Characterization of hepatic epidermal growth factor receptors in the developing rat

Biochimica et Biophysica Acta 930 (1987) 107-113 Elsevier

107

BBA 12070

C h a r a c t e r i z a t i o n of h e p a t i c e p i d e r m a l g r o w t h f a c t o r r e c e p t o r s in t h e d e v e l o p i n g rat Steven B. H o a t h , W i l l i a m L. Pickens, J o h n C. B u c u v a l a s a n d F r e d e r i c k J. S u c h y Division of Neonatology and Division of Pediatric Gastroenterology and Nutrition, Department of Pediatrics, Children's Hospital Medical Center. Cincinnati, OH (U.S.A.) (Received 15 December 1986)

Key words: Epidermal growth factor receptor; Development; (Rat liver)

Binding of 12sI-labeled epidermal growth factor (EGF) was characterized in basolateral plasma membranes prepared from the livers of 21-day gestation fetuses, 14-day-old sucklings and adult Sprague-Dawley rats using a self-generating Percoli gradient method. The membrane preparations employed have been previously assayed in terms of plasma membrane protein yield, enrichment of various marker enzymes and sodium-dependent bile acid and amino acid transport, tzSI-EGF binding was saturable and time and temperature dependent. Equilibrium analyses showed that the suckling period is characterized by a marked decrease in overall hepatic EGF binding capacity (460_ 50 f m o l / m g protein) compared to either the fetal period (1290 + 160 fmol/mg) or to adults of either sex (males= 1540 + 230, females 1010+ 130 fmol/mg). Treatment of the suckling rat with parenteral EGF resulted in a 78% reduction in the observed binding capacity when assessed 2 h after growth factor administration. Comparison of binding affinities revealed no significant difference between the suckling and adult preparations (K d = 0.40 5=0.03 vs. 0.39 + 0.02 nM, respectively); however, both preparations differed significantly from the fetal group which exhibited a decreased affinity of binding with a higher overall dissociation constant (K d = 0.68 5=0.06 nM). Thus, it appears that major ontogenetic changes occur in the rat hepatic ligand/receptor system for epidermal growth factor. These changes are discussed in the context of transitional events in mammalian development such as birth and weaning.

Introduction

Epidermal growth factor (EGF) is a small (molecular weight 6000), single-chained, globular protein which has been isolated from a number of mammalian species including humans, rats and mice [1,2]. In vitro, a number of differing cell

Abbreviation: EGF, epidermal growth factor. Correspondence: S.B. Hoath, Division of Neonatology, Department of Pediatrics, Children's Hospital Medical Center, Cincinnati, OH 45267, U.S.A.

types possess EGF receptors and respond to EGF in the culture medium by entering the cell cycle and undergoing division - hence, the designation 'growth factor' [1,2]. In newborn rodents, EGF also appears to act as a 'differentiation factor' accelerating specific events of integumental maturation such as unfusion of the eyelids [3]. In addition, EGF has been shown to be a potent inhibitor of gastric acid secretion suggesting a role for this factor in the regulation of digestion [4]. Recently, we have become interested in the possible role of EGF in perinatal liver development. In the adult rat, the liver is an apparent

0167-4889/87/$03.50 © 1987 Elsevier Science Publishers B.V. (Biomedical Division)

108

primary target organ for EGF, insofar as rat liver exhibits extraordinarily high numbers of E G F receptors [5]. The ontogeny of E G F binding to rat liver, however, has not been characterized using highly purified hepatic membrane preparations. Our primary purpose in the present study was to determine the ontogeny of E G F binding characteristics in rat liver membranes. Secondarily, we were interested in the presence of sex differences and the effect of exogenous E G F exposure on E G F binding characteristics. Methods

Plasma membrane preparation To examine E G F binding we employed a selfgenerating Percoll gradient technique described by Blitzer and Donovan for isolation of basolateral (sinusoidal) plasma membranes [6]. Relative membrane marker enzyme activities for each membrane preparation were determined and compared for all ontogenic ages studied as previously described [7,8]. The basolateral membranes, were enriched approx. 30-times in the plasma membrane marker (Na + + K +)-ATPase for each group, 5 to 7-fold enriched in the canalicular markers alkaline phosphatase and Mg 2+-ATPase, and not enriched in marker enzyme activities for intracellular organelles. Membrane protein content was determined by the method of Lowry [9]. All binding assays were performed on membranes prepared from multiple animals and, therefore, data points are representative of group averages at each age rather than individual variation. In general, in order to obtain adequate membrane for analysis, the following numbers of animals were combined at each age studied: adults, 2-5; 14-day-old sucklings, 35-40; and 21 day fetuses, 100-120. Each m e m b r a n e preparation was then examined in 2 - 3 separate equilibrium analyses. Binding assay Time and temperature dependence of E G F binding as well as the effect of varying the protein concentration and the magnitude of nonspecific binding were determined for all membrane preparations at all ages. Based on this preliminary work, the following conditions were chosen for the E G F binding assay. The assay was performed at

2 2 ° C using 25 ~g membrane protein per assay tube with a total incubation time of 50 min. Assay tubes contained a total volume of 300 /~1 of 50 m M phosphate buffer (pH 7.3) containing 150 mM NaCI and 0.1% bovine serum albumin. After equilibration, the samples were diluted with 4 ml of ice-cold buffer, harvested by filtration on Millipore E G W P cellulose acetate filters using a multiple manifold apparatus (Hoefer Scientific) then rapidly washed with three 4-ml aliquots of ice-cold buffer. Radioactivity retained by the filters was then measured by crystal scintillation counting (LKB Gamma). Nonspecific binding was assessed in the presence of 2.5 /~g of unlabeled E G F and was invariably less than 2% of the total counts added. E G F used for standards and radiolabeled tracer was purified from mouse submandibular glands by a modification of Cohen's procedure [10] utilizing HPLC as the final purification step [11]. E G F was iodinated by reaction with chloramine T essentially as described previously [12]. Biological specific activity was determined by relative binding of increasing amounts of mouse label compared to fractional displacement of tracer by unlabeled standard using plasma membrane E G F receptors isolated from human term placenta as the binding protein [13]. The spec. act. of the iodinated tracer was 500-700 c p m / p g . Possible degradation of radiolabeled E G F was assessed under standard assay conditions by two methods. First, approx. 150000 cpm of 125I-EGF was incubated with 25 /,g of membrane protein from the various ontogenetic age groups. Tracer binding was performed as described above at room temperature for a total incubation of 60 rain. Following incubation~ the tubes were spun at 8000 × g for 10 min, the supernatants were aspirated and the pellet were counted by gamma scintillation spectrometry. The tracer present in the supernatant fraction was reincubated with membrane protein (150000 c p m / 2 5 /,g protein) and the separation of bound and free moieties was determined following equilibrium binding assay. As a second method for assessing tracer degradation, total trichloroacetic acid-precipitable protein was determined following equilibrium binding of 150000 cpm of radiolabeled E G F with 25 btg membrane protein. Following 60 mill in-

109

cubation under standard assay conditions, the assay was terminated by the addition of sufficient trichloroacetic acid to yield a final concentration of 10% trichloroacetic acid. The tubes were then vortexed and centrifuged at 4000 rpm in a Sorvall RC-3B centrifuge for 15 min. The supernatants were aspirated and the pellets were counted by g a m m a scintillation spectrometry.

Quantitative analysis E G F binding was assessed at equilibrium in the presence of increasing amounts of unlabeled E G F and analyzed by two conventional plotting methods. Plotting the data according to Scatchard [14] allowed conclusions regarding the maximum binding capacity of the membrane preparation as determined from the x intercept and conclusions regarding the dissociation constant or K d of the reaction from the negative reciprocal of the slope. To justify these extrapolations, the criteria set forth by Klotz [15] were satisfied; namely, that the plot of the logarithm of the free ligand concentration versus the amount of ligand bound was curvilinear with the bound ligand approach a plateau and with an apparent inflection point midway through the plot. All assays were examined according to both plotting methods prior to calculation of binding characteristics. Statistical analyses of membrane binding characteristics were performed using either Student's t-test for comparison of group means or one-way analysis of variance with the Student-Newman-Keul's test for multiple sample comparison.

incubation at 22 ° C. Similar results were found in studies of the percent of counts precipitable in 10% trichloroacetic acid after incubation of 125IE G F in the presence and absence of the various m e m b r a n e fractions. Thus, compared to assay tubes containing no membranes (equivalent to maximal or 100% precipitable cpm), tubes containing the various membrane fractions showed a minimum of 96% of all counts added were precipitable. This method of assessing tracer integrity would indicate that less than 4% of the radiolabeled E G F was degraded during the time course of the assay.

Ontogenetic differences in EGF binding characteristics E G F binding as a function of time was examined in all m e m b r a n e groups studied in order to determine equilibrium binding conditions. Fig. 1 shows a representive binding experiment containing membranes from each of the major ontogenetic groups studied. The assay conditions are 7

Adult Male I

I

6

I Fetal

I-

I

~n

5

I

o

4

I

0 _ ~

3

I Adult Female

uckhng

Results o

Assessment of tracer integrity The possibility of degradation of radiolabeled E G F by the various membrane fractions was assessed by studies of the percent of rebinding of tracer to hepatic membranes following incubation under standard assay conditions. No difference was noted between the various ontogenetic groups with all fractions exhibiting a minimum of 98% of tracer rebound to the various m e m b r a n e fractions following initial incubation and separation of bound and free moieties. These studies would indicate that less than 2% of the radiolabeled growth factor was degraded during the 60-min

u.

1

0

20

40

60

80

Incubation Time (minutes)

Fig. 1. Representative experiment showing ontogenetic differences in rate of EGF binding to purified rat hepatic plasma membranes. Each data point represents the mean + S.E. of four individual determinations corrected for nonspecific binding. Analysis of variance performed at assay conclusion (90 min incubation) showed significant differences among all groups ( P < 0.01). Each assay tube contained 25 #g of membrane protein and 121000 cpm of 125I-EGF (196 pg). Details of the binding assay and membrane preparation are given in the text.

110

81

o o

O

5

"~

L 2/ ~0 1~ ii 0 0

resentative experiment. All major experimental groups showed parallelism of binding with increasing protein concentration. Routine binding studies were subsequently performed with 25 #g m e m b r a n e protein. Scatchard and Klotz plots were generated for adult males, 14-day-old postnatal animals of mixed sex and 21-day-fetal pups of mixed sex. Representative E G F binding data for adult and fetal rat hepatic plasma m e m b r a n e preparations are shown in Figs. 3 and 4. D a t a in the Scatchard plot is given as the ratio of b o u n d / f r e e ligand versus p M E G F bound. Linearity of the Scatchard data was assumed when the correlation coefficient was between - 0 . 9 6 and - 1 . 0 (this criterion was satisfied for all m e m b r a n e preparations at all ages). The same data is given in the Klotz plot as fmol E G F b o u n d per assay tube versus log of the free concentration of E G F in fmol. Table I exhibits the mean binding characteristics obtained from four separate fetal preparations of mixed sex, four preparations from 14-day-old animals of mixed sex and 6 adult male preparations. Significantly, the binding capacities of the fetal and adult m e m b r a n e s were not different while there was a marked reduction in the binding capacity of the membranes prepared from the

[]

~ j j ~ ..~SJ~

o Adult Female o 2! pay Fetal ~ 14 Day Suckling

20

40

Membrane Protein (ug) Fig. 2. Effect of increasing membrane protein concentration on 125I-EGF binding at different ontogenetic ages. Aliquots of hepatic basolateral membrane protein ranging from 1 to 50 btg were incubated for 50 min in the presence of 136000 cpm 125I-EGF (244 pg). Results represent the means of four separate determinations at each protein concentration corrected for nonspecific binding. S.E. are incorporated within the data labels.

explained in the figure legend. In general, equilibrium was reached between 3 0 - 4 0 min of incubation. Similarly, the effect of increasing hepatic m e m b r a n e protein on specific :25I-EGF binding was examined. Fig. 2 shows the result of a repSCATCHARD PLOT

KLOTZ PLOT

.16,

30

.12, U. m

tu

20'

.08,

10"

.04'

2"5

5"0

7"5

1 ()0

1.4

2:2

3:0

3:8

pM Bound Log (fmoles EGF Free) Fig. 3. Scatchard and Klotz plot analyses of 1251-EGF binding to fetal rat hepatic plasma membranes. Plasma membranes were prepared utilizing a self-generating Percoll gradient technique as described by Blitzer and Donovan [6]. Binding assays wen performed at 22°C with 25 #g of membrane protein per assay tube after 50 rain incubation (see Methods). The membrane preparation used for this analysis was derived from livers pooled from 120 fetal rats killed on gestational day 21.

111 SCATCHARD PLOT

KLOTZ PLOT

50. .40"0 e0 m

.30 ¸

40.

U.

o

UJ u)

m

_e .20

30'

0

20 .10 10 40

80

120

160

pM Bound

1.2

2.0

218

3.6

Log (fmoles EGF Free)

Fig. 4. Scatchard and Klotz plot analyses of 125I-EGFbinding to adult male rat hepatic plasma membranes. Binding conditions were identical to those described in Fig. 1. Membranes were derived from five adult male rats of over 60 days of age. 14-day-old suckling animals. I n contrast, b i n d i n g affinities were similar for the p o s t n a t a l groups while the fetal p r e p a r a t i o n s exhibited significantly lower affinity of b i n d i n g as shown b y the higher dissociation constant. A similar study using hepatic m e m b r a n e s derived from 7-day-old rats revealed

TABLE I ONTOGENY OF EGF BINDING CHARACTERISTICS IN RAT HEPATIC PLASMA MEMBRANE PREPARATIONS Summary results comparing binding characteristics of rat hepatic membrane preparations from 21-day-old fetuses of mixed sex, 14-day-old sucklings of mixed sex and adult male rats. Data display variations in group means obtained from 2-4 separate equilibrium binding assays performed on individual membrane pools derived from four fetal preparations (440 animals), four suckling preparations (150 animals) and six adult preparations (24 animals). The derived means were subjected to ANOVA with the above significances noted. Conditions of the binding assay were identical for each ontogenetic age and are described in the text. Fetal (4)

14-day (4)

Adult (6)

Binding capacity (fmoles/mgprotein) 1290+_160 460+_50 " 1540+_230 Kd (nM) 0.68+_0.06 a 0.40+_0.03 0.39_+0.02 a p < 0.01.

b i n d i n g characteristics essentially identical to the 14-day-old group (data n o t shown).

Effect of parenteral EGF on EGF binding characteristics in suckling rats In a separate experiment, 28 suckling rats were treated s u b c u t a n e o u s l y with 200 ng per g b o d y weight of exogenous E G F every 12 h for a total of six doses. 28 littermate controls received an equal v o l u m e of water. A n i m a l s were killed 2 h after the last dose of E G F on p o s t n a t a l day 15 a n d their livers were i m m e d i a t e l y processed as described in the m e t h o d s section. This t r e a t m e n t regimen resulted in a m a r k e d l y reduced E G F b i n d i n g capacity (controls, 310 fmol E G F / m g m e m b r a n e protein vs. E G F ; treated, 70 f m o l / m g protein). N o differences in b i n d i n g affinities were noted between groups (controls, K d = 0.39 n M vs. E G F treated, K d = 0.35 nM).

Sexual dimorphism of EGF binding C o m p a r i s o n of adult female basolateral memb r a n e s with adult males showed a higher b i n d i n g capacity in the male group (1500 vs. 1010 fmol E G F / m g m e m b r a n e protein, respectively). These findings s u b s t a n t i a t e the previous report by Benveniste who used a slightly cruder p l a s m a m e m b r a n e p r e p a r a t i o n to exhibit a sexual d i m o r p h i s m in E G F b i n d i n g in adult rat liver [16].

112 Discussion

Our data indicate a marked change in overall E G F binding characteristics following birth. Compared to the 14-day-old suckling rat, plasma membranes prepared from 21-day-old fetuses exhibit a lower affinity and a much higher overall binding capacity. We have examined also several membrane preparations from pups of 7 days postnatal age with essentially similar results to the 14-dayold animals (data not shown). These findings differ from those of Adamson et al. [17] who reported more avid binding of 125I-EGF to crude fetal mouse liver membranes than to adult preparations. Our data, therefore, do not support the concept of an inverse correlation between binding affinity and developmental age. One well known characteristic of the fetal liver is its high proportion of hematopoietic tissue [18]. This blood-forming component of the liver gradually disappears over the first 2 weeks of postnatal life in the rat [18]. In this study, E G F binding data are presented at different ontogenetic ages for membrane pools normalized for standard marker enzymes such as ( N a + + K+)-ATPase. Comparable marker enzyme activities, enrichments and recovery studies would suggest minimal contamination of liver plasma membrane with membrane of hematopoietic cells [7,8]. To our knowledge, the E G F receptor has not been associated with erythropoietic stem cell lines in mammals [19]. It is known, however, that the oncogene product of the avian erythroblastosis virus represents a truncated form of the E G F receptor, suggesting a possible role for E G F in the regulation of erythropoiesis in lower species [20]. In addition to the caveat that the same organ may contain differing tissue sub-types at different ontogenetic stages, developmental binding studies are confronted with a variety of other interesting problems. In the adult and post-weanling rodent, 'higher level' endocrine influences are likely to be of major importance in determining the status of growth factor-receptor systems. For example, thyroid hormone modulates E G F receptor levels in adult rat liver [21] and neonatal mouse skin [22] as well as E G F tissue concentrations in the immature and adult mouse submandibular gland [23,24]. In the developing rat, circulating levels of both

thyroid hormone and glucocorticoids peak during the 14th-17th day of postnatal life [25]. The subsequent period between 17 and 21 days is characterized by several possibly interrelated events: first, under the influence of both thyroid hormones and glucocorticoids, the rodent intestine 'closes' to uptake of macromolecules [26,27], second, the animal weans from milk to a solid diet and third, submandibular gland EGF concentrations exhibit onset of a steady, logarithmic increase [28]. The data presented in this study demonstrate an increase in hepatic E G F binding in the post-suckling rat and, viewed in the context of the above events, we suggest that this increase may be in part a hormonally mediated, post-closure phenomenon. While in humans the exact tissue of origin of E G F is still unclear, it is known that various biological fluids in humans, such as milk, contain high concentrations of EGF ranging from 0.1 to 1.0 /zg per ml [29-31]. Moreover, the neonatal intestine, particularly in the premature human infant and the pre-weanling rodent, is known to absorb intact protein molecules [32]. The data presented in this study are consistent with the hypothesis of milk-derived E G F leading to the observed low E G F binding capacity during the suckling period. While this statement must be regarded for the present as unproven, it is known that oral administration of labeled E G F to the suckling rat results in sequestration of label in liver, skin and other organs [33]. Consequently, E G F may prove useful as an endogenous molecular probe to investigate the role of milk proteins on extraintestinal organ growth and maturation. This hypothesis is also consistent with a recent report by Toyoda et al. [34] demonstrating that E G F binding is higher in fetal than in postnatal rat enterocytes and that the fetal binding capacity exhibits a dose-dependent decrease following exposure to mature rat milk. As demonstrated in the present work, exogenous E G F given parenterally decreases markedly the binding capacity of suckling rat basolateral membranes. Whether this effect is secondary to down regulation of receptors or to masking of tracer binding is undetermined. Further experiments are in progress to evaluate the effect of oral administration of EGF on rat hepatic EGF binding.

113

The binding studies utilizing selectively purified male and female rat hepatic plasma membranes confirm the previous report by Benveniste of a sexual dimorphism of EGF receptor binding in the adult. No attempt was made in the present study to distinguish sex effects in the younger age groups. Clearly, however, the decreased binding capacity of the suckling rats versus adults cannot be explained by sex differences alone as both adult male and female binding capacities are significantly higher than in the suckling group. In summary, we have utilized highly purified hepatic plasma membrane preparations to generate data on the relative affinities and binding capacities of 125I-EGF at various developmental ages in the rat. These data have been discussed in the context of the ontogeny of chemical messenger systems in the rat and the timing of major transitional events in mammalian development such as birth and weaning.

Acknowledgments This work was supported by a grant from the Children's Hospital Research Foundation and by National Institutes of Child Health and Human Development grants HD-20784, HD-00597 and HD-20632. References 1 Carpenter, G. and Cohen, S. (1979) Ann. Rev. Biochem. 48, 193-216 2 Hollenberg, M.D. (1979) Vitamins and Hormones, 37, 69-110 3 Cohen, S. (1962) J. Biol. Chem. 237, 1555-1562 4 Elder, J.B., Ganguli, P.C., Gillespie, I.E., Gerring, E.L. and Gregory, H. (1975) Gut 16, 887-893 5 O'Keefe, E., Hollenberg, M.D. and Cuatrecasas, P. (1974) Arch. Biochem. Biophys. 164, 518-526 6 Blitzer, B.L. and Donovan, C.B. (1984) J. Biol. Chem. 259, 9295-9301

7 Suchy, F.J., Bucuvalas, J.C., Goodrich, A.L., Moyer, M.S. and Blitzer, B.L. (1986) Am. J. Physiol. 251, G665-G673 8 Suchy, F.J., Courchene, S.M. and Blitzer, B.L. (1985) Am. J. Physiol. 248, G648-G654 9 Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, R.J. (1951) J. Biol. Chem. 193, 265-275 10 Savage, C.R. and Cohen, S. (1972) J. Biol. Chem. 247, 7609- 7611 11 Smith, J.A., Ham, J., Winslow, D.P., O'Hane, J.M. and Rudland, P.S. (1984) J. Chromatog. 305, 295-308 12 Carpenter, G. and Cohen, S. (1976) J. Cell Biol. 71,159-171 13 Hock, R.A. and Hollenberg, M.D. (1980) J. Biol. Chem. 255, 10731-10736 14 Scatchard, G. (1947) Ann. N.Y. Acad. Sci. 51,660 672 15 Klotz, I.M. (1982) Science 217, 1247-1249 16 Benveniste, R. and Carson, S.A. (1985) Mol. Cell. Endocrinol. 41,147-151 17 Adamson, E.D. and Meek, J. (1984) Develop. Biol. 103, 62-70 18 Greengard, O., Federman, M. and Knox, W.E. (1972) J. Cell Biol. 52, 261-272 19 Goustin, A.C., Leof, E.B., Shipley, G.D. and Moses, H.L. (1986) Cancer Res. 466, 1015-1029 20 Hunter, T. (1984) Nature 311,414-416 21 Mukku, V.R. (1984) J. Biol. Chem. 259, 6543 6547 22 Hoath, S.B., Lakshmanan, J. and Fisher, D.A. (1985) Pediatr. Res. 19, 277-281 23 Gresik, E.W. and Barka, T. (1980) Am. J. Anat. 159, 177-185 24 Walker, P., Weichsel, M.E., Hoath, S.B., Poland, R.E. and Fisherr, D.A. (1981) Endocrinology 109, 582-587 25 Henning, S.J. (1981) Am. J. Physiol. 241, G199-G214 26 Clark, S.P. (1959) J. Biophys. Biochem. Cyto. 5, 41-50 27 Moog, F. and Yeh, K.Y. (1979) Develop. Biol. 69, 159-169 28 Byyny, R.L., Orth, D.N. and Cohen, S. (1972) Endocrinology 90, 1261-1266 29 Read, L.C., Francis, G.L., Wallace, J.C. and Ballard, F.J. (1985) J. Develop. Physiol. 7, 135-145 30 Beardmore, J.M. and Richards, R.C. (1983) J. Endocrinol. 96, 287-292 31 Shing, Y.W. and Klagsbrun, M. (1984) Endocrinology 115, 273-282 32 Walker, W.A. (1979) in Development of Mammalian Absorptive Processes, Ciba Foundation Symposium 70, pp. 201-219, London 33 Thornburg, W., Matrisian, L. Magun, B. and Koldovsky, O. (1984) Am. J. Physiol. 246, G80-G85 34 Toyoda, S., lee, P.C., Lebenthal, E. (1986) Biochim. Biophys. Acta 886, 295-301