phosphorylation of the hepatic egf receptor with camp-dependent protein kinase

Molecular and Cellular Endocrinology, 34 (1984) 113- 119 Ekevier Scientific Publishers Ireland, Ltd.

113

MCE 01100

Phosphorylation of the hepatic EGF receptor with CAMP-dependent protein kinase

Wayne R. Rackoff, Richard A. Rubin and H. Shelton Earp Cancer Cell Biology Program, Cancer Research Center, and Department of Medicine, University of North Carolina School of Medicine,

Chapel Hill, NC 27514 fU.S.A.)

(Received 9 August 1983; accepted 7 November 1983)

Keywords: serine phosphorylation; tyrosine phosphorylation; growth regulation; rat liver regeneration.

The 170000 dalton hepatic epidermal growth factor (EGF) receptor is phosphorylated on serine and tyrosine residues. The evidence indicates that distinct protein kinases are involved. Since EGF and agents that elevate CAMP are believed to participate in the regulation of liver regeneration, we tested whether or not the catalytic subunit of CAMP-dependent protein kinase (catalytic subunit), a known serine kinase, would utilize the EGF receptor as a substrate. The catalytic subunit increased phosphorylation of the EGF receptor in purified rat liver plasma membranes. The serine specificity of the catalytic subunit was established by phosphoamino acid analysis of electrophoretically purified EGF receptor. The result was confirmed by catalytic subunit phosphorylation of affinity purified preparations of the EGF receptor. The rates of dephosphorylation of the membrane-associated EGF receptor phosphorylated on different residues were compared. Dephospho~lation of serine residues (after catalytic subunit phospho~lation) was considerably slower (t,/, > 120 set) than the removal of phosphotyrosine after stimulation with EGF

(t ,,2 c 30 see).

Epidermal growth factor (EGF) stimulates DNA synthesis in hepatocytes (Richman et al., 1976; Bucher et al., 1978) and many other cell types (Carpenter and Cohen, 1979). A number of other EGF-mediated effects precede DNA synthesis, including increased glycolytic activity, stimulation of amino acid and sugar transport, and increased rates of protein and RNA synthesis (see Carpenter and Cohen, 1979, for a review). The initial step in EGF action, binding to its 170000 dalton plasma membrane receptor (p170), results in activation of a recently discovered tyrosine- specific protein kinase, and phosphorylation of the receptor (Carpenter et al., 1978; S. Cohen et al., 1980, 1982; Ushiro and Cohen, 1980). The tyro-

sine-specific kinase activity copurifies with the receptor and may be intrinsic to the receptor (S. Cohen et al., 1980, 1982). Tyrosine phosphorylation appears to be one of the earliest biochemical events leading to the physiological effects of EGF.

We have been studying a well-characterized in vivo growth process, rat liver regeneration after partial hepatectomy, which may be regulated by EGF (Earp and O’Keefe, 1981; Rubin et al., 1982) and hormones that increase the intracellular CAMP concentration (Bucher and Swafield, 1976). In crude hepatic membranes the EGF receptor is phosphorylated on both tyrosine and serine residues, but tyrosine phosphorylation is predominant in the presence of EGF (Rubin et al., 1982).

0303-7207/84/$03.00 0 1984 Elsevier Scientific Publishers Ireland, Ltd.

114

However, Hunter and Cooper (1981) have re-

ported that phosphoserine is the predominant phosphoa~no acid in the EGF receptor of EGF- stimulated intact A431 cells.

Several lines of evidence have suggested that the

serine kinase is distinct from the intrinsic tyrosine

kinase activity. First, hepatic membrane preparations solubilized with Triton X-100 exhibit much

less serine kinase activity even though tyrosine kinase activity is enhanced (Rubin and Earp,

1983b). Second, in plasma membranes, EGF and the polar solvent dimethylsulfoxide (DMSO) selectively stimulate tyrosine-residue phosphorylation of the EGF receptor, while serine phosphorylation

is unaffected (Rubin and Earp, 1983a). These data, and a rise in the protein kinase modulator CAMP

during liver regeneration (MacManus et al., 1972; Koide et al., 1978), lead us to the present study of whether or not the active catalytic subunit of CAMP-dependent protein kinase, a known serine kinase, could phospho~~ate serine residues on the

EGF receptor.

Materials and methods

Materials Male Sprague-Dawley rats (200-500 g) were

obtained from Charles River Breeding Laborato- ries (Boston, MA) and were fed and watered ad libitum. [ y- 32 P]ATP (1000-3000 Ci/mmole) was

purchased from New England Nuclear (Boston, MA). Acrylamide, his-acryla~de, phosphoamino

acid standards and bovine heart protein kinase catalytic subunit were obtained from Sigma Chem- ical Co. (St. Louis, MO). Purified bovine heart catalytic subunit prepared by the method described in Sugden et al. (1976) was generously provided by J. Corbin of Vanderbilt University. The activity of the catalytic subunit obtained from J. Corbin is 3 ~moles/min/mg protein of enzyme. The activity of the catalytic subunit obtained from Sigma is 25 nM phosphate incorporated into casein/~n/mg of enzyme. Precoated cellulose thin-layer chromato~aphy plates (0.1 mm) were purchased from E. Merck (Darmstadt, F.R.G.). Mouse EGF was purified from submaxillary glands by the method of Savage and Cohen (1972). Wheat germ agglutinin agarose was purchased from E-Y Laboratories (San Mateo, CA).

Membrane preparation Freshly prepared membranes were used for all

experiments except those performed with purified EGF receptor. Male Sprague-Dawley rats were anesthetized with ether and the livers excised be-

tween 8 and 9 a.m. Tissues were homogenized in 0.25 M sucrose/l0 mM Tris-HCl, pH 8.0, and

plasma membranes were prepared by the method

of Touster et al. (1970). The plasma membranes were collected, centrifuged at 105000 x g and re- suspended in 20 mM Pipes buffer, pH 7.0. Protein

concentration was determined by the method of Lowry et al. (1951).

Ahquots of membrane suspensions (35-50 pg protein) were normally phosphorylated in 50 ~1 with 50 mM Pipes buffer, 0.5 mM MnCl,, EGF or catalytic subunit at the indicated concentrations, and 1 FM [Y-~‘P]ATP (usually 5 pCi/tube). Mem- branes were preincubated with EGF for 10 min at

0°C or with catalytic subunit for 1 min at 30°C before the addition of ATP. The reaction was

initiated with radiolabeled ATP and stopped at indicated times. Aliquots (50 ~1) were subjected to electrophoresis on SDS/S% polyac~l~de slab gels with a 3% stacking gel, essentially as described by Rubin and Earp (1983b). Gels were stained with Coomassie Brilliant Blue, destained, dried, and autoradiographed on Kodak XAR-5 film. Ex- posure times varied from 5 h to 3 days at - 70°C. Autoradiographs were scanned with a Kontes densitometer, and peak integration was performed on-line by a Hewlett-Packard reporting integrator.

Phosphoamino acid analysis Analysis of the EGF receptor (~170) was car-

ried out by phospho~lating membranes (150 rg) with 1 PM [Y-~~P]ATP (50 pCi/tube) in 250 pi for

45 sec. The samples were preincubated for 1 min at 0°C with 1.25 fig/sample of EGF or catalytic subunit. 0.5 mM MnCl, was present throughout. Following electrophoresis, the ~170 band was cut out of the gel, el~trophoreticalIy eluted, and subjected to hydrolysis with HCl in vacua as previ- ously described (Rubin and Earp, 1983b). The amino acids were separated on thin-layer chromatography plates by electrophoresis and ascend- ing chromatography, according to the method of

115

Hunter and Sefton (1980). Plates were autoradiographed to identify phosphorylated residues, and the ninhydrin-labeled phosphoamino acid standard spots were scraped from the plates and counted in a liquid scintillation counter.

EGF receptor preparations purified by wheat germ agglutinin affinity and EGF-affi~ty chromatography were also studied. EGF receptor was purified from Triton-solubilized plasma membranes on wheat germ agglutinin agarose. Mem- branes and gel were mixed for 20 mm at 4”C, the gel was washed 3 times with 20 ~01s. of buffer, and elution was performed by shaking with 1 vol. of 0.1 M N-acetylglucosamine for 15 mm at 4°C. 0.5% Triton, 5% glycerol, 20 mM Pipes, pH 7.0, and 0.15 M NaCl was present throughout. EGF- affinity chromatography was performed using Tri- ton-solubilized membranes according to the method of S. Cohen et al. (1982). The experiments with these preparations were carried out as described above, except that 24.5-40.0 ~1 of the various eluates were used, and preincubation was extended to 30 min at 21°C.

Dephosphorylation studies were performed after the standard phosphorylation protocol. Excess un- labeled ATP (1 mM final concentration) was added to the reaction mixture in order to attenuate the radiolabeling process. Samples were then in- cubated at 30°C for the indicated times and the reaction was stopped. Samples were analyzed by electrophoresis, autoradiography and densitometry as described above.

Results

Plasma membranes (35 pg protein) were phosphorylated with catalytic subunit (1 pg/tube) with various concentrations of MgCl, and MnCl,. The densitometry readings from the autoradiogram are plotted in Fig. 1 and demonstrate the ionic cofactor requirement of the reaction. Mn” supported the reaction to a greater extent than Mg2+. Opti- mal stimulation was obtained at a concentration of 3 mM Mn2+. However, other experiments demonstrated that l-10 mM Mn2” raised the basal as well as the catalytic subu~t-stimulated phosphorylation (data not shown). Therefore, subsequent phosphorylations with catalytic subunit were performed at 0.5 mM Mn”.

Fig. 1. Ionic cofactor requirement for the phosphorylation of hepatic EGF receptor with catalytic subunit. Plasma membranes (JO pg) were phosphotylated for 1 min (as described in Materials and Methods) in different concentrations of MgCl, or MnCl,. The samples were subjected to So-~lya~ia~de elcctrophoresis and autoradiograph~ for 24 h. The data plotted here are the densitometry readings from the autoradiograph. The control readings (no catalytic subunit added) were ap- proximately 3 for both Mg2+ and Mn2+. Densitometry readings are expressed in arbitrary units.

The time-course of EGF receptor phosphorylation was studied in the presence and absence of catalytic subunit (Fig. 2). The final ATP concentration was increased to 5 PM (10 pCi [y32PJATP/tube) to be sure that ATP depletion was not responsible for cessation of the reaction at

TIME I Secondsl

Fig. 2. Time-course for the reaction of hepatic EGF receptor with catalytic subunit. Plasma membranes (50 pg) were phosphorylated with 1 pg total of catalytic subu~t/s~pIe for the indicated time (as described in Materials and Methods), subjected to SDS-poiyacrylamide electrophoresis, autoradiography (45 h) and densitometry.

116

longer time-points. Catalytic subunit (1 pg/tube)

stimulated EGF receptor phosphorylation at all time-points tested. Maximal phosphorylation was

Fig. 3. Effects of EGF stimulation and catalytic subunit phos-

phorylation on a preparation of wheat germ agglutinin affinity

purified EGF receptor. Purified receptor was phosphorylated

(as described in Materials and Methods) in the presence of

EGF (0.5 pg/sample) and catalytic subunit (1 p&sample) as

indicated after a 30 min preincubation at 21% The samples

were subjected to SDS-polyacrylamide electrophoresis and autoradiography (8 h). The arrow indicates the 170000 dalton

EGF receptor.

seen at 45 sec. Maximal stimulation over the control (100% increase) was seen at 60 sec. In the light of these findings, phosphorylations were per-

formed at times between 30 and 60 sec. Fig. 3 demonstrates the effect of EGF stimula-

tion and catalytic subunit phosphorylation on a preparation of the EGF receptor (29.5 pi/assay) that was purified using a wheat germ agglutinin affinity resin. The combined effects of EGF and catalytic subunit were also examined. The receptor preparation was preincubated with EGF and catalytic subunit for 30 min at 21°C before the addi-

tion of ATP. Phosphorylations were carried out at 30°C for 1 min. Basal phosphorylation was negli-

gible. Densitometry readings reveal that both EGF and catalytic subunit increased phosphorylation. The combined effect of EGF and catalytic subunit was nearly additive (Table 1). The effect of catalytic subunit appears to be independent of interac- tion between EGF and its receptor.

Phosphorylation of the EGF receptor by catalytic subunit was also observed using EGF receptor that had been purified using EGF-affinity chromatography. This preparation had only two bands (~170 and ~43) upon Coomassie blue stain- ing. Fig. 4 is the autoradiograph of an experiment that compares phosphorylation of the EGF receptor with two preparations of catalytic subunit. The first two lanes are an assay of commercially obtained catalytic subunit (preparation A) phos-

TABLE 1

PHOSPHORYLATION OF WHEAT GERM AGGLUTININ

AFFINITY PURIFIED EGF RECEPTOR

Densitometry

(arbitrary units)

-Fold increase

Control

+ EGF

+ Catalytic subunit

+ EGF and catalytic

subunit

1.9

9.0 3.7

5.2 1.7

16.5 7.7

Phosphorylation of wheat germ agglutinin affinity purified

EGF receptor in the presence of EGF and catalytic subunit.

The autradiograph in Fig. 3 was scanned using a Kontes

densitometer, and peak integration was performed on-line by a

Hewlett-Packard reporting integrator. Densitometry readings

are expressed in arbitrary units since they are relative measures. -Fold increase is calculated by dividing each reading by the

control reading and subtracting 1.

Fig. 4. Phosphorylation of EGF-affinity purified hepatic EGF receptor with two preparations of catalytic subunit. Phosphory-

lations were carried out without (lanes 1 and 3) and with (lanes

2 and 4) purified receptor. The top arrow is the ~170 EGF

receptor; the bottom arrow is ~43. Samples were phosphorylated as described in Materials and Methods, and subjected to

SDS-polyacrylamide electrophoresis and autoradiography (3

days). 1 pg of catalytic subunit was present in each assay.

Preparation A is commercially obtained; preparation B is a

purified preparation (see Materials).

TABLE 2

PHOSPHOAMINO ACID ANALYSIS (COUNTS PER MINUTE)

117

phorylated without (lane 1) and with (lane 2)

addition of EGF-affinity purified EGF receptor. Lanes 3 and 4 are the results of an identical assay in which a highly purified preparation of catalytic

subunit (preparation B) has been substituted for

the commercial preparation. The comparison between commercially prepared catalytic subunit and

a highly purified preparation indicated that some phosphorylated bands seen in Fig. 3 were contami- nants from the commercial catalytic subunit. The major contaminant in the purified catalytic subunit is bovine serum albumin. Figs. 3 and 4 confirmed the initial finding that in plasma membranes the ~170 phosphorylated by catalytic subunit was, indeed, the hepatic EGF receptor.

Phosphoamino acid analysis demonstrates the serine specificity of the phosphorylation with catalytic subunit. The data in Table 2 indicate that

catalytic subunit selectively increased the incor-

poration of phosphate into serine residues relative to the untreated control (21.6-fold increase). In contrast, EGF increased the incorporation of phosphate into tyrosine residues relative to the untreated control (4.1-fold increase).

After establishing the activity and specificity of catalytic subunit, the dephosphorylation reaction was studied. [ 32P]Phosphoseryl EGF receptor (phosphorylated with catalytic subunit) and [ 32P]phosphotyrosyl EGF receptor (phosphorylated in response to EGF stimulation) were dephosphorylated at markedly different rates (Fig. 5). The t,,, dephosphorylation for p-Ser was over 2 min. In’contrast, the t,,, dephosphorylation

p-Tyr was less than 30 sec. for

Control (30°C) Catalytic subunit (30°C)

Control (OV)

EGF (0°C)

Ser

73.8 1667.8

80.8

30.3

(-Fold change)

(22)

(0.37)

Tyr

532.5 454.2

354.9

1818.2

(-Fold change)

(0.85)

(5.1)

Thr

16.1

44.2

5.0

30.5

(-Fold change)

(2.7)

(6.1)

Phosphoamino acid analysis of purified hepatic EGF receptor phosphorylated in the presence of EGF and catalytic subunit. Plasma

membranes were phosphorylated with EGF or catalytic subunit, submitted to SDS-polyacrylamide electrophoresis, and the 170000 dalton hepatic EGF receptor band cut out and electroeluted. The purified receptor was acid hydrolyzed and analyzed for phosphoamino acid content by two-dimensional thin-layer chromatography. Labeled spots were scraped from the plates and counted in a scintillation counter to obtain the data that are presented here. See Materials and Methods for details.

118

I I I 1 4

30 60 120 240

TIME (Seconds)

Fig. 5. Dephosphorylation time-course after phosphorylation of

the hepatic EGF receptor with EGF and catalytic subunit.

Plasma membranes were phosphorylated and dephosphorylated

as described in Materials and Methods, and stopped at the

indicated time-points. Either 0.5 pg EGF or 0.5 pg catalytic

subunit were used in each assay. The samples were subjected to

SDS-polyacrylamide electrophoresis, autoradiography (8 h) and

densitometry.

Discussion

Recent reports suggest that CAMP-dependent protein kinase has a role in the phosphorylation of

membrane-bound substrates in a physiological manner. Costa et al. (1982) have demonstrated the phosphorylation of the a-subunit of the sodium channel in rat brain. Huganir and Greengard (1983) have demonstrated the phosphorylation of the nicotinic acetylcholine receptor in postsynaptic membranes isolated from the electric organ of Tu~~e~o ~~I~or~i~~. Both studies suggest that the reactions take place in a physiological manner.

The present study investigates the possibility that CAMP-dependent protein kinase may be involved in growth regulation through phosphorylation of another membrane-bound substrate, the hepatic EGF receptor. CAMP levels are elevated early during the course of liver regeneration (Mac- Manus et al., 1972; Thrower and Ord, 1974; Koide et al., 1978); EGF has been implicated in liver regeneration because of its ability to stimulate hepatic DNA synthesis (Richman et al., 1976;

Bucher et al., 1978), and ‘down-regulation’ of the EGF receptor during liver regeneration (Earp and O’Keefe, 1981). Leof et al. (1982) have observed that increases in intracellular CAMP potentiate the ‘down-regulation’ of the EGF receptor when Balb/c-3T3 cells are stimulated with platelet-de- rived growth factor (PDGF). In addition, Pessin et al. (1983) have observed that P-adrenergic agonists cause a marked decrease in EGF binding in intact

rat adipocytes. The P-adrenergic effect, presuma- bly mediated by the CAMP-CAMP-dependent pro-

tein kinase system, is due to a decrease in the apparent affinity of the EGF receptor. The in vitro phosphorylation of the EGF receptor by the pro-

tein kinase catalytic subunit reported here fulfills an initial criterion that is necessary to directly link the CAMP and EGF regulatory systems.

The phosphoamino acid analyses of plasma membrane-associated EGF receptor confirmed the specificity of EGF stimulation for tyrosine-residue phosphorylation and demonstrated the specificity of catalytic subunit for serine-residue phosphorylation. Threonine residues were minimally phosphorylated in all of the preparations (Table 2).

The relative specificities of EGF and catalytic subunit allowed us to study the rates of dephosphorylation of EGF receptor p-Tyr and p-Ser (Fig. 5). The dephosphorylation data suggest that p-Ser provides a lingering signal on the EGF receptor while p-Tyr provides a quick one. This may par-

tially explain the observations of those studying the EGF receptor and the insulin receptor in vivo. They have seen a predominance of p-Ser upon phosphoamino acid analyses of the respective re- ceptors (Hunter and Cooper, 1981; Kasuga et al., 1982).

Multi-site phosphorylation is a well-established and simple mechanism for increasing the regulatory potential of an enzyme. P. Cohen (1982) notes that multi-site phosphorylation is the rule rather than the exception for enzymes regulated by phosphorylation. The src kinase (the transforming gene product of the Rous sarcoma virus) can also be phospho~lated by both a tyrosine-specific kinase and CAMP-dependent protein kinase (Erikson et

al., 1980). The significance of this dual-site phosphorylation is not yet clear. Thus, serine phosphorylation could be a pathway through which CAMP levels are integrated with the signal pro-

119

vided by tyrosine kinase activity.

The results reported in this paper address only in vitro serine phosphorylation of the EGF receptor. The physiological significance of these results awaits elucidation. The recovery and analysis of the EGF receptor from whole cells treated with CAMP should provide important data on the in vivo role of CAMP-dependent protein kinase in the regulation of the EGF receptor and in the broader process of growth regulation.

Acknowledgements

We gratefully acknowledge the skillful technical assistance of Cindy Edwards, Kathy Austin, Ruth Dy and Joyce Blaisdell. The study was supported

by a grant from the National Institutes of Health (AM-30002) and a grant-in-aid from the American Heart Association. W.R.R. was supported by a

Biomedical Research Support Grant (No. 2-SO7- RRO-5-406-22). R.A.R. was a National Cancer Institute postdoctoral fellow (5T32CA09.56). H.S.E. is an established investigator of the Ameri- can Heart Association, supported in part with funds provided by the North Carolina Heart As- sociation.

References

Bucher, N.L.R. and Swafield, M.N. (1976) Adv. Enzyme Regul.

13, 281-293.

Bucher, N.L.R., Patel, U. and Cohen, S. (1978) Adv. Enzyme

Regul. 16, 205-213.

Carpenter, G. and Cohen, S. (1979) Annu. Rev. B&hem. 48,

193-216.

Carpenter, G., King, L. and Cohen, S. (1978) Nature (London)

276, 409-410.

Cohen, P. (1982) Nature (London) 296, 613-620.

Cohen, S., Carpenter, G. and King, L. (1980) J. Biol. Chem.

255, 4834-4842.

Cohen, S., Ushiro, H., Stoscheck, C. and Chinkers, M. (1982) J.

Biol. Chem. 257, 1523-1531.

Costa, M.R.C., Casnellie, J.E. and Catterall, W.A. (1982) J.

Biol. Chem. 257, 7918-7921.

Earp, H.S. (1980) J. Biol. Chem. 255, 8979-8982.

Earp, H.S. and O’Keefe, E.J. (1981) J. Clin. Invest. 67,

1580-1583.

Erikson, R.L., Purchio, A.F., Erikson, E., Collett, M.S. and

Brugge, J.S. (1980) J. Cell Biol. 87, 319-325.

Huganir, R.L. and Greengard, P. (1983) Proc. Natl. Acad. Sci.

(U.S.A.) 80, 1130-1134.

Hunter, T. and Cooper, J.A. (1981) Cell 24, 742-752.

Hunter, T. and Sefton, B. (1980) Proc. Natl. Acad. Sci. (U.S.A.)

77,1311-1315.

Kasuga, M., Zick, Y., Blith, D.L., Karlsson, F.A., Hating, H.U.

and Kahn, C.R. (1982) J. Biol. Chem. 257, 9891-9894.

Koide, Y., Earp, H.S., Ong, S. and Steiner, A.L. (1978) J. Biol.

Chem. 253, 4439-4445.

Leof, E.B., Olashaw, N.E., Pledger, W.J. and G’Keefe, E.J.

(1982) Biochem. Biophys. Res. Commun. 109, 83-91.

Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, R.J.

(1951) J. Biol. Chem. 193, 265-275.

MacManus, J.P., Franks, D.J., Yondale, T. and Braceland,

B.M. (1972) Biochem. Biophys. Res. Commun. 49, 1201-

1207.

Richman, R.A., Claus, T.H., Pilkis, S.J. and Friedman, D.L.

(1976) Proc. Natl. Acad. Sci. (U.S.A.) 73, 3589-3593.

Rubin, R.A. and Earp, H.S. (1983a) Science 219,60-63.

Rubin, R.A. and Eat-p, H.S. (1983b) J. Biol. Chem. 258,

5177-5182.

Rubin, R.A., G’Keefe, E.J. and Earp, H.S. (1982) Proc. Natl.

Acad. Sci. (U.S.A.) 79, 776-780.

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

7609-7611.

Sugden, P.H., Holladag, L.A., Reimann, E.M. and Corbin, J.D.

(1976) B&hem. J. 159, 409-422.

Thrower, S. and Ord, M.G. (1974) B&hem. J. 144, 361-369.

Touster, O., Aronson Jr., N.N., Dulaney, J.T. and Hendrick-

son, H. (1970) J. Cell Biol. 47, 604-617.

Ushiro, H. and Cohen, S. (1980) J. Biol. Chem. 255,8363-8365.

phosphorylation of the hepatic egf receptor with camp-dependent protein kinase

Documents