INTRODUCTION

The transforming protein of Rous sarcoma virus, pp60v-src,and its cellular homolog, pp60c-src, are cotranslationallymyristoylated at the amino-terminal glycine 2 residueduring protein synthesis (Buss et al., 1984; Buss and Sefton,1985; Wilcox et al., 1987; Deichaite et al., 1988). Mostpp60v-src molecules are so modified although it has beenestimated that as much as 16% of pp60v-src may not bemyristoylated (Buss and Sefton, 1985). For both proteins,myristoylation is required for association with the plasmamembrane (Buss et al., 1986; Schuh and Brugge, 1988;Reynolds et al., 1989) but not all myristoylated proteins areassociated with the plasma membrane (for review, Seftonand Buss, 1987; McIlhinney, 1990). Independent domainsof Src may cooperate with myristoylation to specify asso-ciation with distinctive cellular membranes (Kaplan et al.,

1990). In particular, significant fractions of myristoylatedpp60c-src and pp60v-src have been found to associate withperinuclear membranes (Resh and Erikson, 1985) andmyristoylated, yet cytosolic, variants of pp60v-src have beendescribed (Garber et al., 1985). On the other hand, wild-type myristoylated pp60c-src overexpressed in NIH 3T3 cellshas been shown to partition between the plasma membraneand the centrosomal area in interphase cells and colocalizewith endocytosed concanavalin A (ConA) at all stages ofthe ConA-induced endocytotic process (David-Pfeuty andNouvian-Dooghe, 1990). More recently, pp60c-src was alsoshown to be enriched in a population of late endosomes inRat-1 c-Src overexpresser cells (Kaplan et al., 1992) and inPC12 synaptic vesicles (Lindstedt et al., 1992).

Many v-src mutations that prevent myristoylation andplasma membrane association also abrograte v-src trans-forming activity (for review, Jove and Hanafusa, 1987); this

613Journal of Cell Science 105, 613-628 (1993)Printed in Great Britain © The Company of Biologists Limited 1993

Myristoylation of pp60src is required for its membraneattachment and transforming activity. The mouse mon-oclonal antibody, mAb327, which recognizes bothnormal, myristoylated pp60c-src and a nonmyristoylatedmutant, pp60c-src/myr−, has been used to compare theeffects of preventing myristoylation on the localizationof c-Src in NIH 3T3-derived overexpresser cells usingimmunofluorescence microscopy. During interphase,pp60c-src partitions between the plasma membrane andthe centrosome, while pp60c-src/myr− is predominantlycytoplasmic but also partly nuclear. The cytoplasmic,but not the nuclear, staining can be readily washed outby brief pretritonization of the cells before fixation, indi-cating that the cytoplasmic pool of pp60c-src/myr−, in con-trast with the nuclear one, does not associate tightly withstructures that are insoluble in the presence of nonionicdetergents. We have previously shown that during G2phase, pp60c-src leaves the plasma membrane and isredistributed diffusely throughout the cytoplasm and totwo clusters of patches surrounding the two separ-ating centriole pairs. In contrast, we now find thatpp60c-src/myr− translocates to the nucleus in late G2 or

early prophase prior to there being any clear evidence ofnuclear membrane breakdown or nuclear lamina dis-assembly. Similar nuclear translocation of pp60c - s r c / m y r−,but not of pp60c-src, is also observed when cells arearrested in G0 or at the G1/S transition. Furthermore,during mitosis, pp60c-src is found primarily in diffuseand patchy structures dispersed throughout the cyto-plasm while pp60c-src/myr− more specifically associateswith the main components of the spindle apparatus(poles and fibers) and inside the interchromosomalspace.

These results suggest that a possible role for myris-toylation might be to prevent unregulated nuclear trans-port of proteins whose nonmyristoylated counterpartsare readily moved into the nucleus. They also raise thepossibility that a subfraction of wild-type pp60c-src maybehave, at specific times, like its nonmyristoylated coun-terpart, and may translocate to the nucleus and exertspecific functions in that location.

Key words: c-Src, c-Src overexpresser cells, myristoylation, cellcycle

SUMMARY

Differential localization patterns of myristoylated and nonmyristoylated

c-Src proteins in interphase and mitotic c-Src overexpresser cells

Thérèse David-Pfeuty1,*, Shubha Bagrodia2 and David Shalloway2

1Institut Curie-Biologie, Centre Universitaire, Bâtiment 110, 91405 Orsay Cédex, France2Section of Biochemistry, Molecular and Cell Biology, Cornell University, Ithaca, NY 14853, USA

*Author for correspondence

614

supports the hypothesis that important cellular targets ofSrc are located at the plasma membrane. However, non-myristoylated, transformation-defective, cytosolic v-Srcproteins still have mitogenic properties, promoting growthof cells to high densities and growth of some normally non-dividing cells of neuronal origin (Kamps et al., 1986;Calothy et al., 1987). Myristoylation and plasma membraneassociation are not absolutely required for src transformingactivity: two nonmyristoylated Src proteins encoded byrecovered avian sarcoma viruses rASV157 and rASV1702fractionate as soluble cytosolic enzymes yet are transfor-mation-competent (Krueger et al., 1982, 1984). In this case,the transforming activity of these unusual mutants dependson the presence of env signal sequences at the amino ter-mini of the retrovirus-encoded fusion proteins (Garber andHanafusa, 1987). Moreover, a slow, time-dependent cellu-lar transformation of cells infected with temperature-sensi-tive mutants of RSV has been reported to occur when theinfected cells are continuously grown at the nonpermissivetemperature, under conditions in which the ts-pp60v-src isimpaired in its capacity to associate with the plasma mem-brane and, consequently, accumulates in the cytoplasm, par-ticularly in the centrosomal area (David-Pfeuty and Nou-vian-Dooghe, 1992). The activities of cytoplasmic (eithermyristoylated or nonmyristoylated) mutants could reflectnon-specific low-level Src-mediated phosphorylation ofplasma membrane-associated targets. Alternatively, it mayindicate that Src also acts at cellular sites distinct from theplasma membrane.

The observation that the activity of c-Src is transientlystimulated during mitosis (Chackalaparampil and Shal-loway, 1988), partly as an indirect result of phosphoryla-tion by p34cdc2 or the related kinase (Morgan et al., 1989;Shenoy et al., 1989, 1992; Bagrodia et al., 1991; Kaech etal., 1991), suggests that c-Src may have special functionsin this phase of the cell cycle (for review, see Taylor andShalloway, 1993). This observation may be related to thefact that wt pp60c-src in NIH 3T3 overexpresser cells delo-calizes from the plasma membrane in late G2 to condensearound the two separating centriole pairs and then to dis-perse in a diffuse and patchy way in the cytoplasm through-out mitosis (David-Pfeuty and Nouvian-Dooghe, 1990).These considerations prompted us to perform a detailedcomparison, using immunofluorescence techniques, of thesubcellular localization of wild-type (wt) and myristoyla-tion-defective (myr−) c-Src proteins throughout the cellcycle both to determine which aspects of the cell cycle-dependent relocalization of c-Src depend on myristoylationand to get an insight into the behavior of the potential, smallsubfraction of nonmyristoylated Src molecules present innormal cells. This study revealed unexpected localizationproperties of myr− c-Src that differed from those of wt c-src: (1) although primarily cytoplasmic, myr− c-Src exhibitsdetectable nuclear localization during the G1 and S phasesof the cycle; (2) it translocates to the nucleus during lateG2 or early prophase; and (3) it concentrates in the spindleapparatus and interchromosomal space during mitosis.These findings raise the possibility that nonmyristoylatedforms of other normally myristoylated proteins might alsobe targeted to the nucleus.

MATERIALS AND METHODS

Cells and plasmids Plasmid-transfected NIH 3T3 cells that overexpress wild-typechicken c-Src (NIH(pMc-src/focus)B cells) have been previouslydescribed (Johnson et al., 1985). To generate cells expressing themyr− mutant, NIH 3T3 cells were cotransfected with pcLN, achimeric plasmid containing Moloney murine leukemia virus longterminal repeats and the Src coding sequence from plasmidpRSVc-SrcLN (Schuh and Brugge, 1988), and G418-resistanceplasmid pSV2neo (Southern and Berg, 1982) and subjected toG418 selection and cloning as described (Kmiecik and Shalloway,1987). A mass-culture [NIH(pcLN/ pSV2neo/MC)C] derived fromabout 50-100 G418-resistant colonies was used.

Cell culture and synchronization Cells were cultured in 35 mm dishes on 22 mm2 coverslips inDulbecco’s modified Eagle’s medium containing 5% newborn calfserum and antibiotics in 10% CO2, at 37°C, for at least two daysbefore immunofluorescence observation. The average cell densitywas 2×103 to 104 cells cm−2.

Cells were synchronized at the G1/S boundary by a thymidine-aphidicolin double-block; cells were grown in normal medium fortwo days, then sequentially incubated with 2.5 mM thymidine (16h), normal medium (8 h) and 5 µg/ml aphidicolin (16 h). Immuno-fluorescence with anti-tubulin antibody and DAPI (4′,6′-diamidino-2-phenylindole) showed that G2 cells started to appear6 h after release from the double-block and were most prevalent7 h after release. Mitotic cells were most prevalent 7-8 h afterrelease; a second peak of mitotic cells appeared 27-29 h afterrelease.

Immunofluorescence and antibodies Monoclonal antibody mAb327 (Lipsich et al., 1983) was fromClinisciences (Paris, France). Visualization of intracellularcytoskeletal structures was achieved using rat monoclonal anti-tubulin (Biosys, France) for microtubules and nitrobenzoxadiazole(NBD)-phallacidin (Molecular Probes, Inc., Beaverton, OR) formicrofilaments.

Anti-p60 polyclonal rabbit antiserum (Resh and Erikson, 1985),monoclonal antibody GD11 (Parsons et al., 1984), human autoan-tibody to lamin B (Guilly et al., 1987) and anti-clathrin polyclonalrabbit antiserum were generous gifts from Marilyn Resh, SarahParsons, Françoise Danon and Paul-Henri Mangeat, respectively.

Secondary antibodies were goat IgG Fab fragments from anti-bodies directed against either mouse or rat IgG. (The anti-rat goatIgG was preabsorbed against a mouse IgG column to eliminatecross-reactivity; a reciprocal preabsorption was used for the anti-mouse goat IgG.) These fragments were conjugated with eitherfluorescein isothiocyanate (FITC; Interchim, France) or tetram-ethyl rhodamine isothiocyanate (TRITC; Interchim, France).These secondary antibodies gave essentially undetectable back-ground fluorescence.

Single- or double-fluorescence cell labeling was performed asdescribed (David-Pfeuty and Singer, 1980) following fixation with3% formaldehyde and subsequent permeabilization by treatmentwith 0.1% Triton X-100 at room temperature. When specified,cells were pre-Tritonized by treatment with PHEM buffer (0.1%Triton X-100, 45 mM PIPES, 45 mM HEPES, 10 mM EGTA, 5mM MgCl2, 1 mM PMSF, pH 6.9) for 30-45 s before fixation(Bailly et al., 1989). Fluorescence microscopy was performed witha Leitz microscope equipped with fluorescein, rhodamine andDAPI filters using a ×40 oil objective. Photographs were takenwith Kodak TMax 400 film.

T. David-Pfeuty, S. Bagrodia and D. Shalloway

615Localization of wt and myr− c-Src

RESULTS

Differential localization of myristoylated andnonmyristoylated c-Src proteins in asynchronouscellsNIH 3T3-derived cells, which express wild-type and non-myristoylated chicken c-Src from transfected chimeric plas-mids containing Moloney murine leukemia virus long ter-minal repeats, and wt and mutant src genes, were used forall studies. They expressed Src at levels 10- to 20-foldabove the level of endogenous c-Src (data not shown), soalmost all immunofluorescently visualized protein was plas-mid-expressed. The mutant gene used for expression ofmyr− Src contained codons for four additional residues atthe amino terminus of the protein. It encodes the amino-terminal sequence Met-Ala-Ala-Ala-Met-Gly-... (where theunderline denotes the first two amino acids encoded by thewt src gene). This mutant has been studied previously inchicken embryo fibroblasts by Schuh and Brugge (1988),who found, as expected, that it was not localized to theplasma membrane but had normal specific protein-tyrosinekinase activity. This was confirmed in NIH 3T3 cells (datanot shown). Anti-Src monoclonal antibodies mAb327 (Lip-sich et al., 1983) and, occasionally, anti-p60 polyclonalrabbit antiserum (Resh and Erikson, 1985) or mAb GD11(Parsons et al., 1984), which react with a broad range of v-Src and chicken, mouse and human c-Src variants, wereused in these studies.

In asynchronous wt c-Src overexpresser cells, mAb327revealed the three characteristic, previously documented(David-Pfeuty and Nouvian-Dooghe, 1990) distributions:(i) uniformly dispersed at the inner face of the plasma mem-brane (open arrows, Fig. 1C); (ii) inside a dense spot situ-ated at the focal point of the interphasic microtubules in thecentriolar area (thin arrows, Fig. 1C and D); (iii) inside acluster of patches surrounding the nucleus and embeddedwithin a microtubule meshwork (arrowheads, Fig. 1C andD). In G2 cells, the mAb327 staining typically remainedconcentrated in part near the two separating centriole pairs(thin arrows, Fig. 1H and I).

A very different pp60c-src localization pattern wasobserved in asynchronous myr− c-Src overexpresser cells.Here, mAb327 gave a rather intense and uniformly distrib-uted cytoplasmic staining (arrowheads, Fig. 1A) identicalto that described by Reynolds et al. (1989). However, wealso detected a nuclear staining (arrows, Fig. 1A), whichwas low (compared to the cytoplasmic staining) in themajority of the cells, but conspicuous in a minor cell pop-ulation (~2%). Triple labeling of the cells with anti-Src,anti-tubulin antibodies and DAPI (large arrows, Fig. 1E-G)clearly indicates that prominent nuclear staining occurs inlate G2 cells (these can be unambiguously distinguished bythe presence of the two centriole pairs brightly labeled withthe anti-tubulin antibodies (thin arrows, Fig. 1F) and theappearance of the typically punctate chromosomal DAPIstaining (large arrow, Fig. 1G) that signals the beginningof chromosome condensation). In addition, anti-Src stain-ing concentrated in the spindle apparatus of mitotic cells(arrowheads, Fig. 1E-G).

These first observations indicated that pp60c-src/myr− is

located predominantly in the cytoplasm of interphase cellsbut that it translocates to the nucleus, presumably in G2phase.

Localization of pp60c-src/myr in synchronized cellsTranslocation of myr−c-Src during first S and G2following release from G1/S blockmyr− and wt c-Src overexpresser cells were synchronizedat the G1/S boundary by thymidine-aphidicolin double-block (see Materials and Methods) and their distributionswere observed at various times (tR) after release from thedouble-block. Unexpectedly, at tR = 0 (Fig. 2A and B) avery intense nuclear staining (generally partly excludedfrom the nucleolar region; small arrows), largely supplantedthe cytoplasmic staining in the majority of the myr− c-Srccells. The percentage of cells exhibiting this feature variedbetween 70 and 100% from one experiment to the otherand decreased as cells were released from the block at latertimes. This abundant nuclear accumulation of pp60c-src/myr−

did not reappear at the following G1/S phase transition,which occurred about 20 h later. We infer that pp60c-src/myr−

collects intensely in the nucleus during the abnormally longarrest in G1 phase caused by thymidine-aphidicolin double-block but not in normal G1 phases. A similar, abnormal,myr− c-Src nuclear accumulation also occurred during arrestin G0 induced by 48 h serum starvation; no nuclear accu-mulation of wt c-Src was observed after release from adouble-block (Fig. 2E and F) or during a G0 arrest.

The strong nuclear concentration of pp60c-src/myr− did notprevent passage into G2, which started approximately 7 hafter release from the double-block. A natural controlindicating that the observed nuclear accumulation ofpp60c-src/myr− is not an artifact was provided by the obser-vation that anti-Src staining was totally excluded from thenucleus, but not from the cytoplasm, of confluent cells(which do not progress into G2 after release from the block;Fig. 2C and D).

G1 and S phases following the first round of mitosisafter release from a G1/S blockThe first round of mitosis occurred 7 to 9 h after releasefrom the thymidine-aphidicolin double-block. During thefollowing G1 and S phases, that is between 9 and 27 h afterrelease, mAb327 staining in wt and myr− c-Src overex-presser cells did not differ significantly from that exhibitedby unsynchronized cell populations (cf. Figs 1A to 3A, andFigs 1C to 3E).

Brief permeabilization of the cells with Triton X-100 for30-45 s (pre-Tritonization) before fixation (see Materialsand Methods) washed out almost all of the cytoplasmicmAb327 staining but preserved the nuclear one in subcon-fluent cultures (arrows, Fig. 3C). Such treatment did notseverely alter the interphasic microtubule network (Fig.3D), or the organization of actin into microfilament bun-dles (not shown). This experiment showed that the cyto-plasmic pool of pp60c-src/myr− does not associate tightly withstructures that are insoluble in the presence of nonionicdetergents. In contrast, it showed that the nuclear myr− c-Src does interact with nuclear structures that are insolublein Triton X-100.

616

For comparison, the effect of brief pre-Tritonizationbefore fixation on wt c-Src overexpresser cells is shown inFig. 3E-H. When the cells were fixed without pre-Tri-tonization, c-Src partitioned between the centriolar area

(thin arrows, Fig. 3E), the perinuclear region (arrowheads,Fig. 3E) and, uniformly at the inner face of the plasmamembrane (open arrow, Fig. 3E). Strikingly different c-Srcfeatures appeared when the cells were permeabilized with

T. David-Pfeuty, S. Bagrodia and D. Shalloway

Fig. 1. Double immunolabeling of asynchronous wt (C, D, H-J) and myr− (A, B, E-G) c-Src overexpresser cells with anti-Src (A, C, E, H)and anti-tubulin (B, D, F, I) antibodies. Each row of panels shows the same cells visualized by immunofluorescence with different filters.Cells in (G) and (J) display chromosomal DAPI fluorescence following triple staining. Black arrows and arrowheads in A identify nuclearand diffuse cytoplasmic concentrations of myr− c-Src, respectively. Small arrows, arrowheads and open arrows in (C) and (D) designatepericentriolar, perinuclear and plasma membrane-associated pp60c-src, respectively. A selected region containing late G2 cell (largearrows) from each of these cultures is displayed in E, F, G (myr−) and H, I, J (wt). The small black arrows in F, H, I identify the twoseparating centriole pairs in the G2 cells; the arrowheads in (E, F, G) identify a metaphase cell.

617Localization of wt and myr− c-Src

Triton X-100 before fixation (Fig. 3G): a residual pericen-triolar concentration of c-Src occasionally persisted (thinarrows, Fig. 3G), but most often, the perinuclear pp60c-src

patches were no longer present. In addition, the previouslyuniform labeling of c-Src at the inner cell surface wasreplaced by a peculiar pattern of a multitude of roundedpatches or vesicles. The relatively large size of thesepatches and double-labeling experiments with anti-Src andanti-clathrin (the major coat protein of coated pits) anti-bodies indicated that these structures were not coated pits(not shown). They could represent early endosomes. Westress that pre-Tritonizing the wt c-Src overexpresser cellsdid not reveal any nuclear c-Src staining.

Second round of G2 and mitosis after release from aG1/S block A second peak of G2 and mitotic cells appeared 27 to 29h after release from the thymidine-aphidicolin double-block. At tR = 27 h, a significant population of late G2 cells(up to 25%) were present. These were characterized by thepresence of centriole pairs brightly labeled with anti-tubu-lin antibodies (arrows, Fig. 4J), the appearance of a typi-cally punctate DAPI staining pattern (Fig. 4G and K) and

prominent myr− c-Src nuclear staining (Fig. 4E and I). Aswell as this population of cells that were clearly in late G2or already in early prophase, a significant number of cells(between 40 and 50%) were observed in which the cyto-plasmic myr− c-Src was not uniformly distributed but wasdiffusely concentrated around the nucleus (arrow, Fig. 4A)or in which myr− c-Src appeared to accumulate within thedense microtubule meshwork surrounding the nucleus(arrowheads, Fig. 4A and B).

In the late G2 cells, the conspicuous mAb327 nuclearstaining was not seen in the early condensing chromosomesand instead filled the interchromosomal space, which wasunstained by DAPI (small arrows, Fig. 4E, G, I and K).Phase-contrast microscopy (Fig. 4D, H and L) indicated thatthe nuclear accumulation of pp60c-src/myr− occurred beforeclear evidence of nuclear membrane breakdown and beforechromosomes became clearly distinguishable in phase con-trast. As shown by double-immunolabeling of the cells withanti-Src and anti-lamin B antibodies (Fig. 4F), it alsooccurred before nuclear lamina disassembly.

Treatment of these synchronized cells, starting at tR = 25h, with 10 µg/ml of nocadozole for 90 min did not preventthe translocation of pp60c-src/myr− to the nucleus in late G2

Fig. 2. Immunolabeling of wt (E) and myr− (A, C) c-Src overexpresser cells with anti-Src antibodies following release from a thymidine-aphidicolin double-block. Time following double-block release (tR) is 0. Prominent nuclear staining with anti-Src antibodies of a majorityof myr− c-Src overexpresser cells is evident at tR = 0 (A), except in plate areas in which cells reach confluency (C). Under identicalexperimental conditions, wt c-Src never exhibits a detectable nuclear location (E). (B) and (D) show the phase-contrast pictures of cells in(A) and (C), respectively. The tiny arrows in (A) and (B) point out nucleoli from which the anti-Src staining is partly excluded. (F) showschromosomal DAPI staining of cells in (E).

618

or early prophase cells (arrowheads, Fig. 5C and D). Thissuggests that the interphasic microtubules are not involvedin the late G2 phase-dependent translocation of myr− c-Srcfrom the cytoplasm to the nucleus. In addition, nocadozoletreatment had no effect on the cytoplasmic distribution ofmyr− c-Src (cf. Fig. 5A and C). In contrast, disruption ofthe interphasic microtubules by nocadozole treatment of wtc-Src overexpresser cells (Fig. 5E-H) induced a redistribu-tion of pp60c-src into patches that were scattered throughthe cytoplasm (Fig. 5G) as previously reported.

Differential localization of myristoylated andnonmyristoylated c-Src proteins during thevarious phases of mitosisA chronological sequence of the variations in wt and myr−

c-Src localization during progression through mitosis wasreconstructed from observations on selected cells (from cellpopulations 27 to 29 h after double-block release), whosepositions in the cycle were identified by analysis of theirtubulin and DAPI staining patterns. We found that the strik-ingly contrasting localization patterns of overexpressed

T. David-Pfeuty, S. Bagrodia and D. Shalloway

Fig. 3. Double immunolabeling ofwt (E-H) and myr− (A-D) c-Srcoverexpresser cells with anti-Src(A, C, E, G) and anti-tubulin (B,D, F, H), following release from athymidine-aphidicolin double-block (between 9 and 25 h). Thecells in (C, D, G, H) were brieflypretritonized before fixation asdescribed in Materials andMethods. The arrows in (A, C)point out the nuclear concentrationof myr− c-Src and the arrowheadsin (A), the cytoplasmic diffusemyr− c-Src distribution that isabsent in pre-Tritonized cells (C).Thin arrows, arrowheads and openarrow in (E, F) designate thepericentriolar, perinuclear andplasma membrane-associateddistributions, respectively, of wtpp60c-src. The thin black arrows in(G, H) show a residualpericentriolar concentration of c-Src in pre-Tritonized c-Srcoverexpresser cells, whichotherwise exhibit a peculiarpattern of a multitude of roundedpatches at the level of their uppercell surface.

619Localization of wt and myr− c-Src

620

pp60c-src and pp60c-src/myr− still persist during mitosis. Atthe end of the G2 phase, immediately before nuclear mem-brane breakdown, pp60c-src-containing patches typicallycondense around the two centriole pairs, which are now

symmetrically located with respect to the nucleus (arrows,Fig. 6A and B). c-Src appears to be clearly excluded fromthe nucleus at this time. This contrasts with myr− c-Src,which is predominantly intranuclear. However, myr− c-Src

T. David-Pfeuty, S. Bagrodia and D. Shalloway

Fig. 5. Double labeling of asynchronous wt (E-H) and myr− (A-D) c-Src overexpresser cells, untreated (A, B, E, F) or treated withnocadozole (C, D, G, H) with anti-Src (A, C, E, G) and anti-tubulin (B, D, F, H). The big arrowheads point out late G2 or early prophasecells before nuclear membrane breakdown. Nocadozole treatment does not at all perturb the subcellular distribution of myr− c-Src even inmitotic cells (compare A and C) but it greatly affects the wt pp60c-src distribution (compare E and G).

621Localization of wt and myr− c-Src

also starts, slightly but detectably, to accumulate at the levelof the two diplosomes, which are freed from the depoly-merized interphasic microtubules (arrows, Fig. 6D-F).

In prometaphase, during the process of chromosome

rearrangement orchestrated by the complex dynamics of thespindle apparatus, the wt pp60c-src-containing patches,which were previously concentrated around the two centri-ole pairs, partly dissolve from the centriole pairs, which

Fig. 6. Triple labeling of wt (A-C, J-L) and myr− (D-I, M-O) c-Src overexpresser cells with anti-Src (A, D, G, J, M), anti-tubulin (B, E, H,K, N) and DAPI (C, F, I, L, O) during different subphases of mitosis. Rows of panels show late G2 or early prophase (A-F), prometaphase(G-O) cells with appropriate filters. The long arrows point to the location of the two centriole pairs in G2 and spindle poles in prophaseand prometaphase. The small arrows in G and I indicate interchromosomal space that is heavily stained by anti-Src antibody.

622

now serve as spindle poles (arrows, Fig. 6J and K). In con-trast, association of pp60c-src/myr− with the spindle poles isgreatly increased (large arrows, Fig. 6G, H, M and N).pp60c-src/myr− continues to fill the interchromosomal space(thin arrows, Fig. 6G and I).

The features observed in prometaphase became evenmore apparent in metaphase: wt pp60c-src was mostlyexcluded from the space occupied by the components ofthe spindle apparatus and metaphase plate (Fig. 7a-c and g-i) although faint staining of the spindle fibers and spindlepoles was occasionally observed (thin arrows and arrow-heads, respectively, Fig. 7a, b, g and h). In contrast, all thecomponents of the spindle apparatus, spindle poles andspindle fibers (arrowheads and thin arrows, respectively,Fig. 7d, e, j and k), the residual interchromosomal spaceand the contour of the metaphase plate (Fig. 7d and j), wereheavily labeled by pp60c-src/myr−.

In anaphase, wt pp60c-src is diffusely distributed throughthe cytoplasm (Fig. 7m) whereas pp60c-src/myr− remains pref-erentially concentrated at the spindle poles (small arrows,Fig. 7p), along the interpolar microtubules, and in the mid-zone separating the two sets of highly condensed chromo-somes (open arrow, Fig. 7p).

In early telophase (Fig. 8A-F) wt pp60c-src started becom-ing concentrated again in the centrosomal area (arrowheads,Fig. 8A and B) and at the cell surface of the two partingdaughter cells (thin arrow, Fig. 8A); some pp60c-src wasstill distributed diffusely through the cytoplasm (openarrow, Fig. 8A). In contrast, the pp60c-src/myr−-associatedfluorescence reappeared strongly and uniformly throughoutthe cytoplasm (open arrow, Fig. 8D).

In late telophase (Fig. 8G-L), the residual diffuse cyto-plasmic pp60c-src-associated fluorescence vanished and thetwo main areas of pp60c-src distribution in the separatedcells were at the inner cell surface and in the centrosomalarea (long arrows and black arrowhead, respectively, Fig.8G and H). pp60c-src was also concentrated at intercellularcontact areas (white arrowhead, Fig. 8G). At the same stage,the pp60c-src/myr− distribution had not yet returned to itsinterphase distribution (predominently in the cytoplasm).As long as the DAPI staining retained its punctate pattern,myr− c-Src remained concentrated mostly in the nucleus(Fig. 8J-L). It subsequently reverted to the distribution typ-ical of interphase cells (lower cell, Fig. 8J-L).

DISCUSSION

The same anti-Src mAb, mAb327 (Lipsich et al., 1983), hasbeen used to compare the subcellular localization of myris-toylated (wt) and nonmyristoylated (myr−) c-Src in NIH3T3-derived overexpresser cells (both expressing similarlevels of wt and mutant chicken pp60c-src, ~15- to 20-foldhigher than the level of endogeneous mouse pp60c-src)during the various phases of the cell cycle. Similar obser-vations have been made with a second anti-Src mAb, GD11(Parsons et al., 1984), and with a polyclonal anti-Src anti-body, αp60 (Resh and Erikson, 1985). A summary of theresults obtained is presented in Fig. 9. We recall the mainobservations: (1) in G1 and S phases, pp60c-src partitionsprimarily between the inner face of the plasma membrane

and the centrosome (with somewhat higher pericentriolarconcentration) while pp60c-src/myr− is predominantly distrib-uted diffusely in the cytoplasm with a small amount in thenucleus; (2) in G2 phase pp60c-src leaves the plasma mem-brane and is redistributed diffusely throughout the cyto-plasm and into two clusters of patches surrounding the twoseparating centriole pairs; in contrast, pp60c-src/myr− translo-cates predominantly to the nucleus prior to any clear evi-dence of nuclear membrane breakdown or nuclear laminadisassembly; (3) during early mitosis, up to anaphase,pp60c-src is found primarily in diffuse and patchy structuresdispersed throughout the cytoplasm whereas pp60c-src/myr−

strongly associates with specific structures involved inmitosis: that is, with the main constituents of the spindleapparatus and the segregating chromosomes; (4) intelophase pp60c-src gradually resumes its characteristicinterphase, partitioning between the plasma membrane andthe centrosome, but also remains concentrated in part at theintercellular contact area between the two still-attacheddaughter cells; apparently more time is required forpp60c-src/myr− to recover its typical interphase distribution,since myr− c-Src remains concentrated in the nucleus aslong as the two daughter cells are connected.

For technical reasons the studies described here were per-formed using cell lines expressing abnormally high levelsof the Src proteins and the possibility of artifacts must beconsidered. For example, saturating the highest-affinity,physiological sites of interaction could possibly result insecondary association with unsaturated, lower-affinity sub-cellular sites. We believe however that such a risk is min-imal and, in any case, needs to be assumed when one isworking with proteins (like c-Src) that are normally presentat very low levels that are undetectable by ordinary avail-able techniques. Because of the possibility of artifacts, oneshould be especially cautious if the overexpressed proteinapparently colocalizes and/or copurifies with a major cel-lular constituent such as a cytoskeletal protein like tubulinor myosin or a plasma membrane-associated receptor. Butthe finding that overexpressed Src apparently colocalizeswith minor cellular components such as the pericentriolarmaterial or a nuclear antigen is, to us, more likely to reflecta true situation, which, without the artifice of overexpres-sion, would have remained undiscovered. Confidence inthese results obtained with overexpressed Src is supportedto the extent that they overlap previous studies indicatingalso that wt pp60c-src and naturally occurring nonmyristoy-lated forms of Src exhibit major localization sites at theplasma membrane (Loeb et al., 1987) and in the cytoplasm(Krueger et al., 1982, 1984), respectively. Furthermore,variations between the expression levels of Src in individ-ual cells provided an internal check that indicated the lackof any significant effects of dose on subcellular distribu-tion: even in cells in which the level of overexpression waslowest (but still detectable), the anti-Src antibodies showedthat pp60c-src preferentially occupied a centrosomal pooland was secondarily located at the plasma membrane, whilepp60c-src/myr− preferentially occupied a nuclear pool at theend of G2/early prophase and was secondarily located inthe cytoplasm. We infer that this indicates the presence ofhigh-affinity, physiological sites of interaction for pp60c-src

in the centrosomal area and for pp60c-src/myr− in the nucleus.

T. David-Pfeuty, S. Bagrodia and D. Shalloway

623Localization of wt and myr− c-Src

Fig. 7. Triple or double labeling of wt (a-c, g-i, m-o) and myr− (d-f, j-l, p-q) c-Src overexpresser cells with anti-Src (a,d,g,j,m,p), anti-tubulin (b,e,h,k,n,q) and DAPI (c,f,i,l,o) in metaphase and anaphase. Rows are displays with different filters of metaphase (a-l) andanaphase (m-r) cells. r, phase-contrast image of the anaphase cell in (p, q). The positions of the spindle poles are indicated by arrowheadsin metaphase cells (a,b,d,e,g,h,j and k) and by thin arrows in anaphase cells (n, p and q). The thin arrows (e, h and k) identify spindlefibers. These fibers were only weakly stained by anti-Src antibody in wt c-Src overexpresser cells (thin arrow in g) but strongly stained inmyr− c-Src overexpresser cells (thin arrows in d and j). The open arrows in (n,p and q) show interpolar microtubules that are associatedwith myr− c-Src (p) but not with wt c-Src (m) during anaphase.

624

Also, the significant differences observed between the cellcycle-dependent subcellular distributions of the wild-typeand nonmyristoylated forms suggest that the patternsobserved reflect specific associations and not merely non-specific low-affinity sites of interaction.

Many cases are known where protein mutation leads tomislocalization, particularly when the mutation affects aregion involved in the specification of localization (as for

the myristoylation site in pp60 src) or when it induces a con-formational change that exposes cryptic localizationsequences (as for v-Sis, lacking its signal sequence; Lee etal., 1987). These changes in protein localization very fre-quently correlate with changes in the biological activity ofthe proteins, emphasizing the influence of the subcellularenvironment on the physiological function of an enzyme.These remarks are especially relevant to the Src field and

T. David-Pfeuty, S. Bagrodia and D. Shalloway

Fig. 8. Triple labeling of wt (A-C, G-I) and myr− (D-F, J-L) c-Src overexpresser cells with anti-Src (A, D, G, J), anti-tubulin (B, E, H, K)and DAPI (C, F, I, L) in early (A-F) and late (G-L) telophase. The black arrowheads in (A, B, G and H) point out newly clusteringcentrosomal patches of pp60c-src; thin arrows in (A) and (G), identify plasma membrane-associated wt c-Src; open arrows identify diffusecytoplasmic c-Src (A) or myr− c-Src (D and J); larger arrows in (G) and (J) show nuclei either devoid of c-Src (G) or staining for myr− c-Src (J); the white arrowhead in (G) marks an intercellular contact area where pp60c-src has accumulated.

625Localization of wt and myr− c-Src

further justify our present investigation, since it has previ-ously been shown that myristoylation is required for plasmamembrane localization of pp60src (Buss et al., 1986) andthat it augments the transforming potential of the viral andmutated cellular enzymes (Cross et al., 1984; Pellman etal., 1985; Schuh and Brugge, 1988) even though it does notaffect their intrinsic protein-tyrosine kinase-specific activi-ties (Schultz et al., 1985; Kamps et al., 1986; Reynolds etal., 1989). However, it has often been tacitly assumed thatabrogation of myristoylation simply results in loss of a

specific (plasma membrane) localization, not that it maylead to a highly specific, alternative pattern of localization.The detailed comparative immunofluorescence study pre-sented here clearly illustrates the complexity of the factorsinfluencing the localization of pp60c-src. In addition toinducing a gain of function (plasma membrane localiza-tion), myristoylation induces three losses of function. It: (1)prevents pp60c-src translocation to the nucleus duringgrowth arrest; (2) prevents nuclear translocation during G2in a normal cell cycle; and (3) strongly reduces the affin-

Fig. 9. Summary of cell cycle-dependent localization of wt pp60c-src (s, left panel) and pp60c-src/myr− (-, right panel). First row: theinterphase distribution of the two proteins is plasma membrane-associated and centrosomal for wt pp60c-src, and cytoplasmic diffuse andweakly nuclear for pp60c-src/myr−. Second row: in late G2, before nuclear membrane breakdown, wt pp60c-src condenses around the twocentriole pairs and pp60c-src/myr− concentrates in the nucleus. Third and fourth rows: in metaphase and anaphase, wt pp60c-src exhibits adiffuse or patchy distribution through the cytoplasm; pp60c-src/myr− condenses on mitosis-specific structures. Fifth row: in telophase, wtpp60c-src resumes very fast its interphasic distribution and concentrates in intercellular contact areas; pp60c-src/myr− remains concentrated inthe nucleus in early G1 before recovering its typical interphasic (mainly cytoplasmic) distribution.

626

ity of pp60c-src for components of the spindle apparatus andfor the interchromosomal space during mitosis. (Among adozen overexpressed c-src mutants studied, the myristoy-lation-defective one is the only one that induced a nucleartranslocation of the protein and an association with themitotic apparatus.) This diversity of effect is consistent withthe view, arising from studies of a variety of myristoylatedviral and cellular proteins (McIlhinney, 1990, for review),that myristoylation may mediate protein-protein interac-tions rather than protein-membrane interactions. Even theeffect of myristoylation on plasma membrane associationmay be mediated via effects on protein-protein interactions,since pp60src must interact with saturable high-affinity bind-ing sites for membrane binding (Resh, 1989; Goddard etal., 1989; Resh and Ling, 1990).

It is noteworthy that wt pp60c-src has a strong affinity forspecific structures that are well developed during interphasebefore nuclear envelope breakdown (i.e. the plasma mem-brane, the centrosome and (during G2) the two pericentri-olar areas). Beginning at mitosis, when the interphasicmicrotubules start depolymerizing, pp60c-src is redistributedinto diffuse and patchy structures throughout the cytoplasm,an effect that can be mimicked by artificial disruption ofthe interphasic microtubules through nocadozole treatment.These observations: first, indicate that both the maintenanceof pp60c-src association with the plasma membrane and itsaccumulation in the centrosomal area depend on theintegrity of the interphasic microtubule network; and,second, they also suggest that the redistribution of wtpp60c-src starting at the end of G2 could simply result fromthe natural change in the state of organization and the sub-sequent depolymerization of the interphasic microtubules.During mitosis, wt pp60c-src does not display particularlyhigh affinity for any specific structures participating inmitotic events. These observations are in direct contrastwith those for pp60c-src/myr−; its interphase distribution doesnot depend at all on the presence of a well-developed micro-tubule network but, starting at late G2 phase and duringmitosis it exhibits a strong affinity for many componentsthat participate in mitotic events (spindle poles and fibers,interchromosomal material).

p p 6 0s r c has been linked to control of a wide variety ofevents occurring throughout the cell, from the plasmamembrane to the nucleus (for review, Cooper, 1990).While its varied activities might be mediated by multiplestatic subpopulations that participate in (tyrosine) phos-phorylation cascades, it is interesting to consider the pos-sibility that c-Src participates in intracellular signalling asa physically transported messenger between the plasmamembrane and the nucleus. We have previously shown thatwt pp60 c - s r c apparently associates with vesicles containingConA-receptor complexes (endosomes) throughout theentire ConA-receptor-mediated endocytotic process(David-Pfeuty and Nouvian-Dooghe, 1990). This led us tosuggest that the accumulation of pericentriolar and perin-uclear pp60c - s r c-containing patches during interphase mightrepresent plasma membrane-derived vesicles or endosomesin transit between the plasma membrane and the centro-some. This hypothesis is supported, on one hand, by thefinding reported here that a subpopulation of plasma mem-brane-associated wt pp60c - s r c appears to be a component

of vesicular structures that are insoluble in the presence ofnonionic detergents (possibly early endosomes) and, on theother hand, by a biochemical fractionation study (Kaplanet al., 1992) showing that a subpopulation of wt pp60c - s r c

cofractionates with late endosomes. The report (Linstedt etal., 1992) that pp60c - s r c also associates specifically withsynaptic vesicles in the neuroendocrine PC12 cell line isalso consistent with such an hypothesis. We find that thelack of myristoylation of pp60c - s r c / m y r− blocks associationnot only with the plasma membrane, but also with ConA-induced endocytotic vesicles and endosomes (unpublishedresults). This is consistent with the observation thatp p 6 0c - s r c / m y r− does not accumulate like wt pp60c - s r c at thecentrosome in interphase cells. It also implies thatp p 6 0c - s r c / m y r− must be transported to the nucleus by an inde-pendent mechanism.

A central question raised by this work is whether thebehavior of overexpressed pp60c-src/myr− could possiblyreflect that of a naturally occurring subset of pp60c-src innormal cells. Buss and Sefton (1985) have estimated thatas much as 16% of pp60v-src may actually not be myris-toylated. If an equivalent percentage of pp60c-src is non-myristoylated in normal cells, it would be undetectable byconventional techniques. Overexpression of wt pp60c-src bya factor of 10 to 20 would even be insufficient to raise sucha nonmyristoylated subpopulation above the threshold levelof detectability by immunofluorescence (which is a few foldhigher than the endogeneous level of pp60c-src); this mayaccount for the fact that we did not detect such a subpop-ulation in wt c-Src overexpresser cells. We do not know ifthe molecules in this putative non-myristoylated populationpossess an N-terminal methionine or if, as for wild-typepp60a-40Vc-Srca+40V, it is removed by aminopeptidaseactivity. While there is no reason to believe that the pres-ence or absence of this residue (or the additional alaninespresent in the pcLN mutant) plays any significant role orthat the localization site(s) that govern nuclear transloca-tion are at the amino end of the molecule, the possibilitythat small amino acid changes in this region could haveeffects on localization cannot be excluded. Interestingly,abundant nuclear concentration of myr− c-Src in lateG2/early prophase cells and within the mitotic apparatusdoes not interfere with mitotic progression; nor is progres-sion through the cell cycle prevented by the massive nuclearaccumulation of myr− c-Src that occurs in cells that havebeen blocked at the G1/S transition by thymidine-aphidi-colin treatment. These observations imply that a naturallyoccurring subset of pp60c-src/myr− would probably not exertan inhibitory effect on cell cycle and mitotic progression.It is also possible that the behavior of pp60c-src/myr− couldmimic to some extent that of a subpopulation of wtpp60c-src. Indeed allosteric modifications of pp60c-src, per-haps resulting from phosphorylations or dephosphorylationswithin the amino-proximal region of the molecule, mightinterfere with the protein-protein interaction required forplasma membrane localization and could generate cytoso-lic enzymes capable of being transported to the nucleus.This is consistent with the report that myristate is presentin soluble cytoplasmic as well as membrane-bound, pp60src

(Buss et al., 1984). Such hypothetical motions would prob-ably be regulated in a more refined manner that would pre-

T. David-Pfeuty, S. Bagrodia and D. Shalloway

627Localization of wt and myr− c-Src

vent the excessive nuclear accumulation observed here withthe constitutive mutant myr− c-Src. Within this model, nopredictions can be made regarding the issue of whether anaturally occurring subset of c-Src proteins (whose behav-iour would only be partially mimicked by mutantpp60c-src/myr−) would exert an inhibitory or positive effecton cell cycle and mitotic progression.

It is worthwhile to mention here two recent observationsthat are relevant to our observations. First, Zhao et al.(1992) have reported that in vitro calcium-induced ker-atinocyte differentiation occurs concomitant with a markedincrease in nuclear phosphotyrosine content and along witha translocation of c-Src to the nucleus. Second, C. Willman(personal communication) has shown that a naturally occur-ring variant of c-Fgr lacking the amino-terminal domainarises by alternative internal translation initiation inmyeloid cells; this variant has a nuclear location in myeloidcells and in stably transfected fibroblasts. These data sup-port the hypothesis that naturally occurring forms ofpp60c-src and of other members of the Src family couldindeed move into the nucleus and exert a physiologicalfunction in that location. Our results also raise the intrigu-ing suggestion that myristoylation might be required to pre-vent wt pp60c-src from interacting with nuclear constituentsduring G2 phase and with the mitotic apparatus during mito-sis. Interestingly, another well-described myristoylatedenzyme is the catalytic subunit of cAMP-dependent proteinkinase II (cAMP-dPKII), which translocates to the nucleussolely following cAMP-induced dissociation from theGolgi-associated cAMP-dPKII regulatory subunit (Nigg etal., 1985). Since many protein kinases (including Src)exhibit little specificity in in vitro assays, catalyzing thephosphorylation of a wide variety of substrates that appar-ently are never phosphorylated under physiological con-ditions, it is easy to conceive that the spatial and temporaldistributions of such enzymes need to be subjected to verystrict control in vivo in order to avoid unscheduled proteinphosphorylation that could be catastrophic for cell survival.

We thank Yolande Nouvian-Dooghe for her excellent techni-cal assistance, Mesdames Irène Gaspard and Colette Pouget fortheir expert artwork, Eric Bailly for his helpful advices, Sarah Par-sons, Marilyn Resh, Françoise Danon and Paul-Henri Mangeat forproviding antisera and Joan Brugge for plasmid pRSVc-srcLN.We also thank Madame Françoise Arnouilh for typing the manu-script. This study was supported by the Centre National pour laRecherche Scientifique, the Curie Institute and the Associationpour la Recherche sur le Cancer (France) and by grants CA32317and CA47333 and RCDA CA01139 from the National Institutesof Health (USA).

REFERENCES

Bagrodia, S., Chackalaparampil, I., Kmiecik, T. E. and Shalloway, D.(1991). Altered tyrosine 527 phosphorylation and mitotic activation ofp60src. Nature 349, 172-175.

Bailly, E., Dorée, M., Nurse, P. and Bornens, M. (1989). p34cdc2 is locatedin both nucleus and cytoplasm; part is centrosomally associated at G2/Mand enters vesicles at anaphase. EMBO J. 8, 3985-3995.

Buss, J. E., Kamps, M. P., Gould, K. and Sefton B. M. (1986). Theabsence of myristic acid decreases membrane binding of pp60src but doesnot affect tyrosine protein kinase activity. J. Virol. 58, 468-474.

Buss, J. E., Kamps, M. P. and Sefton, B. M. (1984). Myristic acid is

attached to the transforming protein of Rous sarcoma virus during orimmediately after synthesis and is present in both soluble and membrane-bound forms of the protein. Mol. Cell. Biol. 4, 2697-2704.

Buss, J. E. and Sefton, B. M. (1985). The rare fatty acid, myristic acid, isthe lipid attached to the transforming protein of Rous sarcoma virus andits cellular homologue. J. Virol. 53, 7-12.

Calothy, G., Laugier, D., Cross, F. R., Jove, R., Hanafusa, T. andHanafusa, H. (1987). The membrane-binding domain and myristylationof pp60v-src is not essential for stimulation of cell proliferation. J. Virol.61, 1678-1681.

Chackalaparampil, I. and Shalloway, D. (1988). Altered phosphorylationand activation of pp60c-src during fibroblast mitosis. Cell 52, 801-810.

Cooper, J. A. (1990). The Src family of protein tyrosine kinases. InPeptides and Proteins Phosphorylations (ed. B. Kamp and P. F.Alewood) pp. 85-113. Boca Raton: CRC Press, Inc..

Cross, F. R., Garber, E. A., Pellman, D. and Hanafusa, H. (1984). A shortsequence in the pp60 src N terminus is required for pp60src myristylationand membrane association and for cell transformation. Mol. Cell. Biol. 4,1834-1842.

David-Pfeuty, T. and Nouvian-Dooghe,Y. (1990). Immunolocalization ofthe cellular src protein in interphase and mitotic NIH c-src overexpressercells. J. Cell Biol. 111, 3097-3116.

David-Pfeuty, T. and Nouvian-Dooghe, Y. (1992). Slow time dependentcellular transformation induced at restrictive temperature by ts-srcmutants. Oncogene 7, 1611-1623.

David-Pfeuty, T. and Singer, J. S. (1980). Altered distributions of thecytoskeletal proteins vinculin and a-actinin in cultured fibroblaststransformed by Rous sarcoma virus. Proc. Nat. Acad. Sci. USA77, 6687-6691.

Deichaite, I., Casson, L. P., Ling, H. P. and Resh, M. D. (1988). In vitrosynthesis of pp60v-src: myristylation in a cell-free system. Mol. Cell. Biol.8, 4295-4301.

Garber, E. A., Cross, F. R. and Hanafusa, H. (1985). Processing ofpp60v-src to its myristylated membrane-bound form. Mol. Cell. Biol. 5,2781-2788.

Garber, E. A. and Hanafusa, H. (1987). NH2-terminal sequences of twosrc proteins that cause aberrant transformation. Proc. Nat. Acad. Sci. USA84, 80-84.

Goddard, C., Arnold, S. T. and Felsted, R. L. (1989). High affinitybinding of an N-terminal myristoylated p60src peptide. J. Biol. Chem. 264,15173-15176.

Guilly, M. N., Danon, F., Brouet, J. C., Bornens, M. and Courvalin, J. C.(1987). Autoantibodies to nuclear lamin B in a patient with thrombopenia.Eur. J. Cell Biol. 43, 266-272.

Jove, R. and Hanafusa, H. (1987). Cell transformation by the viral srconcogene. Annu. Rev. Cell. Biol. 3, 31-56.

Johnson, P. J., Coussens, P. M., Danko, A. V. and Shalloway,D. (1985).Overexpressed pp60c-src can induce focus formation without completetransformation of NIH3T3 cells. Mol. Cell. Biol. 5, 1073-1083.

Kaech, S., Covic, L., Wyss, A. and Ballmer-Hofer,K. (1991). Associationof pp60c-src with polyoma virus middle-T abrogates mitosis-specificactivation. Nature 350, 431-433.

Kamps, M. P., Buss, J. E. and Sefton, B. M. (1986). Rous sarcoma virustransforming protein lacking myristic acid phosphorylates knownpolypeptides substrates without inducing transformation. Cell 45, 105-112.

Kaplan, K. B., Swedlow, J. R., Varmus, H. E. and Morgan,D. O. (1992).Association of pp60c-src with endosomal membranes in mammalianfibroblasts. J. Cell Biol. 118, 321-333.

Kaplan, J. M., Varmus, H. E. and Bishop, J. M. (1990). The src proteincontains multiple domains for specific attachment to membranes. Mol.Cell. Biol. 10, 1000-1009.

Kmiecik, T. E. and Shalloway, D. (1987). Activation and suppression ofpp60c-src transforming ability by mutation of its primary sites of tyrosinephosphorylation. Cell 49, 65-73.

Krueger, J. G., Garber, E. A., Chin, S. S. M., Hanafusa, H. andGoldberg, A. R. (1984). Six-variant pp60src proteins of recovered aviansarcoma viruses interact with adhesion plaques as peripheral membraneproteins: effects on cell transformation. Mol. Cell. Biol. 4, 454-467.

Krueger, J. G., Garber, E. A., Goldberg, A. R. and Hanafusa,H. (1982).Changes in amino-terminal sequences of pp60src lead to decreasedmembrane association and decreased in vivo tumorigenicity. Cell 28,889-896.

Lee, B. A., Maher, D. W., Hannink, M. and Donoghue, D. J. (1987).

628

Identification of a signal for nuclear targeting in plateled-derived-growth-factor-related molecules. Mol. Cell. Biol. 7, 3527-3537.

Linstedt, A. D., Vetter, M. L., Bishop, J. M. and Kelly, R. B. (1992).Specific association of the proto-oncogene product pp60c-src with anintracellular organelle, the PC12 synaptic vesicle. J. Cell Biol. 177, 1077-1084.

Lipsich, L. A.,Lewis, A. J. and Brugge, J. S. (1983). Isolation ofmonoclonal antibodies that recognize the transforming protein of aviansarcoma viruses. J. Virol. 48, 352-360.

Loeb, D. M., Woolford, J. and Beemon, K. (1987). pp60c-src has lessaffinity for the detergent insoluble cellular matrix than do pp60v-src andother viral protein-tyrosine kinases. J. Virol. 61, 2420-2427.

McIlhinney, R. A. J. (1990). The fats of life: the importance and function ofprotein acylation. Trends Biochem. Sci. 15, 387-391.

Morgan, D. O., Kaplan, J. M., Bishop, J. M. and Varmus,H. E. (1989).Mitosis-specific phosphorylation of pp60c-src by p34cdc2-associatedprotein kinase. Cell 57, 775-786.

Nigg, E. A., Hilz, H., Eppenberger, H. M. and Dutly,F. (1985). Rapid andreversible translocation of the catalytic subunit of cAMP-dependentprotein kinase II from the Golgi complex to the nucleus. EMBO J. 4,2801-2806.

Parsons, S. J., McCarley, D. J., Ely, C. M., Benjamin, D. C. and Parsons,J. T. (1984). Monoclonal antibodies to Rous sarcoma virus pp60src reactwith enzymatically active pp60src of avian and mammalian origin. J.Virol. 51, 272-282.

Pellman, D., Garber, E. A., Cross, F. R. and Hanafusa, H. (1985). Finestructural mapping of a critical NH2-terminal region of p60src. Proc. Nat.Acad. Sci. USA 82, 1623-1627.

Resh, M. D. (1989). Specific and saturable binding of pp60v-src to plasmamembranes: evidence for a myristyl-src receptor. Cell 58, 281-286.

Resh, M. D. and Erikson, R. L. (1985). Highly specific antibody to Roussarcoma virus src gene product recognize a novel population of pp60v-src

and pp60c-src molecules. J. Cell Biol. 100, 409-417. Resh, M. D. and Ling, H. (1990). Identification of a 32 K plasma membrane

protein that binds to the myristylated amino terminal sequence ofpp60v-src. Nature 346, 84-86.

Reynolds, A. B., Roesel, D. J., Kanner, S. B. and Parsons, J. T. (1989).Transformation-specific tyrosine phosphorylation of a novel cellularprotein in chicken cells expressing oncogenic variants of the aviancellular src gene. Mol. Cell. Biol. 9, 629-638.

Schuh, S. M. and Brugge, J. S. (1988). Investigations of factors thatinfluence phosphorylation of pp60src on tyrosine 527. Mol. Cell. Biol. 8,2465-2471.

Schultz, A. R., Henderson, L. E., Oroszlan, S., Garber, E. A. andHanafusa, H. (1985). Amino terminal myristylation of the protein kinasep60src, a retroviral transforming protein. Science 227, 427-429.

Sefton, B. M. and Buss, J. E. (1987). The covalent modification ofeukaryotic proteins with lipid. J. Cell Biol. 104, 1449-1453.

Shenoy, S. J., Choi, J. K., Baprodia, S., Copeland, T. D., Maller, J. L.and Shalloway, D. (1989). Purified maturation promoting factorphosphorylates pp60c-src at the sites phosphorylated during fibroblastmitosis. Cell 57, 763-774.

Shenoy, S., Chackalaparampil, I., Bagrodia, S., Lin, P. J. andShalloway, D. (1992). Role of p34cdc2-mediated phosphorylations intwo-step activation of pp60c-src during mitosis. Proc. Nat. Acad. Sci. USA89, 7237-7241.

Southern, P. J. and Berg, P. (1982). Transformation of mammalian cells toantibiotic resistance with a bacterial gene under control of the SV40 earlyregion promoter. J. Mol. Appl. Genet. 1, 327-341.

Taylor, S. J. and Shalloway, D. (1993). The cell cycle and c-Src. Curr.Opin. Genet. Dev. 3, 26-33.

Wilcox, C., Hu, J. S. and Olson, E. N. (1987). Acylation of proteins withmyristic acid occurs cotranslationally. Science 238, 275-278.

Zhao, Y., Sudol, M., Hanafusa, H. and Krueger, J. (1992). Increasedtyrosine kinase activity of c-Src during calcium-induced keratinocytedifferentiation. Proc. Nat. Acad. Sci. USA 89, 8298-8302.

(Received 18 December 1992 - Accepted 14 March 1993)

T. David-Pfeuty, S. Bagrodia and D. Shalloway