetic plots show that chronic exposure of A431 cells to EGF shortens the time required for half-replacement of light receptor with heavy receptor from about 16 h (Fig. 1OA) to about 4 h (Fig.  10B). The half-lives for the rates of inactivation of light receptors in control and EGF down-regulated cells, determined from semilogarithmic plots (Fig. lOC), were 16 and 4.5 h, respectively. EGF-induced down-regulation had little effect on the zero order rate constant for the synthesis of heavy receptor, ks = 46,000 and 42,000 receptor sites/cell/h for control and down-regulated A431 cells, respectively. It is concluded, therefore, that control of EGF receptor level by EGF in A431 cells is exerted at the level of receptor inactivation, rather than receptor synthesis.

DISCUSSION
When maintained under comparable cell culture conditions different cell types have been found to express widely differing levels of EGF receptor (Refs. 23-25 and Table II), ranging from 50,000 to several million receptor sites/cell. In addition, the steady state level of EGF receptor for a particular cell type can be modulated by changes in nutritional/physiological state. For example, exposure of cells to EGF causes ligandinduced down-regulation of EGF receptor level . Differences such as these in the level of cell surface receptors are thought to be an important factor in determining the responsiveness of the cell to EGF. Few investigations, however, have focused on the mechanism(s) by which the level of active EGF receptors is varied. At steady state the total number of functional EGF receptors in a cell will be determined by the relative rates of synthesis and decay, ks and k~, of active receptor. Thus, a change in either rate constant will produce an alteration of cellular receptor level.
The specific objectives of the present investigation were to ascertain: 1) how the levels of EGF receptors are maintained at 1.5 million/cell in human A431 cells but at only 70,00O/cell in human fibroblasts when cultured under comparable conditions and 2) whether EGF-induced down-regulation of EGF receptor number in A431 cells is due to altered rates of synthesis or degradation of receptor.
The heavy isotope density-shift method, employed in these studies to measure rates of active receptor synthesis and degradation, has several important attributes. Labeling of cells with heavy amino acids to measure simultaneously the synthesis and turnover of receptor is a benign procedure and does not require the use of inhibitors. Although inhibitors of protein synthesis are commonly used to follow receptor decay in the absence of receptor synthesis, the inhibition of protein synthesis is known to cause other impairments of cellular function. We have shown (26), for example, that inhibitors of protein synthesis, such as cycloheximide or puromycin, reduce the rate of insulin receptor inactivation in 3T3-Ll adipocytes by 3-4-fold. It appears that the continued synthesis of a shortlived protein(s) is required for the rate-limiting step in the turnover of the insulin receptor (26).
It should be emphasized that the density-shift method follows the rate at which EGF-binding activity of the receptor is lost and, thus, measures the rate-limiting step in the receptor inactivation process. This step is of particular physiological importance in the pathway leading to the degradation of receptor protein, since it is at this point that receptor function is lost. It remains to be determined whether EGF receptor inactivation is synonymous with degradation of receptor protein. This does not seem likely, however, since blocking lysosomal proteolysis with chloroquine: which causes the intracellular accumulation of the ligand-binding activity of certain other receptors (27,28), has virtually no effect on cellular EGF-binding capacity. These results are similar to those observed with the insulin receptor in chick heptocytes (29).
The success of the density-shift technique is dependent upon quantitative assay of the EGF receptor. Two major problems were encountered in preliminary experiments, however. First, the EGF-binding activity was found to be unstable in crude cellular extracts, possibly due to proteolysis of receptor. Secondly, detectable activity was greatly reduced in the presence of a variety of salts, including CsCl, and thus it seemed that the isopycnic banding procedure could not be applied for the EGF receptor. We observed, however, that both of these problems could be circumvented by including Con A in the EGF-binding assay mixture. This modification of the immobilized-lectin assay of Nexo et al. (30) allowed for the quantitative estimation of EGF-binding activity, even in the presence of CsC1. A number of experiments indicated that Con A probably served to increase the precipitability of the 4 M . N. Krupp, D. T. Connolly, and M. D. Lane, unpublished observations. 125 I-EGF-receptor complexes. It is of interest that in whole cells, both Con A (31) and succinylated Con A (32) were effective inhibitors of EGF binding to the receptor. Thus, the lectin, whether in solution or immobilized (30), evidently exerts opposite effects on the EGF receptor depending upon whether the receptor is in the cell membrane or in solution.
The results of our present investigation on the control of the synthesis and inactivation/decay of the EGF receptor using the heavy isotope density-shift technique are summarized in Table 11. The 20-fold greater number of cellular EGF receptors in human A431 carcinoma cells than in human fibroblasts was found to be due both to a 12-fold higher rate of receptor synthesis and to a slightly lower, -l.S-fold, rate of receptor inactivation. The estimate of the synthetic rate constant, ks, was determined either by measurement of the limiting slopes of the progress curves (Fig. 6, A and B ) for heavy receptor synthesis (6,300 and 72,000 site/h/cell) or was calculated from the equation R = ks/ku (5,320 and 64,500 sites/ h/cell). Although the absolute values obtained by the two methods of calculation are slightly different, both methods imply that the A431 cells synthesize receptor a t a 12-fold faster rate than the fibroblasts.
Several investigators have reported turnover rates for the EGF receptor. Aharonov et al. (8) estimated that EGF-binding activity in 3T3 fibroblasts decays with a half-life of 6 h when protein synthesis is blocked by cycloheximide. Carpenter (33) has shown that in human fibroblasts starved for histidine in medium supplemented with L-histidinol, the EGF receptor turns over with a half-life of 14.5 h. Bhargava and Makman (34) observed a receptor half-life of 6 h in IMR-90 human lung fibroblasts when protein synthesis was blocked with cycloheximide or when glycosylation was inhibited with tunicamycin. Treatment of the cells with either actinomycin D or captothesin to inhibit transcription caused the receptor to decay with a 23-h half-life. As pointed out above, the validity of turnover studies conducted in the presence of inhibitors is open to question.
The EGF receptor, like many other peptide hormone receptors (8-10) undergoes ligand-induced down-regulation of cellsurface EGF-binding capacity. In A431 cells, the down-regulation is rapid ( t 1 / 2 = 2-3 h; Fig. 7A), and EGF concentrationdependent (ED, = 10 nM, Fig. 7 B ) . The extent of downregulation varied from 70 to 83% in five different experiments (some results not shown). These results are in accord with those of others (9, 10) for A431 cells and other cell types. Wrann and Fox (9) have shown that down-regulation is rapid and EGF concentration-dependent in A.431 cells. Others have shown that in 3T3 fibroblasts (8), human fibroblasts (35), and adult rat liver parenchymal cells (36) that EGF receptor downregulation occurs with a time course similar to the A431 cells.
It is of interest that EGF-induced down-regulation in the steady state led not only to a reduction of cell surface receptor level, but also to a reduction in the number of total cellular receptors, i.e. receptors extracted with Triton X-100 from total cellular membranes (Fig. 7 A ) . Thus, down-regulation of EGF receptors in A431 cells, like down-regulation of insulin receptors in 3T3-Ll adipocytes (15) and in 3T3-C2 fibroblasts (16), is due to a decrease in the number of active receptors/ cell rather than to the redistribution of receptors from the cell surface to an intracellular compartment (14). The greatly reduced level of EGF receptors in the down-regulated state in A431 cells is, therefore, entirely attributable to an increased rate of receptor inactivation.
The heavy isotope density-shift method used in this investigation has identified receptor inactivation as the rate-limiting step modulating EGF-induced receptor down-regulation. As shown in Table 11, little change in the rate of receptor synthesis (ks = 43,000-46,000 sites/cell/h for control cells and ks = 37,000-42,000 sites/cell/h for down-regulated cells) occurred as a consequence of down-regulation. However, EGFinduced down-regulation caused a shortening of the half-life for receptor inactivation from 16 to 4.5 h. Thus, chronic exposure of A431 cells to EGF causes an increased rate of receptor inactivation/decay. The mechanism by which receptor inactivation is increased remains obscure. As discussed above the initial rate-limiting step in receptor inactivation probably does not involve lysosomal proteolysis, since chloroquine, which blocks lysosomal protease action, does not lead to the accumulation of cellular receptor. In view of the fact that the EGF receptor is subject to reversible EGF-activated phosphorylation (37), it is possible that the phosphorylation state of the receptor may determine its turnover rate. Further investigations will be necessary to determine the exact mechanism by which EGF induces EGF receptor inactivation.