Molecular and Cellular Endocrinology, 18 (1980) 189-199

0 Elsevier/North-Holland Scientific Publishers, Ltd.

189

PROTEIN PHOSPHORYLATION IN HUMAN PLACENTA

STIMULATION BY EPIDERMAL GROWTH FACTOR

Graham CARPENTER ‘,‘, Lonnie POLINER and Lloyd KING Jr. ’ Departments of 1 Medicine (Dermatology) and 2 Biochemistry, Vanderbilt University School of Medicine, and 1 Veteran’s Administration Hospital, Nashville, TN 37232 (U.S.A.)

Received 7 December 1979; accepted 11 February 1980

Membranes prepared from normal human term placenta possess a protein kinase activity

which phosphorylates endogenous substrates in the presence of [T-~~P]ATP. This kinase activ-

ity requires either Mg2+ or Mn2+, IS enhanced by glycerol and appears to be cyclic-nucleotide-

independent. Addition of epidermal growth factor to the placental membrane preparation

increases the level of total phosphorylation by approx. 35%. When analyzed by sodium dodecyl

sulfate gel electrophoresis and autoradiography, the placental proteins whose degree of phos-

phorylation was enhanced by epidermal growth factor had apparent molecular weights of

170 000, 150 000, and 25 000. This report documents the presence of protein kinase activity in

human placenta and demonstrates that epidermal growth factor can enhance protein phos-

phorylation in normal human tissue.

Keywords: polypeptide hormones; mechanism of action.

Epidermal growth factor (EGF) is a small polypeptide, molecular weight 6045, isolated from both mouse (Cohen, 1962) and human (Cohen and Carpenter, 1975) sources and which stimulates cell proliferation in vivo and in vitro (see Carpenter and Cohen, 1978, 1979 for reviews). In addition to its mitogenic properties EGF can also act to inhibit gastric acid secretion in vivo (Bower et al., 1975) and to enhance the production of human chorionic gonadotropin in cultured choriocar- cinema cells (Benveniste et al., 1978). Experiments utilizing cultured cells have demonstrated biochemically (Carpenter and Cohen, 1976) and morphologically (Haigler et al., 1978, 1979) that the binding of EGF to receptors on the cell surface is rapidly followed by clustering of the hormone-receptor complexes, internaliza- tion of these complexes by an endocytotic mechanism, and degradation of the growth factor within lysosomes. Although it has not been determined whether the internalization and/or degradation processes are important for the biological activ- ity of EGF, biochemical signals must be generated at some point in this sequence of events to influence cell function. Recently, Carpenter et al. (1978, 1979) have utilized membrane preparations from a unique cultured tumor-cell line (A-431) to

190 Crakam Carpenter, Lonnie Poliner, Lloyd King Jr.

demonstrate that EGF can provoke enhanced protein p~losphorylation in vitro. If this finding is to be considered as a part of the mechanism of action of EGF on all target cells, then enhanced protein phosphorylation should be demonstrable in other target cells and tissues. In this report we have investigated the influence of EGF on protein phosphorylation in preparations of human placental membranes.

METHODS AND MATERIALS

Materials. Mouse-derived EGF, prepared as described elsewhere (Savage and Cohen, 1972), was generously provided by Dr. Stanley Cohen, Vanderbilt Univer- sity. Cyclic AMP, cyclic GMP, dibutyryl cyc1i.c AMP, ATP, h&tone (‘type II-A), and bovine insulin were obtained from Sigma. [y-32PIATP was purchased from ICN. The phosvitin and casein were gifts of Dr. T. Soderling, Vanderbilt University. Prolactin, growth hormone, follicle-stimulating hormone, and thyrotropin were NIH preparations, while pitressin was from MRC. Glucagon was a gift from Dr. John Exton, Vanderbilt University.

~~e~~2~~a~e p~e~a~at~~n. Term human placentae were obtained within 20 min after delivery, placed on ice, cut into small pieces (approx. 4 g), and washed exten- sively with cold saline to remove blood. The tissue was then packaged in aluminum foil, frozen on dry ice, and stored at -65°C. A crude membrane fraction from pla- cental tissue was prepared essentially by the procedure described by Cuatrecasas (3971) and stored in small aliquots at -65” in 10 mM HEPES buffer, pH 7.4. Pro- tein was determined by the procedure of Bradford (1976) using y-globulin as a standard.

~~~os~~l~~~~at~5~ assay. The reaction mixtures contained, unless otherwise stated

in the text, in a 60-~1 volume the following: 20 mM HEPES buffer, pH 7.4, 0.1% bovine serum albumin, 10% (v/v) glycerol, 30 mM NaCl, 50 mM MgCi,, placental membrane (100 pg protein), and [-Y-~~P]ATP (15 FM, approx. 7 X lo5 cpm). In experiments in which EGF was added, the growth factor (100 ng) was preincu- bated in the reaction mixture in the absence of labeled ATP for 1 h at 25”. All phosphorylation assays were performed by rapidly cooling the incubation mixtures to 0” in an ice-water bath, adding [y-32P]ATP, and 3 min later terminating the reac- tion. The phosphorylatio~ reaction was stopped by quickly pipetting a SO+1 ali- quot of each reaction mixture onto a square (2 cm’) of Whatmann 3 MM filter paper and immediately dropping it into a beaker of ice-cold 10% tr~c~oroacetic acid containing 10 mM sodium pyrophosphate. The filters were washed extensively (4 times) with shaking in the trichloroacetic acid plus pyrophosphate solution at room temperature. Finally, the filters were extracted once with alcohol, once with ether, dried, and the rad~oactivjty measured in a Nuclear Chicago gas ilow counter.

Protein phospho@ation and epidermal growth factor 191

Gel electrophoresis and audioradiography. Components of the placental mem-

brane preparation were separated by sodium dodecyl sulfate (SDS) gel electro- phoresis (7.5% acrylamide) by the method of Laemmli (1970). Molecular-weight standards used to calibrate the gels were myosin 200 000; fl-galactosidase 130 000; phosphorylase B 94 000; bovine serum albumin 68 000; ovalbumin 43 000. The gels were fixed and stained with Coomassie Blue or periodic acid-Schiff (PAS) as described by King and Morrison (1976). The slab gels were dried under vacuum and autoradiography was performed using Kodak RP Royal X-O Mat film.

RESULTS

The data in Fig. 1 demonstrate the requirement of the placental membrane phos- phorylation activity for the divalent cations Mg2+ or Mn*‘. Although Mn2+ was more potent at low concentrations (below 5 mM), high levels of Mg2+ (50- 100 mM) produced approx. 50% more activity than Mn2+. At 50-100 mM concen- trations of Mn2+ the level of phosphorylation activity was not higher than that achieved in the presence of 5 mM Mn2+ (data not shown). Also, the addition of cal- cium (l-10 mM) did not increase the phosphorylation activity in placental mem- branes.

To demonstrate that the observed radioactivity in the phosphorylated mem-

hkYe?k- 0 100

Mg’+(mM)

I I I I I

0 I 2 3 4 5

Mn2+(mM)

Fig. 1. Effect of divalent cations on protein-phosphorylation activity in human placenta. Ali-

quots of placental membranes (100 pg) were incubated in the standard phosphorylation assay described in Methods, with increasing concentrations of MgC12 OI MnC12.

192 Graham Carpenter, Lonnie Poliner, Lloyd King Jr.

brane components is derived from [T-~‘P]ATP an excess (1 mM) of unlabeled ATP was added to the reaction mixture prior to the addition of membranes. This proce- dure reduced the incorporation of radioactivity into trichloroacetic acid-insoluble material by 86%. Furthermore, the radioactivity incorporated into endogenous membrane substrates could be “chased” by the addition of unlabeled ATP as shown in Fig. 2. The dephosphorylation of labeled substrates proceeded rapidly as 40% of the incorporated radioactivity was “chased” after a I-min incubation with un- labeled ATP. Approx. 30% of the incorporated 32P however, could not be removed

upon prolonged incubation (90 min) with unlabeled ATP. The time course of the phosphorylation of placental membranes in the presence

and absence of EGF is shown in Fig. 3. Maximal phosphorylation activity was achieved after 3-5 min incubation at 0” with [T-~~P]ATP. Prior incubation of the membranes with EGF increased the level of phosphorylation activity by approx.

L I I 1 ,

0 15 30 45 %b TIME (MINI

Fig. 2. Dephosphorylation of placental membranes. Aliquots of placental membranes (100 pg)

were phosphorylated by incubation at 0” for 10 min in the standard reaction mixture which contained 15 PM [T-~~P]ATP. At this time (t 0) one set of tubes received unlabeled ATP (final concentration 600 JLM, l ) and the control tubes (0) received no additions. At the indicated

times the level of 32P remaining in the membranes was determined as described in Methods.

Protein phosphorylation and epidermal growth factor 193

0 I 2 3 4 5

TIME (MIN)

Fig. 3. Effect of time on the phosphorylation of placental membranes in the presence and

absence of EGF. Reaction mixtures (minus labeled ATP) were preincubated for 1 h at room

temperature in the presence (0) or absence (o) of EGF. The reaction mixtures were then rapidly

cooled to O”, [r-32P]ATP was added, and at the indicated times the level of radioactivity incor-

porated was determined as described in Methods.

35%. The ability of EGF to enhance phosphorylation activity was consistently ob-

served at varying concentrations of placental membranes (25-l 50 pg protein) and of ATP (5-60 PM).

The results shown in Fig. 4 indicate that both the basal and EGF-stimulated phosphorylation activities are enhanced by the addition of glycerol to the reaction mixture. In this experiment the basal level of phosphorylation was increased 55% by the presence of 10% glycerol, while the EGF-stimulated activity was increased 125%. The magnitude of EGF-stimulation of phosphorylation activity was markedly affected by glycerol. The stimulation of phosphorylation by the growth factor was negligible in the absence of glycerol, but was increased to 20, 33 and 50% at glycerol concentrations of 2, 5 and IO%, resp.

Several aspects of the phosphorylation activity in placental membranes are

194 Graham Carpenter, Lonnie Poliner, Lloyd King Jr.

Table 1

Effect of various additions on placental membrane phosphorylation activity

Expt.

No. Additions 32P incorporated (cpm)

-EGF +EGF

1 None 1619 2129

Cyclic GMP (1 O+ M) 1659 1774

Cyclic AMP (1O-5 M) 1648 1837

Dibutyryl cyclic AMP (10m5 M) 1752 1806

None 1199 _

Histone (60 pg) 1449 _

Casein (60 fig) 1466 _

Phosvitin (60 pgj 1538 _

None

Insulin (100 ng) Thyrotropin (100 ngj

Glucagon (100 ng)

Pitressin (100 ng)

FSH (100 ng)

Growth hormone (100 ng)

Prolactin (100 ng)

2050 2413 2193 -

2012 _

1791 _

2082 _

2179 _

2116 _

2103 _

Standard phosphorylation assays (3 min at 0”) were performed as described in Methods after

preincubations with the indicated additions under the following conditions. Expt. 1, the pre-

incubation was for 10 mm at 25” ; Expt. 3, all hormones were preincubated at 25” for 60 min.

In Expt. 2 the phosphorylation assay was carried out for 15 min at 0”.

Table 2

Heat inactivation of placental membrane phophorylation activity

Temperature

eo

s2P incorporated (cpm)

-EGF +EGF

25” 1702 (100%) 2256 (100%)

45” 1229 (72%) 1744 (77%) 50” 1097 (63%) 1382 (61%)

55” 915 (54%) 1080 (48%)

60” 244 (14%) 345 (15%)

Placental membranes were incubated for 10 min at the indicated temperatures and quickly

cooled by dilution into ice-cold buffer (20 mM HEPES, pH 7.4). Aliquots (100 pg) of the

heated membranes were then assayed for phosphorylation activity at O”, as described in

Methods, in the presence or absence of EGF. Values in parentheses indicate the percentage of

activity remaining relative to the unheated control (25’).

Protein phosphorylation and epidermal growth factor 195

I I I

0 2 5 IO

GLYCEROL (percent)

Fig. 4. Influence of glycerol on phosphorylation activity in the presence and absence of EGF.

Placental membranes (100 pg) were preincubated with (0) or without (0) EGF for 1 h at room

temperature in reaction mixtures containing increasing concentrations of glycerol. The mixtures

were rapidly cooled and phosphorylation activity was measured (3 min at 0”) as described in

Methods.

reported in Table 1. The data reported in Expt. 1 indicate that the phosphoryla- tion activity is apparently cyclic nucleotide independent. The ability of the placen- tal phosphorylation system to utilize exogenous substrates is shown in Expt. 2. All 3 exogenous substrates increased the total level of labeled 32P incorporated by a relatively small amount (approx. 30%). Compared to other systems (Carpenter et al., 1979) none of the exogenous proteins is well utilized as a substrate in this phos- phorylation system. The data presented in Expt. 3 shows that several other poly- peptide hormones did not increase phosphorylation in placental membranes under these experimental conditions.

We have examined the heat sensitivity of the phosphorylation activities present in placental membranes (Table 2). Both the basal level of phosphorylation and EGF-stimulated activity were decreased to approx. the same extent following

lo-min incubations at progressively higher temperatures. To further characterize the EGF-stimulated phosphorylation of the endogenous

placental membrane preparation we have subjected the phosphorylated substrates

196 Graham Carpenter, Lonnie Poliner, Lloyd King Jr.

T.D.

I ORIGIN 170K 80K

01 I I L I I I I I I I 0 I 2 3 4 5 6 7 8 9

cm FROM ORIGIN

Protein phosphorylation and epidermal growth factor 197

to SDS-gel electrophoresis and autoradiography. The results shown in Fig. 5A, 5B indicate that the phosphorylation of at least 2 components is significantly enhanced by EGF. These substrates have apparent molecular weights of 170 000 and 2.5 000 and do not correspond to major protein bands detected by Coomassie blue staining. The 170 000 molecular weight phosphorylated component does correspond to a membrane component which is stainable by the PAS method suggesting it may be a sialoglycoprotein. The phosphorylation of a third membrane component having a molecular weight of approx. 150 000 also is increased by EGF, but to a lesser extent than the 2 major phosphorylated bands. Interestingly, the level of radio- activity in a phosphorylated band having a molecular weight of approx. 80000 is

not increased or decreased significantly by EGF.

DISCUSSION

Previously, Carpenter et al. (1978, 1979) demonstrated the ability of EGF to enhance protein phosphorylation in membranes prepared from cultured A431 cells. This cell line, derived from a human epidermoid carcinoma, is unique as it possesses a high concentration of EGF receptors (approx. 2 X 106) per cell (Haigler et al., 1978). O’Keefe et al. (1974) have shown that placental membranes also con- tain a high level of EGF receptor activity. Membranes prepared from human placen- tal tissue and A-431 cells bind, resp., 8 picomoles and 16 picomoles of EGF per

mg membrane protein (O’Keefe et al., 1974; Carpenter et al., 1979). It is thought that membranes which contain high levels of receptor activity may amplify bio- chemical signals generated as a consequence of hormone-receptor interactions. The data presented previously (Carpenter et al., 1978, 1979) and in this report demonstrate that in 2 different systems protein phosphorylation activity is increased by the formation of EGF-receptor complexes.

The EGF-sensitive protein-phosphorylation systems detected in A-431 and pla- cental membranes have several similar characteristics: (1) the requirements for Mg” or Mn2+ are quantitatively similar, (2) both systems are not affected by exo- genous cyclic nucleotides, and (3) in both instances a membrane component having a molecular weight of approx. 170 000 is heavily phosphorylated in the

Fig. 5. Gel electrophoresis of placental membranes phosphorylated in the presence or absence

of EGF. Aliquots of membrane protein (100 pg) were incubated at 0” for 3 min with [Y-~~P]-

ATP in the presence or absence of EGF as described in Methods. The reaction was stopped by

the addition of 40 ~1 of Laemmli SDS sample buffer (Laemmli, 1970). The mixture was boiled for 5 mm and subjected to SDS-gel electrophoresis (7.5% acrylamide) according to the method

of Laemmli (1970). The gels were stained with Coomassie blue or PAS (King and Morrison, 1976), vacuum-dried, and the presence of 32P detected by autoradiography. Fig. SA: lane 1,

molecular weight markers; lane 2, Coomassie blue stain; lane 3, PAS stain; lanes 4 and 5, auto-

radiographs of membrane components phosphorylated in absence and presence of EGF, resp.

Fig. 5B: densitometric scans (Corning Model 740) of autoradiographs.

198 Graham Carpenter, Lonnie Poliner, Lloyd King Jr.

presence of EGF. This component is of interest as studies in which EGF has been covalently crosslinked to membrane proteins have suggested approx. molecular weights for the EGF receptor of 180 000 and 160000 in placenta (Hock et al., 1979) and 190000 in 3T3 cells (Das et al., 1977). Also, the 170000 phos- phorylated component detected in this study is stained by PAS suggesting the presence of a glycoprotein. Previously, studies utilizing lectins have indicated that

the EGF receptor in both placenta (Hock et al., 1979) and human fibroblasts (Car- penter and Cohen, 1977) is glycoprotein in nature. Further studies are clearly necessary to determine whether phosphorylation of the receptor is a consequence of growth-factor binding.

The EGF-sensitive protein-phosphorylation activity detected in placenta differs from that characterized in A-431 membranes in terms of: (1) ability to phosphor- ylate exogenous substrates -which is decreased in the placental system, and (2) heat-sensitivity. The phosphorylation activity in the tumor-cell membrane was rapidly lost at 45” (Carpenter et al., 1979) whereas heating to 60” was required for extensive loss of activity in the placental membranes.

EGF is known to affect the proliferation and differentiation of many fetal tis- sues including skin (Cohen and Elliott, 1963) palate (Hassell, 1975) and lung, esophagus, and trachea epithelia (Sundell et al., 1975). A biological effect of EGF in placental tissue has not been reported. The growth factor does, however, stim- ulate the production of human chorionic gonadotropin in cultured human chorio- carcinoma cells (Benveniste et al., 1978). Also, EGF is present in many human tis- sues including serum (Dailey et al., 1978) and amniotic fluid (Barka et al., 1978). Nevertheless, it is not possible at this time to link EGF receptors and the EGF- sensitive phosphorylation system with a biologic effect in placental tissue. In other systems, utilizing intact cells or.tissues, increased protein phosphorylation has been demonstrated upon the application of different mitogenic agents (Mastro and Rozengurt, 1976; Yu et al., 1978; Costa et al., 1978). The addition of EGF to chick embryos (Huff and Guroff, 1978) or 3T3 cells (Smith et al., 1979) has produced increased phosphorylation, resp., of a nuclear protein and a ribosomal protein. It

appears, therefore, that the phosphorylation of many cellular proteins is a conse- quence of EGF binding which may be related to increased cell proliferation.

ACKNOWLEDGEMENTS

The authors wish to thank Linda Coats and Elane West for their technical assis- tance and Dr. Stanley Cohen for his advice. This work was supported by a Basil O’Connor Research Grant from the National Foundation-March of Dimes (G.C.); a grant CA24071 from the National Cancer Institute (G.C.); a Veterans Administra- tion Clinical Investigator Award (L.K.).

Protein phosphorylation and epidermal growth factor 199

REFERENCES

Barka, T., van der Noen, H., Gresik, E.W., and Kerenyi, T. (1978) Mt. Sinai .I. Med. 45, 679-

684.

Benveniste, R., Speeg, K.V., Carpenter, G., Cohen, S., Lindner, J., and Rabinowitz, D. (1978)

J. Clin. Endocrinol. Metabolism. 46, 169-172.

Bower, J.M., Gamble, R., Gregory, H., Gerring, E.L., and Willshire, I.R. (1975) Experientia 15, 825-826.

Bradford, M.M. (1976) Anal. Biochem. 72, 248-254. Carpenter, G., and Cohen, S. (1976) J. Cell Biol. 71, 159-171.

Carpenter, G., and Cohen, S. (1977) Biochem. Biophys. Res. Commun. 79,545-552.

Carpenter, G., and Cohen, S. (1978) in: Biochemical Actions of Hormones, Vol. V, Ed. G. Lit- wack (Academic Press, New York) pp. 203-247.

Carpenter, G., and Cohen, S. (1979) Annu. Rev. Biochem. 48, 193-216.

Carpenter, G., Kin::, L., and Cohen, S. (1978) Nature (London) 276, 409-410.

Carpenter, G., King, L., and Cohen, S. (1979) J. Biol. Chem. 254,4884-4891.

Cohen, S. (1962) J. Biol. Chem. 237, 1555-1562.

Cohen, S., and Carpenter, G. (1975) Proc. Natl. Acad. Sci. (U.S.A.) 72, 1317-1321.

Cohen, S., and Elliott, G.A. (1963) J. Invest. Dermatol. 40, l-5.

Costa, M., Gerner, E.W., and Russell, D.H. (1978) Biochim. Biophys. Acta 538, l-10. Cuatrecasas, P. (1971) Proc. Natl. Acad. Sci. (U.S.A.) 68, 1264-1268.

Dailey, G.E., Kraus, J.W., and Orth, D.N. (1978) J. Clin. Endocrinol. Metabolism. 46, 929- 936.

Das, M., Miyakawa, T., Fox, C.F., Pruss, R.M., Aharonov, A., and Herschman, H. (1977) Proc. Natl. Acad. Sci. (U.S.A.) 74, 2790-2794.

Haigler, H., Ash, J.F., Singer, S.J., and Cohen, S. (1978) Proc. Natl. Acad. Sci. (U.S.A.) 75, 5025-5029.

Haigler, H.T., McKanna, J.A., and Cohen, S. (1979) J. Cell Biol. 81, 382-395.

Hassell, J.R. (1975) Develop. Biol. 45, 90-102.

Hock, R.A., Nexo, E., and Hollenberg, M.D. (1979) Nature (London) 277,403-405.

Huff, K.R., and Guroff, G. (1978) Biochem. Biophys. Res. Commun. 85,464-472.

King, L.E., and Morrison, M. (1976) Anal. Biochem. 71, 223-230. Laemmh, V.K. (1970) Nature (London) 227,680-685.

Mastro, A., and Rozengurt, E. (1976) J. Biol. Chem. 251,7899-7906.

O’Keefe, E., Hollenberg, M.D., and Cuatrecasas, P. (1974) Arch. Biochem. Biophys. 164,518- 526.

Savage, C.R., and Cohen, S. (1972) J. Biol. Chem. 247, 7609-7611.

Smith, C.J., Wejksnora, P.J., Warner, J.R., Rubin, C.S., and Rosen, O.M. (1979) Proc. Natl. Acad. Sci. (U.S.A.) 76, 2725-2729.

Sundell, H., Serenius, F.S., Barthe, P., Friedman, Z., Kanarek, KS., Escobedo, M.B., Orth, D.N., and Stahlman, M.T. (1975) Pediat. Res. 9, 371.

Yu, M.W., Hori, S., Tolson, N., Huff, K., and Guroff, G. (1978) Biochem. Biophys. Res. Com- mun. 81,941-946.