Mitogenic Signaling from theEGF Receptor Is Attenuated by a ...

Molecular Biology of the Cell Vol. 7, 871-881, June 1996

Mitogenic Signaling from the EGF Receptor Is Attenuated by a Phospholipase C-y/Protein Kinase C Feedback Mechanism Philip Chen, Heng Xie, and Alan Wells* Department of Pathology, University of Alabama at Birmingham, and Veterans Administration Medical Center, Birmingham, Alabama 35294-0007 Submitted December 29, 1995; Accepted March 28, 1996 Monitoring Editor: Joseph Schlessinger

We recently demonstrated that epidermal growth factor receptor (EGFR)-mediated signaling of cell motility and mitogenesis diverge at the immediate post-receptor level. How these two mutually exclusive cell responses cross-communicate is not known. We investigated a possible role for a phospholipase C (PLC)-dependent feedback mechanism that attenuates EGF-induced mitogenesis. Inhibition of PLCy activation by U73122 (1 JIM) augmented the EGF-induced [ Hithymidine incorporation by 23-55% in two transduced NR6 fibroblast lines expressing motility-responsive EGFR; increased cell division and mitosis was observed in parallel. The time dependence of this increase revealed that it was due to an increase in maximal incorporation and not a foreshortened cell cycle. Motility-responsive cell lines expressing a dominant-negative PLCy fragment (PLCz) also demonstrated augmented mitogenic responses by 25-68% when compared with control cells. PLCz- or U73122-augmented mitogenesis was not observed in three nonPLCOy activating, nonmotility-responsive EGFR-expressing cell lines. Protein kinase C (PKC), which may be activated by PLC-generated second messengers, has been proposed as mediating feedback attenuation due to its capacity to phosphorylate EGFR and inhibit the receptor's tyrosine kinase activity. Inhibition of PKC by Calphostin C (0.05 JIM) resulted in a 57% augmentation in the fold of EGF-induced thymidine incorporation. To further establish PKC's role in this feedback attenuation mechanism, an EGFR point mutation, in which the PKC target threonine654 was replaced by alanine, was expressed. Cells expressing these PKC-resistant EGFR constructs demonstrated EGF-induced motility comparable to cells expressing the threonine-containing EGFR. However, when these cells were treated with U73122 or Calphostin C, the mitogenic responses are not enhanced. These findings suggest a model in which PKC activation subsequent to triggering of motility-associated PLCy activity attenuates the EGFR mitogenic response. INTRODUCTION Marikovsky et al., 1993; Miettinen et al., 1995; Sibilia and Wagner, 1995). Many reports have demonstrated Activation of the epidermal growth factor receptor (EGFR) elicits multiple cellular and biological re- that growth factor-induced cell motility and mitogenic sponses in vitro and in vivo, including mitogenesis responses can be signaled independently (Nister et al., and cell movement. Both cell proliferation and move- 1988; Hartmann et al., 1992; Faletto et al., 1993; Chen et and Bade, al., 1994a;t Manske ment are required functionsrecentlys such ashaves ment~ etule ~ ~for important ~ ~1993).an ~cellWel ~~~~~~~ptwy (Rynld a!.,1994), via distinct signalet ing pathways (Reynolds al., 1993). We recently have development and wound repair (Adamson, 1990; demonstrated that the divergence of the pathways Corresponding author: LHRB 531, Department of Pathology

University of Alabama at Birmingham, Birmingham, AL 35294-

0007.

( 1996 by The American Society for Cell Biology

to EGFR-elicited motility and mitogenic responses occurs at the immediate post-receptor level (Chen et al., 1994a,b). How the balance of signaling

leading

871

P. Chen et al.

between these two pathways is resolved remains to be elucidated. Functional tyrosine kinase activity is required for both mitogenic and motility responses elicited by EGFR (Chen et al., 1987, 1994a). Receptor autophosphorylation and subsequent activation of phospholipase Cy (PLCy) are required for EGF-induced cell movement but not for mitogenesis (Wells et al., 1990; Decker, 1993; Chen et al., 1994a,b). PLCy activation by EGFR (Margolis et al., 1990a) produces diacylglycerol (DAG) and inositol trisphosphate, which activate the ser/thr kinase protein kinase C (PKC). PKC represents a large gene family of at least 12 isoforms differing in their structure, tissue distribution, subcellular localization, mode of activation, and substrate specificity (Dekker and Parker, 1994). PKCs phosphorylate a wide variety of substrates including proteins involved in signal transduction (including Ras, GAP, and Raf) (Hug and Sarre, 1993) as well as motility-associated cytoskeletal modulators (including fak, profilin, and MARCKS) (Hansson et al., 1988; Aderem, 1992). These characteristics suggest a role for PKC isoforms in mediating specificity in intracellular signal transduction pathways. PKCs also attenuate signaling from growth factor receptors including EGFR (Whitely and Glaser, 1986; Bowen et al., 1991; Welsh et al., 1991). PKC phosphorylates the threonine654 residue (T654) on EGFR (Welsh et al., 1991), decreasing binding affinity of the receptor for ligands (Jimenez-deAsua and Goin, 1992; Morrison et al., 1993) and diminishing tyrosine kinase activity of the receptor (Lund et al., 1990). Phosphorylation at this site on the receptor by PMA-activated PKC has been shown to block EGF-induced lamellipodia retraction (Welsh et al., 1991). In addition, serine (S654) substitution causes constitutive phosphorylation at this site with a concomitant decrease in thymidine incorporation (Bowen et al., 1991). These observations suggest a linear, physiological feedback attenuation mechanism in which the activation of the motility-associated PLCy and subsequent PKC lead to attenuation of the mitogenic response elicited by EGFR. However, the physiological importance of this biochemical connection is yet to be demonstrated. Dissection of this putative pathway will determine whether this serves as a decision point by which a cell can direct a pluripotential extracellular stimulus into a specific biological response. In this study, we utilized signaling-restricted EGFR expressed on receptor-devoid NR6 fibroblast cells to differentially activate the EGF-induced motility pathway and this putative downstream mitogenesis-attenuation mechanism. We also used pharmacological as well as molecular agents to disrupt this motility-associated pathway, and we assessed the affects on EGFR-mediated mitogenic response. 872

MATERIALS AND METHODS Generation of NR6 Cells Expressing EGFR Constructs Construction of the EGFR and stable expression in NR6 cells were by standard methods, and have been described previously (Wells, 1990; Welsh et al., 1991; Chen et al., 1994a). Briefly, WT EGFR is a full-length cDNA derived from a human placental cDNA library. c'973, c'991, and c'1000 represent EGFR in which stop codons are introduced just distal to the amino acid number indicated. c'1000F992 was created from c'1000 by replacing the sole remaining autophosphorylation site at y992 with a phenylalanine (F92). WTA61 is a full length EGFR in which the target site of PKC phosphorylation T5 was replaced with alanine (A654) by site-directed mutagenesis (Welsh et al., 1991). c'1000A654 was created from c'1000 with A654 replacing T6'. These EGFR are shown schematically in Table 1. The constructs were expressed on NR6 cells, 3T3-derivatives that lack endogenous receptors (Pruss and Herschman, 1977). This was accomplished by retroviral-mediated transduction as previously described (Chen et al., 1994a). Polyclonal lines were established by selection in 300 jig/ml G418 (Life Technologies, Gaithersburg, MD). The infectant cell lines presented physiologic levels of receptors (60,000-250,000 EGF binding sites per cell) with similar dissociation constants (Kd were 0.2 nM to 0.7 nM). All of the EGFR possessed kinase activity; the cells that presented the kinase-active EGFR all demonstrated a mitogenic response to EGF (Table 1).

Thymidine Incorporation Assay EGF-induced mitogenesis was assessed by incorporation of [3H]thymidine in the target cells (Chen et al., 1994a). Cells were plated on plastic and grown to confluence in MEMa with 7.5% fetal bovine serum (FBS). The cells were then switched to media containing 1% dialyzed FBS for 24 h. The cells were subsequently treated with or without EGF (25 nM) and incubated at 37°C for 16 h. [3H]thymidine (5 ,uCi/ml) was added and incubation continued for another 10 h. In the time-course experiments, [3H]thymidine was added at the times indicated to determine incorporation over the subsequent 3-h interval. The cells were then washed with ice-cold phosphate-buffered saline (PBS) twice and then treated with 5% trichloroacetic acid at 40C for 30 min. After two more washes with PBS the cells' incorporated [3H]thymidine was solubilized with 0.2N NaOH and measured by scintillation counter. The pharmacological agent, U73122 (1-(6-((17,3-3-methoxyestra1,3,5(10)-trien-17-yl)amino)hexyl)-lH-pyrrole-2,5-dione) (BIOMOL) was added to the cells to inhibit EGF-induced PLC activity. It was introduced at 1 ,uM into the media in the [3H]thymidine incorporation assays 15 min prior to the addition of EGF. The inhibitory effects on EGF-induced PLC activity by this compound were determined previously (Chen et al., 1994b); U73122 inhibits PLC activities but not phospholipase A2 or phospholipase D activities (Bleasdale et al., 1990; Powis et al., 1992; our unpublished observations). The inactive congener of U73122, U73343 (1-(6((17,3-3-methoxy-

estra-1,3,5(10)-trien-17-yl)amino)hexyl)-2,5-pyrrolidine-dione), was added at the same concentration in parallel and served as a control. The pharmacological agent Calphostin C (BIOMOL) was added to the cells to inhibit PKC activity; Calphostin C was introduced at 0.05 ,tM into the media in the [3H]thymidine incorporation assays 15 min prior to the addition of EGF.

Cell Proliferation Assays Two alternative methods to assess EGF-induced mitogenesis were used to correlate thymidine incorporation with cell division. Cell proliferation determinations by counting were described previously (Chen et al., 1994a). Briefly, NR6 cells (-5,000 cells/well) were plated in a 12-well tissue culture plate. Immediately after attachment on the plastic, the cells were switched to media containing 1 % Molecular Biology of the Cell

EGF Receptor Feedback Inhibition by PLC/PKC

Table 1. EGF-induced PLCy activity and cell migration in NR6 cells expressing EGFR constructs

Construct Parental

Schematic of EGF receptor no EGF receptor T 654

958

y992 y1 068

1148

Migrationa 101±4

PLCy activityb 97±2

225±41 170±11 121±14 105±18 114±4 177±25 215±8 213±28

337±74 335±55 121±4 110±10 101±11 240±75 n.d. n.d.

1173

WT c' 1 000

E -1 EI-

c'1 000F9 9 2

c'991 c'973 c' 1186 F3

_ 0I/IMZI

FPMFIMF!tl

WTA654 c'1 000A654

O

-

Extracellular ligand-binding domain

-

Transmembrane domain

Tyrosine kinase domain Carboxy-terminal regulatory domain - Domain around autophosphorylatable tyrosine y992 - Domain around autophosphorylatable tyrosine y1068 - Domain around autophosphorylatable tyrosine y 1148 - Domain around autophosphorylatable tyrosine y 1173 -

F, phenylalanine replacing the autophosphorylated tyrosines at the indicated sites; A654, alanine replacing the PKC phosphorylating site T654. EGF-induced migration after 24 h of 25 nM EGF treatment; expressed as percentage of non-EGF treated cells, mean + SEM, n = 4-12. b EGF-induced PLCy activity measured as 3H-phosphotidylinositol turnover after 30 min treatment of 25 nM EGF; expressed as percentage of non-EGF treated cells, mean + SEM, n = 3-7. a

dialyzed FBS for 24 h. The cells were subsequently treated with or without EGF (10 nM) and incubated at 37°C for 3 days (Chen et al., 1994a; Reddy et al., 1994). At the end of incubation, cells were trypsinized and counted by Coulter counter. Alternatively, the fraction of cells entering mitosis was assessed by incorporation of bromo-deoxyuridine utilizing commercially available reagents (BrdU Staining Kit, Oncogene Sciences). BrdU incorporation was determined from 14-24 h post EGF stimulation as with thymidine incorporation. The effect of PLC inhibition on cell proliferation was demonstrated by introduction of the selective PLC inhibitor U73122. This pharmacological agent was added at 1 ,AM 15 min prior to the addition of EGF; U73343 served as the control.

an SV40 promoter serving as a were transfected into WT and

selectable marker. These c'1000 EGFR-expressing NR6 cells and selected with 400 nM methotrexate (Chen et al., 1994b). Induced expression of PLCz in the pDexMTX transfectants was accomplished by introducing 2 ,uM dexamethasone into the media 16 h before the thymidine incorporation assay. Dexamethasone remained in the media throughout the assay period. The expression of constitutive and dexamethasone-induced PLCz expression was verified by immuno-blot analyses described below. control of

constructs

Cell Motility Assay

Cloning and Expression of Dominant Negative PLC y-1 in NR6 Cells

EGF-induced migration was assessed by the ability of the cells to move into an acellular area as previously described (Chen et al., 1994a). U73122 (and U73343) or Calphostin C was introduced at 1 p.M or 0.05 ,uM, respectively, into the media 15 min prior to the addition of EGF and remained throughout the assays.

A dominant-negative PLCy-1 gene fragment that encodes the Z region SH2 and SH3 domains (amino acids 517-901, designated as PLCz; Homma and Takenawa, 1992) of this enzyme was cloned into the pXf vector (Chen et al., 1994b). The PLCz fragment also was cloned into pDexMTX, the pXf vector with the 1.2-kb MMTV promoter in place of the SV40 early promoter, for inducible expression. These PLCz-containing vectors contain a DHFR gene under the

Effects of TPA (12-O-tetradecanoylphorbol 13-acetate) on the tyrosine kinase activity of EGFR were demonstrated by immuno-blot analyses. Cells (4 x 106) were treated with 10 nM TPA 10 min prior to EGF (10 nM) treatment and total cell lysates were loaded on each

Vol. 7, June 1996

Western Blot Analysis

on

Whole Cell Lysate

873

P. Chen et al. lane. Calphostin C was introduced 15 min prior to TPA treatment to inhibit PKC activation. The immuno-blot was probed with a monoclonal anti-phosphotyrosine antibody (PY-20, Transduction Laboratory, Lexington, KY) and visualized by probing with AP-conjugated secondary antibodies followed by development with a colorimetric method (Promega, Madison, WI). The expression of dexamethasone-induced, MMTV-driven PLCz was verified by Western blot analyses. Cells (4 x 106) were treated with or without 2 ,uM dexamethasone for 16 h before cell lysis. Whole cell lysates were separated on 7.5% SDS-PAGE. The blot was probed by mixed monoclonal anti-PLCy-1 antibody (05-163; UBI, Lake Placid, NY) and visualized as described above.

RESULTS Inhibition of PLCy Augments EGF-induced Thymidine Incorporation We previously have demonstrated that EGF-induced cell motility requires PLCy activity. The selective pharmacologic inhibitor U73122 decreased both EGFinduced PLC activity (but not EGF-induced PLD activity) and the motility response (Chen et al., 1994b). To assess the effects on mitogenesis by inhibiting PLC activation, we introduced this agent at 1 ,uM in thymidine incorporation assays and observed an augmentation of 23-55% in EGF-induced incorporation response in two transduced NR6 cell lines expressing WT and c'1000 EGFR (Figure 1A), which are motilityresponsive and activate PLCy upon EGF stimulation (Table 1). The basal level of thymidine incorporation was increased in the presence of U73122, but to a lesser extent than the EGF-induced increase. Although the absolute level of EGF-induced thymidine incorporation varied somewhat between the infectant cell lines, there was no correlation with EGFR level. To correlate thymidine incorporation response with cell division, we performed cell proliferation assays as an alternative method to assess mitogenesis. Similar to what is observed in thymidine incorporation response, EGF-induced cell proliferation in WT EGFRexpressing NR6 cells is also augmented by 31% in the presence of 1 ,tM U73122 (Figure 1B). As assessed by BrdU incorporation, the percentage of cells stimulated to enter mitosis increased 23% in the face of U73122 (60 ± 1% versus 74 ± 5% in U73343). Interestingly, U73122 also increased basal thymidine incorporation and cell proliferation (Figure 1, A and B), although the augmentation in the absence of EGF was less than that in the presence of EGF. The observed augmentation in both basal and EGF-induced mitogenesis suggests that a physiologically active inhibitory mechanism is attenuated by U73122 treatment. To determine whether this augmentation is due to an increase in maximal incorporation or altered kinetics of mitogenesis, we determined thymidine incorporation at various times. In cells expressing WT and c'1000 EGFR, U73122 resulted in increased thymidine incorporation throughout the entire period of the mitogenic burst elicited by EGF (Figure 1C), further sup874

porting an increased percentage of cells being recruited into the proliferative response rather than a foreshortened cell cycle. To demonstrate that this phenomenon is PLC'y specific and motility-associated, we examined the EGFinduced thymidine incorporation response in cells expressing c'973, c'991, and c'100OF992 EGFR that do not activate PLCy nor elicit motility response upon ligand stimulation (Table 1). In these cells, treatment with 1 ,uM U73122 did not result in increased EGF-induced thymidine incorporation (Figure 2), although basal thymidine incorporation was increased in two of three cell lines. These data suggest that basal PLC activity suppresses low level mitogenic stimuli. We further demonstrated the specificity of PLC/s involvement in this mitogenic attenuation mechanism by introducing a dominant negative PLCy-1 fragment (PLCz) in two motility-responsive cell lines. PLCz encodes the SH2- and SH3-containing Z region of PLC-y-1 and overexpression of this peptide inhibits activation of the endogenous PLCy by EGFR (Chen et al., 1994b) and other receptor tyrosine kinases (Homma and Takenawa, 1992). Expression of PLCz in NR6 cells expressing two motility-responsive EGFR, c'1000 and c'l 186F3, augmented both basal and EGF mitogenic responses (Figure 3). Because the constitutive expression of PLCz may preferentially select a subpopulation of fast growing cells, we expressed PLCz acutely under the control of an MMTV promoter. WT EGFR-expressing NR6 cells were treated with or without 2 ,uM dexamethasone to induce PLCz expression, as verified by immuno-blot (Figure 4A). Concomitantly, the mitogenic response of this expression was tested; treatment with dexamethasone augmented the EGF-induced thymidine incorporation by 58%, while decreasing the basal rate of thymidine incorporation (Figure 4B). Conversely, treatment of non-PLCz-transfectant WT EGFR-expressing NR6 cells with dexamethasone (2 ,uM) decreased basal thymidine incorporation by 35-50%, although the EGF-induced thymidine incorporation was not affected. Together, these data suggest that a physiological feedback attenuation mechanism on EGF-induced mitogenic response is mediated by receptor-activated, motility-associated PLCy.

Pharmacological Inhibition of PKC Augments EGF-induced Thymidine Incorporation Activation of PKC by nonphysiological phorbol esters has been shown to inhibit the tyrosine kinase activity of the EGFR (Downward et al., 1985; Bowen et al., 1991) as well as the biological responses it mediates (Welsh et al., 1991). To determine whether PKC participates in the physiological feedback attenuation mechanism described, we employed the selective PKC inhibitor CalMolecular Biology of the Cell


A

B

4000

WT EGFR-expressing NR6 Cells 6

o

c

0 50

0 .

N.0

c

0

o

o

E

40

0s

o)

a

L.

30

= o)

0

CE

0 0 .0

I--

20

E 1 0

-

c'1 000

WT

0.

EGFR Construct Expressed in NR6 Cells

U73122

control

Treatment

Figure 1. Effect of the PLC-specific inhibitor U73122 on EGF-induced [3Hlthymidine incorporation (A), cell proliferation (B), and the time course of [3Hlthymidine incorporation (C) in infectant NR6 cell lines expressing motility-responsive

EGFR.

tested.

E

E Q

I0owo

.X

800000

X L_

60/00l o

c

4000/0

4/00 l

e

O

X E

E

200000

> l0000

incorporated expressed as ac-

The

[ Hithymidine is

c'1000 EGFR-expressing NR6 Cells

WT EGFR-expressing NR6 Cells 1000000

EGF-in-

[Hithymidine incorporation was determined for each cell line tested in the presence of U73122 (1 ,aM) or its inactive congener U73343 (1 ,uM). Equivalent numbers of cells (-100,000 cells) duced

were

C

, tual counts. *, basal incorpora2 3 0 21 2 4 15 8 1 2 tion without EGF; (III), basal incorporation in the presence of (hrs) Time Time (hrs) U73122; (1), EGF-stimulated incorporation (25 nM EGF); and SD for three determinations; p < 0.01 between (D), EGF-stimulated incorporation in the presence of U73122. Shown are mean U73122-treated and U73343-treated cells. Cell proliferation were expressed as cell number after 3 days of 10 nM EGF exposure. (-), basal proliferation; (3), EGF-stimulated proliferation (25 nM EGF); Shown are mean + SD for three determinations. 9

1

phostin C and assessed its effects on EGF-induced mitogenic response. Treatment with TPA (10 nM) dramatically decreased the EGF-induced tyrosine phosphorylation on EGFR as demonstrated by immunoblot analysis. The presence of Calphostin C (as low as 0.01 ,uM) reversed this inhibition (Figure 5). Treatment with Calphostin C (0.05 ,uM) resulted in a 57% augmentation in the fold of EGF-induced mitogenic response in WT EGFR-expressing cells (Figure 6). Calphostin C treatment decreased the absolute level of both the basal and EGF-induced thymidine incorporation. Replacing the PKC Phosphorylation Site P54 on EGFR with A654 Abrogates the PLCy and PKC-mediated Attenuation Mechanism Activated PKC inhibits EGFR's tyrosine kinase activity by phosphorylating T654 on the receptor. Others Vol. 7, June 1996

7

9

1

2

1

5

1

8

21

2 4

2 7

3 0

have reported that replacing this residue with serine (S654) has resulted in constitutive phosphorylation on this site and has dramatically decreased EGFR's signaling and the mitogenic response it mediates (Bowen et al., 1991). On the other side, replacing this residue with alanine (A654) makes this site resistant to phosphorylation by activated PKC; this has resulted in augmented EGFR-mediated biological responses such as lamellipodia retraction and mitogenesis (Bowen et al., 1991; Welsh et al., 1991). To determine whether phosphorylation on T654 is secondary to PLC-mediated feedback attenuation mechanism, we expressed on NR6 cells, WT, and c'1000 EGFR in which T654 was replaced by alanine, WTA654, and c'1000A654, respectively. We examined the EGF-induced mitogenic response in the presence of the selective PLC inhibitor U73122 and assessed the attenuation response within each line independently. Although U73122 caused an 875

P. Chen et al.

0

'~3000 0 00

000CL

0E00X 0

c'l00OF992

c'991

c'973


Figure 2. Effect of the PLC-specific inhibitor U73122 on EGF-induced [3H]thymidine incorporation in infectant NR6 cell lines expressing nonmotility-responsive EGFR. EGF-induced [3Hithymidine incorporation was determined for each cell line tested in the presence of U73122 (1 ,M) or its inactive congener U73343 (1 ,uM). Equivalent numbers of cells (--100,000 cells) were tested. The incorporated [3Hlthymidine is expressed as actual counts. *, basal incorporation without EGF; (WI), basal incorporation in the presence of U73122; (0), EGF-stimulated incorporation (25 nM EGF); (O), EGFstimulated incorporation in the presence of U73122. Shown are mean ± SD for three determinations.

augmentation in thymidine incorporation in WT and c'1000 EGFR-expressing cells (Figure 1A), the mutation to A654 totally abrogated this augmentation (Figure 7A). Similar to thymidine incorporation, U73122enhanced cell proliferation also is abolished by alanine replacement (Figures 1B and 7B). The time course 4000 -+ 668%

+ 25%

0 0

h.

3000-

C0 x

2000

'Z~

E

I-C

1000

c'l 000

c'lOOOPLCz

c'1186F3 c'1186F3PLCz

EGFR and PLCz Constructs Expressed In NR6 Cells

Figure 3. Effect of constitutively expressed dominant-negative PLCz on EGF-induced [3H]thymidine incorporation in two motilityresponsive EGFR NR6 cell lines with or without the expression of PLCz. Equivalent numbers of cells (-100,000 cells) were tested. The incorporated [3Hlthymidine is expressed as actual counts. *, basal incorporation with no treatment; (1, EGF-stimulated incorporation (25 nM EGF). Shown are mean ± SD for six determinations. 876

experiments demonstrated that thymidine incorporation occurred at similar levels throughout the entire mitogenic burst period regardless of the presence of U73122 (Figure 7C). The alanine replacement did not affect the upstream PLC-mediated motility response (Figure 8 and Table 1). To further establish this site in the feedback pathway, we examined the mitogenic response in the presence of PKC inhibitor Calphostin C. In the presence of 0.05 ,uM Calphostin C, NR6 cells expressing WTA654 no longer exhibit the augmented fold-increase in mitogenic response as observed in WT EGFR-expressing cells (Figure 6). Together, these observations placed the T654 phosphorylation site in EGFR downstream in the motility-associated, PLC-mediated feedback attenuation mechanism on EGF-induced mitogenesis. DISCUSSION Mitogenesis and cell movement are assumed to be mutually exclusive biological responses. However, many growth factors are capable of triggering both. The mechanism by which a cell chooses one response is unknown. Signaling pathways leading to EGFRmediated mitogenic and motility responses diverge at the immediate post-receptor level. EGFR-activated PLC-y activity is required for the motility response but is not necessary for mitogenesis (Chen et al., 1994b). MAP kinase activity, which has been implicated in mediating the mitogenic response, is not sufficient to elicit the motility response (Chen et al., 1994b). Whether the two signaling pathways leading to distinct biological responses communicate downstream has not been determined previously. Herein we delineated a motility-associated feedback attenuation mechanism affecting EGF-induced mitogenic response. Inhibition of the PLCy activation by a selective pharmacological agent, U73122, or the expression of a dominant-negative molecular inhibitor, PLCz, augmented the EGF-induced mitogenic response. By using both a pharmacologic agent, which selectively blocks PLC-mediated PIP2 hydrolysis, and a molecular approach to prevent EGFR-signaled activation of PLCy, we can definitively implicate a physiological mitogenic attenuation mechanism mediated by motility-associated PLCy. Downstream of PLCy, second messengers DAG and inositol trisphosphate elicit various intracellular biochemical processes including the activation of PKC. PKC was a candidate for mediating this attenuation mechanism because its activation by phorbol esters has been shown to inhibit EGFR's intrinsic tyrosine kinase activity secondary to phosphorylation of EGFR (Whitely and Glaser, 1986; Lund et al., 1990; Morrison et al., 1993). Inhibition of PKC by Calphostin C augmented the EGF-induced mitogenic reMolecular Biology of the Cell


A

Figure 4. (A) Expression of dexamethasone-inducible PLCz and (B) its effect on EGF-induced [3Hlthymidine incorporation in WT EGFR-expressing NR6 cells. WT/mmtv-PLCz indicates WT EGFR-expressing NR6 cells transfected with the pDexMTX-PLCz construct (see text). Cells were exposed to dexamethasone (2 ,uM) for 16 h before whole cell immuno-blot for PLCy; the band at -130 kDa is the endogenous PLCy and the band at -50 kDa represents the PLCz fragment. The incorporated [3H]thymidine is expressed as actual counts. Shown are mean + SD for three determinations.

B

Construct: Treatment:

WT/mmtv-PLCz cdr dex c

:

1500

200

PLCg-1 00

97

x 0

._ E 69

E

500

> I.

PLCz

46

Unstimulated

Dexamethasone

Treatment of WT EGFR NR6 cells Expressing pDexMTX-PLCz

sponse. Replacement of the PKC target site on EGFR abrogated U73122 and Calphostin C augmentation in mitogenic response. These data place PKC downstream of PLCy in mediating this feedback mechanism. Together, an intracellular molecular mechanism is demonstrated for the EGFR-activated motility pathway to inhibit the mitogenic response. PKC exerts pleiotropic effects on multiple intracellular biochemical pathways as well as cellular biological responses. Activated PKC has been demonstrated to inhibit EGFR signaling by phosphory-

lating the EGFR at T654 (Lund et al., 1990; Bowen et al., 1991; Welsh et al., 1991; Seedorf et al., 1995). This PLC,y/PKC feedback inhibition of signaling may be common for other members of the receptor tyrosine kinase (RPTK) family because PKC has also been observed to inhibit insulin receptor signaling (Seedorf et al., 1995) as well as to attenuate Grb2 phosphorylation by PDGF receptor (Benjamin et al., 1994). Furthermore, these pathways may enable various RPTKs to cross communicate; PDGF, probably c

0

_

25nM EGF lOnM TPA

_

_

Calphostin C (uM)

-

-

+

-

+

+

+

+

+

+

+

+

+

+

+

+

0V1

0o

0o 10.s

c

8

00 0_

7-

0. a.

00

EGFR ->

.

L

ILl

3-

2-

C

1-

WT

WTA654

EGFR Construct Figure 5. Calphostin C reversal of TPA-induced inhibition of EGFR tyrosine kinase activity. Cells were exposed to EGF (25 nM) in the presence of TPA (10 nM) for 5 min to demonstrate PKC-mediated inhibition of EGFR kinase activity. Various concentrations of Calphostin C were added to the cells, concurrent with TPA and EGF administration, to determine the concentration of Calphostin C that would prevent TPA inhibition of EGF-induced kinase. EGFR kinase activity was determined by autophosphorylation and phosphorylation of other cellular substrates. Equal amounts of whole cell lysates were loaded on each lane. The immuno-blot was probed with an anti-phosphotyrosine antibody.

Vol. 7, June 1996

Figure 6. Effect of Calphostin C on EGF-induced [3Hlthymidine incorporation in WT and WTA654 EGFR-expressing NR6 cells. EGFinduced [3H]thymidine incorporation was determined in the presence or absence of 0.05 ,uM Calphostin C. Equivalent numbers of cells (-100,000 cells) were tested. The incorporated [3H]thymidine is expressed as percent incorporation in the absence of EGF treatment, with the basal (100%) being depicted as a horizontal line. (u), EGF-stimulated incorporation (25 nM EGF); and (Ol), EGF-stimulated incorporation in the presence of Calphostin C. Shown are mean ± SD for three determinations. 877

P. Chen et al. WTA654 EGFR-expressing NR6 Cells

B

A 300

70-

0

6 0-

co0

a.-. 2000-

r 0

010

0

-w

._

504 0-

x

3 0-

:6 01000

= C

0b0

E

20

U

ai

1 00-

c'1 00OA654

WTA654

C

WTA654

EGFR-expressing

NR6 Cells

800000-

Treatment

c'1OOOA654 EGFR-expressing

NR6 Cells

Figure 7. Effect of the PLCspecific inhibitor U73122 on

EGF-induced [3Hlthymidine incorporation (A), cell prolifera-

E

a

U73122

control


o

tion

a.

time

course

of

expressing motility-response

6000

0

m

(B), and

[3Hlthymidine incorporation ff /) X I\ I0 t(C) in infectant NR6 cell lines // " q :7 '>L

600000-

0L EGFR in which the PKC phos\^I S ffi N cNn.. 065 400000phorylation site T654 was rewith A654. EGF-induced 400000 \'K /placed .

E

m

200000

ff

\9

s

s:

'\\

0-

1 2 15 1 8 21 24 27 3 0

tested in the presence of U73122 (1 ,uM) or its inactive congener

. _d_

_

Xl*

[3H]thymidine incorporation determined for each cell line

was

U73343

b ers

0-

of

tested.

18 21 1 2 1 5 309

24

27

30

uM).Equivalent

cells (- 1 00 ,000 cells) The

incorporated

numw ere

[3H]thy-

midine is expressed as actual

N, basal incorporation without EGF; (Ill), basal incorporation in the presence of U73122; (0), EGF-stimulated incorporation (25 nM EGF); and 0, EGF-stimulated incorporation in the presence of U73122. Shown are mean ± SD for three determinations. Cell proliferation (B) was expressed as cell number after 3 days of 10 nM EGF exposure. Shown are mean ± SD for three determinations. Time (hrs)

via PLCy activation (Margolis et al., 1990b), has been shown to induce the T654 phosphorylation on EGFR by PKC (Davis and Czech, 1987). The physiological importance of these putative biochemical processes is not clear. However, one phenomenon observed here may be attributed to such crosstalk. When the feedback attenuation pathway described herein was disrupted at the level of PLCy, we observed an increase in basal mitogenic response (Figure 1, A and B, and Figure 2). This may be due to a background signaling from various RPTKs in addition to EGFR to activate PLCy in cells in the basal state; this increment in basal proliferation is noted in cells expressing both motility-responsive and nonmotility-responsive EGFR. As PLCy is inhibited by U73122 or PLCz (Figures 3 and 4), the downstream feedback inhibition on mitogenesis is relieved, thus cells expressing WT and c'1000 exhibit augmented 878

Time (hrs)

counts.

basal and EGF-induced mitogenic responses. However, in cells expressing EGFR that do not activate PLC-y (Figure 2), only the basal mitogenic rate is altered, implying that EGFR-mediated mitogenesis is already maximally induced. This model is further buttressed by the lack of effect of U73122 on the basal or EGF-induced mitogenic rate in cells expressing EGFR resistant to PKC transmodulation (Figure 7). These findings suggest that low level PLCy activity may play an important physiologic role in preventing mitogenesis in cells either exposed to suboptimal levels of growth factors or secondary to nonspecific activation of RPTKs by oligomerization (Lemmon and Schlessinger, 1994). TPA-activated PKC has been shown to decrease the high affinity binding between EGFR and its ligands. This transmodulation may contribute to the mitogenic attenuation. However, replacement of the PKC target Molecular Biology of the Cell


0

1-

u

.

]

T

0

'=200

co

=°i0

0;

E

150

La-

CD w 50 WT

WTA654


Figure 8. EGF-induced cell motility in the face of disruption of the PKC-mediated attenuation signal by replacing the PKC phosphorylation site T654 on EGFR with A654 and treatment with Calphostin C (0.05 ,kM). EGF-induced cell migration response is expressed as percent response in the absence of EGF treatment. *, basal cell migration with no treatment; (I1), basal migration in the presence of Calphostin C; (12), EGF-stimulated migration (25 nM EGF); and El, EGF-stimulated incorporation in the presence of Calphostin C. Shown are mean ± SD for three determinations.

site T654 with acidic glutamate did not alter the EGF

binding properties (Morrison et al., 1993). Bowen et al. has also reported that constitutive phosphorylation on this site by S654 replacement did not affect TPA-activated transmodulation on EGFR, nor did replacement by the nonphosphorylatable A654 (Bowen et al., 1991). These observations suggest that the PKC-mediated feedback attenuation mechanism is probably not due to decreased ligand binding but rather by altering the intrinsic tyrosine kinase activity of the receptor (Figure 5). The exact mechanism of how the phosphorylation on this site blocks EGFR signaling still remains elusive. PKC is a family of ser/thr kinases with a large number of isotypes, each differs greatly in structure, function, subcellular localization, tissue distribution, and substrate specificity (Dekker and Parker, 1994). Activation of PKC has been shown to mediate numerous biological responses (Gladhaug et al., 1992; Hug and Sarre, 1993; Dekker and Parker, 1994). These responses likely result from the diversity of these isotypes and often demonstrate isotypeand tissue/cell-type specificity. The isoform(s) involved in this feedback attenuation pathway has been determined to be a classical, calcium-dependent isotype(s) (Welsh et al., 1991). We further speculate that it is a caveolae-associated isoform, such as PKCa, because EGFR and PKC colocalize within caveolae (Mineo et al., 1995). This attenuation mechanism likely acts locally in a submembrane localization, in conjunction with the EGFR-mediated PIP2 hydrolysis that actuates the motility signal (Chen et Vol. 7, June 1996

al., 1996). Blocking PKC activation with Calphostin C inhibits most isoforms whose activation depends on DAG. Although this treatment augmented the fold-increase in EGF-induced mitogenic response (Figure 6), it decreased the absolute basal mitogenic responses by 50-60%. This indicates that in addition to the isoform(s) involved in the feedback attenuation pathway, other PKC isoforms may mediate at least part of the mitogenic response. This is consistent with previous findings demonstrating some PKC isoforms have the capacity to induce mitogenesis (Ullrich et al., 1986; Hug and Sarre, 1993). Furthermore, the A654 point mutation that replaced the PKC target site in EGFR established the specific effects on mitogenesis of PKC isoform(s) involved in the feedback pathway (Figure 7, A-C). Identification of which specific isoform(s) will involve specific inhibitors (i.e., anti-sense treatment) and will clarify and differentiate the mixed effects PKC activity exerts on mitogenesis. Phosphorylation of T654 on EGFR can serve as a general feedback mechanism in EGF-induced biological responses. It has been demonstrated that lamellipodia retraction response, like mitogenesis, is subjected to this feedback attenuation mechanism (Welsh et al., 1991). In keratinocytes, activation of PKC by nonphysiologic phorbol esters resulted in inhibition of EGF-induced cell migration, and inhibition of PKC by H7 led to a slight increase in the migration response (Ando et al., 1993). In our system, however, replacing T654 with A654 or treating cells with Calphostin C (Figure 8) to disrupt the feedback attenuation loop did not significantly affect the EGF-induced motility response. These observations suggest that different biological responses elicited by EGFR in different cells/ tissues may have differential sensitivity to this physiological feedback attenuation mechanism. Again, the tissue/cell type-specific and isoform-specific characteristics of PKC actions most likely play a role as do the subcellular localizations of PKC and PLC'y. These findings also suggest that the motility response is not sensitive to intermittent disruption of EGFR signaling by feedback attenuation, while the mitogenic response may require prolonged, constant EGFR activity. As more signaling pathways are dissected, communications among pathways become more apparent and complex (Pawson, 1995). Often multiple pathways converge to control one biological response. It is likely that subtly altering the balance of the signaling strength among these pathways dictates which response predominates. We propose that a specific PKC isoform or subcellular localization of PKC serves this role in determining whether a cell moves or proliferates upon EGF stimulation. 879

P. Chen et al.

ACKNOWLEDGMENTS We thank Drs. J. Murphy-Ullrich, J. Kudlow, T. Tumer, and M. Van Epps-Fung for helpful discussions and comments, and Kiran Gupta for technical assistance and suggestions. This work was supported by the American Cancer Society (CB-118). P.C. was supported by the University of Alabama at Birmingham School of Medicine Medical Scientist Training Program.

REFERENCES Adamson, E.D. (1990). EGF receptor activities in mammalian development. Mol. Reprod. Dev. 27, 16-22. Aderem, A. (1992). Signal transduction and the actin cytoskeleton: the roles of MARCKS and profilin. Trends Biol. Sci. 17, 438-443. Ando, Y., Lazarus, G.S., and Jensen, P.J. (1993). Activation of PKC inhibits human keratinocyte migration. J. Cell Physiol. 156,487-496. Benjamin, C.W., Linseman, D.A., and Jone, D.A. (1994). Plateletderived growth factor stimulates phosphorylation of growth factor receptor-binding protein-2 in vascular smooth muscle cells. J. Biol. Chem. 229, 31346-31349. Bleasdale, J.E., Thakur, N.R., Gremban, R.S., Bundy, G.L., Fitzpatrick, F.A., Smith, R.J., and Bunting, S. (1990). Selective inhibition of receptor-coupled phospholipase C-dependent processes in human platelets and polymorphonuclear neutrophils. J. Pharmacol. Exp. Ther. 255, 756-768. Bowen, S., Stanley, K., Selva, E., and Davis, R.J. (1991). Constitutive phosphorylation of the epidermal growth factor receptor blocks mitogenic signal transduction. J. Biol. Chem. 266, 1162-1169. Chen, P., Gupta, K., and Wells, A. (1994a). Cell movement elicited by epidermal growth factor receptor requires kinase and autophosphorylation but is separable from mitogenesis. J. Cell Biol. 124, 547-555. Chen, P., Murphy-Ullrich, J., and Wells, A. (1996). A role for gelsolin in actuating EGF receptor-mediated cell motility. J. Cell Biol. (in press). Chen, P., Xie, H., Sekar, M.C., Gupta, K.B., and Wells, A. (1994b). Epidermal growth factor receptor-mediated cell motility: phospholipase C activity is required, but MAP kinase activity is not sufficient for induced cell movement. J. Cell Biol. 127, 847-857. Chen, W.S., Lazar, C.S., Poenie, M., Tsien, R.Y., Gill, G.N., and Rosenfeld, M.G. (1987). Requirement for intrinsic protein tyrosine kinase in the immediate and late actions of the EGF receptor. Nature 328, 820-823. Davis, R.J., and Czech, M.P. (1987). Stimulation of epidermal growth factor receptor threonine 654 phosphorylation by platelet-derived growth factor in protein kinase C-deficient human fibroblasts. J. Biol. Chem. 262, 6832-6841. Decker, S.J. (1993). Transmembrane signaling by epidermal growth factor receptors lacking autophosphorylation sites. J. Biol. Chem. 268, 9176-9179. Dekker, L.V., and Parker, P.J. (1994). Protein kinase C: a question of specificity. Trends Biol. Sci. 19, 73-77. Downward, J., Waterfield, M.D., and Parker, P. (1985). Autophosphorylation and protein kinase C phosphorylation of the epidermal growth factor receptor: effect on tyrosine kinase activity and ligand binding affinity. J. Biol. Chem. 260, 14538-14546. Faletto, D.L., Kaplan, D.R., Halverson, D.O., Rosen, E.M., and VandeWoude, G.F. (1993). Signal transduction in c-met mediated motogenesis. EXS 65, 107-130. Gladhaug, I.P., Refsnes, M., and Christoffersen, T. (1992). Regulation of surface expression of high-affinity receptors for epidermal growth 880

factor (EGF) in hepatocytes by hormones, differentiating agents, and phorbol ester. Digest. Dis. Sci. 37, 233-239. Hansson, A., Skoglund, G., Lassing, I., Lindberg, U., and IngelmanSundberg, M. (1988). PKC-dependent phosphorylation of profilin is specifically stimulated by PIP2. Biochem. Biophys. Res. Commun. 150, 526-531. Hartmann, G., Naldini, L., Weidner, K.M., Sachs, M., Vigna, E., Comoglio, P.M., and Birchmeier, W. (1992). A functional domain in the heavy chain of scatter factor/hepatocyte growth factor binds the c-Met receptor and induces cell dissociation but not mitogenesis. Proc. Natl. Acad. Sci. USA 89, 11574-11578. Homma, Y., and Takenawa, T. (1992). Inhibitory effect of src homology (SH) 2/SH3 fragments of phospholipase C-y on the catalytic activity of phospholipase C isoforms: identification of a novel phospholipase C inhibitor region. J. Biol. Chem. 267, 21844-21849. Hug, H., and Sarre, T.F. (1993). Protein kinase C isoenzymes: divergence in signal transduction. Biochem. J. 291, 329-343. Jimenez-deAsua, L., and Goin, M. (1992). Prostaglandin F2 alpha decreases the affinity of epidermal growth factor receptors in Swiss mouse 3T3 cells via protein kinase C activation. FEBS Lett. 299, 235-238. Lemmon, M.A., and Schlessinger, J. (1994). Regulation of signal transduction and signal diversity by receptor oligomerization. Trends Biol. Sci. 19, 459-463. Lund, K.A., Lazar, C.S., Chen, W.S., Walsh, B.J., Welsh, J.B., Herbst, J.J., Walton, G.M., Rosenfeld, M.G., Gill, G.N., and Wiley, H.S. (1990). Phosphorylation of the epidermal growth factor receptor at threonine 654 inhibits ligand-induced intemalization and downregulation. J. Biol. Chem. 265, 20517-20523. Manske, M., and Bade, E.G. (1994). Growth factor-induced cell migration: biology and methods of analysis. Int. Rev. Cytol. 155, 49-96. Margolis, B., Rhee, S., Felder, S., Mervic, M., Lyall, R., Levitzki, A., Ullrich, A., Zilberstein, A., and Schlessinger, J. (1990a). EGF induces tyrosine phosphorylation of phospholipase C-II: a potential mechanism for EGF receptor signaling. Cell 57, 1101-1107. Margolis, B., Zilberstein, A., Franks, C., Felder, S., Kremer, S., Ullrich, A., Rhee, S.G., Skorecki, K., and Schlessinger, J. (1990b). Effect of phospholipase C-gamma overexpression on PDGF-induced second messengers and mitogenesis. Science 248, 607-610. Marikovsky, M., Breuing, K., Lui, P.Y., Eriksson, E., Higashiyama, S., Farber, P., Abraham, J., and Klagsbrun, M. (1993). Appearance of heparin-binding EGF-like growth factor in wound fluid as a response to injury. Proc. Natl. Acad. Sci. USA 90, 3889-3893. Miettinen, P.J., Berger, J.E., Meneses, J., Phung, Y., Pedersen, R.A., Werb, Z., and Derynck, R. (1995). Epithelial immaturity and multiorgan failure in mice lacking epidermal growth factor receptor. Nature 376, 337-341. Mineo, C., James, G.L., Smart, E.J., and Anderson, R.G.W. (1995). EGF stimulates the recruitment of raf to caveloae. Mol. Biol. Cell 6 (suppl) 10a. Morrison, P., Takishima, K., and Rosner, M.R. (1993). Role of threonine residues in regulation of the epidermal growth factor receptor by protein kinase C and mitogen-activated protein kinase. J. Biol. Chem. 268, 15536-15343. Nister, M., Hammacher, A., Mellstrom, K., Siegbahn, A., Ronnstrand, L., Westermark, B., and Heldin, C.-H. (1988). A gliomaderived PDGF A chain homodimer has different functional activity from a PDGF AB heterodimer purified from human platelets. Cell 52, 791-799. Pawson, T. (1995). Protein modules and signalling networks. Nature 373, 573-579. Molecular Biology of the Cell

EGF Receptor Feedback Inhibition by PLC/PKC Powis, G., Seewald, M.J., Gratas, C., Melder, D., Riebow, J., and Modest, E.J. (1992). Selective inhibition of phosphatidylinositol phospholipase C by cytotoxic ether lipid analogues. Cancer Res. 52, 2835-2840. Pruss, R.M., and Herschman, H.R. (1977). Variants of 3T3 cells lacking mitogenic response to epidermal growth factor. Proc. Natl. Acad. Sci. USA 74, 3918-3921. Reddy, C.C., Wells, A., and Lauffenburger, D. (1994). Proliferative response of fibroblasts expressing intemalization-deficient EGF receptors is altered via differential EGF depletion effects. Biotechnol. Prog. 10, 377-384. Reynolds, N.J., Talwar, H.S., Baldassare, J.J., Henderson, P.A., Elder, J.T., Voorhees, J.J., and Fisher, G.J. (1993). Differential induction of phosphatidylcholine hydrolysis, diacylglycerol formation, and PKC activation by EGF and TGF-alpha in normal human skin fibroblasts and keratinocytes. Biochem. J. 294, 535-544. Seedorf, K., Shearman, M., and Ullrich, A. (1995). Rapid and long term effects of protein kinase C on receptor tyrosine kinase phosphorylation and degradation. J. Biol. Chem. 270, 18953-18960.

Vol. 7, June 1996

Sibilia, M., and Wagner, E.F. (1995). Strain-dependent epithelial defects in mice lacking the EGF receptor. Science 269, 234-238. Ullrich, A., Riedel, H., Yarden, Y., Coussens, L., Gray, A., Dull, T.J., Schlessinger, J., Waterfield, M.D., and Parker, P. (1986). Protein kinases in cellular signal transduction: tyrosine kinase growth factor receptors and protein kinase C. Cold Spring Harbor Symp. Quant. Biol. 51, 713-724. Wells, A., Welsh, J.B., Lazar, C.S., Wiley, H.S., Gill, G.N., and Rosenfeld, M.G. (1990). Ligand-induced transformation by a noninternalizing EGF receptor. Science 247, 962-964.

Welsh, J.B., Gill, G.N., Rosenfeld, M.G., and Wells, A. (1991). A negative feedback loop attenuates EGF-induced morphological changes. J. Cell Biol. 114, 533-543.

Whitely, B., and Glaser, L. (1986). Epidermal growth factor (EGF) promotes phosphorylation at threonine-654 of the EGF receptor: possible role of protein kinase C in homologous regulation of the EGF receptor. J. Cell Biol. 103, 1355-1362.

881