Transforming Growth Factor PI Inhibits Epidermal Growth Factor Receptor Endocytosis and Down-regulation in Cultured Fetal Rat Hepatocytes*

Incubation of fetal rat hepatocytes (FRH) with trans- forming growth factor Dl (TGF-D1) resulted in growth arrest and a biphasic effect on epidermal growth factor (EGF) receptor. After 2 h of exposure, EGF receptor (EGFR) was reduced by 43%. From 6 to 24 h, TGF-Dl exposure resulted in progressive increase in EGFR up to 74% over control. The increased binding was due to increase in high affinity EGF binding sites. FRH grown in medium containing EGF exhibited down-regulated EGFR with loss of high affinity EGF binding sites. With TGF-O1 exposure, high affinity EGFR was not down-regulated by EGF. Since down-regulation of EGFR involves internalization, the kinetics of EGF receptor-mediated endocytosis were examined. In TGF-D1-exposed FRH, EGF endocytosis was inhibited, with a reduction in the first order rate constant for the process from 0.078 to 0.043 min”. Despite inhibition of growth, receptor down-regulation, and EGF endocytosis after TGF-O1 exposure, EGF-induced receptor autophosphorylation was preserved as demonstrated by [32P]phosphate-labeling of immunoprecipitated EGFR. These observations provide direct evidence that TGF-D1 regulates growth of fetal cells. Further, they suggest that TGF-D1 regulates endocytosis of EGF and possibly of other ligands. values experiments.

Transforming growth factor beta (TGF-P)' is a ubiquitous, multifunctional growth factor initially identified by its ability to induce growth and a transformed phenotype in fibroblasts (1,2). For epithelial cells or cell lines, including hepatocytes, TGF-P acts as a growth inhibitor (3). TGF-/3 appears to regulate hepatocyte growth during fetal and neonatal liver development and during liver regeneration. During fetal liver * This work was supported by grants from the Alcoholic Beverages Manufacturers Research Council, the Department of Veterans Affairs (Research Service, Audie L. Murphy Memorial Department of Veterans Affairs Medical Center, San Antonio, TX), and by Research Scientist Development Award 5K02AA00121-03 from the National Institute on Alcohol Abuse and Alcoholism (GH). 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.
j To whom correspondence should be addressed Dept. of Medicine, The University of Texas Health Science Center, 7703 Floyd Curl Dr., San Antonio, TX 78284-7878.
The abbreviations used are: TGF, transforming growth factor; EGF, epidermal growth factor; EGFR, epidermal growth factor receptor; FRH, fetal rat hepatocytes; HEPES, 4-(2-hydroxyethyl)-l-piperazineethanesulfonic acid; K,, endocytotic rate constant (first order rate constant for endocytosis); SDS, sodium dodecyl sulfate; WE, William's E medium. development, TGF-/3 mRNA is detectable as early as day 10, but is confined to mesenchymal cells such as marrow elements (4-7). In vivo models of liver regeneration after partial hepatectomy or chemical hepatocellular necrosis exhibit TGF-P mRNA induction with a peak at the time of maximal cell growth (8,9). Exogenous TGF-0' acts to inhibit such regeneration (10). As in the fetal liver, the cells which express the message are nonparenchymal cells (8,9). Thus, TGF-p appears to function as an externally produced modulator of hepatocyte growth. TGF-/3 also is a growth inhibitor for primary cultures of hepatocytes, both from normal and regenerating liver (3, 11,12). We have used primary cultures of fetal rat hepatocytes (FRH) as an in vitro model for investigating regulation of liver cell growth (13). In contrast to adult hepatocytes, FRH grow rapidly in culture without prior treatment needed to prime the cells for growth. FRH also are maintained in defined medium, dependent only on added epidermal growth factor (EGF), which simplifies the demonstration of growth-promoting and -inhibiting influences. The in. vivo observations of TGF-P mRNA expression in the fetal liver and its proposed role as a negative regulator of liver cell growth during liver regeneration prompted these investigations on the effect of TGF-Pl on fetal hepatocyte growth in uitro and on the metabolism of EGF in these cells.

MATERIALS AND METHODS
Cell Culture, Measurement of Cell Growth, EGF Binding to Fetal Liver Cells, EGF Internalization, and EGF Receptor Up-regulation-Cell culture, measurement of cell growth, EGF binding to fetal liver cells, EGF internalization, and EGF receptor up-regulation were performed as previously described (13). TGF-0, (Collaborative Research Inc., Bedford, MA) was included in the culture medium at various concentrations, usually 2 ng/ml, either with or without EGF, for periods up to 24 h as described for the individual experiments.
Measurement of DNA Synthesis-Replicative (semiconservative) DNA synthesis was assayed as hydroxyurea-inhibitable [3H]thymidine incorporation into DNA. Paired groups of 35-mm dishes of fetal hepatocytes were incubated in quintuplicate with culture medium containing 20 pCi/ml [3H]thymidine (ICN Radiochemicals, Imine, CA) with or without 5 mM hydroxyurea for 4 h at 37 "C. The cells were lysed by scraping in cold 5 mM Tris buffer, pH 7.0, followed by sonication at 30 watts for 2 min (Fisher Sonic Dismembranator). A portion of the sonicate was assayed fluorimetrically for DNA content using 4',6-diamidino-2-phenylindole dihydrochloride (14) and a portion subjected to precipitation with 10% trichloroacetic acid at 0 "C overnight, followed by four washes with 10% trichloroacetic acid. 4',6-Diamidino-2-phenylindole dihydrochloride was obtained from Aldrich. The resulting pellet was dissolved in Redi-Solve (Amersham Corp.) and counted in a liquid scintillation counter (Beckman Instruments). DNA synthesis was calculated as total minus hydroxyureainhibitable disintegrations/min of 3H incorporated per pg of DNA.
Results are expressed as percent of control values to allow comparison of different experiments.
Autophosphorylation and fzPlP04 Labeling of EGF Receptor-

TGF-8 and EGF Endocytosis 13239
Fetal hepatocytes were grown in 75-cmZ Primaria'" flasks in WE plus 2 ng/ml EGF with or without TGF-PI exposure for 24 h. The culture medium was removed and replaced with phosphate-free RPMI medium containing 50 pCi/ml [3zP]P04 (Amersham) lacking EGF, either with or without 2 ng/ml TGF-P1, and the cells were incubated at 37 "C for 2 h. 0.5 ml of WE medium with or without EGF (to achieve a final concentration of 100 ng/nl) was then added for an additional 2-30 min at 37 "C. Subsequently, the cells were washed twice with phosphate-buffered saline (5 mM sodium phosphate, 150 mM NaCl, pH 7.4) and were collected in 1 ml of cold lysis buffer (50 mM HEPES, 5 mM EDTA, 1% Nonidet P-40, 0.1 mM phenylmethylsulfonyl fluoride, 10 mM sodium pyrophosphate, 100 p M NaVO,, 1 mM NazATP (Sigma), pH 7.0) by scraping with a rubber scraper. The cell lysate was briefly sonicated, rocked at 4 "C for 15 min, then centrifuged in an Eppendorf centrifuge for 5 min to remove undissolved material. An aliquot was removed for protein assay, and the supernatant was then used for immunoprecipitation. Immunoprecipitation of EGF Receptor-Immunoprecipitation was performed essentially as previously described (13). 10 p1 of 10% bovine serum albumin was added per ml of cell lysate, then 1 pg/ml nonimmune mouse IgG and 100 pl of a 1:l slurry of Affi-Gel/protein A in lysis buffer were added. The solution was rocked for 1 h at room temperature. The beads were removed by centrifugation in an Eppendorf centrifuge for 2 min. The supernatant was transferred to a new tube. Monoclonal anti-EGFR (EGF-R1, Amersham) or monoclonal anti-phosphotyrosine (Py-20, ICN Immunobiologicals) was added to a final concentration of 1 pg/ml in lysis buffer and incubated at 4 "C for 2 h, followed by addition of 100 p1 of a 1:l slurry of Affi-Gel/ protein A beads in lysis buffer. The mixture was rocked overnight at 4 "C. The beads were washed twice with the lysis buffer, then three times with phosphate-buffered saline. The immunoprecipitated protein was eluted by boiling the beads with 2% SDS, 2% 2-mercaptoethanol, 60 mM Tris-HC1, pH 6.8, for 3-5 min (for EGF-R1) or with 2% SDS, 2% 2-mercaptoethanol, 100 mM phenylphosphate, 60 mM Tris-HC1, pH 6.8, for 3-5 min (for Py-20), and then subjected to electrophoresis on 6% polyacrylamide gels in Laemmli buffer system (15). The gel was fixed in 10% acetic acid and 25% methanol, dried, and autoradiographed at -70 "C with Kodak RP X-Omat film. Densitometry was performed by scanning the autoradiogram on a flatbed scanner (Microtek MSF-300G, Microtek Labs, Torrance, CA), and the digitized image was edited using "Digital Darkroom" (Silicon Beach Software, San Diego, CA). The integration of the peaks was performed with the program "Image 1.05" (shareware). All manipulations were performed on a Macintosh IIx (Apple Computer, Sunnyvale, CA).
General Methods-Protein determinations were by the method of Lowry (16). Differences in experimental data were detected by a oneway ANOVA analysis with p < 0.05 considered to be significant.
Curve fitting of binding isotherms and internalization experiments was with a generalized least square method for one-component, twocomponent, or first order rate model.

RESULTS
T h e effect of TGF-P1 on EGF-stimulated growth of FRH was dependent on the concentration of TGF-0, used. At very low concentrations (5-10 pg/ml, 0.2-0.4 PM), TGF-P, appeared to enhance EGF-induced cell replication (Fig. 1). Although statistically significant increases in cell growth were seen in two experiments, the pooled results of all experiments did not achieve statistical significance ( p = 0.08). As TGF-& concentrations were increased above 100 pg/ml, growth was significantly inhibited. Maximal growth inhibition was seen at TGF-PI concentrations of 1.5 ng/ml (60 PM) or greater. Addition of increased amounts of EGF up to 50 ng/ml (25 nM) did not overcome this growth inhibition. Assay of the effects of TGF-PI on DNA synthesis as assayed by hydroxyurea-inhibitable [3H]thymidine incorporation into DNA paralleled the results obtained with cell counting (data not shown). A trend toward stimulation of cellular DNA synthesis was seen at TGF-P1 concentrations less than 1 PM, but this did not reach statistical significance. As TGF-P1 concentration was increased above 1 ng/ml (40 p~) , inhibition of DNA synthesis was observed. At 2 ng/ml TGF-P,, [3H]thymidine incorporation was 43% of control, while the same concentra- Jere cultured in WE medium with 2 ng/ml EGF for 24 h, then exposed to the same medium also containing TGF-P, in a concentration of 0.1 ng/ml (3.8 p M ) or 2 ng/ml (76.8 pM). After varying periods of time, the cells were removed from the incubator and washed, and binding of lZ5I-EGF was measured in triplicate samples as previously described (13). The final concentration of EGF in the binding medium was 2 ng/ml (0.33 nM). Parallel plates, treated identically except that '"I-EGF was not added, were counted to determine cell number/plate. Results were calculated as femtomoles Exposure of FRH to TGF-P1 affected EGF binding as well as EGF-mediated replication. EGF binding was assayed at a n EGF concentration of 2 ng/ml (0.33 nM). At concentrations of TGF-P1 which did not affect growth (0.1 ng/ml), EGF binding to FRH was not significantly altered during a 24-h exposure (Fig. 2). With 2 ng/ml TGF-PI, a biphasic temporal course was observed. EGF binding transiently decreased up to 40% with a minimum occurring after 2 h of exposure. EGF binding then progressively increased, exceeding control values after 6 h of exposure (Fig. 2). After 24 h of TGF-PI exposure, the average amount of EGF binding was 174 +-24% of control ( n = 6, p < 0.05). The role of increased receptor number versus change in receptor affinity in this observed increase in EGF binding was investigated by measuring the binding isotherm of EGF on FRH. During culture in WE medium con-taining EGF, EGF binding is down-regulated, characterized by loss of high affinity receptors (13). On withdrawal of EGF, there is a prompt, cycloheximide-inhibitable increase in EGF binding associated with the appearance of a high affinity receptor population (13). A Scatchard plot of the binding curves for FRH cultured in the presence of EGF showed that control cells exhibited the down-regulated pattern. FRH exposed to TGF-& for 24 h showed a pattern similar to the upregulated state (Fig. 3A). This was confirmed by withdrawing EGF from the medium. After EGF withdrawal, the high affinity binding sites for EGF to control FRH were upregulated, while only a slight increase in high affinity EGF binding to FRH cultured in the presence of TGF-B1 was observed (Fig. 3B). Results of fit for high affinity sites in up-0 = TGF-P exposed 0 =Control were cultured in WE medium with 2 ng/ml EGF for 24 h. The hepatocytes were then cultured in WE medium with 2 ng/ml EGF with (TGF-P, exposed) or without (control) 2 ng/ml TGF-P, for an additional 24 h. The medium in each well was then replaced with 1 ml of cold (4 "C) fresh WE medium containing 10 mg/ml bovine serum albumin and 0.1-0.2 pCi of T -E G F with varying amounts of unlabeled EGF to achieve final EGF concentrations of 0.1-50 ng/ml. After 3 h, the cells were washed and bound Iz5I-EGF was measured in triplicate samples as previously described (13). Nonspecific binding, assayed as T -E G F binding in the presence of 500-fold excess unlabeled EGF, was subtracted from the binding values obtained with each sample. Cell number/plate was determined in parallel. Each point is the average of three replicate samples. The line drawn through the points is a smoothed curve linking the points and is not derived from the values of the two component fit to the data. 0, TGF-6,exposed fetal hepatocytes; 0, control fetal hepatocytes (no TGF-01 exposure). Panel B, up-regulated cells. Fetal rat hepatocytes were cultured with 2 ng/ml EGF for 24 h followed by an additional 24 h with (TGF-8, exposed) or without (control) 2 ng/ml TGF-PI. To upregulate the cells, fresh WE medium without EGF (with or without TGF-PI) was added, followed by incubation at 37 "C for 1 h. Binding of Iz5I-EGF was then determined as for panel A . 0, TGF-&-exposed fetal hepatocytes; 0, control fetal hepatocytes (no TGF-0, exposure). regulated control FRH were Kd = 0.84 nM and n = 29,000 sites/cell, and for TGF-P1-exposed FRH were K d = 0.66 nM and n = 26,000 sites/cell. Results of fit for high affinity sites in down-regulated TGF-&-exposed FRH were K d = 1.2 nM and n = 22,000 sites/cell. No binding sites were found on down-regulated control FRH with a two-component model with a K d set between 0.1 and 2 nM. The increase in high affinity EGF binding sites on EGF withdrawal from TGF-&exposed FRH was not statistically significant, nor were there any significant differences between up-regulated control and TGF-&-exposed FRH.
Down-regulation of EGF binding on EGF exposure involves internalization and degradation of EGFR (17). Since EGFR down-regulation appeared to be inhibited, the kinetics of EGF internalization in FRH exposed to TGF-P1 for 24 h were investigated. FRH were cultured under standard conditions (in WE medium with 2 ng/ml EGF for 24 h followed by 24 h of culture in the same medium with or without 2 ng/ml TGF-Pl), then surface EGFR was up-regulated by culture for 1 h in WE without EGF (continuing with or without TGF-P]). Once the external EGF binding sites had been loaded by incubation of the cells at 4 "C in WE containing Iz5I-EGF, internalization was initiated by a temperature jump to 25 "C.
Internal and external EGF were distinguished by the sensitivity of Iz5I-EGF to stripping by acid medium. The initial rate of internalization was slower for the TGF-P1-exposed experimental group, and a much smaller fraction of the surface EGF was internalized in TGF-&-exposed cells at any given time point (Fig. 4). The data shown are "single-cohort'' internalization rates for control and TGF-P1-exposed cells for an initial concentration of EGF of 0.13 nM. Similar results were seen for internalization at EGF concentrations of 0.03 and 0.37 nM. The data were analyzed by fitting the internalization data to a first order rate model, and the first order rate constant (K,) was derived (19,20). For an EGF concentration of 0.13 nM, the K, was 0.078 f 0.012 for control FRH and 0.041 k 0.008 for TGF-&-exposed FRH. The apparent reduction in K, by TGF-pl exposure was confirmed by formal measurement of K, at several different initial EGF concentrations and levels of surface receptor occupancy (Fig. 5). Based on this analysis, TGF-Pl exposure resulted in a 48% reduction in the rate of EGF internalization (K, for control cells = 0.077 f 0.004 uersus 0.048 f 0.007 for TGF-&-exposed cells, p = 0.036).
Internalization of EGFR is triggered by ligand binding, rapidly followed by receptor aggregation and localization in the coated pit. An early event in this sequence is autophosphorylation of EGFR on tyrosine residues, which may act to enhance the rate of internalization (21). The effect of TGF-p1 on EGF-stimulated EGFR tyrosine phosphorylation was measured to see if this could provide an explanation for the reduction in efficiency of receptor internalization. FRH were grown in the presence or absence of TGF-PI for 24 h, then briefly labeled with [3zP]P04 in the absence of EGF. The cells were then exposed to EGF, followed by solubilization of the cells, immunoprecipitation with monoclonal anti-EGFR antibody (EGF-R1) or anti-phosphotyrosine (Py-20), polyacrylamide gel electrophoresis, and autoradiography. The results of these experiments are shown in Fig. 6. EGFR autophosphorylation was not demonstrably different in the 180-kDa band derived from control or TGF-&-exposed cells following immunoprecipitation with Py-20 or EGF-R1. This was confirmed by quantitative densitometry. The amount of basal 32P incorporation in EGFR recovered by immunoprecipitation by Py-20 or EGF-R1 seemed to be slightly more in the TGF-P1exposed cells as compared with the control cells. Thus, EGFstimulated EGFR phosphorylation was comparable in control

FIG. 4. Effect of EGF concentration on the rate of endocytosis of EGF by cultured FRH.
Internalization experiments were performed as described previously (13,18). Fetal rat hepatocytes were cultured in WE medium with 2 ng/ml EGF for 24 h, then cultured in WE medium with 2 ng/ml EGF with or without 2 ng/ml TGF-& for an additional 24 h. EGFR was up-regulated by incubation of the FRH for 1 h in WE without EGF (with or without TGF-pl), followed by incubation of the cells in WE containing Iz5I-EGF for 3 h a t 4 "C. Cells were washed with cold phosphate-buffered saline, and internalization was initiated by the addition of 2 ml of warmed (25 "C) WE medium without EGF, followed by incubation for various times a t 25 "C. Internalized EGF was determined by the acid stripping method, and total (surface bound plus internalized) EGF was measured in parallel plates not subjected to acid stripping (13,18). Parallel plates incubated with 500-fold excess cold EGF were used to estimate nonspecific binding and fluid phase pinocytosis. The data shown are single-cohort internalization rates for control and TGF-&-exposed cells for an initial concentration of EGF of 0.13 nM and represent the average of three determinations. The data were analyzed by fitting the internalization data to a first order rate model and the first order rate constant (K.) was derived from a generalized least squares fit. For an EGF concentration of 0.13 nM, the K, was 0.078 f 0.012 for control FRH and 0.041 f 0.008 for TGF-&-exposed FRH. Panel A , control FRH; panel B, TGF-&-exposed FRH. a, surface EGF (in femtomoles/104 cells); 0, internal (acid-resistant) EGF. and TGF-PI-exposed cells. An increase in basal EGFR phosphorylation, which could represent alteration in regulatory phosphorylation of EGFR, was found in TGF-P,-exposed cells by this technique.

DISCUSSION
Identification of mRNAs for growth factors in fetal tissues suggests that transforming growth factor-a (TGF-a) and TGF-P are physiologic ligands in rodent fetal development from day 10 through birth (4-7). EGF only appears late in gestation, and this may be to be maternal in origin. Thus, in the experimental model used in this study, in uitro culture of fetal liver cells, EGF was used as a surrogate for the physiologic ligand TGF-a. In cultures of transformed cells or primary cultures of cells derived from adult animals, TGF-PI has pleomorphic effects, inhibiting cellular proliferation of epithelial tissues, promoting mesenchymal-derived cell growth, and regulating expression of differentiated function in cultured cells (7,22). Based on such observations, a role for FIG. 6. Effect of transforming growth factor j3, exposure on

[32P]P04 incorporation into immunopurified EGF receptor.
Fetal hepatocytes were grown in WE medium containing 2 ng/ml EGF, with or without 2 ng/ml TGF-& for 24 h. The cells were labeled for 2 h in phosphate-free RPMI medium containing 50 pCi/ml of ["P]P04 (Amersham) lacking EGF, and either with or without 2 ng/ ml TGF-OI. Cells were then incubated in fresh WE alone or WE containing 100 ng/ml EGF for 5 min a t 37 "C. After EGF exposure, the cells were centrifuged and the supernatant was used for immunoprecipitation as described. The immunoprecipitated protein was electrophoresed on 6% polyacrylamide gels and autoradiographed at -70 "C. Left panel, immunoprecipitation with anti-phosphotyrosine (Py-20). Following immunoprecipitation with monoclonal anti-phosphotyrosine (Py-20, ICN Immunobiologicals) in a final concentration of 1 pg/ml, the immunoprecipitated protein was eluted by boiling the beads with 2% SDS, 2% 2-mercaptoethanol, 100 mM phenylphosphate, 60 mM Tris-HC1, pH 6.8, for 3-5 min. Identity of lanes (from the left): control cells without added EGF, control FRH plus 100 ng/ ml EGF, TGF-&-exposed FRH without EGF, TGF-&-exposed cells after addition of EGF. Right panel, immunoprecipitation with anti-EGF receptor (EGF-Rl). Following immunoprecipitation with monoclonal anti-EGFR (EGF-R1, Amersham) in a final concentration of 1 pg/ml, the immunoprecipitated protein was eluted by boiling the beads with 2% SDS, 2% 2-mercaptoethanol, 60 mM Tris-HCI, pH 6.8, for 3-5 min. Identity of lanes (from the left): control cells without added EGF, control FRH plus 100 ng/ml EGF, TGF-&exposed FRH without EGF, TGF-&-exposed cells after addition of EGF.
TGF-PI in growth regulation and induction of differentiation during fetal and postnatal development has been proposed (7). However, such a growth-regulatory role has not been directly demonstrated in fetal cells. The role of TGF-PI in rat fetal cell growth was assessed in vitro with primary cultures of FRH taken from 19-day fetuses. In this system, incubation of the fetal liver cells with TGF-PI resulted in prompt and complete growth arrest both as assessed by cell number and by measurements of DNA synthesis. The concentration providing maximal growth inhibition was 2 ng/ml or 80 p~. Growth of adult hepatocytes and other epithelial cells is inhibited by similar sub-nanomolar concentrations (3,11,12).

TGF-p and EGF Endocytosis
However, this is the first demonstration of such an effect in fetal cells.
With increasing time of incubation with TGF-P1, a biphasic effect on surface epidermal growth factor receptor levels was observed. With exposure times up to 2 h, EGF receptor levels were reduced approximately 40%. With exposure times of 6-24 h, TGF-(3 exposure resulted in gradual increase in EGF binding to FRH. A similar biphasic effect on EGF binding by TGF-@ has been reported for cultured fibroblasts (23). In other reports, TGF-P, exposure has resulted variably in decreased EGF binding (24,25) and increased EGF binding (26,27). The predominant effects reported have been manifest by alteration of levels of high affinity EGF receptors (23)(24)(25)(26). In TGF-@,-exposed FRH, the increase in binding was also found to be due to an increase in high affinity EGF receptors, as demonstrated by Scatchard analysis of the EGF binding isotherm. Under control conditions, EGF receptors on FRH (in EGF-containing medium) are down-regulated with loss of the high affinity population of EGF binding sites as compared with up-regulated cells following EGF withdrawal. After a 24h TGF-P1 exposure, FRH cultured in the presence of EGF exhibited EGF binding isotherms similar to those for upregulated cells despite the presence of sufficient EGF in the medium to completely down-regulate control FRH (13). On withdrawal of EGF, no increase in EGF binding at high affinity sites occurred in the TGF-@,-exposed FRH. Thus, after 24 h of TGF-Pl exposure, ligand-mediated down-regulation of EGFR was inhibited. In adult hepatocytes, no effect of TGF-@, exposure on EGFR down-regulation has been reported, but under the experimental conditions of that study the cells were already maximally down-regulated, making such measurements difficult (11).
Homologous down-regulation of the EGF receptor involves ligand-induced receptor internalization and degradation (17). To quantitate the effects of TGF-@, exposure on internalization, the kinetics of EGF internalization were examined. The measurement of single-cohort internalization is an experimental design which isolates experimental effects on binding kinetics from those on kinetics of internalization (19,20). After fitting the data to a simple first order model, the apparent rate constant for internalization, K,, was determined (19,28). This experiment was repeated for cells exposed to varying concentrations of EGF which resulted in varying concentrations of occupied surface receptor, the substrate for internalization. This exercise confirmed that the rate of internalization appeared to be first order with respect to initial surface EGF concentration and that the K, values derived at various concentrations of EGF were in good agreement. In TGF-@-exposed cells, internalization of EGF was consistently inhibited with an overall 48% reduction in K, from 0.077 to 0.048 min".
Despite the inhibition of mitogenesis, receptor down-regulation, and internalization, early steps in EGF-related signal transduction did not appear to be affected. EGF-induced receptor autophosphorylation was preserved as detected by labeling of immunoprecipitated EGFR with [32P]orthophosphate after EGF exposure in both control and TGF-Pl-exposed FRH. These observations extend previous reports of a lack of effect of TGF-PI on EGFR autophosphorylation in nonreplicating adult hepatocytes to fetal liver cells undergoing active growth (11).
The mechanism of inhibition of EGFR endocytosis by TGF-P1 is not defined by these studies. Autophosphorylation of EGFR appears to promote endocytosis of occupied receptor (21), but our results do not support reduced EGF-stimulated EGFR phosphorylation as the mechanism leading to reduced EGF endocytosis. Results from the immunoprecipitation experiments suggest that TGF-P1 exposure may result in activation of protein kinase(s) that recognize EGFR as a substrate. In TGF-P1-exposed cells, basal [32P]P04 content of immunoprecipitated EGFR was increased, whether prepared with anti-EGFR or anti-phosphotyrosine. Although phosphoaminoacid determination was not performed, this suggests that this [32P]P04 is incorporated in EGFR at sites other than tyrosine. Such regulatory phosphorylation of EGFR could interfere with interaction of occupied receptor with the coated pit, resulting in reduced endocytosis. Alternately, the inhibition of endocytosis could be due to phosphorylation of components of the endocytotic apparatus at sites which control interaction of the receptor with the coated pit or generation of the endosome. Further characterization of the effects of TGF-P, exposure on EGFR phosphoaminoacid content and the effect of TGF-@, on fluid phase and receptor-mediated endocytosis of a variety of ligands is being pursued.