Myristylation Is Required for Tyr-527 Dephosphorylation and ...

MOLECULAR AND CELLULAR BIOLOGY, Mar. 1993, p. 1464-1470 0270-7306/93/031464-07$02.00/0 Copyright © 1993, American Society for Microbiology

Vol. 13, No. 3

Myristylation Is Required for Tyr-527 Dephosphorylation and Activation of pp6Ocsrc in Mitosis SHUBHA BAGRODIA, STEPHEN J. TAYLOR, AND DAVID SHALLOWAY* Section of Biochemistry, Molecular and Cell Biology and Cancer Biology Laboratories, Department of Pathology, Cornell University, Ithaca, New York 14853 Received 10 July 1992/Returned for modification 13 August 1992/Accepted 1 December 1992

The chicken proto-oncoprotein c-Src is phosphorylated by p34cdc2 during mitosis concomitant with increased c-Src tyrosine kinase activity. On the basis of indirect evidence, we previously suggested that this is caused by partial dephosphorylation at Tyr-527, the phosphorylation of which suppresses c-Src kinase activity. In support of this hypothesis, we now show that treatment of cells with a protein tyrosine phosphatase inhibitor, sodium vanadate, blocks the mitotic increase in Src kinase activity. Also, we show that an amino-terminal mutation that prevents myristylation (and membrane localization) of c-Src does not interfere with the p34cdc2-mediated phosphorylations but blocks both mitotic dephosphorylation of Tyr-527 (in kinase-defective Src) and stimulation of c-Src kinase activity. Furthermore, in unsynchronized cells, the kinase activity of nonmyristylated c-Src is suppressed by 60% relative to wild-type c-Src, presumably because of increased Tyr-527 phosphorylation. Consistent with this, the Tyr-527 dephosphorylation rate measured in cell homogenates is much higher for wild-type, myristylated c-Src than for nonmyristylated c-Src. Tyr-527 phosphatase activity was primarily associated with the nonsoluble subcellular fraction. These findings suggest that the phosphatase(s) that acts on Tyr-527 is membrane bound and indicate that membrane localization of c-Src is necessary for its mitotic activation by dephosphorylation of Tyr-527.

Tyr-527 phosphorylation and down-regulation of its kinase activity either directly, i.e., by autophosphorylation (8, 9, 35), or indirectly by activation of C-terminal Src kinase (for further discussion, see references 35 and 37). In any case, we believe that this decrease in Tyr-527 phosphorylation in c-Src(R295) reflects a smaller decrease in c-Src phosphorylation. To further examine this model, we studied the effects of a protein tyrosine phosphatase inhibitor on mitotic activation of c-Src. We also studied the effects of myristylation on Tyr-527 phosphorylation level and Src kinase activity in mitosis and the rates of Tyr-527 dephosphorylation in fractionated cell homogenates to determine whether altered subcellular localization influences c-Src interaction with Tyr-527 kinases and phosphatases.

Chicken c-src is the cellular homolog of the Rous sarcoma virus transforming gene v-src. Both genes encode membrane-associated 60-kDa phosphoproteins with tyrosine kinase activity (25). Association of v-Src and c-Src with cellular membranes requires covalent attachment of myristate to Gly at position 2, and mutation of this residue prevents myristylation and membrane binding (3, 12, 31, 33, 34). Other amino-terminal sequence determinants are also involved in membrane association (18), and at least for v-Src, association may involve a 32-kDa Src receptor protein (29, 30). Myristylation is not required for the mitogenic effect of v-Src (4) but is required for transformation of chicken embryo fibroblasts by v-Src (10, 17) and activated c-Src (31, 41) (although our preliminary results suggest that nonmyristylated, activated c-Src is weakly transforming in mouse fibroblasts). During mitosis, c-Src is phosphorylated by p34cdc2 or a related kinase at three amino-terminal serine and threonine residues, and c-Src kinase activity is increased (6, 22, 35, 38, 40). These mitosis-specific phosphorylations (MSPs) retard c-Src electrophoretic mobility but, by themselves, do not stimulate its kinase activity (22, 38). C-Src kinase activity can be negatively regulated by phosphorylation of its C-terminal Tyr-527 residue (5, 20, 28), and we (1, 37) and others (16) have proposed that a 10 to 20% decrease in Tyr-527 phosphorylation is the direct cause of its mitotic activation. Although we cannot reliably directly detect such a small decrease, the kinase activities of c-Src mutants lacking Tyr-527 are not further increased during mitosis (1, 16), and Tyr-527 phosphorylation in kinase-defective c-Src(R295) (containing a Lys-295--+Arg substitution) decreases 70% in mitosis (1). The difference in Tyr-527 phosphorylation between wild-type (wt) and kinase-defective c-Src might reflect a negative feedback loop in which activated c-Src promotes *

MATERIALS AND METHODS

Plasmid constructions. Plasmids pM5HHB5 (expressing wt c-Src) (20) and pcR295 [expressing c-Src(R295)] (38), which contain Moloney murine leukemia virus long terminal repeats for efficient expression of src coding sequences, have been previously described. Plasmids pcLN and pcR295/LN were constructed by cleaving expression vectors pM5HHB5 and pcR295, respectively, at their unique Sall (1,187 bp upstream of the c-src coding region) and MluI (within the src coding region) sites and replacing the 1,962-bp c-src N-terminal coding fragment with the corresponding SalI-MluI fragment from pRSVcsrcLN (a gift from J. Brugge [33]). Cell culture technique. Cells were grown at 37°C and 10% CO2 in Dulbecco's modified Eagle's medium (GIBCO-BRL) supplemented with 5% calf serum. Mass cultures were created by cotransfection with 1 ,ug of circular c-src plasmid and 0.1 ,ug of pSV2neo expression plasmid (which confers G418 resistance [39]) into NIH 3T3 cells as described previously (20, 36). Coselected mass cultures contained 35 to 50 G418-resistant colonies. Cells were plated at 105 cells per 100-mm plate prior to initiation of all experiments. Mitotic

Corresponding author. 1464

VOL. 13, 1993

MYRISTYLATION AND MITOTIC ACTIVATION OF pp6OC-Src

cells were collected by treatment with 0.4 [ug of nocodazole (Sigma) per ml, an inhibitor of microtubule formation, for 6.5 to 7.5 h and then mechanical shake-off as described previously (38). Vanadate treatment of cells. A 0.1 M stock solution was freshly prepared by dissolving Na3VO4 (Fisher Scientific) in 0.1 M unbuffered HEPES (N-2-hydroxyethylpiperazine-N'2-ethanesulfonic acid) for each experiment (2). Unsynchronized cells were treated with 500 pM Na vanadate for 30 min before lysis. In pilot experiments, we found that vanadate treatment prevents NIH 3T3 cells from entering mitosis (possibly by suppressing activation of p34cdc2), so it was necessary to study its effects after collecting mitotic cells. That is, nocodazole-arrested mitotic cells were collected by shake-off, replated, and incubated with 500 FxM vanadate for 30 min before lysis in the continued presence of nocodazole. Immunoprecipitation and phosphoprotein analyses. Equilibrium pp6QY' expression levels were determined as previously described (44) except that cells were labelled with 90% methionine-free essential medium (compounded from Selectamine kits; GIBCO-BRL), 10% Dulbecco's modified Eagle's medium, and 10% calf serum. Procedures for immunoprecipitation and tryptic mapping of 32P-labelled Src proteins from unsynchronized and mitotic c-Src overexpresser cells have been previously described (38). Western blots were immunoprobed with anti-Src monoclonal antibody (MAb) 327 (21) or antiphosphotyrosine (anti-PTyr) MAb 4G10 (UBI) and 1251I-protein A (ICN) as described previously (1) except that transfer membranes were blocked with 2% bovine serum albumin and 0.1% Tween 20 was included in all wash buffers. Specific kinase activity assay. Unlabelled unsynchronized and mitotic cells were washed twice with STE (0.15 M NaCl, 0.02 M Tris [pH 7.2], 1 mM EDTA) or phosphate-buffered saline (PBS; 0.14 M NaCl, 2.6 mM KCl, 10 mM Na2HPO4 [pH 7.4]) and lysed in modified RIPA buffer (1% Triton X-100, 1% Na deoxycholate, 0.1% sodium dodecyl sulfate [SDS], 0.15 M NaCl, 0.02 M NaH2PO4 [pH 7.2], 2 mM EDTA, 30 mM Na4P2O7, 50 mM NaF, 1 mM phenylmethylsulfonyl fluoride, 10 ,ug of aprotinin per ml, 0.2 to 0.5 mM Na3VO4, 1 mM ZnCl2, 1 mM Na2MoO4) or 100 mM MOPS (morpholinepropanesulfonic acid) (pH 7.4)-100 mM NaCl1% Nonidet P-40 (vol/vol-1 mM EDTA-50 mM NaF-10 jig of aprotinin per ml-10 [Lg of leupeptin per ml-1 mM phenylmethylsulfonyl fluoride-0.5 mM Na3VO4. For immunoprecipitation, lysates were normalized to contain approximately equal amounts of Src. Kinase assays were performed by incubating 6% of the Src immune complex with 1 puM [_y-32P]ATP and 8 ,ug of acid-denatured rabbit muscle enolase (Sigma) for 4 min at 23°C as previously described (1). A second portion (40%) of the Src immune complex was Western blotted (immunoblotted) and immunoprobed with anti-Src antibody MAb 327 and 2s1-protein A as described above to determine the amount of Src protein in the kinase reaction. Specific kinase activities were calculated by dividing radioactivities incorporated into the 32P-labelled enolase bands (determined by Betascope [Betagen, Waltham, Mass.] analysis) by those in the 125I-labelled Src bands (determined by gamma counting and/or Phosphorimager [Molecular Dynamics, Sunnyvale, Calif.] analysis). Tyrosine 527 phosphatase assay. Cells overexpressing wt c-Src or nonmyristylated c-Src(LN) were washed on 100-mm plates twice with ice-cold PBS and then scraped into 1 ml of PBS per plate. Cells from four plates were pelleted, resuspended in 1 ml of hypotonic buffer (10 mM HEPES [pH 7.5], 1 mM MgCl2, 10 pug of leupeptin per ml, 10 pg of

1465

aprotinin per ml) and incubated on ice for 5 min. After homogenization with a Dounce homogenizer (40 strokes), 1 ml of homogenate was added to 200 pI of a 6x concentrated solution to give (final concentrations) 10 mM HEPES (pH 7.5), 100 mM NaCl, 5 mM EDTA, 10 pug of aprotinin per ml, and 10 pug of leupeptin per ml. Homogenate (400 pul) was removed and kept on ice, and the remaining 800 p.1 was centrifuged at 105,000 x g for 20 min. The supernatant was removed, and the pellet was resuspended in 800 RI of buffer of the same composition as the adjusted homogenate. Portions (400 p.1 each) of the homogenate, supernatant (soluble fraction), and resuspended pellet (particulate fraction) were mixed with 100 pI of 250 mM imidazole (pH 7.0)-10 mM dithiothreitol and incubated at 370C. At the indicated time points, 60 p.1 was removed into 60 p.1 of (2x) electrophoresis sample buffer and immediately boiled. Samples were subjected to SDS-polyacrylamide gel electrophoresis (PAGE) and anti-PTyr/anti-Src immunoblotting as described above, and the relative PTyr content of Src was determined by Phosphorimager analysis. RESULTS Vanadate suppresses the mitosis-specific increase in kinase activity of c-Src mutated at Tyr-416. Our previous studies have indirectly suggested that stimulation of c-Src kinase activity during mitosis is due to Tyr-527 dephosphorylation (1). To directly test this hypothesis, we examined the effect of stabilizing Tyr-527 phosphorylation on this stimulation by treating unsynchronized and mitotic cells overexpressing c-Src with sodium vanadate for 30 min and then by an immune complex kinase assay using acid-denatured enolase as an exogenous substrate. Vanadate stabilizes phosphorylation at both Tyr-416 and Tyr-527 in wt c-Src; these phosphorylations have competing effects on Src kinase activity (19). To eliminate this competition (and to more properly model mitotic c-Src, which is not phosphorylated at Tyr-416 [7]), we analyzed c-Src(F416) which contains a Tyr-416--Phe substitution (Fig. 1). As previously shown (19), vanadate treatment suppressed c-Src(F416) specific kinase activity in unsynchronized cells (cf. lanes 1 and 3), probably by causing a small, undetectable increase in Tyr527 phosphorylation. In the absence of vanadate, the specific kinase activity of c-Src(F416) from mitotic cells was approximately twofold higher than that from unsynchronized cells (Fig. 1, cf. lanes 1 and 2); this stimulation was completely suppressed by vanadate treatment (Fig. 1, lanes 3 and 4; Table 1), implying that a change in Tyr-527 phosphorylation was required for activation. Inclusion of up to 500 p.M vanadate in the kinase assay did not noticeably affect phosphorylation of enolase (data not shown), confirming that the effect of vanadate resulted from its activity in vivo. Mitosis-specific Tyr-527 dephosphorylation and kinase activation is blocked in nonmyristylated c-Src. To examine the role of myristylation (and membrane localization) in mitosisspecific Cdc2-mediated phosphorylation of c-Src, Tyr-527 dephosphorylation, and activation of Src kinase activity, we modified previously described src expression plasmids pM5HHB5 (15) and pcR295 (38), which express wt c-Src and kinase-defective c-Src(R295), to create plasmids pcLN and pcR295/LN, which encode nonmyristylated kinase-active or kinase-defective variants, respectively. The src proteins encoded by these mutants contain four additional amino acids (Met Ala Ala Ala) at their amino termini and are not myristylated at Gly-2 nor localized to the plasma membrane (33). Plasmids were cotransfected into NIH 3T3 cells with

1466

BAGRODIA ET AL. _

MOL. CELL. BIOL.

F416

U

M

1

2

U

M

+

+

3

4

VAN

A U

1

M 2

In

U1

U)

asct-:

c ccm:

U

3

M

4

U

5

az

M U 6 7

M

8

U 9

10

M

8

9

10

A anti- PY

__

_

-a

_

_

"

EN

2

1

3

4

5

6

7

B anti-SRC

B

1

2

3

_SRC

FIG. 1. Effect of vanadate treatment on the relative specific kinase activity of mutant c-Src from unsynchronized and mitotic cells. c-Src(F416) was immunoprecipitated with MAb 327 from untreated (-) or vanadate (VAN)-treated (+) unsynchronized (U) (lane 1 and 3) and mitotic (M) (lanes 2 and 4) NIH 3T3-derived Src overexpresser cells. (A) A portion (6%) of each immunoprecipitate was incubated with acid-denatured enolase (EN) as an exogenous substrate and [,y-32P]ATP, and products were analyzed by SDS7.5% PAGE and autoradiography (Table 1). (B) Another portion (40%) of the immunoprecipitate was Western blotted and probed with anti-Src antibody and 1251I-protein A.

coselection plasmid pSV2neo, and mass cultures of G418resistant colonies were isolated. Immunoprecipitation with anti-Src MAb 327 (21) after equilibrium metabolic labelling with [35S]methionine showed that the Src proteins were expressed at least 10 times above the level of endogenous mouse c-Src. In agreement with Schuh and Brugge (33), we confirmed that wt c-Src was mainly localized in the particulate fraction and nonmyristylated c-Src(LN) in the soluble fraction of cell homogenates (see Fig. 5). We have previously shown that Tyr-527 phosphorylation of c-Src(R295) is decreased in mitosis (1). Here we compared Tyr-527 phosphorylation in c-Src(R295) and c-Src(R295/LN) to study the role of myristylation in this phenomenon. Tyr-416 is not phosphorylated in unsynchronized and mitotic cells (see Fig. 4 and references 1, 6, and 38), so PTyr-527 is the only significantly phosphorylated tyrosine in vivo and its level can be measured with anti-PTyr antibodies. wt c-Src, TABLE 1. Relative specific kinase activity of c-Src(F416) following vanadate treatment of cells Expressed

protein

protein

c-Src(F416) c-Src(F416)

Vanadate -

+

Id

k.

-~~~~O

;

4

Relative specific kinase

activity'

Uns

Mit

Mit/Uns

1.0 0.6

1.8 0.6

1.8 x. 1.04 1.0 _x 1.06

a Specific kinase activity of c-Src mutant from unsynchronized (Uns) and mitotic (Mit) cells untreated (-) or treated with vanadate (+) were calculated from the ratio of 32P incorporated into enolase relative to the amounts of immunoblotted c-Src in experiments of the type shown in Fig. 1. Values are geometric means fractional standard errors of the mean from three experiments and are normalized to the kinase activity of c-Src(F416) from unsynchronized cells untreated with vanadate.

FIG. 2. Effect of myristylation on Tyr-527 phosphorylation of wt and kinase-defective c-Src mutants from unsynchronized and mitotic cells. Duplicate portions of detergent cell lysates containing wt or mutant c-Src proteins from unsynchronized (U) and mitotic (M) Src overexpresser cells were separated by SDS-10% PAGE, subjected to Western blotting, and immunoprobed with anti-Src MAb 327 or anti-PTyr MAb 4G10 (anti-PY) and 125I-protein A and autoradiography (Table 2). Lanes: 1 and 2, wt c-Src; 3 and 4, kinase-defective c-Src(R295) from a clonal cell line; 5 an 6, nonmyristylated and kinase-defective c-Src(R295/LN); 7 and 8, kinasedefective c-Src from a mass-cultured cell line.

nonmyristylated c-Src(LN), kinase-defective c-Src(R295), and kinase-defective, nonmyristylated c-Src(R295/LN) were immunoprecipitated from unlabelled unsynchronized and mitotic cells, and duplicate immunoblots were probed with anti-Src and anti-PTyr antibodies. Alternatively, whole-cell lysates were probed with anti-Src and anti-PTyr antibodies (Fig. 2 and Table 2; Src was the predominant PTyr-containing protein in these overexpresser cells, allowing quantitation of its tyrosine phosphorylation without immunoprecipitation). The characteristic mitotic decrease in Tyr-527 phosphorylation was observed in two different kinase-defective c-Src(R295) overexpresser cell lines during mitosis (Fig. 2, lanes 3, 4, and 9, 10), and the lack of myristylation in kinase-defective c-Src(R295/LN) blocked the mitosis-speTABLE 2. Relative kinase activities and phosphorylation levels Relative specific kinase

c-Src c-Src(LN) (myr-) c-Src

c-Src(LN) c-Src(R295) c-Src(R295/LN)

Relative PTyr

activity'

Expressed protein Uns

Mit

1.0 0.4

2.5 0.5

Mit/Uns

2.5 1.2

x X

Uns

Mit

1.0 1.0 0.9 0.8

1.0 0.9 0.4 1.0

levels'

Mit/Uns

1.08 1.09 1.0 0.9 0.4 1.2

X x X x

1.10 1.05 1.10 1.12

a Specific kinase activities of wt and nonmyristylated c-Src (myr-) from unsynchronized (Uns) and mitotic (Mit) cells were calculated from the ratio of 32p incorporated into enolase relative to the amounts of immunoblotted c-Src in experiments of the type shown in Fig. 3. Values are geometric mean x fractional standard error of the mean from five experiments and are normalized to the kinase activity of c-Src from unsynchronized cells. b The relative levels of tyrosine phosphorylations in wt and mutant c-Src proteins from unsynchronized (Uns) and mitotic (Mit) cells were determined from gamma counting of double-immunoblotting experiments like that shown in Fig. 2. Values are geometric mean fractional standard error of the mean from four experiments.

VOL. 13, 1993

MYRISTYLATION AND MITOTIC ACTIVATION OF pp60C-Src wt

U 1

M 2

wt

mvrU M 3 4

B

A

-

mvr

U

M U M

1

2

3

4

EN

FIG. 3. Relative specific kinase activity of nonmyristylated c-Src mutant from unsynchronized and mitotic cells. Unlabelled wt (lanes

1 and 2) and nonmyristylated c-Src (myr-, lanes 3 and 4) from unsynchronized (U) and mitotic (M) NIH 3T3-derived Src overexpresser cells were immunoprecipitated with MAb 327, and specific kinase activities were assayed as described in the legend to Fig. 1. (A) Enolase (EN) phosphorylation by Src. (B) Western blots immunoprobed with with anti-Src antibody and 1"I-protein A showing the relative amounts of Src used in the kinase reaction.

cific Tyr-527 dephosphorylation (Fig. 2, lanes 7 and 8). The absence of mitosis-specific Tyr-527 dephosphorylation in c-Src(R295/LN) was confirmed by partial Staphylococcus aureus V-8 proteolytic analysis (data not shown). The predicted 10 to 20% decrease in Tyr-527 phosphorylation in wt c-Src is below experimental accuracy, so it was not possible to determine whether Tyr-527 phosphorylation was affected in nonmyristylated c-Src(LN) by either immunoblotting or V-8 analysis.

1467

As mentioned earlier, we have argued that the mitosisspecific decrease in Tyr-527 phosphorylation observed in kinase-defective c-Src is indicative of a smaller decrease in Tyr-527 phosphorylation in wt c-Src which is responsible for the stimulation of c-Src kinase activity (1, 37). Given the lack of Tyr-527 dephosphorylation in nonmyristylated, kinasedefective c-Src, this model predicts that the specific kinase activity of nonmyristylated c-Src would not be increased during mitosis. To test this, we immunoprecipitated wt and nonmyristylated c-Src proteins from unsynchronized and mitotic cells, and their specific kinase activities were measured in vitro (Fig. 3 and Table 2). As previously shown (1, 6, 16), the specific kinase activity of wt c-Src was stimulated two- to threefold in mitotic cells. However, the absence of myristylation significantly inhibited, by 80%, the mitotic activation of c-Src(LN) (Fig. 3, lanes 2 and 4). Furthermore, its kinase activity was suppressed in unsynchronized cells by 60% relative to wt c-Src (Fig. 3, lanes 1 and 3). Thus, the lack of myristylation is associated with a decrease in the specific kinase activity of c-Src during interphase and suppression of its activation during mitosis. Myristylation is not required for MSPs of c-Src. Since the MSP sites play a role in stimulating c-Src kinase activity during mitosis (37), we examined the effect of myristylation on these p34cdc2-mediated phosphorylations to see whether the absence of phosphorylation at these MSP sites could account for the suppression of stimulation in nonmyristylated c-Src kinase activity. Tryptic phosphopeptide maps of 32P-labelled wt c-Src immunoprecipitated from unsynchronized cells displayed previously characterized spots (Fig.

LOG

7*1

7f

A~~~~~~~~ E

G

ii~~~s}I MITOTIC UPPER

" s

__

_

_

_

s**-~~~~~0310I

I|

FIG. 4. Tryptic phosphopeptide analysis of wt, nonmyristylated, and kinase-defective c-Src mutants from unsynchronized and mitotic cells. 32P-labelled wt and mutant c-Src proteins were immunoprecipitated from unsynchronized and mitotic NIH 3T3-derived Src overexpresser cells and eluted, and tryptic phosphopeptides were mapped as described in Materials and Methods. In panels A to D, proteins were isolated from unsynchronized cells in log-phase growth (Log), and in panels E to H, proteins were isolated from the strongly retarded c-Src band from mitotic cells (Mitotic Upper). Panels: A and E, wt c-Src; B and F, myristylation-defective c-Src(LN); C and G, kinase-defective c-Src(R295); D and H, kinase-defective and myristylation-defective c-Src(R295/LN). Spot numbering is described in the text.

MOL. CELL. BIOL.

BAGRODIA ET AL.

1468

myr-

wt

A HOMOGENATE

Time (min.) 0 4 8 16 32 48 0

4

8 16 32 48

HOMOGENATE [ I

SOLUBLE

PARTICULATE

0 4

8 16 32 48 H

0

PARTICULATE

SOLUBLE

rI

|l

4 8 16 32 48 0

4 8 16 32 48

8 16 32 48 H

0 4

Anti-PTyr

Anti-Src

_

_I

_

-

4

B 0

a

I

0 A

0.1 0

1

0

20

30

40

50

Time ,(r nutes:

4A): spot 1 (protein kinase A-phosphorylated phosphoserine 17 [27]), spots 3 and 4 (PTyr-527 [7]), spots 5 and 6 (protein kinase C-phosphorylated phosphoserine 12 [14, 43)], and spot 7 (protein kinase C-phosphorylated phosphoserine 48 [14, 43]). c-Src(LN), c-Src(R295), and c-Src(R295/LN) from unsynchronized cells displayed similar maps except that spots 5, 6, and 7 (Ser-12 and Ser-48) were not observed in the nonmyristylated mutants (Fig. 4B and D), even after pretreating cells with 12-O-tetradecanoylphorbol-acetate (TPA) (data not shown). In mitotic cells, wt c-Src displayed previously observed additional mitosis-specific spots arising from p34cdc2-mediated phosphorylations (Fig. 4E): spots 8 and 11 (phosphothreonine 46), spot 9 (phosphoserine 72), spot 10 (phosphoserine), and spot 12 (phosphoserine 17 and phosphothreonine 34) (38). Spot 7 (phosphoserine 48) was not observed in lysates from mitotic cells (Fig. 4B), even after TPA treatment to stimulate protein kinase C (data not shown). c-Src(LN), c-Src(R295), and c-Src(R295/LN) from mitotic cells displayed similar maps except that spots 5 and 6 were again absent in the nonmyristylated c-Src mutants (Fig. 4F and H). The intensities of spots in tryptic phosphopeptide maps showed significant variations between experiments (due to various factors such as differential solubility of phosphopeptides or differential cleavage), so they cannot be used to assess quantitative changes in (e.g., Tyr-527) phosphorylation levels. However, they are sufficiently accurate for qualitative purposes, as the complete absence of detectable Tyr-416 phosphorylation supports our assumption that the PTyr immunoblots (in Fig. 2) reflect c-Src Tyr-527 phosphorylation alone. Relative rates of Tyr-527 dephosphorylation in soluble and particulate subcellular fractions. Detergent-free homogenates prepared from cells overexpressing wt c-Src or nonmyristylated c-Src(LN) were separated into particulate and soluble fractions (at 4°C to minimize dephosphorylation during this

FIG. 5. Relative rates of tyrosine dephosphorylation of wt and nonmyristylated c-Src in soluble and particulate subcellular fractions. (A) Homogenates from wt and nonmyristylated (myr-) c-Src overexpresser cells were fractionated by centrifugation into soluble and particulate components (at 40C) in the presence of 5 mM EDTA (to inhibit kinase activity) as described in Materials and Methods. To measure the rate of tyrosine dephosphorylation (at t = 0), fractions (normalized to equivalent volume) were shifted to 370C and incubated for the indicated durations. Proteins were separated by SDS-10% PAGE, and blots were probed with anti-PTyr or anti-Src antibodies and 125I-protein A and were quantitated by Phosphorimager analysis. Lanes marked H represent aliquots of homogenate that were boiled in electrophoresis sample buffer immediately following homogenization. Longer exposure of the autoradiographs revealed a signal, probably due to endogenous mouse c-Src, in the particular fraction from myrc-Src overexpresser cells. (B) Relative PTyr contents were determined from the PTyr/Src ratio at each time point and plotted (values are relative to the phosphorylation level of the wt homogenate). Note that the ordinate is only a relative scale and does not indicate absolute phosphorylation levels. Symbols: 0, wt homogenate; A, wt particulate fraction; wt soluble fraction; 0, nonmyristylated c-Src homogenate; El, nonmyristylated c-Src soluble fraction. Values are averages from two independent experiments. U,

30-min preparatory phase). These fractions and the unfractionated cell homogenate were then incubated at 370C in the presence of EDTA (to inhibit Mg2+-dependent phosphorylation), and the rate of tyrosine (i.e., Tyr-527) dephosphorylation was assayed by anti-PTyr and anti-Src immunoblotting (Fig. 5). In homogenates, wt c-Src was dephosphorylated with a t112 of -10 min. However, dephosphorylation of nonmyristylated c-Src(LN), which was predominantly soluble, was not detected even after 48 min (Fig. 5). Similarly, no dephosphorylation of soluble wt c-Src was detected. Tyr-527 dephosphorylation in wt c-Src occurred to a greater extent during the preparatory period in the particulate fraction than in the unfractionated homogenate, possibly reflecting a concentration of proteins in the pelleted fraction. DISCUSSION In unsynchronized cells, nonmyristylated c-Src(LN) had measurably lower specific kinase activity than wt c-Src. This phenomenon, consistent with a previous report (42), suggests that Tyr-527 is phosphorylated to a higher level in c-Src(LN) than in wt c-Src. A small increase in the Tyr-527 phosphorylation level (e.g., from 90 to 95%) would be experimentally undetectable, but would result in a significant fractional decrease in the amount of nonphosphorylated Tyr-527 and, hence, in relative c-Src kinase activity. This could reflect either increased or decreased accessibility of

VOL. 13, 1993

MYRISTYLATION AND MITOTIC ACTIVATION OF pp6oC-Src

nonmyristylated c-Src(LN) to a Tyr-527 kinase or phosphatase. The latter possibility is supported by our demonstration that the rate of Tyr-527 dephosphorylation in cell homogenates is greatly decreased in nonmyristylated cSrc(LN) relative to that of wt c-Src. It is unlikely that this difference merely reflects saturation of a phosphatase by overexpressed nonmyristylated c-Src since wt c-Src was overexpressed at similar levels. The finding that Tyr-527 dephosphorylation requires membrane localization of c-Src suggests that the tyrosine phosphatase that acts on this residue is itself membrane associated. This is further supported by our finding that the rate of Tyr-527 dephosphorylation of wt c-Src in the particulate fraction is increased relative to that in the soluble fraction and is consistent with the observation that many protein tyrosine phosphatases are membrane associated (11). In particular, the receptor-like protein tyrosine phosphatase a, which has been shown to dephosphorylate Tyr-527 of c-Src in vitro and in vivo, is membrane associated (45). In addition, nonmyristylated c-Src may also be more accessible in vivo to a cytosolic Tyr-527 kinase such as C-terminal Src kinase (23, 24, 26, 32). In mitotic cells, stimulation of nonmyristylated c-Src(LN) kinase activity was suppressed relative to that of wt c-Src. This could be explained by reduced accessibility of nonmyristylated c-Src to a relevant mitotically regulated kinase or phosphatase. Consistent with this, nonmyristylated, kinasedefective c-Src(R295/LN) did not display the decrease in Tyr-527 phosphorylation observed in myristylated, kinasedefective c-Src(R295). Although myristylation was required for attenuations in Tyr-527 phosphorylation and, hence, in kinase activity, it was not required for the MSPs of c-Src. As previously reported for v-Src (17), myristylation was required for phosphorylation of Ser-12 and Ser-48 in response to TPA treatment (data not shown). This phenomenon is probably explained by a requirement for membrane localization of Src for it to serve as a good substrate for protein kinase C, which translocates to the membrane when activated and has been shown to phosphorylate Ser-12 and Ser-48 (14). The lack of phosphorylation of Ser-48 during mitosis, even following TPA treatment, may reflect interference with the protein kinase C substrate recognition site by the additional MSP at Thr-46. We have previously shown that mitotic activation of c-Src is partly driven by a two-step process in which (i) MSPs by p34cdc2 or a related kinase sensitize or desensitize c-Src to a

Tyr-527 phosphatase or kinase, respectively, thereby causing (ii) dephosphorylation of Tyr-527 and activation of c-Src (37). Additional MSP-independent events, e.g., modulation of Tyr-527 kinase/phosphatase activity by other mechanisms, also participate (37, 40). Although nonmyristylated c-Src is phosphorylated at the MSP sites, it is evidently not subject to the second step of the two-step process or to the MSP-independent component of Src activation. This is most easily explained if the observed mitotic changes in c-Src result from activation of a membrane-associated Tyr-527 phosphatase, the action of which is evaded by cytosolic, nonmyristylated c-Src(LN) and c-Src(R295/LN). However, since the Tyr-527 phosphorylation level is governed only by the ratio of phosphatase to kinase activity, we cannot exclude the possibility that it is the phosphorylation rate that changes during mitosis (e.g., see Fig. 4 and reference 37). The finding that 30-min vanadate treatment suppressed the mitosis-specific increase in Src kinase activity supports the hypothesis (1, 16, 37) that decreased Tyr-527 phosphoryla-

1469

tion is required for this activation. This effect was probably not due to inactivation of p34"""2 kinase by stabilization of Tyr-15 phosphorylation (13), since no change in the amount of retarded mobility c-Src (i.e., containing p34"d"2-mediated MSPs) was apparent. (Longer vanadate treatment [>1 h] did, however, result in loss of the retarded mobility mitotic c-Src band, indicating dephosphorylation of at least one MSP site [data not shown].) The ability of vanadate to act within 30 min was consistent with our measurement of the Tyr-527 phosphate off-rate in crude homogenates, -10 min-'. The suppression of activation by vanadate and the correlated effects of myristylation on mitotic activation of c-Src and mitotic Tyr-527 dephosphorylation of c-Src(R295/ LN) support the model in which a reduction in the level of Tyr-527 phosphorylation is the direct cause of c-Src activation in mitosis. The membrane-associated protein tyrosine phosphatase(s) that acts on Tyr-527 may also act on other important regulatory proteins, including other src family members. It will be important to determine how the activity of this phosphatase is regulated and whether its activation is responsible for the mitotic stimulation of c-Src kinase activity. ACKNOWLEDGMENTS We thank J. Brugge for MAb 327 and S. Schuh for plasmid

pRSVcsrcLN. This research was supported by Public Health Service grants CA32317 and CA47333 and a research career development award (to D.S.) from the National Cancer Institute. REFERENCES 1. Bagrodia, S., I. Chackalaparampil, T. Kmiecik, and D. Shalloway. 1991. Tyrosine 527 phosphorylation and mitotic activation of p60cs?". Nature (London) 349:172-175. 2. Bernier, M., M. Laird, and D. M. Lane. 1988. Effect of vanadate on the cellular accumulation of ppl5, an apparent product of insulin receptor tyrosine kinase action. J. Biol. Chem. 263: 13626-13634. 3. Buss, J. E., M. P. Kamps, K. Gould, and B. M. Sefton. 1986. The absence of myristic acid decreases membrane binding of p605" but does not affect tyrosine protein kinase activity. J. Virol. 232:468-474. 4. Calothy, G., D. Laugier, F. R. Cross, R. Jove, T. Hanafusa, and H. Hanafusa. 1987. The membrane-binding domain and myristylation of p60`sr are not essential for stimulation of cell proliferation. J. Virol. 61:1678-1681. 5. Cartwright, C. A., W. Eckhart, S. Simon, and P. L. Kaplan. 1987. Cell transformation by pp60Orc mutated in the carboxyterminal regulatory region. Cell 49:83-91. 6. Chackalaparampil, I., and D. Shalloway. 1988. Altered phosphorylation and activation of pp60csrC during fibroblast mitosis. Cell 52:801-810. 7. Cooper, J. A., K. L. Gould, C. A. Cartwright, and T. Hunter. 1986. Tyr527 is phosphorylated in pp60csrC: implications for regulation. Science 231:1431-1434. 8. Cooper, J. A., and C. S. King. 1986. Dephosphorylation or antibody binding to the carboxy terminus stimulates pp6Oc-sr. Mol. Cell. Biol. 6:4467-4477. 9. Cooper, J. A., and A. MacAuley. 1989. Positive and negative regulation of pp6ocs'? by intermolecular autophosphorylation. Proc. Natl. Acad. Sci. USA 85:4232-4236. 10. Cross, F. R., E. A. Garber, D. Pellman, and H. Hanafusa. 1984. A short sequence in the pp60`rc N terminus is required for p6Osrc myristylation and membrane association and for cell transformation. Mol. Cell. Biol. 4:1834-1842. 11. Fisher, E. H., H. Charbonneau, and N. K. Tonks. 1991. Protein tyrosine phosphatases: a diverse family of intracellular and transmembrane enzymes. Science 253:401-405. 12. Garber, E. A., F. R. Cross, and H. Hanafusa. 1985. Processing of pp6o-src to its myristylated membrane-bound form. Mol.

1470

MOL. CELL. BIOL.

BAGRODIA ET AL.

Cell. Biol. 5:2781-2788. 13. Gould, K. L., and P. Nurse. 1989. Tyrosine phosphorylation of the fission yeast cdc2+ protein kinase regulates entry into mitosis. Nature (London) 342:39-45. 14. Gould, K. L., J. R. Woodgett, J. A. Cooper, J. E. Buss, D. Shalloway, and T. Hunter. 1985. Protein kinase C phosphorylates pp6l5rc at a novel site. Cell 42:849-857. 15. Johnson, P. J., P. M. Coussens, A. V. Danko, and D. Shalloway. 1985. Overexpressed pp6Ocs"' can induce focus formation without complete transformation of NIH 3T3 cells. Mol. Cell. Biol. 5:1073-1083. 16. Kaech, S., L. Covic, A. Wyss, and K. Ballmer-Hofer. 1991. Association of pp60csrc with polyomavirus middle-T abrogates mitosis-specific activation. Nature (London) 350:431-433. 17. Kamps, M. P., J. E. Buss, and B. M. Sefton. 1986. Rous sarcoma virus transforming protein lacking myristic acid phosphorylates known polypeptide substrates without inducing transformation. Cell 45:105-112. 18. Kaplan, J. M., H. E. Varmus, and J. M. Bishop. 1990. The src protein contains multiple domains for specific attachment to membranes. Mol. Cell. Biol. 10:1000-1009. 19. Kmiecik, T. E., P. J. Johnson, and D. Shalloway. 1988. Regulation by the autophosphorylation site of overexpressed pp6Ocsrc. Mol. Cell. Biol. 8:4541-4546. 20. Kmiecik, T. E., and D. Shalloway. 1987. Activation and suppression of pp60csrc transforming ability by mutation of its primary sites of tyrosine phosphorylation. Cell 49:65-73. 21. Lipsich, L. A., A. J. Lewis, and J. S. Brugge. 1983. Isolation of monoclonal antibodies that recognize the transforming proteins of avian sarcoma virus. J. Virol. 48:352-360. 22. Morgan, D. O., J. M. Kaplan, J. M. Bishop, and J. E. Varmus. 1989. Mitosis-specific phosphorylation of p60csrc by p34cdc2 associated protein kinase. Cell 57:775-786. 23. Nada, S., M. Okada, A. MacAuley, J. A. Cooper, and H. Nakagawa. 1991. Cloning of a complementary DNA for a protein-tyrosine kinase that specifically phosphorylates a negative regulatory site of p60csrc. Nature (London) 35:69-72. 24. Okada, M., S. Nada, T. Yamanashi, T. Yamamoto, and H. Nakagawa. 1991. CSK: a protein-tyrosine kinase involved in regulation of src family kinases. J. Biol. Chem. 266:2424924252. 25. Parsons, J. T., and M. J. Weber. 1989. Genetics of src: structure and function of a protein tyrosine kinase. Curr. Top. Microbiol. Immunol. 147:79-127. 26. Partanen, J., E. Armstrong, M. Bergman, T. P. Makela, H. Hirvonen, K. Heuber, and K. Alitalo. 1991. cyl encodes a putative cytoplasmic tyrosine kinase lacking the conserved tyrosine autophosphorylation site (Y416src). Oncogene 6:20132018. 27. Patchinsky, T., T. Hunter, and B. M. Sefton. 1986. Phosphorylation of the transforming protein of Rous sarcoma virus: direct demonstration of phosphorylation of serine 17 and identification of an additional site of tyrosine phosphorylation in p6ovrC of Prague sarcoma virus. J. Virol. 59:73-81. 28. Piwnica-Worms, H., K. B. Saunders, T. M. Roberts, A. E. Smith, and S. H. Cheng. 1987. Tyrosine phosphorylation regulates the biochemical and biological properties of pp6Ocsrc. Cell 49:75-82. 29. Resh, M. D. 1989. Specific and saturable binding of pp60vsrc to

30. 31.

32.

33. 34.

35. 36. 37.

38.

39.

40. 41.

42. 43. 44.

45.

plasma membranes: evidence for a myristyl-src receptor. Cell 58:281-286. Resh, M. D., and H.-P. Ling. 1990. Identification of a 32K plasma membrane protein that binds to the myristylated amino terminal sequence of p60vsrc. Nature (London) 346:84-86. Reynolds, A. B., D. J. Roesel, S. B. Kanner, and J. T. Parsons. 1989. Transformation-specific tyrosine phosphorylation of a novel cellular protein in chicken cells expressing oncogenic variants of the avian cellular src gene. Mol. Cell. Biol. 9:629638. Sabe, H., B. Knudsen, M. Okada, S. Nada, H. Nakagawa, and H. Hanafusa. 1992. Molecular cloning and expression of chicken C-terminal Src kinase: lack of stable association with c-Src protein. Proc. Natl. Acad. Sci. USA 89:2190-2194. Schuh, S. M., and J. S. Brugge. 1988. Investigation of factors that influence phosphorylation of pp6ocsrc on tyrosine 527. Mol. Cell. Biol. 8:2465-2471. Schultz, A. M., L. E. Henderson, S. Oroszlan, E. A. Garber, and H. Hanafusa. 1985. Amino terminal myristylation of the protein kinase p60csrc, a retroviral transforming protein. Science 227: 427-429. Shalloway, D., S. Bagrodia, I. Chackalaparampil, P.-H. Lin, and S. J. Taylor. 1992. c-Src and mitosis. Ciba Found. Symp. 170:248-274. Shalloway, D., P. M. Coussens, and P. Yaciuk. 1984. Overexpression of the c-src protein does not induce transformation of NIH 3T3 cells. Proc. Natl. Acad. Sci. USA 81:7071-7075. Shenoy, S., I. Chackalaparampil, S. Bagrodia, P.-H. Lin, and D. Shalloway. 1992. Role of p34cdc2-mediated phosphorylations in two step activation of pp60csrc during mitosis. Proc. Natl. Acad. Sci. USA 89:7237-7241. Shenoy, S., J.-K. Choi, S. Bagrodia, T. D. Copeland, J. L. Mailer, and D. Shalloway. 1989. Purified maturation promoting factor phosphorylates pp60csP` at the sites phosphorylated during fibroblast mitosis. Cell 57:763-774. Southern, P. J., and P. Berg. 1982. Transformation of mammalian cells to antibiotic resistance with a bacterial gene under control of the SV40 early region promoter. J. Mol. Appl. Genet. 1:327-341. Taylor, S. J., and D. Shalloway. c-Src and the cell cycle. Curr. Opinions Genet. Dev., in press. Wilson, L. K., D. K. Luttrell, J. T. Parsons, and S.J. Parsons. 1989. pp6Ocsrc tyrosine kinase, myristylation, and modulatory domains are required for enhanced mitogenic responsiveness to epidermal growth factor seen in cells overexpressing c-src. Mol. Cell. Biol. 9:1536-1544. Wyss, A., S. Kaech, and K. Ballmer-Hofer. 1990. Myristylation of pp6ocsrc is not required for complex formation with polyomavirus middle-T antigen. J. Virol. 64:5163-5166. Yaciuk, P., J. K. Choi, and D. Shalloway. 1989. Mutation in amino acids in pp60csrc that are phosphorylated by protein kinase C and A. Mol. Cell. Biol. 9:2453-2463. Yaciuk, P., and D. Shalloway. 1986. Features of the pp60csrc carboxyl terminus that are required for transformation. Mol. Cell. Biol. 6:2807-2819. Zheng, X. M., Y. Wang, and C. J. Pallen. 1992. Cell transformation and activation of pp60csrc by overexpression of a protein tyrosine phosphatase. Nature (London) 359:336-339.