Prolactin Decreases Epidermal Growth Factor Receptor Kinase Activity via a Phosphorylation-dependent Mechanism*

Previously, we have shown that prolactin inhibits epidermal growth factor (EGF)-induced mitogenesis in mouse mammary epithelial cells without altering the response to other growth promoting agents. This effect has been associated with reduced EGF-induced EGF receptor (EGFR) tyrosine phosphorylation, Grb-2 association, and Ras activation. Our current hypothesis is that prolactin induces an alteration in EGFR kinase activity via a phosphorylation-dependent mechanism. To test this hypothesis, we treated normal murine mammary gland cells with or without 100 ng/ml prolactin. EGFR isolated by wheat germ agglutinin affinity chromatography from nontreated cells exhibited substantial ligand-induced phosphorylation, and EGFR isolated from prolactin-treated cells displayed minimal EGF-induced EGFR phosphorylation, as well as decreased kinase activity toward exogenous substrates. The observed decrease in ligand-induced EGFR phosphorylation could not be attributed to either differential amounts of EGFR, decreased EGF binding affinity, or the presence of a phosphotyrosine phosphatase or ATPase. EGFR isolated from prolactin-treated cells exhibited increased phosphorylation on threonine. Removal of this phosphorylation with alkaline phosphatase restored EGFR kinase activity to levels observed in nontreated cells. Therefore, these results suggest that prolactin antagonizes EGF signaling by increasing EGFR threonine phosphorylation and decreasing EGF-induced EGFR tyrosine phosphorylation.

lular domain, a transmembrane domain, and a cytoplasmic domain that contains a 300-amino acid sequence that is similar to the pp60 c-src tyrosine kinase (3,4). Although derived from a single gene, EGFR has been shown to be present on the membranes of most cell types, including mammary gland epithelium, as both high affinity (k D ϭ 0.1 nM) and a low affinity (k D ϭ 1 nM) sites (5,6). The EGFR has also been shown to bind transforming growth factor-␣ with similar affinities (7).
Upon ligand binding, the EGFR can homo-or heterodimerize with other members of the EGFR family including erbB2, erbB3, and erbB4 (2,8,9). Following dimerization, the receptor undergoes inter-and intramolecular autophosphorylation on tyrosines in the C-terminal cytoplasmic domain (3,10,11,12). These phosphotyrosines serve as docking sites for Src homology 2-containing proteins such as Grb2, phospholipase C-␥, phosphatidylinositol 3-kinase, and Shc (13,14). Upon binding of Src homology 2 proteins, there is a cascade of kinase activation, which leads eventually, in the case of Grb2, to the stimulation of the serine/threonine kinase MAPK, but other Src homology 2 kinase activation mechanisms do exist (10,15,16). Once activated, MAPK can either phosphorylate cytoplasmic substrates such as p90 rsk (16,17) or translocate to the nucleus to induce the production of transcription factors such as c-Fos (16,18).
In addition to tyrosine phosphorylation sites needed for signal transduction, certain serine and threonine residues exist in the cytoplasmic portion of the EGFR that, upon phosphorylation, reduce EGF binding, induce receptor desensitization, or promote down-regulation (10, 19 -21). While some sites have been proposed to regulate one specific event, e.g. down-regulation, some sites may participate in several mechanisms of EGFR modulation (22). Mutation of specific sites renders the receptor insensitive to homologous or heterologous regulation (reviewed in Refs. 21 and 22). The best described of these events is mediated by protein kinase C (19,20), but others have been reported (21,22).
Our laboratory has previously reported Prl's inhibition of EGF-induced DNA synthesis in a dose-dependent manner in mammary epithelial cells without affecting response to other growth-promoting agents such as insulin-like growth factor-1 or cholera toxin (23). Prl also inhibited the ability of EGF to activate the Ras-MAPK pathway, an effect that was correlated with altered EGF receptor tyrosine phosphorylation and EGFR-Grb2 interactions (24). While these observations suggest that Prl may alter EGF receptor kinase activity (25), the results are based on altered EGFR phosphorylation state in vivo, which does not indicate whether Prl alters EGF receptor activity directly. Therefore, the objective of the present study was to determine the effects of Prl on EGF-induced kinase activity of isolated EGFR.

EXPERIMENTAL PROCEDURES
Cell Culture-Normal murine mammary gland epithelium (NMuMG) cells and NIH 3T3 cells were purchased from American Type Culture Collection (ATCC, Rockville, MD). Cells were grown in Falcon * This work was supported by National Institutes of Health Grant HD 31532 and U.S. Department of Agriculture Grant WIS 3769. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
EGF Receptor Isolation-Isolation of EGFR was performed as described by Yarden and Schlessinger (12) with minor modifications. Cells (10 plates each treatment; 3 ϫ 10 7 cells/plate) were serum-starved for 18 h in DMEM and then treated with or without 100 ng/ml ovine prolactin (NIDDK, National Institutes of Health) in DMEM for 15 min. Cells were then washed with cold (4°C) phosphate-buffered saline (PBS) and lysed with 2 ml of cold (4°C) lysis buffer (50 mM HEPES, 1.0% Triton X-100, 100 g/ml aprotinin and 2 mM phenylmethylsulfonyl fluoride, pH 7.6). Lysates were centrifuged at 50,000 ϫ g for 30 min at 4°C, and 3 ml of the cell lysate supernatant was applied to a 2-ml gel bed of agarose-conjugated wheat germ agglutinin (Vector Laboratories, Burlingame, CA) in a 10-ml plastic column (Bio-Rad) in a 4°C refrigerator. The column was allowed to equilibrate for 15 min by stopping the flow of supernatant over the gel bed. The full supernatant volume was then passed over the column three times. The column was then washed with 30 ml of wash buffer (50 mM HEPES and 0.1% Triton X-100, pH 7.6). EGF receptor was eluted off the column by the addition of 3 ml of elution buffer (50 mM HEPES, 0.1% Triton X-100, and 300 mM N-acetylglucosamine, pH 7.6). Protein content was determined either by BCA protein assay (Pierce) or Bradford method (26). Eluent preparations were aliquoted into microcentrifuge tubes and stored at Ϫ70°C until use. Receptors were used within approximately 2 months after isolation, over which time we did not observe any decrease in kinase activity.
Receptor Kinase Assay-Kinase activity of solubilized EGFR was assessed as follows. Equal amounts of EGFR isolated from NMuMG or NIH 3T3 cells treated with or without the indicated concentrations of Prl were incubated with 1 M [␥-32 P]ATP (NEN Life Science Products; specific activity, 300 -500 Ci/mmol), 1 mM MgCl 2 with or without the indicated concentrations (0 -100 ng/ml; 0 -17 nM) of murine EGF (Upstate Biotechnologies, Inc., Lake Placid, NY) at pH 7.6 and vortexed briefly. The reaction was allowed to proceed for the indicated times at room temperature and was terminated by the addition of Laemmli SDS loading dye (27). Reaction mixtures were then heated to 100°C for 5 min and separated on a 7.5% SDS-PAGE gel. Gels were then fixed overnight using 10% acetic acid, 10% methanol, and 80% water, dried, and exposed to x-ray film (Fuji) at Ϫ70°C. Relative band intensities were determined using Collage ® image analysis system (Fotodyne, New Berlin, WI).
Western Blots-To determine if equal amounts of receptor were present in assay experiments or to determine the phosphorylation state of isolated EGF receptor, equivalent amounts of protein from each receptor preparation eluent were separated on a 7.5% SDS-PAGE. Gels were then transferred to Immobilon-P membrane (Millipore, Bedford, MA). Membranes were then blocked in phosphate-buffered saline and 3% Tween 20 (PBS-T) supplemented with 2% bovine serum albumin for 30 min at room temperature or overnight at 4°C. In the case of equal EGFR determination, membranes were then incubated with mouse anti-EGF receptor (Transduction Laboratories, Lexington, KY), while mouse anti-phosphothreonine (Sigma) or mouse anti-phosphoserine (Sigma) were used to determine phosphorylation state of the EGFR. In either case, membranes were incubated with antibodies for 1.5 h, washed three times for 10 min each in PBS-T containing 0.1% bovine serum albumin and then incubated with sheep anti-mouse IgG peroxidase conjugate (Sigma). Blots were then washed again three times for 10 min each in 0.1% PBS-T. Protein was determined by enhanced chemiluminescence per the manufacturer's specifications (NEN Life Science Products), and membranes were exposed to x-ray film. Relative band intensities were determined by Collage ® .
Preparation of 125 I-EGF-Murine receptor grade EGF (Harlan, Madison, WI) was iodinated using IODO-GEN ® (Pierce) protocol with some modifications. Briefly, 20 g of IODO-GEN ® , which was coated on the bottoms of acid-washed 12 ϫ 75-mm borosilicate tubes, was reacted with 5 g of EGF and 1 mCi of Na 125 I (NEN Life Science Products) for 3 min at room temperature. 125 I-EGF was then applied to a Sephadex G25-80 column, and fractions were collected, counted, and stored at Ϫ20°C until use.
125 I-EGF Binding-125 I-EGF binding was performed on solubilized receptors from cells treated with or without Prl as follows; reaction tubes were coated with ϳ200 l of Sigmacote (Sigma) and vortexed briefly, excess liquid was aspirated off, and tubes were allowed to dry at room temperature for 30 min. Reaction mixtures contained ϳ50 g of wheat germ agglutinin eluent protein, variable concentrations of 125 I-EGF (0.1-10 nM) (specific activity, 1.1 Ci/pmol) with or without unlabeled concentrations of EGF at a 100-fold excess for each labeled EGF concentration. Binding was allowed to proceed at room temperature for 2 h and terminated by the addition of ice-cold human gamma globulin (fraction II; Sigma), final concentration 1 mg/ml. To precipitate EGFR-125 I-EGF complexes, ice-cold 6000 molecular weight polyethylene glycol (Sigma) (final concentration, 125 mg/ml) was added and incubated on ice for 10 min. The tubes were centrifuged at 9000 ϫ g, and the pellet was washed with ice-cold 125 mg/ml polyethylene glycol. The tips of the reaction tubes, which contained the precipitate, were excised and counted using a Minaxi-Tricarb 4000 series scintillation counter (Packard/United Technologies, Downers Grove, IL). Receptor affinities were determined using the Ligand ® program (Biosoft, Ferguson, MO).
Peptide Phosphorylation-To determine if Prl decreases EGFR kinase activity toward exogenous substrates, equal amounts of EGFR isolates from NMuMG cells treated with or without Prl were incubated with 20 mM HEPES, 2 mM MnCl 2 , 50 M Na 3 VO 4 , 3 mM NaCl, 0.15 M 2-mercaptoethanol, 24 M EDTA, 20 M [␥-32 P]ATP (20 Ci/mmol), with or without 100 ng/ml EGF, with or without 6 mM Src (RRLIEDADY-AARG) (Sigma and Calbiochem, San Diego, CA) or 10 mM angiotensin II (Sigma) peptides. In each case, control reactions containing all components except EGFR isolates or peptide were utilized. Reaction mixtures were preincubated at room temperature with or without 100 ng/ml of EGF and prior to the addition of peptide. Kinase reactions were initiated by the addition of ATP and allowed to proceed for 12 min, upon which 100 g of bovine serum albumin (Sigma) was added to the mixture followed by trichloroacetic acid (5% final). Reaction mixtures were vortexed and centrifuged at 16000 ϫ g for 10 min. Aliquots of the supernatant were spotted onto P-81 phosphocellulose paper (Whatman Lab Division, Maidstone, United Kingdom). P-81 paper was then washed three times for 5 min each in 85 mM phosphoric acid (pHϳ 1.5), removed, and counted for adsorbed radioactivity in a liquid scintillation counter.
Phosphoamino Acid Analysis-To determine if Prl induces the phosphorylation of the EGFR in the intact cell, 32 P metabolic labeling was utilized. 30 million NMuMG cells were seeded onto Falcon Integrid ® 150-mm dishes (Becton-Dickinson) and grown to near confluency. Cells were serum-starved for 16 h in serum-free DMEM, which was then changed to phosphate-free DMEM supplemented with 10 M phosphate containing [ 32 P]orthophosphoric acid (NEN Life Science Products; 12.5 Ci/mmol) and allowed to incubate at 37°C for 4 h. Cells were then treated with or without 100 ng/ml prolactin for 15 min. Medium was removed, and cells were lysed with 2 ml of cold (4°C) lysis buffer (10 mM Tris-HCl, 1% Triton X-100, 5 mM EDTA, 30 mM sodium pyrophosphate, 50 M sodium fluoride, 100 M sodium orthovanadate, and 1 mM phenylmethylsulfonyl fluoride, pH 7.6). Cell lysates were centrifuged at 16,000 ϫ g for 30 min at room temperature. The supernatant was then removed and placed in a separate tube, and 5 g of sheep polyclonal anti-EGFR (Upstate Biotechnology) with 25 l of agarose-bound protein A/G plus (Santa Cruz Biotechnologies) was added and incubated at room temperature for 1.5 h, manually inverting the tubes every 15 min. Tubes were centrifuged at 300 ϫ g for 5 min and washed three times with lysis buffer. The final precipitate was resuspended in 100 l of SDS-loading dye and separated on a 7.5% SDS-PAGE gel. Gels were wrapped in plastic wrap and exposed to x-ray film at room temperature. The band in the 170-kDa range was excised and put in a 4-ml glass vial with 400 l of high pressure liquid chromatography grade 6 N HCl (Pierce). This mixture was then heated for 60 min at 110°C. The HCl was then removed under vacuum. 100 l of a phosphoamino acid solution, which contained phosphoserine, phosphothreonine, and phosphotyrosine (each at 0.33 mg/ml; all purchased from Sigma) in 5% acetic acid and 0.5% pyridine was used to resuspend the residue. Samples were then spotted onto a 20 ϫ 20 cellulose thin layer chromatography plate (EM Science, Gibbstown, NJ). The plate was then moistened with 5% acetic acid and 0.5% pyridine, and thin layer electrophoresis was performed at 1.3 kV for 3.5 h on a cooling plate. The plate was allowed to dry, sprayed with 0.2% ninhydrin in acetone, and then put in an oven for 15 min to expose the standards. The plate was then exposed to x-ray film at Ϫ70°C.
Dephosphorylation of EGFR-Dephosphorylation of EGFR was performed as described previously (28) with some modifications. Briefly, equal amounts of isolated EGFR from NMuMG cells treated with or without Prl were incubated with agarose-conjugated alkaline phosphatase (50 diethanolamine units; pHϳ 9) (Sigma) for 2 h at room temperature. Afterward, the reaction mixture was centrifuged at 320 ϫ g for 10 min, and a portion of the supernatant was transferred to another tube. To ensure that any residual alkaline phosphatase would not interfere with subsequent kinase reactions, sodium pyrophosphate, final concentration 30 mM, was added at pH 7.8. Kinase reactions were then performed and analyzed as described above.

RESULTS
EGFR from non-Prl-treated NMuMG cells exhibited substantial EGF-induced phosphorylation of a 170,000 molecular weight protein (pp170 kDa; Fig. 1A). Prl treated NMuMG cells exhibited significantly less EGF-induced p170 phosphorylation, which parallels our observations in intact NMuMG cells (24,25). However, EGFR isolated from NIH 3T3 cells, which have been shown to be Prl-unresponsive (29), exhibited no reduction in EGF-induced p170 phosphorylation (Fig. 1B). Western blot analysis indicated that the observed decrease in EGF-induced EGFR phosphorylation induced by Prl could not be attributed to differential amounts of receptor in the kinase assays (Fig. 2). EGFR immunoprecipitation studies revealed that ϳ70% of detectable pp170 was recovered in the pellet fraction. Phosphoamino acid analysis showed that EGF-induced EGFR phosphorylation was predominantly on tyrosines (ϳ90%) and was substantially decreased in EGFR isolated from Prl-treated cells (not shown). Furthermore, only EGFand transforming growth factor-␣ induced a similar increase in EGFR phosphorylation in nontreated NMuMG cells, whereas other growth factors examined, such as insulin-like growth factor-1, bFGF, and relaxin, failed to induce such an increase (not shown).
Time course studies indicated that receptors from NMuMG cells not treated with Prl exhibited a time-dependent EGFinduced increase in EGFR phosphorylation, reaching apparent maximal stimulation by 4 min (Fig. 3). In contrast, receptors isolated from Prl-treated cells could not overcome the decrease in EGF-induced EGFR phosphorylation even at 12 min, and overall they displayed less change in phosphorylation at all time points examined. While receptors isolated from cells not treated with Prl displayed maximal phosphorylation with 0.125 M ATP, receptors from Prl-treated cells failed to reach levels of phosphorylation observed for the control group, even at 2 M ATP (Fig. 4). As expected, receptors isolated from nontreated cells also exhibited a dose-dependent response to increasing concentrations of EGF, which displayed an apparent maximal phosphorylation at 10 ng/ml (1.7 nM) (Fig. 5). In contrast to receptors isolated from cells not treated with Prl, receptors isolated from Prl-treated cells showed less phosphorylation in response to EGF at all concentrations tested up to 100 ng/ml (17 nM) (Fig. 5).
To determine if the Prl-induced decrease in EGF-induced EGFR phosphorylation was due to reduced EGFR affinity for EGF, we performed 125 I-EGF binding analysis of solubilized EGFR from cells treated with or without Prl (Table I). As

FIG. 1. Prl decreases EGF-induced EGFR phosphorylation.
EGFR from NMuMG (A) or NIH3T3 (B) cells treated with or without Prl were incubated with 1 M [␥-32 P]ATP (specific activity, 300 Ci/mmol), 1 mM MgCl 2 , with or without 10 ng/ml EGF. Reaction mixtures were separated by SDS-PAGE, fixed, dried, and exposed to x-ray film. Autoradiograms are representative of three separate experiments. Bar graphs are presented as mean pp170 density relative to control (ϪPrl/ ϪEGF) of three separate experiments Ϯ S.E. *, p Ͻ 0.05 compared with control.

FIG. 2. EGFR Western blot. EGFR from NMuMG cells treated with
or without Prl were separated on a 7.5% SDS-PAGE gel, transferred to membranes, blocked overnight, and probed with anti-EGFR. The Western blot is representative of three separate experiments. The bar graph is presented as mean pp170 density relative to control (ϪPrl) of three separate experiments Ϯ S.E.
reported in intact NMuMG cells (25), we observed high (2 ϫ 10 Ϫ11 to 4 ϫ 10 Ϫ11 pM) and low (8 ϫ 10 Ϫ10 to 9 ϫ 10 Ϫ10 nM) EGF binding affinities from cells treated with or without Prl. Although Prl-treated cells displayed numerically higher dissociation constants (K d ) for both high and low affinities compared with control cells, they were not significantly different from nontreated cells. Paralleling our Western blot data (Fig. 2), there was no significant difference in the concentration of receptors between the treatment groups (p Ͼ .05).
To determine if the observed decrease in EGF-induced EGFR phosphorylation in Prl-treated NMuMG cells was specific to Prl, NMuMG cells were incubated with or without 100 ng/ml of Prl, bovine growth hormone, or mouse U5 anti-prolactin receptor (not shown). Receptors isolated from bovine growth hormone-treated NMuMG cells exhibited EGF-induced EGFR phosphorylation approximately equal to that of nontreated cells. As in our previous studies (24,25), Prl inhibited EGFinduced EGFR phosphorylation. Additionally, the U5 anti-Prl receptor antibody, which was previously shown to act as a Prl agonist by presumably inducing receptor dimerization (30), mimicked the effect of Prl in decreasing EGF-induced EGFR phosphorylation. However, these effects were only observed if Prl was added to intact cells. The addition of Prl to isolated EGFR failed to decrease EGF-induced EGFR phosphorylation (not shown).
To ascertain whether the Prl-induced decrease in EGFR kinase activity is restricted to EGF-induced autophosphorylation, we examined the ability of EGFR to phosphorylate peptide substrates. While we determined that EGFR K m for angiotensin II in isolates from NMuMG cells not treated with Prl was approximately 1.6 mM, a K m for EGFR isolated from Prl-treated cells could not be determined due to the failure to detect an FIG. 3. Prl inhibition of EGFR phosphorylation cannot be overcome with increasing time. EGFR from NMuMG cells treated with or without Prl were incubated with 1 M [␥-32 P]ATP (specific activity, 300 Ci/mmol), 1 mM MgCl 2 , and 10 ng/ml EGF. Reaction mixtures were separated by SDS-PAGE, fixed, dried, and exposed to x-ray film. The autoradiogram is representative of three separate experiments. The line graph is presented as mean pp170 density relative to control (ϪPrl/ϩEGF) of three separate experiments Ϯ S.E.

FIG. 4. Prl inhibition of EGFR phosphorylation cannot be overcome with increasing ATP. EGFR from NMuMG cells treated
with or without Prl were incubated with the indicated concentrations of [␥-32 P]ATP (specific activity, 500 Ci/mmol), 1 mM MgCl 2 , and 10 ng/ml EGF. Reaction mixtures were separated by SDS-PAGE, fixed, dried, and exposed to x-ray film. The autoradiogram is representative of three separate experiments. Line graph data points are presented as mean pp170 density relative to control (ϪPrl/ϩEGF/0.06 M ATP) of three separate experiments Ϯ S.E.

FIG. 5. Prl inhibition of EGFR phosphorylation cannot be overcome with increasing EGF. EGFR from NMuMG cells treated
with or without Prl were incubated with 1 M [␥-32 P]ATP (specific activity, 300 Ci/mmol), 1 mM MgCl 2 with or without the indicated concentrations of EGF. Reaction mixtures were separated by SDS-PAGE gel, fixed overnight, dried, and exposed to x-ray film. The autoradiogram is a representative of three separate experiments. Line graph data points are presented as mean pp170 density relative to control (ϪPrl/ϪEGF) of three separate experiments Ϯ S.E. increase in peptide phosphorylation (not shown). As shown in Fig. 6, EGF induced an increase in angiotensin II (ϳ6-fold, Fig.  6A) and Src peptide phosphorylation (ϳ7-fold, Fig. 6B) from EGFR isolated from NMuMG cells not treated with Prl. However, paralleling our autophosphorylation data, Prl decreased EGF-induced EGFR phosphorylation of both angiotensin II and Src peptides to similar extents.
To determine if Prl induces a phosphatase that could be involved in heterologous regulation of EGFR kinase activity, we assessed phosphotyrosine phosphatase activity in our EGFR isolates by the previously described malachite green method (31). We observed no Prl-induced increase in dephosphorylation of either Src-autophosphorylated site, representing amino acids 412-422 of v-Src (RRLIEDAEpYTARG) (where pY represents phosphorylated Tyr), or an alternative tyrosinephosphorylated peptide (RRLIEDAEpYAARG, not shown). Furthermore, using silica gel thin layer chromatography, we detected no changes in ATPase activity in EGFR isolates from Prl-treated cells (not shown). Therefore, these results suggest that the observed changes were probably due to altered EGFR kinase activity and not due to an increase in EGFR dephosphorylation or decrease in ATP content in the assay.
The fact that many investigators have reported various hormones that can regulate EGFR kinase activity by serine and/or threonine phosphorylation prompted us to examine the phosphorylation state of the EGFR that were isolated from cells treated with or without Prl. While Western blot analysis revealed that receptors isolated from cells not treated with Prl exhibited low but detectable threonine phosphorylation, receptors isolated from Prl-treated cells displayed a significant increase (ϳ4-fold) in threonine phosphorylation (Fig. 7A) but no detectable serine phosphorylation (not shown). To conclusively show that Prl does indeed induce phosphorylation of EGFR, we employed 32 P metabolic labeling followed by thin layer electrophoresis. As shown in Fig. 8, while there were detectable levels of serine phosphorylation in control and Prl-treated cells, NMuMG cells treated with Prl exhibited a clear increase in threonine phosphorylation. However, there was no phosphotyrosine observed in either treatment group.
To determine if the removal of the observed threonine phosphorylation would cause the return of kinase activity of EGFR isolates from Prl-treated cells, we dephosphorylated isolated EGFR and examined kinase activity in response to EGF. Incubation of EGFR isolated from Prl-treated cells with alkaline phosphatase efficiently reduced detectable levels of threonine phosphorylation (Fig. 9A) and resulted in the return of EGFinduced EGFR phosphorylation that was equivalent to cells not treated with Prl (Fig. 9C). Ci/mmol). After the addition of trichloroacetic acid and bovine serum albumin, the reaction mixtures were centrifuged, and supernatant aliquots were spotted onto P-81 phosphocellulose paper. P-81 paper was washed, and adsorbed radioactivity was determined by liquid scintillation counting. Bar graphs represent mean relative to control (ϪEGF/ ϩpeptide) of three separate experiments Ϯ S.E. *, p Ͻ 0.05 compared with control.

DISCUSSION
We have previously reported that treatment of NMuMG cells with Prl prior to EGF treatment inhibits EGF-induced DNA synthesis (23), EGFR tyrosine phosphorylation (24,25), and Ras-MAPK signaling (24) in intact cells. However, whether these effects are due to modulation of EGFR function or some downstream event cannot be determined from those results. In this study, we have also shown that receptors isolated from NMuMG cells treated with Prl exhibit a decrease in EGFinduced EGFR phosphorylation as well as a decrease in the ability of EGFR to phosphorylate angiotensin II or Src peptides, an effect that could not be attributed to differential EGFR concentrations, increasing phosphotyrosine phosphatases, ATPase activity, or varying binding affinities for EGF. NIH 3T3 cells, which have been shown to be unresponsive to Prl (29), presumably because they have no Prl receptors, however, failed to exhibit a Prl-induced decrease in EGF-induced EGFR phosphorylation. The effect of Prl in the present study appeared to be specific, insofar as the U5 anti-prolactin receptor, which has been shown to stimulate Prl receptor signaling (30), also decreased EGF-induced EGFR phosphorylation to levels observed in the Prl-treated group. However, this was not observed in EGFR that were isolated from NMuMG cells treated with bovine growth hormone, which has been reported not to bind to the Prl receptor (32). Increasing concentrations of EGF, ATP, or reaction time did not allow the EGFR isolated from Prl-treated cells to overcome the observed decrease in EGFinduced EGFR phosphorylation. These results suggest that EGFR isolated from Prl-treated cells contain an inhibitory constraint that inhibits kinase activity in response to ligand binding.
It has been shown that various kinases are activated either directly or indirectly in the EGFR signaling pathway (reviewed in Refs. 15 and 33). In addition to phosphorylating cellular substrates that elicit a biological response, these kinases have been reported to phosphorylate the EGFR on serine and/or threonine, which results in either desensitization or down-regulation of the receptor (10, 19 -21) or activation of specific tyrosine phosphatases (34). While the latter has been shown not to be a probable mechanism in our current study, two alternatively proposed modes of EGFR autokinase regulation (22) are more consistent with our data. One model is that upon Ser/Thr phosphorylation there is a conformational change in the receptor disallowing further kinase activity, whereas the second model proposes that the Ser/Thr phosphorylation inhibited EGFR function by preventing the receptor from interacting with other biomolecules within the usual signaling complexes. Researchers have determined that the EGFR is under extensive modulation by other growth factor receptors (transmodulation), such as nerve growth factor (35), platelet-derived growth factor (28), insulin (36), and basic fibroblast growth factor (35), modifying EGFR autokinase activity, EGF binding, or down-regulation. These have been proposed to be via phosphorylation (35) but may also involve phosphorylation-independent mechanisms (28). Although Western blotting and phosphoamino acid analysis results reported here show that EGFR isolated from Prl-treated cells have increased threonine phosphorylation, they do not conclusively show that such a phosphorylation event can decrease EGFR kinase activity. However, the restoration of kinase activity of EGFR isolated from Prl-treated cells by dephosphorylation lends evidence to the hypothesis that threonine phosphorylation of EGFR in response to Prl causes a decrease in EGF-induced EGFR kinase activity.
Prl has been proposed to signal through a variety of mechanisms including Ras (37), MAPK (38), herterotrimeric G proteins (39), protein kinase C (40), and Jak-STAT (41) pathways. Transmodulation of the EGFR by several of these pathways, most notably protein kinase C and MAPK, have been previously reported. In fact, mechanisms of EGFR phosphorylation leading to altered biochemical activity have been loosely categorized as protein kinase C-dependent and -independent. Treatment of cells with phorbol esters decreases EGFR signaling (3,10,19,42) and decreases high affinity binding (10, 20, 43) but fails to affect dimerization (22). While in vivo studies have shown that protein kinase C can phosphorylate Thr-654 (19,20,22), other phosphorylation sites have also been described. EGFR has been shown to be phosphorylated on Thr-669 after treatment with EGF, phorbol esters, or thapsigargin (44 -46). This effect has been shown to be mediated by MAPK (45,46), and investigators have reported that phosphorylation at this site causes a slight reduction in EGF-induced EGFR kinase activity (19). In recent years, several researchers have described serine kinases that may be involved in regulating EGFR function. Kuppuswamy et al. (21) have reported that Cdc2 is capable of phosphorylating EGFR in vitro on serine 1002, an effect that resulted in a time-dependent decrease in kinase activity. Furthermore, the calcium-calmodulin dependent kinase II has been shown to phosphorylate EGFR on serine 1046/1047 and subsequently decrease EGFR kinase activity (22). While a flurry of research has tried to delineate which residue is involved in specific EGFR regulatory functions, e.g. desensitization or down-regulation, mutational analysis studies have reported that neither Thr-654 nor Thr-669 is sufficient to desensitize EGFR to MAPK or protein kinase C (19,22); therefore, other mechanisms of regulating EGFR function must exist. However, serine 1046/1047 may be involved in regulation of EGFR kinase activity (22), while Thr-654/669 may be involved in regulating high affinity binding or EGFR numbers (20,22,42). In the present study, we observed low levels of basal serine phosphorylation that was unaffected by Prl. In contrast, Prl induced a readily detectable increase in threonine phosphorylation. This finding, combined with the observation FIG. 8. Phosphoamino acid analysis. NMuMG cells were plated on 150-mm dishes and metabolically labeled with [ 32 P]orthophosphoric acid. Cells were treated with or without Prl and lysed, and EGFR was immunoprecipitated. Following separation on 7.5% SDS-PAGE gel, the band in the 170-kDa range was excised and hydrolyzed in 6 N HCl at 100°C. The hydrolysis mixture was lyophilized, resuspended in 5% acetic acid, and 0.5% pyridine containing amino acid standards and separated by thin layer electrophoresis. The plate was allowed to dry and exposed to x-ray film. that dephosphorylation of EGFR from Prl-treated cells restores kinase activity, suggests that threonine phosphorylation plays a critical role in Prl-mediated inhibition of EGFR kinase activity.
FIG . 9   FIG. 9. Alkaline phosphatase restores EGFR kinase activity. EGFR from NMuMG cells treated with or without Prl were incubated with or without alkaline phosphatase, separated by SDS-PAGE, transferred to membranes, and probed with anti-phosphothreonine (A) or anti-EGFR (B). C, alkaline phosphatase-treated EGFR from Prl-treated and nontreated cells were incubated with 1 M [␥-32 P]ATP (specific activity, 300 Ci/mmol), 1 mM MgCl 2 , and 10 ng/ml EGF. Reaction mixtures were separated by SDS-PAGE gel, fixed, dried, and exposed to x-ray film. Blots and autoradiograms are representative of three separate experiments. Bar graphs are presented as mean p170 density relative to control (A and B; ϪPrl/ϪALP) or mean pp170 density relative to control (C; ϪPrl/ϩEGF/ϪALP) of three separate experiments Ϯ S.E. *, p Ͻ 0.05 compared with control.