Rous sarcoma virus-transformed fibroblasts adhere primarily at discrete protrusions of the ventral membrane called podosomes

Experimental Cell Research 159 (1985) 141-157

Rous Sarcoma Virus-Transformed Fibroblasts Adhere Primarily at Discrete Protrusions of the Ventral Membrane Called Podosomes GUIDO TARONE,* D A N I E L A CIRILLO, FILIPPO G. GIANCOTTI, PAOLO M. COMOGLIO and PIER CARLO MARCHISIO Institute of Histology and General Embryology, University of Torino School of Medicine, Torino, Italy

Rous sarcoma virus-transformed BHK cells (RSV/B4-BHK) adhere to a fibronectincoated substratum primarily at specific dot-shaped sites. Such sites contain actin and vinculin and represent close contacts with the substratum as revealed by interference reflection microscopy. Only a few adhesion plaques and actin filament bundles can be detected in these cells as compared to untransformed parental fibroblasts. In thin sections examined with transmission electron microscopy (TEM) these adhesion sites correspond to short protrusions of the ventral cell surface that contact the substratum at their apical portion. These structures, which may represent cellular feet, are therefore called podosomes. By screening a number of different transformed fibroblasts plated on a fibronectincoated substratum we find that podosomes are common to mammalian and avian cell lines transformed either by Rous sarcoma virus (RSV) or by Fujinami avian sarcoma virus (FSV), whose oncogenes encode specific tyrosine kinases. Using antibodies reacting with phosphotyrosine in immunofluorescence experiments, we show that phosphotyrosinecontaining molecules are concentrated in podosomes. Podosomes are not detected in fibroblasts transformed by other retroviruses (Snyder-Theilen sarcoma virus, Abelson leukemia virus and Kirsten sarcoma virus) or by DNA tumor viruses (polyoma, SV40), indicating that podosome-mediated adhesion in transformed fibroblasts is related to the peculiar properties of some oncoproteins and possibly to their tropism for adhesion systems. Podosomes and adhesion plaques, although similar in cytoskeletal protein composition, have different mechanisms and kinetics of formation. Assembly of podosomes, in fact (i) does not require fetal calf serum (FCS) in the adhesion medium, that is necessary for the organization of adhesion plaques; (ii) does not require protein synthesis; and (iii) is insensitive to the ionophore monensin, that prevents adhesion plaque formation. Moreover, during attachment to fibronectin-coated dishes, podosomes appear in the initial phase (60 rain) of attachment, while adhesion plaques require a minimum of 180 min. In conclusion podosomes of RSV- and FSV-transformed fibroblasts represent a phenotypic variant of adhesion structures. © 1985AcademicPress, Inc.

Alteration of cell-substratum and cell-cell adhesion upon neoplastic transformation is important for the onset and maintenance of malignancy. The adhesion of fibroblasts to a growth substratum is a complex process that requires the coordinate interaction of the plasma membrane with both the adhesive molecules * To whom offprint requests should be sent. Address: Istituto di Istologia ed Embriologia Generale Cso. Massimo D'Azeglio 52, 10126 Torino, Italy. Copyright ~ 1985 by Academic Press, Inc. All fights of reproduction in any form reserved 0014-4827/85 $03.00

142

Tarone et al.

of the extracellular matrix and with cytoskeletal structures [1, 2]. The interaction of cells with extracellular matrix molecules, such as fibronectin, triggers cytoskeleton reorganization and induces cell spreading [3]. Ultimately this process results in the organization of specialized regions of the ventral plasma membrane, namely 'close contacts' and 'adhesion plaques' [4], which represent discrete substratum adhesion sites. Fibroblast transformation by Rous sarcoma virus has a pleiotropic effect on cell adhesion and induces alteration in the organization of extracellular matrix, plasma membrane and cytoskeleton. In fact, the synthesis of fibronectin and collagen is greatly diminished in RSV-transformed fibroblasts and these molecules are not assembled in an organized extracellular matrix [5-8]. Moreover, actin is no longer organized in prominent stress fibres [9, 10], and adhesion plaques are reduced in number and size [11, 12]. In addition to these alterations peculiar 'dot-shaped' adhesion sites are known to appear in RSV-transformed cells [11, 12]. These sites contain cytoskeletal proteins such as actin, vinculin and alpha-actinin, and are also sites where the pp60 Src oncogene-coded protein is concentrated [11, 13, 14]. These adhesion structures have been regarded as modified adhesion plaques, but their nature is still uncertain and it is not known whether these peculiar adhesion structures occur in all transformed cells. In this paper, we show that dot-shaped adhesion sites, containing actin and vinculin, are found in fibroblasts transformed by Rous or Fujinami sarcoma viruses, whose oncogenes encode tyrosine-kinases known to interact with adhesion structures [11-15]. Using antibodies reacting with phosphotyrosine, we also provide evidence for an overall increase of tyrosine-phosphorylated molecules within these structures. Moreover, analysis of kinetics and mechanisms of formation indicates that the dot-shaped adhesion structures represent an adhesive device different from adhesion plaques.

MATERIALS AND METHODS Cells Baby Hamster Kidney (BHK) fibroblasts transformed by the Bryan high titre strain of Rous sarcoma virus (RSV/B4-BHK) were a gift of Professor L. Warren, Philadelphia, Pa. Untransformed BHK cells were obtained from the stock of Dr I. Macpherson, London. The SR-BALB mouse cell line, transformed by Schmidt-Ruppin D strain of RSV, was kindly provided by Dr J. T. Parsons, Charlottesville, Va. The B77-3T3 mouse cell line transformed by the Bratislava 77 strain of RSV was obtained from Dr J. M. Bishop, San Francisco, Calif. The clonal cell line B77-AA6-3T3 was isolated and characterized in our laboratory [16, 17]. Fisher rat embryo cells transformed by the SR-A strain of RSV (SRA-RAT-1) or by the Fujinami sarcoma virus (FvR) were kindly provided by Dr H. Hanafusa, New York. 16Q- is a continuous quail cell line transformed by the Bryan high titre strain of RSV [18]. ST-FRE are Fisher rat embryo cells transformed by the Snyder-Theilen strain of Feline sarcoma virus (ST-FeSV) and were provided by Dr F. Veronese, Frederick, Md. BALB/c fibroblasts transformed by the Kirsten sarcoma virus (K/BALB) and hamster fibroblasts transformed by Simian virus 40 (PM-3) were provided by Professor C. Monti-Bragadin, Trieste. ANN-1 are NIH mouse fibroblasts transformed by Abelson virus. PY-BHK are BHK fibroblasts transformed by polyoma virus. All cell lines were propagated at 37°C, 5 % CO2 and 100% humidity in Dulbecco's modified Eagle medium Exp Cell Res 159 (1985)

Adhesion mechanisms of RSV-transformed cells

143

(DMEM) containing antibiotics and fungizone and supplemented with 10 % fetal bovine serum. Cells were routinely tested for mycoplasma and found to be negative.

Cell Adhesion Conditions Glass coverslips were extensively acid-washed, and coated with purified plasma fibronectin (I0 ~tg/ml) in 150 mM sodium chloride, 10 mM sodium phosphate buffer pH 7.4 (PBS) for 60 min at room temperature. Residual protein-binding sites on the glass surface were blocked by further incubation with 0.2% bovine serum albumin (BSA) in PBS. Fibronectin was purified from human plasma by affinity chromatography on gelatin-Sepharose as described previously [19]. Cells, harvested from culture dishes by EGTA treatment, were plated in serum-free DMEM. To inhibit protein synthesis cells were incubated with 20 ~tM cycloheximide for 2 h before harvesting and the drug was kept in the medium during adhesion. Monensin (1 ~tM) was directly added to cells plated on coated dishes. Fetal calf serum (FCS) was added to the adhesion medium, when specified.

Antibodies Anti-azobenzyl phosphonate (ABP) antibodies were raised, affinity-purified and characterized, as previously reported [20]. Briefly, ABP antisera were raised in rabbits by three monthly spaced injections of 1 mg of keyhole limpet hemocyanin to which - 3 0 p-azobenzyl phosphonate groups were covalently linked per 100 000 molecular weight units. Antibodies were purified by affinity chromatography on ABP-Bovine Serum Albumin coupled to cyanogen bromide-activated Sepharose 4B (4 mg of protein/g of Sepharose) in 0.1 M sodium bicarbonate. Immunoglobulins from ABP sera were precipitated with 40 % saturated ammonium sulfate and applied to the column at room temperature. After washing the column with PBS, bound antibodies were eluted with 3 M potassium isothiocyanate pH 7 at room temperature, and dialysed against PBS. The antibodies were found to cross-react specifically with phosphotyrosine (P-Tyr) by radioimmuno assay and to precipitate selectively P-Tyr proteins [20]. For immunofluorescence experiments ABP antibodies were used at 125 ~tg/ml [21]. Anti-vinculin serum, raised in rabbits immunized with homogeneous chicken gizzard vinculin, was obtained through the courtesy of Dr B. Geiger, Rehovot, Israel. A monoclonal mouse antibody to vimentin, that reacts with vimentin of most mammalian and non-mammalian cells, was provided by Dr M. Prat, Turin, Italy. Rabbit anti-tubulin antibodies were raised against homogeneous calf brain tubulin and affinity-purified according to [22]. Affinity-purified fluorescein- or rhodamine-isothyocyanate-tagged goat anti-rabbit IgGs were purchased from Kirkegaard & Perry Inc., Gaithersburg, and used at 100 ~tg/ml.

Fluorescence and Interference Reflection Microscopy Coverslips were processed for fluorescence microscopy essentially as described in previous papers [20, 21]. The procedure involved fixation at room temperature in 3.7% formaldehyde in phosphatebuffered saline (PBS) for 10 min. In order to make cell membranes permeable to large molecules, coverslips were treated according to different procedures. For preserving simultaneously microfilaments and adhesion structures, mild permeabilization was achieved at room temperature by soaking coverslips in PBS plus 0.1% Triton X-100 for 1 min. For preserving microtubules, coverslips were immersed in prechilled methanol and acetone at -10°C for 5 min and 5 sec respectively followed by air-drying [22]. Primary antibodies, as well as fluorescein or rhodamine isothyocyanate-labelled phalloidin (F-PHD or R-PHD; a kind gift of Professor Th. Wieland, Heidelberg), were spread over the cells and the coverslips were incubated for 30 min at 37°C in a water-saturated atmosphere. After careful rinsing with PBS the coverslips were treated with the fluorochrome-labelled second antibody and incubated as above. In double-labelling experiments, cells were simultaneously incubated with F-PHD plus the appropriate primary antibody followed by rinsing and second incubation in rhodamine isothiocyanatelabelled second antibody. Finally, coverslips were extensively rinsed and mounted in 50 % glycerolPBS. Mounted coverslips were viewed in a Leitz Diavert inverted microscope equipped with both fluorescence and interference reflection microscopy (IRM) optics. Fluorescence photographs were recorded with Kodak Tri-X films developed in Gradual ST20 developer (Ornano, Milan); IRM pictures were recorded on Agfa Ortho 25 films developed in Agfa Neutol NE. 10-858337

Exp CellRes 159(1985)

144

Tarone et al.

Electron Microscopy For transmission electron microscopy (TEM) cells were fixed, dehydrated and embedded in Epon 812 essentially as described [23]. Thin sections were cut perpendicularly to the contact surface of the cells, stained with uranyl acetate and lead citrate, and examined in a Zeiss EM 109 electron microscope.

RESULTS

Identification of Cell-Substratum Contact Sites in RSV-Transformed BHK Fibroblasts by Interference Reflection and Electron Microscopy Cell-substratum interaction was investigated by interference reflection microscopy (IRM), a technique which allows to analyse the separation distance between the ventral plasma membrane and the substratum surface [24]. By this method it is possible to identify two different types of cell adhesion sites namely, ~close contacts' and 'adhesion plaques'. The former are characterized by a grey image and correspond to broad areas, where the plasma membrane is 30-50 nm apart from the substratum. Adhesion plaques are characterized by a black image and correspond to small (1-2 ~tm), arrowhead-shaped membrane areas closely apposed to the substratum (10-15 nm). In order to evaluate cell-substratum adhesion in standard and reproducible conditions, Rous sarcoma virus-transformed B H K cells (RSV/B4-BHK) were plated on glass coverslips coated with purified plasma fibronectin, and cultured in serum-free medium. Under these conditions, cells adhered rapidly and reached maximal flattening after 60 min at 37°C. Interaction with substratum occurred in broad, grey areas of 'close contact' accounting for almost the whole-cell ventral surface (fig. 1 a), while 'adhesion plaques' were absent; instead, dot-shaped grey areas surrounded by a white ring appeared in all cells (fig. 1 a). According to the interpretation of the IRM images proposed by Izzard & Lochner [24], such pattern indicates that the membrane comes in contact with the substratum in correspondance of the central grey spot and is lifted far from the adhesion surface (more than 50 nm) in the immediately adjacent clear area. TEM analysis of sections perpendicular to the adhesion surface revealed that cells contacted the substratum by means of protrusions of the ventral plasma membrane, similar to short villi. These structures are likely to correspond to the

Fig. 1. Adhesion pattern and actin distribution in RSV/B4-BHK fibroblasts. Cells, plated on fibronectin-coated glass coverslips were fixed, permeabilized and stained with R-PHD and observed by (a, c, f, h) IRM; (b, d, e, g, i) fluorescence microscopy. One hour after plating in the absence of serum cells display multiple punctate sites of adhesion of the close contact type surrounded by clear areas (a, c at arrow). These adhesion sites contain (b, d) F-actin. Following 3 h incubation in the presence of 2 % serum, these adhesion sites tend to progressively cluster and form ring- or crescent-shaped structures (e). These structures (f, g), as well as disperse dots (h, i), remain attached to the substratum and can be detected after removing cells from coverslips by a jet of buffer. (a, b , f - i ) ×1 200; (c, d) x2400; (e) x 650. Exp Cell Res 159 (1985)

Adhesion mechanisms of RSV-transformed cells 145

Exp Cell Res 159 (1985)

146

Tarone et al.

Fig. 2. TEM analysis of (a, b) RSV/B4-BHK; (c) untransformed BHK fibroblasts sectioned perpendicularly to the substratum. Cells were cultured on fibronectin-coated plastic dishes in the (a) absence; or (b, c) presence of serum. Short protrusions of the ventral plasma membrane (arrowheads) contact the substratum and appear to contain a meshwork of microfilaments (a); these structures are likely to correspond to dot-shaped adhesion sites described in fig. 1 a-d. In the presence of serum compact areas of adhesion containing a microfilament meshwork appear to contact the substratum (b, at asterisks), while individual protrusions (e.g., at arrowhead) are seldom observed. Typical adhesion plaques associated with filamentous material converging toward the membrane (c, at arrows) are found in untransformed BHK fibroblasts. (a, b) ×26000; (c) × 19000.

dot-shaped adhesion sites seen by IRM for the following reasons: (i) their shape is expected to generate a dark spot with the white halo by IRM; (ii) they contain a meshwork of filaments likely to represent F-actin stained by phalloidin (see next paragraph); (iii) their size varies between 200 and 400 nm and is compatible with the size of the adhesion sites detected by IRM (300-500 nm) (fig. 2a). The morphology of these substratum contact sites is clearly different from that of adhesion plaques of control B H K fibroblasts (fig. 2 c). Due to the characteristic structure, such dot-shaped adhesion sites can be regarded as cellular feet and in view of this similarity we suggest to call them 'podosomes' (from the greek root, ~ro6-, foot). The adhesive role of podosomes was further demonstrated by the fact that when adherent cells were removed from the growth surface by a jet of buffer, these structures remained attached to the substratum (fig. l f-i). Adhesive structures similar to those above described have been previously described in other RSV-transformed cells [I 1-13, 25]. Exp Cell Res 159 (1985)

Adhesion mechanisms of RSV-transformed cells 147

Fig. 3. Double-label fluorescence of RSV/B4-BHK fibroblasts stained for (a, b) F-actin and vinculin;

(c, d) F-actin and vimentin. (a, b) Correspondance between the location of F-actin and vinculin at ring- or crescent-shaped adhesion structures. (d) Intermediate filaments of the vimentin type have a distribution totally independent from that of F-actin at (c) podosomes, x 1 200.

Organization of Cytoskeletal Proteins in Podosomes In order to visualize microfilament organization, RSV/B4-BHK fibroblasts were stained with rhodamine-labeUed phalloidin (R-PHD), which selectively binds to F-actin [26]. As shown in fig. 1 b, microfilament bundles of the 'stress fiber' type, that represent the prominent actin-containing structures in untransformed fibroblasts (see fig. 4 d), were absent in RSV/B4-BHK cells. Instead Factin was concentrated in spots coincident with dot-shaped cell-substratum contact sites, as revealed by simultaneous immunofluorescence and IRM observation of the same cells (fig. 1 a-d, f-i). Moreover, vinculin, a cytoskeletal protein selectively associated to cell-substratum adhesion sites [4], was found within podosomes, as well as in rosette-shaped structures representing clustered podosomes (see next paragraph) (fig. 3 b). On the other hand, vimentin (fig. 3 d) and tubulin (not shown), two cytoskeletal proteins not directly involved in substratum anchorage structures, were organized according to their typical fibrous pattern and were not coincident with podosomes. Exp Cell Res 159 (1985)

148

Tarone et al.

Exp Ceil Res 159 (1985)

Adhesion mechanisms of RSV-transformed cells 149 Analysis of Podosome and Adhesion Plaque Formation In the attempt to test whether podosomes have features similar to those of adhesion plaques, the formation of cell-substratum contacts in both untransformed and RSV-transformed B H K fibroblasts was studied in different experimental conditions, namely in culture media with or without serum or in the presence of metabolic inhibitors. Untransformed B H K cells plated on fibronectin-coated dishes in serum-free DMEM, i.e., under conditions identical with those used for RSV/B4-BHK, appeared to interact with the substratum at diffuse grey areas corresponding to 'close contacts', but never formed podosomes as assessed by IRM- and F-actin staining (fig. 4 a, b). This was true also for a number of untransformed fibroblasts including secondary cultures of chick embryo cells, 3T3 mouse cells and N R K rat cells (not shown). Under these conditions untransformed B H K cells organize only very few adhesion plaques and actin bundles, as indicated by R-PHD staining (fig. 4 a, b). When calf serum was added to the medium, the adhesion pattern of B H K cells changed dramatically. The close contact area progressively decreased and a large number of sharp IRM black areas corresponding to adhesion plaques appeared (fig. 4 c, d). Similar results were reported by Thom et al. [27]. Concomitantly, actin microfilaments were assembled in prominent bundles ending at adhesion plaques and lying close to the ventral cell surface (fig. 4 d). In no case podosomes were observed. The organization of adhesion plaques and actin bundles took at least 3 h of incubation at 37°C and appeared to reach a maximum at 2 % serum concentration. The mechanism(s) whereby serum promotes the organization of adhesion plaques are unknown. We found, however, that organization of adhesion plaques and actin cables could also be induced by incubating cells in serumfree medium supplemented with growth factors, such as insulin, transferrin and selenium. This suggests that the organization of adhesion plaques requires some sort of signal or metabolic stimuli provided by some of the above-mentioned factors. Moreover, we also found that cells must be biosynthetically active, since focal adhesion did not appear if cells were treated with cyclohexymide (20 ~tM) to inhibit protein synthesis (fig. 4 e, f), or with the ionophore monensin (1 ~tM) (fig. 4g, h), known to block export of secretory proteins [28]. The gray dots surrounded by white rings seen by IRM in fig. 4 g do not represent

Fig. 4. IRM adhesion pattern and F-actin distribution in untransformed BHK fibroblasts in different

experimental conditions. Cells were plated on fibronectin-coated glass coverslips. The omission of serum from the medium prevents the formation of (a) discrete adhesion plaques, as well as (b) organization of microfilament bundles upon plating. When serum is added (c) adhesion plaques and (d) microfilament bundles of the stress fiber type appear in 3 h after plating. Broad close contact areas appear but neither stress fibers nor adhesion plaques are formed when fibroblasts growing in the presence of serum are treated with (e,f) cycloheximide (20 ~tM) or (g, h) monensin (1 ~tM) that block protein synthesis and secretion respectively. ×900. Exp Cell Res 159 (1985)

150

Tarone et al.

Fig. 5. IRM adhesion pattern and actin distribution in RSV/B4-BHK cells plated in the presence of

2 % serum and treated with (a, b) 20 ~tM cycloheximide or (c, d) 1 ~tM monensin. Neither drugs affect the formation of podosomes and their clustering. Adhesion occurs predominantly at (a) circular or (c) crescent-shaped close contacts (arrows). Close contacts correspond precisely to the location of (b, d) compact F-actin-containing rings. ×900.

podosomes, since they do not obviously correspond to F actin-containing sites (fig. 4h). This kind of IRM image is likely to be generated by a secondary interference produced by cytoplasmic vesicles and plasma membrane [29]. Monensin-treated cells, in fact, accumulate a large number of Golgi vesicles as the result of blocking the secretory pathway [28]. In conclusion, the organization of focal adhesion and actin bundles in normal B H K cells adherent on fibronectincoated dishes requires the presence of serum or growth factors and considerably long time during which synthesis and secretion of proteins must normally proceed. The formation of podosomes in RSV-transformed B H K cells seems to be regulated by various mechanisms. In fact, podosomes were assembled in the absence of serum and had fully developed within 1 h after plating (fig. 1 a-d). When RSV/B4-BHK cells were incubated in the presence of serum, the distribution of these adhesion sites was typically reshuffled. Under these culture condiExp Cell Res 159 (1985)

Adhesion mechanisms of RSV-transformed cells 151 tions, cells displayed peculiar ring- or crescent-shaped structure(s) (fig. l f, g) that stained intensely with R-PHD and corresponded to adhesion sites as shown by IRM. These are likely to correspond to compact and extended adhesion areas detected in thin cell sections by TEM (fig. 2 b). The ring-shaped structures are likely to be generated by a clustering of podosomes. When serum was present, in fact, podosomes were not uniformly dispersed on the ventral cell surface but tended to be concentrated in areas located under the cell nucleus (fig. 1 e). Moreover, in cells displaying a prominent ring-shaped adhesion site, podosomes were less numerous (figs If, g and 5). In some cases careful observation of RPHD-stained cells indicated that crescent-shaped adhesion sites consisted of packed dots (figs 1 e, and 7c); this fine structure cannot, however, always be easily documented by photography due to the high concentration of actin giving rise to a strong fluorescence signal. In the presence of serum RSV/B4-BHK cells also showed a small number of adhesion plaques with associated microfilament bundles (fig. 1 e). These adhesion plaques were usually confined to cell edges with a distribution different from that found in normal cells (e.g., see fig. 4 c, d). Podosome formation does not require protein synthesis or secretion since it also occurred in the presence of cycloheximide (20 ~tM) and monensin (I txM) (fig. 5). Coating the substratum with an adhesion molecule, such as fibronectin, seemed to favor podosome formation. In fact, while in standard culture conditions (10 % FCS and uncoated substratum) the number of cells displaying podosomes varied considerably in different experiments, on fibronectin-coated surfaces virtually all cells formed podosomes in a reproducible fashion. In conclusion, these results indicate that podosomes do not share either mechanisms or kinetics of formation with adhesion plaques. They must, then, be considered as a different class of adhesion structures.

Cell-Substratum Adhesion and Actin Organization in Fibroblasts Transformed by Different Oncoviruses A number of fibroblastic cells transformed either in vivo or in vitro by different oncoviruses have been analyzed for the presence of podosomes by staining with R-PHD and IRM. As shown in fig. 6 and table 1, not all transformed cells have detectable podosomes. Among positive cells we found cells of hamster, mouse, rat and quail origin transformed by different strains of Rous sarcoma virus. Rat cells transformed by Fujinami sarcoma virus (FSV) also formed a large number of podosomes (fig. 6 e). On the other hand, fibroblasts transformed by other retroviruses, such as Abelson leukemia virus, Snyder-Theilen feline sarcoma virus, Kirsten sarcoma virus did not attach to the coated dish by means of podosomes. Similarly, podosomes were not detected in polyoma or SV40-transformed cells. Most cells devoid of podosome displayed also a considerable number of thick actin bundles and adhesion plaques (fig. 6 b, d, f), suggesting inverse correlation Exp Cell Res 159 (1985)

152

Tarone et at.

Table I. Occurrence of podosomes in different transformed cell lines Name

Animal species

Transforming agent

Podosomes

RSV/B4-BHK B77-3T3 B77-AA6-3T3 SR-BALB SRA-RAT-I FvR 16QANN-1 ST-FRE K/BALB PY-BHK PM-3

Hamster Mouse Mouse Mouse Rat Rat Quail Mouse Rat Mouse Hamster Hamster

RSV-BH4 RSV-B77 RSV-B77 RSV-SRD RSV-SRA FSV RSV-BH4 Abelson leuk. virus ST feline sarc. virus Kirsten sarc. virus Polyoma virus SV40

+ + + + + + + -

between adhesion plaques and podosomes. Then occurrence of podosomes seems to depend on the nature of the transforming agent.

Detection of Phosphotyrosine-Containing Molecules within Podosomes The fact that podosomes occurred in fibroblasts transformed by Rous or Fujinami sarcoma viruses, whose oncogene products are tyrosine kinases with a peculiar tropism for cell adhesion structures [11-15], suggested that podosomes might result from the action of these tyrosine kinases upon specific targets located within adhesion structures. To test this possibility, immunofluorescence experiments were performed using affinity-purified antibodies cross-reacting with phosphotyrosine (P-Tyr). These antibodies were prepared against the synthetic hapten azobenzylphosphonate (ABP) and reacted specifically with P-Tyr, as previously documented by phospho-amino acid analysis of proteins immunoprecipitated by these antibodies [20]. As shown in fig. 7 a, b, ABP antibodies reacted with podosomes as revealed by the correspondence of ABP antibodies positive structures with actin dots. A clear staining can also be observed in RSV/B4-BHK cells plated on fibronectin-coated dishes in the presence of serum, where podosomes are packed in typical ring-shaped, structures (fig. 7 c, d). ABP antibodies usually stained an area slightly wider than that stained by F-PHD, indicating that molecules containing P-Tyr are also present in structures immediately surrounding these adhesion sites. The immunofluorescence staining was inhibited by incubating ABP antibodies with excess P-Tyr, demonstrating the specificity of the reaction. In addition to podosomes, residual adhesion plaques (fig. 7 b, arrowheads) and cytoplasm were stained with ABP antibodies in agreement with previously reported data [20, 21]. Exp Cell Res 159 (1985)

Adhesion mechanisms of RSV-transformed cells 153

Fig. 6. Distribution of F actin-containing structures in fibroblasts transformed by different oncovi-

ruses. Podosomes are present in (a) SR/BALB; (c) B77-3T3 RSV-transformed mouse cells and (e) in FyR rat ceils transformed by RSV. (b) PM-3 cells transformed by SV40; (d) K/BALB cells transformed by Kirsten sarcoma virus and (./")ANN-1 cells transformed by Abelson virus showed alteration of the microfilament distribution but no podosomes. ×900.

Exp Cell Res 159 (1985)

154

Tarone et al.

Fig. 7. Double-label fluorescence staining of RSV/B4-BHK cells with (a, c) F-PHD; (b, d) ABP antibodies reacting with P-Tyr. (b, d) Phosphotyrosine-containing material (large arrows) was concentrated in correspondence of (a, c) F-PHD stained podosomes and ring-shaped adhesion structures. In (b) (at arrowheads) residual adhesion plaques located at the cell periphery are also decorated by ABP antibodies, x900.

DISCUSSION In this paper we define as podosomes dot-shaped adhesion sites detectable by IRM and containing both F-actin and vinculin. These sites, which by TEM produce a pattern of substratum contacting feet, represent major adhesion structures of RSV-transformed B H K fibroblasts plated on fibronectin. Dot-like adhesion structures of similar shape and size have previously been described in RSVtransformed cells and have often been interpreted as modified adhesion plaques [11, 13, 15]. This interpretation was based mainly on the observation that dot-like sites and adhesion plaques have a similar composition in cytoskeletal proteins. The data presented in this paper, however, indicates that podosomes and adhesion plaques are likely to represent two distinct adhesion devices. This conclusion is based on the following observations. (a) Both IRM and TEM images are consistent with the idea that podosomes correspond to short cylindrical protrusions of the ventral plasma membrane with a diameter of about 400 nm and contacting the substratum only at their tips. Exp Cell Res 159 (1985)

Adhesion mechanisms of RSV-transformed cells 155 Adhesion plaques, on the contrary, correspond to a flattened portion of the ventral plasma membrane with a considerable surface area. (b) Formation of podosomes and adhesion plaques is controlled by different mechanisms. Adhesion plaques do not appear spontaneously when cells are plated on fibronectin-coated dishes, since their organization requires extracellular stimuli provided by serum or by a mixture of growth factors. Plaques are formed only when cells respond to such factors by synthesizing and secreting new components. Conversely, podosome formation seems to rely on apparently simpler and relatively independent mechanism as they appear in the absence of serum factors and protein synthesis. Moreover, occurrence of podosomes is likely to be favored or even induced by the presence of fibronectin on the substratum surface, since cells plated in conventional culture conditions form variable amounts of such adhesive structures. (c) The possibility that podosomes represent a precursor form of adhesion plaques seems unlikely. In fact, treatment of untransformed fibroblasts in conditions that prevent adhesion plaque formation (absence of serum, block of protein synthesis or addition of monensin) does not result in accumulation of podosomes. Moreover, podosomes do not seem to be derived from the dismantling of adhesion plaques in round actively dividing cells, such as transformed cells, since they were never observed in untransformed B H K fibroblasts during mitosis. In conclusion we suggest that podosomes are alternative structures providing some transformed fibroblasts of an adhesion system different from that accounted for by adhesion plaques also insofar their molecular architecture is concerned. Among cell tested we can detect podosomes in fibroblasts transformed by Rous or Fujinami sarcoma viruses. These retroviruses carry oncogenes, named src and fps respectively, coding for proteins with tyrosine kinase activity and with molecular weight of 60000 (pp60 src) and 130000 (pl30fps) [30-33]. Subcellular localization studies indicated that the pp60 src of RSV is localized both at the inner face of the plasma membrane and at cell-substratum adhesion sites, namely at dot-shaped sites and residual adhesion plaques [11-14]. While membrane localization is due to hydrophobic bonding of the pp60 to the lipid bilayer [34--36], the interaction with adhesive structures occurs through electrostatic interactions [37]. The pl30fps of Fujinami sarcoma virus is also associated to subcellular components by means of electrostatic interactions, and a fraction of these molecules is localized at cell adhesion structures [15]. The abl oncogene product, that also has tyrosine kinase activity, interacts hydrophobically with the plasma membrane [38-40]. In ANN-1 mouse cells, that do not form podosomes, the abl protein has a diffuse distribution on the cell membrane and is not found at cell substratum contact sites [38-40]. In some Abelson-transformed cells, namely rat fibroblasts, the abl protein was found to be localized at cell-substratum adhesion sites [41]; in these case cells display dot-shaped adhesion by all means comparable to podosomes. Podosomes are absent in fibroblasts transformed by a number of other oncoviruses that induce cell transformation by mechanisms different Exp Cell Res 158 (1985)

156

Tarone et al.

from tyrosine phosphorylation. These considerations suggest that the induction of podosomes is related to tyrosine-kinase activity and might be the result of the unique interaction of the oncogene-coded proteins with specific submembrane adhesion structures. This hypothesis is also supported by the finding that podosomes of RSV/B4-BHK cells contain a relatively high concentration of tyrosinephosphorylated molecules compared with the surrounding cytoplasm. We show, in fact, that podosomes are decorated by antibodies reacting with P-Tyr. The nature of the molecules detected by ABP antibodies in podosomes is not known at present. Pp60 src itself [42] and vinculin [43] are known to contain P-Tyr and can contribute to the positive signal in immunofluorescence. However, previous data from our laboratory indicate that, in addition to pp60 src, P-Tyr-containing molecules with molecular weight of 130000 and 70000 and with still unknown functions are also present at the level of substratum-bound cytoskeletal structures [20, 21]. Whatever the nature of the target molecules, immunofluorescence experiments indicate a relevant point. In fact, to evaluate the possible implications of tyrosine phosphorylation it is important to know where phosphorylated proteins are localized within the cell. The overall increase of P-Tyr in podosomes in living RSV-transformed cells suggests that the activity pp60 src takes place mainly at these adhesive structures. The possible functional significance of podosomes remains to be established. Wang et al. [44] reported that these structures contain gelsolin, a protein that modulates F-actin polymerization and suggested that podosomes may represent forms of transient association between the cell and the substratum. Moreover Chen et al. [45] have recently shown that the dot-like adhesion sites of RSV-transformed cells correspond to areas of localized proteolytic degradation of extracellular fibronectin and might, then, be important in remodeling the extracellular matrix and possibly in tumor cell invasion during metastasis. We have shown that adhesion structures very similar, if not identical, to the podosomes of RSV-transformed cells, occur in monocytes and osteoclasts [46, 47]. These latter are non-transformed cells deputed to resorption of bone matrix. These findings raise the possibility that podosomes might represent specialized adhesion structures occurring in cells displaying a peculiar interaction with their environment. We would like to thank Drs B. Geiger, M. Prat and Th. Wieland for generously providing reagents employed in these study. The excellent technical assistance of M. R. Amedeo and P. Rossino is gratefully acknowledged. This research was supported by grants of Progetto Finalizzato "Oncologia" Consiglio Nazionale delle Ricerche and Associazione Italiana per la Ricerca sul Cancro.

REFERENCES 1. 2. 3. 4.

Grinnell, F, Int rev cytol 53 (1978) 64. Rees, D A, Lloyd, C W & Thorn, D, Nature 267 (1977) 124. Hynes, R O, Destree, A T & Wagner, D D, Cold Spring Harbor syrup quant biol 46 (1981) 659. Geiger, B, Cold Spring Harbor syrup quant biol 46 (1981) 671.

Exp Cell Res 159 (1985)

A d h e s i o n m e c h a n i s m s o f R S V - t r a n s f o r m e d cells

157

5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20.

Vaheri, A & Mosher, D F, Biochim biophys acta 516 (1978) 1. Yamada, K M & Olden, K, Nature 275 (1978) 179. Vaheri, A, Kurkinen, M, Lehto, V, Linder, E & Timpl, R, Proc natl acad sci US 75 (1978) 4944. Hynes, R O & Yamada, K M, J cell biol 95 (1982) 369. Edelman, G M & Yahara, I, Proc natl acad sci US 73 (1976) 2047. Wang, E & Goldberg, A R, Proc natl acad sci US 73 (1976) 4065. Rohrschneider, L R, Proc natl acad sci US 77 (1980) 3514. David-Pfeuty, T & Singer, S J, Proc natl acad sci US 77 (1980) 6687. Shriver, K & Rohrschneider, L R, J cell biol 89 (1981) 525. Nigg, E A, Sefton, B M, Hunter, T, Walter, G & Singer, S J, Proc natl acad sci US 79 (1982) 5322. Feldman, R A, Wang, E & Hanafusa, H, J virol 45 (1983) 782. Di Renzo, M F & Bretti, S, Intj cancer 30 (1982) 751. Di Renzo, M F, Doneda, L, Larizza, L & Comoglio, P M, Intj cancer 31 (1983) 455. Murphy, H, Virology 77 (1977) 705. Tarone, G, Galetto, G, Prat, M & Comoglio, P M, J cell biol 94 (1982) 179. Comoglio, P M, Di Renzo, M F, Tarone, G, Giancotti, F G, Naldini, L & Marchisio, P C, EMBO j 3 (1984) 483. 21. Marchisio, P C, Di Renzo, M F & Comoglio, P M, Exp cell res 154 (1984) 112. 22. Weber, K, Rathke, P C, Osborn, M & Franke, W W, Exp cell res 102 (1976) 285. 23. Cirillo, D, Comoglio, P M, Naldini, L & Marchisio, P C, Cell biol int rep 8 (1984) 959. 24. Izzard, C S & Lochner, L R, J cell sci 21 (1976) 129. 25. Marchisio, P C, Capasso, O, Nitsch, L, Cancedda, R & Gionti, E, Exp cell res 151 (1984) 332. 26. Wulf, E, Deboben, A, Bautz, F A, Faulstich, H & Wieland, Th, Proc natl acad sci US 76 (1979) 4498. 27. Thorn, D, Powell, A J & Rees, D A, J cell sci 35 (1979) 281. 28. Uchida, N, Smilowitz, H, Ledger, P W & Tanzer, M L, J biol chem 255 (1980) 8638. 29. Gingell, D, Todd, I & Owens, N, J cell sci 54 (1982) 287. 30. Collet, M S & Erikson, R L, Proc natl acad sci US 75 (1978) 2021. 311 Levinson, A D, Opperman, H E, Varmus, H E & Bishop, M J, Cell 15 (1978) 561. 32. Feldman, R A, Hanafusa, T & Hanafusa, H, Cell 22 (1980) 757. 33. Pawson, T, Guyden, J, Kung, T H, Radke, K, Gillmore, T & Martin, G S, Cell 22 (1980) 767. 34. Kreuger, J G, Wang, E & Goldberg, A R, Virology 101 (1980) 25. 35. Levinson, A D, Courtneidge, S A & Bishop, J M, Proc natl acad sci US 78 (1981) 1624. 36. Garber, E A, Kreuger,', J G, Hanafusa, H & Goldberg, A R, Nature 302 (1983) 161. 37. Kreuger, J G, Garber, E A, Chin, S M, Hanafusa, H & Goldberg, A R, Mol cell biol 4 (1984) 454. 38. Witte, O N, Rosenberg, N & Baltimore, D, J virol 31 (1979) 776. 39. Boss, M A, Dreyfuss, G & Baltimore, D, J virol 40 (1981) 472. 40. Sefton, B M, Trowbridge, I S & Cooper, J A, Cell 31 (1982) 465. 41. Rohrschneider, L R & Najita, L M, J virol 51 (1984) 547. 42. Collett, M S, Purchio, A F & Erikson, R L, Nature 285 (1980) 167. 43. Sefton, B M, Hunter, T, Ball, E H & Singer, S J, Cell 24 (1981) 165. 44. Wang, E, Yin, H E, Kreuger, G J, Caliguri, L A & Tamm, I, J cell biol 98 (1984) 761. 45. Chen, W T, Olden, K, Bernard, B A & Fong Chu, W, J cell biol 98 (1984) 1546. 46. Zambonin-Zallone,A, Teti, A, Primavera, M V, Naldini, L & Marchisio, P C, J anat 137 (1983) 57. 47. Marchisio, P C, Cirillo, D, Naldini, L, Primavera, M V, Teti, A & Zambonin-Zallone, A, J cell biol 99 (1984) 1696. Received October 17, 1984

Printedin Sweden

Exp CellRes 159 (1985)