JOURNAL OF CELLULAR PHYSIOLOGY 147:182-190 (1991)
Restoration of Adhesive Potentials of Ehrlich Ascites Carcinoma Cells by Modification of Plasma Membrane ALEXEI A. BOCDANOV, Jr.,* LARISA V. GORDEEVA, BORIS A. BAIBAKOV, LEONID B. MARCOLIS, AND VLADlMlR P. TORCHlLlN Institute of Experimental Cardiology, USSR Cardiology Research Center, 12 7552 Moscow (A.A.B., V.P.T.) and A.N. Belozersky Laboratory of Molecular Biology and Bioorganic Chemistry, Moscow State University, 1 19899 Moscow (L.V.C., B.A.B., L.B.M.), USSR
A novel technique for modulating the spreading of ascites cells has been developed. Plasma membranes of Ehrlich ascites carcinoma cells were modified in two different ways: 1 ) biotin residues were covalently coupled to membrane components; 2) biotinylated lipid was introduced into plasma membranes. Adhesion and spreading of modified cells on avidin-coated substrates were studied and compared to those of non-modified cells. Both types of membrane alteration were shown to induce specific (biotin-dependent) interaction with immobilized avidin with resultant cell spreading. Spread cells attained cpithelioid-like morphology with the formation of wide thin lamellae, focal contacts with substrate, and circular actin bundles. The process of spreading was shown to be energy-dependent: it could be blocked by metabolic inhibitors and by low temperature. Formation of extended lamellae was prevented by preincubation of cells in the presence of cytochalasin B. The effects of metabolic poisons, low temperature, and microfilament-disruptive drugs were reversible and after the restoration of physiological conditions the cells resumed the spreading process. lmmunoprecipitation of biotinylated cell lysates with antiserum to cytoplasmic domain of P,-integrin subunit revealed a major 110 kD avidin-binding component. We conclude that lack of spreading of ascites carcinoma cells may be explained by the lack of functionally active adhesion- and spreading-competent cell-surface receptors, but may not be attributed to the defects in intracellular function or organization. lntracellular machinery of cell spreading is preserved in these ascites cells and could be turned on by cell attachment to the substrate via artificial adhesive site incorporated into plasma membrane.
Numerous intercellular processes and cell functions are dependent on a complex process of cell-substrate interaction, known as adhesion. Attachment and spreading of normal eucariotic cells onto solid substrate result in the initiation of mRNA translation (Benecke et al., 1978; Farmer et al., 1978) recovery of mRNA synthesis and cell-shape controlled expression of cytoskeletal genes (Benecke et al., 1980; Ben-Ze’evet al., 1980; 1984),alteration of fluidity of plasma membrane (Nakache et al., 1985), and activation of Na+/H+antiporters with the concomitant increase of intercellular pH (Margolis et al., 1988; Schwartz et al., 1990). As a result, cellular proliferation (Folkman and Greenspan, 1975; Folkman and Moscona, 1978) and migratory activity (Vlodavskyet al., 1980)are potentiated. Therefore, it is reasonable to consider solid substrate as a complex “growth/activation factor” for many different cells. Various transformed cells exhibit partial or complete release of cell proliferation from anchorage dependence. This is accompanied by the reduced ability of transformed cells to spread on a solid substrate. Ascites cells may serve as an extreme example of transformed 0 1991 WILEY-LISS, INC.
cells which have lost completely the anchorage dependent proliferation. For example, Ehrlich ascites carcinoma cells do not adhere tightly to either clean or t o serum-coated glass. The inability of these cells to adhere and to spread on a solid substrate may be explained in two different ways. On the one hand, transformation may affect the level of expression of integrin subunits involved in cell-substrate adhesion and spreading (Piantefaber and Hynes, 1989). On the other hand, cytoskeletal genes may be expressed in an aberrant way as a result of transformation. In the latter Received June 21,1990;accepted November 15, 1990. *To whom correspondence should be addressed. Alexei A. Bogdanov is now a t Webster Center for Biological Sciences Amherst College, Amherst MA 01002. Abbreviations used: BAH-PE, N-[(6-biotinylamido)hexanoyll phosphatidylethanolamine dioleoyl; HBS, Hepes buffered saline; HBSS, Hanks’ balanced salts saline; NHS-LC-B, N-hydroxysulfosuccinimidyl(6-biotiny1amido)hexanoate; NHS-SS-B, N-hydroxysulfosuccinimidyl 2-(biotinamido~ethyl-1,3-dithiopropionate; SEM, scanning electron microscopy.
ADHESION OF ASCITES CELLS
case the assembly of normal cytoskeleton necessary for cell spreading is altered (Vasiliev, 1985). Several model systems designed for controlled alteration of cell adhesion and spreading have been reported (Folkman and Moscona, 1978; Schnaar, 1984). For example, coating of plastic with increasing amounts of poly(2-hydroxyethy1)methacrylic acid results in progressive decrease of spreading of cultured cells (Folkman and Moscona, 1978). Several artificial polyacrylamide supports containing charged groups (Brandley et al., 1987a) or carbohydrates (Brandley et al., 1987b; Schnaar et al., 1978)were shown to increase hepatocyte and lymphocyte adhesion. It was demonstrated that plastic substrate covered with polyclonal (Grinnell and Hays, 1978) or monoclonal antibodies to cell surface proteins (Grinnell et al., 1988)promote attachment and spreading of BHK cells. However, none of the substrates reported so far were used for modulation of attachment and spreading of ascites cells. We also are not aware of any attempts to modify the surface of these cells to facilitate adhesion. Here we demonstrate, that modification of plasma membrane of Ehrlich ascites carcinoma cells with adhesion-promoting molecules not only facilitates the adhesion to a model substrate, but also initiates normal cell spreading, which involves rearrangement of the cytoskeleton, organization of focal contacts, and attainment of epithelioid morphology. MATERIALS AND METHODS Model substrates Avidin-coated glass coverslips were prepared as follows: coverslips were treated by 4% HF in 50% HNO, for several seconds, washed in water, and dehydrated in acetone. A solution of 1,l'-carbonyldiimidazole in dimethylformamide (100 mg/ml) was applied to coverslips and reaction proceeded for 1 h in a inert atmosphere with stirring. After washing with cold water coverslips were treated overnight with a solution of egg white avidin (Sigma, 2 mg/ml) in 15 mM sodium tetraborate, pH 8.0. Coverslips were washed extensively in 10 mM Hepes, 0.15 M NaC1, pH 7.5 (HBS) and used the same day. Alternatively, avidin-coated coverslips were prepared by avidin attachment to the biotinylated glass. Glass coverslips were treated with 7% 3-(triethoxysily1)-propylamine (Merck) in benzene (1.5 h at SOT) and washed with n-hexane. The coverslips were incubated in a solution of N-hydroxysulfosuccinimidyl(6biotinylamido) hexanoate (NHS-LC-B, Pierce) (1 mgi ml, in 20 mM Hepes, 0.15 M NaC1, pH 7.5) for 2 h at 20°C. After extensive washing with water coverslips were incubated in avidin solution (1.5 mg/ml in Hanks' balanced salts solution (HBSS). Other types of substrates were prepared by immobilization on carbonyldiimidazole-activatedglass of bovine serum albumin (fraction V, Sigma), 10% fetal calf serum (Gibco), or poly-L-lysine (100 kD, Sigma), respectively. Biotinylation of cells Ehrlich ascites carcinoma cells, propagated in C3HA mice, were washed from ascites fluid with HBSS and suspended at concentration of 1 x lo7 celldm1 in 20 mM
183
Hepes, 0.15 M NaC1, and 7 mM glucose (pH 7.4). Suspended cells (0.5 ml) were mixed with the equal volume of 1 mgiml NHS-LC-B or 1 mgiml N-hydroxysulfosuccinimidyl 2-(biotinamido)ethyl-1,3-dithiopropionate (NHS-SS-B, Pierce) in HBS and incubated for 20 min at 4°C. Five to six washings of modified cells were performed by sedimentation through 30% FicollPlaque (Pharmacia) solution in HBSS (2 min at 8OOg). Cells were resuspended in HBSS and used in further experiments. Alternatively biotinylated lipid N-[(G-biotiny1amido)hexanoyl] phosphatidylethanolamine dioleoyl (BAHPE, prepared as in (Bayer et al., 1979) was introduced into plasma membranes of cells. Then 10 p,l of BAH-PE solution (1 mgiml in ethanol) were added to 1 ml of HBSS (Ca, Mg free, Gibco), containing 5 x lo5 cells. After incubation for 30 min at 4"C, cells were washed 2 times by centrifugation through 30% Ficoll-Paque in HBSS (2 min at 8OOg).Washed cells were susgended in serum-free MEM without phenol red at 4 x 10 cellslml. Incubation conditions In different experiments 3-5 x lo5 cellsiml were added in 50 p,l aliquots t o coverslips (8x8 mm) and incubated for 15-180 min at 37°C in humid chambers. In some ex eriments sodium azide (1.3 mg/ml in glucose-free H SS) or cytochalasin B (10 pgiml), respectively, were added in the incubation mixture. All the additives were present in washing solutions throughout the experiments. In control experiments cells were incubated with substrates coated with BSA or with avidin-coated substrates preincubated with free biotin (1mgiml). In some experiments cells were incubated at 4°C. Cells modified with NHS-SS-B were treated with 2 mM cysteine in MEM, pH 7.6, at different stages of incubation. After 4 washings with HBSS, cells were examined by light, phase-contrast, interference-reflection microscopies either live or fixed with 4% formaldehyde in PBS. Microscopy For fluorescent microscopy biotinylated cells were stained with FITC-avidin (10 pg/ml prepared as in Heggeness and Ash (1977). To visualize polymerized actin, formaldehyde-fixed cells were extracted with 1% Triton X-100 in PBS, fixed with 2% formaldehyde, and stained with TRITC-labeled phalloidin (2 pg/ml). Fluorescence and interference-reflection images were observed using an Opton Photomicroscope I11 equipped with HBO mercury lamp and standard glass filters. Scanning electron microscopy (SEM) was performed on a PSEM 500 (Phillips) instrument operated at 35 kV. Samples were fixed with 2.5% glutaraldehyde in PBS, washed, dehydrated in acetone series, dried at C02 critical point, and shadowed with platinum. Preparation of cell lysates, electrophoresis, a n d blotting A total of 1 x lo7 biotinylated or control cells were lysed in 1%Nonidet NP-40,lO mM Tris, 1mM EDTA, 1 mM EGTA, 50 pg/ml pancreatic trypsin inhibitor, 2 mM PMSF (pH 7.5) for 5 min on ice and passed several times through a blue pipette tip (Gilson). Nuclei were separated at 1,0008-for 5 min. Lysate was treated with
i
184
BOGDANOV ET AL.
A
B
2 209 7*
60-
45-
Fig. 1. Staining of biotinylated ascites carcinoma cells with FITCavidin in suspension. Suspended cells were stained with FITC-avidin (20 Fgirnl) for 5 min at 4°C. A Fluorescent. B Phase-contrast images. Bar = 10 pm.
5% trichloroacetic acid, sediment was collected by centrifugation and redissolved in 0.5% SDS in water. Protein concentration was determined using BCA protein assay reagent (Pierce) according to manufacturer's recommendations. SDS-polyacrylamide gel electrophoresis was run in 7.5-15% gradient gels or in 10% gels using Laemmli discontinuous buffer system (Laemmli, 1970). Gels were electroblotted on nitrocellulose sheets (0.2 p,m, Schliecher and Schull) and biotinylated proteins were visualized by ABC-kit (Vector).Alternatively, radioiodinated avidin was used for autoradiographical visualization as described below. Immunoprecipitation A total of 5-106 x lo7 cells were lysed and lysates obtained were immunoprecipitated with the use of antiserum to pl-integrin subunit (a generous gift of Dr. R.O. Hynes) as described in Marcantonio and Hynes (1988). Proteins eluted from protein A-Sepharose were separated by electrophoresis and electroblotted as described before. Nitrocellulose membranes were blocked
18-
Fig. 2. Visualization of avidin-binding proteins in modified ascites carcinoma cells. A Cells were lysed and proteins were separated by SDS electrophoresis in 10%' polyacrylamide gels. Gels were electroblotted on nitrocellulose. Biotinylated proteins were detected by incubation with ['251]-labeledavidin and visualized by autoradiography. B: Cell lysates were immunoprecipitated with anti+, antiserum (5 pgiml) and protein A-Sepharose.Eluted proteins were separated by SDS-electrophoresisin 10%polyacrylamide gels, electroblotted, incubated with [12511-avidin,and subjected to autoradiography. To the left indicated are the positions of molecular weight markers (kD). Details are given in Materials and methods.
by 3% gelatin, 1%BSA in PBS (pH 7.5) overnight, washed, and incubated with [1251]-labeledavidin (labeling was performed as described in Markwell and Fox (1978), 2 ~ 1 cpmiml 0 ~ in 0.1% Tween-20, 1% BSA in PBS (pH 7.5). Extensively washed and dried membranes were exposed to XAR-5 film (Kodak) without intensifying screen for 3-6 days. Determination of biotin in cell lysates Biotin content in cell lysates was determined by colorimetric method (Green, 1965),exactly as described in Pierce manual to NHC-LC-B (No. 21335).
ADHESION OF ASCITES CELLS
185
Fig. 3. Interaction of ascites cells with non-derivatizedglass substrate. A: Phase contrast image of cells adsorbed on glass, bar= 10 km. B SEM micrograph of attached cell; bar= 1 pm.
RESULTS Ehrlich ascites carcinoma cells were modified by water-soluble membrane-impermeant activated biotin derivative (NHS-LC-B). Biotin residues on the cell surface were detected by staining of modified cells with FITC-labeled avidin. All biotinylated cells exhibited fluorescent staining (Fig. 1). In the presence of free biotin no fluorescent avidin was bound to cells. Biotinylated proteins were identified by radioiodinated avidin or avidin-peroxidase on blots of total cell lysates, separated by SDS-electrophoresis. One ma’or 110 kD band and five minor ones with molecu ar masses of approximately 80, 66, 62, 42, and 28 kD, respectively, were visualized (Fig. 2A). We performed immunoprecipitation of cell lysates with rabbit antiserum t o p,-subunit of integrin (Markantonio and Hynes, 1988). Antibody-bound biotinylated proteins were detected by autoradiography using radioiodinated avidin as a probe. We found that the major 110 kD band was at least partially composed of biotinylated integrin subunits. Though cell lysis and immunoprecipitation were carried out in the presence of divalent cations, no other biotinylated integrin subunits of higher molecular weight were found to associate with isolated subunits (Fig. 2B). To examine the amount of biotin residues introduced by covalent modification, cell lysates were treated with pronase and protein fragments were tested for 4’-hydroxyazobenzene-2-carboxylic acid displacement from the active center of avidin. In most experiments approximately 160 pg of biotin were coupled per pg of total cell protein. Viability of modified cells exceeded 99% as determined by trypan-blue exclusion test. Morphology of cells was studied by phase-contrast and scanning electron microscopy. Non-modified carcinoma cells were washed to remove ascites fluid by HBSS and cell suspension was incubated over a glass coverslip at 37°C for up to 180 min. A t the end of incubation period cells were visualized on a substrate. However, after washings only few cells remained on the glass. In phase-contrast micrographs these cells appear to be
!
spherical (Fig. 3A). SEM examination of the same specimen showed that those cells which remain on the substrate are anchored by cytoplasmic protrusions (Fig. 3B). If the suspension of washed cells was incubated with the serum-covered coverslip, the number of cells adhering to glass was even lower, than in the case of clean (acid-washed) glass. Similar results were obtained when a clean glass coverslip was used as substrate and serum was added to the incubation mixture. In contrast to results of experiments carried out with non-modified Ehrlich carcinoma cells, the number of adhered cells significantly increased if biotinylated cells were allowed to attach to avidin-coated glass. Shortly after the round suspended cells start to attach to the substrate they begin to spread: in 30 min at 37°C thin transparent lamellae were formed. In 1.5-2 h after the incubation had been started, 80-90% of attached cells become spread. The morphology of cells varied: some cells exhibited spherical central (perinuclear) areas and flat circular lamella. In some cells perinuclear area was more spread and lamella was narrow (Fig. 4A), Occasional elongated (polarized) cells were also present. Usually they were restricted from both sides by adjacent cells. Some preparations contained cells of several different morphologies. Examination of cells spread on avidin-coated glass by SEM revealed wide smooth thin circular lamellae and numerous microvilli in the central region of cells with wide thin lamellae (Fig. 4B). The latter were smooth and free of microvilli. The modified cells attached to the avidin-coated coverslips remain spread in HBSS another 3-5 h. If the solution was changed for the serum-supplemented media, cells proliferate remaining attached and spread for at least another 36 h. Several kinds of control experiments have been performed to prove the actual involvement of biotin-avidin interaction in cell spreading as described below. In corresponding control experiments biotinylated cells were incubated either with non-modified glass, or with BSA-covered glass. In 1-3 h of incubation at 37°C cells attached to substrate, but did not spread. Such
186
BOGDANOV ET AL.
Fig. 4. Spreading of biotinylated cells on avidin-coated glass coverslips. Cells (5 x lo6 cellsiml) were treated with 1mg/ml NHS-LC-B in 20 mM Hepes, 0.15 M NaC1,7 mM glucose (pH 7.4) for 20 min at 4"C, washed, and incubated over avidin immobilized on glass coverslip. A Phase contrast micrography;bar= 10 pm.B SEM micrograph of attached cells; bar= 1 pm.
cells remained spherical if avidin-coated coverslips were preincubated with the excess of free biotin before the addition of cells (Fig. 5). In both cases cells remained round-shaped after prolonged incubation, with only a few percent (1-3% of the total cell number) weakly spread. Non-modified cells when applied to avidin-coated coverslips remained spherical for 1-3 h of incubation period. On the polylysine-coated glass a fraction of non-modified cells developed small fan-shaped lamellae. However, perinuclear area of cells always remained spherical and these cells have never achieved epithelioid shape. A total of 95-97% of non-modified cells adhered to avidin-coated glass did not spread at all. The same is true for biotinylated ones incubated over immobilized polylysine. To find out, whether biotin residues on cell surface directly keep the cells spread on avidinylated substrate, cells were modified with NHS-SS-B. This reagent contains disulfide bridge, which can be easily cleaved with thiol-containing membrane-impermeant agents in physiological conditions. When modified cells attached to avidin-coated glass (in 10 min of incubation), cysteine was added to reduce disulfide bonds. None of adherent cells were spread after the addition of cysteine. The amount of adherent cells was noticeably reduced. In control experiment modified cells, if allowed to spread on avidin-coated glass, attained epithelioid morphology in spite of the presence of cysteine. The involvement of biotinylated plasma membrane com onents in the formation of contacts with immobilize avidin have been proved by staining with FITCavidin. Modified cells were intensively stained in suspension or shortly after attachment to glass (Fig. 1). However, after spreading they do not bind avidin any more and remain non-stained with fluorescent avidin.
B
Fig. 5. Attachment of biotinylated cells to avidin-coated glass, preincubated in the presence of free biotin (1 mg/ml). Bar= 10 pm.
Spreading of biotinylated cells on immobilized avidin at 4°C or in the presence of sodium azide or cytochalasin B was tested. After incubation for 1-2 h at 4°C the cells remained spherical. The same was true when azide was present for 50-60 min. The inhibition of spreading was reversible: 1 h after attached cells were transferred to 37°C they began to flatten. The same was true when NaN,containing solution was replaced for HBSS. In this case SEM revealed ruffles and well-developed lamellae on the edges of cells in 1.5 h after the restoration of physiological conditions. The addition of cytochalasin B to suspended biotinylated cells inhibited the formation of wide lamellae after
ADHESION OF ASCITES CELLS
187
Fig. 6. Effect of cytochalasin B on spreading of biotinylated cells on avidin. A, C, D: Phase-contrast images. B: SEM image. A: Attachment of cells to avidin in the presence of 10 pgiml of cytochalasin B; bar= 10 km. B: SEM image of cell presented in A; bar= 1 pm. C: Cells pretreated with cytochalasin B and then allowed to spread in drugfree media; bar= 10 pm.D: Effect of cytochalasin B on biotin-modified cells spread on avidin-coated coverslip. Cytochalasin B was added after cells were spread.
their attachment to immobilized avidin. Instead only long thin spikes were formed for 50-60 min of incubation at 37°C (Fig. 6A,B). When cytochalasin-treated cells were transferred to drug-free media numerous small lamellae were detected around the long protrusions; the central area of many cells was flattened (Fig. 6C). Treatment with cytochalasin B also affected already spread cells. One hour after addition of drug, cellular lamellae retracted and cells became arborized (Fig. 6D). Biotinylated cells spread on avidin were studied by interference reflection microscopy. The majority of cells had a ring of focal (black) contacts with substrate alongside the external edge. Focal contacts were located in lamellar regions of cells (Fig. 7A,B). To study the morphology of the actin cortex of the spread ascites cells, fixed cells after Triton extraction were stained by fluorescent labeled phalloidin. A cir-
cular microfilament bundle was visualized alongside the margin of many spread cells (Fig. 8). An alternative way of plasma membrane modification with biotin residues was tested. Suspended ascites carcinoma cells were incubated with biotinylated lipid (BAH-PE) at 4°C. Staining with FITC-avidin revealed biotinylated lipid uniformly incorporated andlor attached to plasma membranes (Fig. 9A). Suspended cells were washed and tested for the ability to spread on immobilized avidin. In 10-15 min at 37°C cells attached to substrate and developed well-defined epithelioid morphology (Fig. 9B). In the case of biotin-PE-mediated adhesion the cells remain less spread than when the total cell surface was biotinylated. DISCUSSION Mouse ascites carcinoma cells proliferate in suspension and attach only weakly to glass and cultural
188
BOGDANOV ET AL.
Fig. 7. Formation of focal contacts with immobilized avidin. A Interference-reflection image. B: Phase-contrast image of the same field. Bar= 5 pm.
Fig. 8. Visualization of polymerized actin in spread ascites cells. Cells were extracted with Triton X-100, fixed with 2 6 formaldehyde and stained with TRITC-phalloidin (2 pgiml).
plastic. In contrast to normal epithelia from which these cells have been derived 70 years ago they are unable to spread on these surfaces. The reason for the lack of the features which are typical for normal phenotype is not clear. The main question to be answered is whether ascites carcinoma cells lack functionally active receptors specific for spreading-competent molecules, or the defects in spreading are caused by some alterations of the intracellular machinery. We attempted to discriminate between these two possibilities by modification of cell membrane with low-molecular weight molecules which should promote cell adhesion to specially prepared model substrate. In our experiments we used glass coated with immobilized avidin as a substrate for cells, subjected to mild modi-
fication with either membrane-impermeant activated biotin analogs or with biotinylated lipid, artificially inserted into plasma membrane. Both types of plasma membrane modification dramatically increased the number of cells attached to avidin-coated substrate. This is a trivial result which can be predicted taking into account high-affinity binding of biotin molecules with avidin (K, of avidin-biotin complex have been reported as M) (Green, 1963).What is non-trivial is that cell surface modification also promoted cell spreading with the development of epithelioid morphology. Involvement of avidin-biotin interactions in ascites cell spreading was proved in control experiments, where non-modified cells were used or when biotinbinding sites of immobilized avidin were blocked with the excess of free biotin. Thus, the adhesion and spreading of modified ascites cells is not determined solely by electrostatic interactions due to the difference in charge between cationic avidin (PI=10.5; Green, 1965) and cells, bearing a net negative charge. This conclusion was confirmed by experiments with positively charged poly-L-lysine-coatedglass. The majority of cells attached to this substrate were spherical and only some of them revealed cytoplasmic processes. However, even these cells have never developed epithelioid mor hology. When kotin was linked to plasma membrane components with thiol-cleavable linker, spreading process could be blocked at the early stages of development by membrane-impermeant cysteine. Thus, biotin, either coupled to plasma membrane components (mainly t o 110-kD protein(s) or incorporated into the membrane via the lipid anchor, can serve as an “artificial receptor,” which stimulates changes in cell behavior with respect to the model adhesive substrate. The modified cells remain alive as tested with dyeexclusion test. They also remain alive after the spread-
ADHESION OF ASCITES CELLS
189
Fig. 9. Ascites cells modified by preincubation with biotinylated lipid. A: Fluorescent micrograph of suspended cells, stained with FITC-avidin (20 pgirnl). B: Spreading of modified cells on avidin, a phase-contrast image. Bar= 10 km.
ing for another few hours if they remain in salt solution (that is approximately the period other cultures can survive in similar conditions). If the solution is changed for a serum-supplemented culture media the cells remain spread on the substrate and look morphologically normal for at least another 36 hours. However, at that time we did not check whether the cells are still bound to the substrate by biotin-avidin linkage. The very process of cell spreading is also evidence that cells are alive. However, to exclude the possibility of non-physiological spreading (e.g., energy-independent flattening of cells without the involvement of intracellular structures), we incubated modified cells with immobilized avidin in the presence of inhibitor of respiration (sodium azide), a microfilament-disruptive drug (cytochalasin B) or at low temperature. All these experiments have demonstrated the requirement of energy and microfilament reorganization in biotinmodulated cell spreading. The effects of sodium azide, cytochalasin B, and low temperature were reversible, since cells resumed with spreading when transferred to physiological medium or if warmed to 37°C. By staining of filamentous actin in spread cells were have demonstrated circular organization of microfilament bundles characteristic of normal epithelial cells (Domnina et al., 1985).The process of cell spreading is hypothesized to be governed by cytoskeletal rearrangements caused by the interactions of cell-surface receptors (integrins) (Hynes, 1987) with matrix components. These receptors supposedly interact with the cytoskeleton via cytoplasmic C-terminal domain (Chen et al., 1985; Marcantonio and Hynes, 1988). Alternatively, they are able to form a linkage with cytoskeletal elements (including actin bundles) via transmembrane proteins such as talin (Horwitz et al., 1986). In the case of ascites cells the lack of cell spreading may be due t o the defects in integrin structure or in the organization of its subunits. Related phenomenon was observed for some transformed cells, which exhibit reduced ability to spread if compared to their normal counterparts (Piantefaber
and Hynes, 1989). We found that antibodies to cytoplasmic domain of PI integrin subunit (Marcantonio and Hynes, 1988) precipitate a biotinylated 110 Kd protein. Thus, we may hypothesize that biotinylation of cell surface proteins leads to modification of P1-subunits, which initiates spreading when bound to avidincoated substrate. However, we do not exclude that biotinylated proteins other than integrin subunit may provide some intracellular signal which activate adhesion of modified ascites cells since there are at least five major bands visualized with '"I-labeled avidin (Fig. 2A). Evidence showing that attachment and spreading of fibroblasts may be initiated without the involvement of integrins (e.g. fibronectin receptors) has been recently presented in (Curtis, 1987; Grinnell et al., 1988). Moreover, our results apparently do not contradict with mechanochemical explanation of processes underlying cell adhesion and motility (reviewed in Oster, 1989). From the above described experiments one can conclude that the role played by integrins or other adhesion-competent proteins may be simulated by a simpler molecule, that is biotinylated PE. In the latter case cells were modified with biotinylated lipid introduced into plasma membrane. Although these cells remain less spread than those subjected to biotinylation, they also flattened and attained epithelioid morphology (Fig. 9B). It seems unlikely that in this case plasma membrane proteins could be anchored to the substrate directly and thus facilitate the formation of polymerized actin bundles. Nevertheless, on the basis of experiments described we could assume that the driving force of spreading is high-affinity interaction of biotinylated plasma membrane components with avidin. The assembly of cytoskeleton is presumably a secondorder process initiated by cell attachment to the substrate and is the prerequisite for lamellae to be formed. It should be noted, in conclusion, that the results obtained indicate that loss of substrate dependence by ascites carcinoma cells is concomitant with the loss of functionally active adhesion- and spreading-competent receptors. The intercellular processes responsible for
190
BOGDAN(>V ET AL.
adhesion and for the attainment of normal epithelioid morphology are definitely not altered in the course of tumor progression of these cells.
LITERATURE CITED Bayer, E.A., Rivnay, B., and Skutelsky, E. (1979) On the mode of liposome-cell interactions. Biotin-conjugated lipids as ultrastructural probes. Biochim. Biophys. Acta, 550:464-473. Benecke, B.J., Ben-Ze’ev, A., and Penman, S. (1978) The control of mRNA production, translation and turnover in suspended and reattached anchorage-dependent fibroblasts. Cell, 14:931-939. Benecke, B.J., Ben-Ze’ev, A,, and Penman, S. (1980)The regulation of RNA metabolism in suspended and reattached anchorage-dependent 3T6 fibroblasts. J. Cell. Physiol., 103:247-254. Ben-Ze’ev,A. (1984) Differential control of cytokeratins and vimentin synthesis by cell-cell contact and cell spreading in cultured epithelial cells. J. Cell Biol., 99:1424-1433. Ben-Ze’ev,A., Farmer, S.R., and Penman, S. (1980) Protein synthesis requires cell-surface contact while nuclear events respond to cell shape in anchorage-dependent fibroblasts. Cell, 21:365-372. Brandley, B.K., Weisz, O.A., and Schnaar, R.L. (1987a) Cell attachment and long-term growth on derivatizable polyacrylamide surfaces. J. Biol. Chem., 262:6431-6437. Brandley, B.K., Ross, T.S., and Schnaar, R.L. (1987b) Multiple carbohydrate receptors on lymphocytes revealed by adhesion to immobilized polysaccharides. J. Cell Biol., 105:991-997. Chen, W.T., Hasegawa, T., Hasegawa, C., Weinstock, C., andYamada, K.M. (1985) Development of cell surface linkage complexes in cultivated fibroblasts. J. Cell Biol., 100:1103-1114. Curtis, A. (1987) Cell activation and adhesion. J. Cell Sci., 87:60% 611. Domnina, L.V., Rovensky, Yu.A., Vasiliev, Yu.M., and Gelfand, I.M. (1985)Effects of microtubule-destroying drugs on the spreading and shape of cultured epithelial cells. J. Cell Sci., 74:267-282. Farmer, S.R., Ben-Ze’ev,A., and Penman, S. (1978) Altered translatability of messenger RNA from suspended anchorage-dependent fibroblasts: reversal upon cell attachment to a surface. Cell, 15:627637. Folkman, J., and Greenspan, H.P. (1975) Influence of geometry on control of cell growth. Biochim. Biophys. Acta, 41 7:211-245. Folkman, J., and Moscona, A. (1978) Role of cell shape in growth control. Nature, 273:34L349. Green, N.M. (1963) The use of biotin-14Cfor kinetic studies and for assay. Biochem. J., 89585-591. Green, N.M. (1965) A spectrophotometric assay for avidin and biotin based on binding dyes by avidin. Biochem. J., 94:23c, 24c.
Grinnell, F., and Hays, D. (1978)Induction of active cell spreading by substratum adsorbed ligands directed against the cell surface. Exp. Cell Res., 116:175-190. Grinnell, F., Ho, C.H., and Tuan, T.L. (1988) Cell adhesion and phagocytosis promoted by monoclonal antibodies not directed against fibronectin receptors. J. Cell Sci., 90:201-214. Heggeness, M.H., and Ash, J.F. (1977) Use of the avidin-biotin comolex = ~ - - - - for the localization of actin and mvosin with fluorescent microscopy. J. Cell Biol., 73:783-788. Horwitz, A,, Duggan, K., Buck, C., Beckerle, M.C., and Burridge, K. (1986) .~,Interaction of olasma membrane fibronectin receDtor with talin: a transmembraie linkage. Nature, 320531-533. Hynes, R.O. (1987)Integrins: a family of cell surface receptors. Cell, 48.549-554. Laemmli, U.K. (1970) Cleavage of structural groteins during assembly of the head of bacteriophage T4. Nature, 227:680-685. Marcantonio, E.E., and Hynes, R.O. (1988) Antibodies to the conserved cytoplasmatic domain of the integrin PI subunit react with proteins in vertebrates, invertebrates and fungi. J. Cell Biol., 106:1765-1772. Margolis, L.B., Rozovskaya,I.A.,and Cragoe, E.J. (1988)Intercellular pH and cell adhesion to solid substrate. FEBS Lett., 234:449, 450. Markwell, M.A.K., and Fox, C.F. (1978)Surface-specificiodination of membrane proteins of viruses and eucariotic cells using 1,3,4,6tetrachloro-3u-6a-diphenylglycoluril. Biochemistry, 1 7:4807-4817. Nakache, M., Schrieber, A.B., Gaub, H., and McConnell, H.M. (1985) Heterogeneity of membrane phospholipid mobility in endothelial cells depends on cell substrate. Nature, 317:75-77. Oster, G. (1989) Cell motility and tissue morphogenesis. In: Cell Shape: Determinants, Regulation, and Regulatory Role. W.F. Stein and F. Bronner, eds. Academic Press, San Diego, pp. 33-61. Piantefaber, L.C., and Hynes, R.O. (1989) Changes in integrin receptors on oncogenically transformed cells. Cell, 56:281-290. Schnaar, R.L. (1984)Immobilized glycoconjugatesfor cell recognition studies. Anal. Biochem., 143:l-13. Schnaar, R.L., Weigel, P.H., Kuhlenschmidt, M.S., Lee, Y.C., and Roseman, S. (1978)Adhesion of chicken hepatocytes to polyacrylamide gels derivatized with N-acetylglucosamine. J. Biol. Chem., 253:7940-7951. Schwartz, M.A., Cragoe, E.J., Jr., and Lechene, C.P. (1990) pH regulation in spread cells and round cells. J. Biol. Chem., 265:13271332. Vasiliev. Y.M. 119851 Soreadine of transformed and non-transformed cells. Biochim. Biophis. Act; 780:21-65. Vlodavskv. I., Lui. G.M., and Gospodarowicz,D. (19801 MorDholom. amearance. erowth behavior and mimatorv activitv of human t;mor cells miintained on extracellula~matr~x ver8u8 plastic. Cell, 19:607-617. ~~
~~~
~
Restoration of Adhesive Potentials of Ehrlich Ascites Carcinoma Cells by Modification of Plasma Membrane ALEXEI A. BOCDANOV, Jr.,* LARISA V. GORDEEVA, BORIS A. BAIBAKOV, LEONID B. MARCOLIS, AND VLADlMlR P. TORCHlLlN Institute of Experimental Cardiology, USSR Cardiology Research Center, 12 7552 Moscow (A.A.B., V.P.T.) and A.N. Belozersky Laboratory of Molecular Biology and Bioorganic Chemistry, Moscow State University, 1 19899 Moscow (L.V.C., B.A.B., L.B.M.), USSR
A novel technique for modulating the spreading of ascites cells has been developed. Plasma membranes of Ehrlich ascites carcinoma cells were modified in two different ways: 1 ) biotin residues were covalently coupled to membrane components; 2) biotinylated lipid was introduced into plasma membranes. Adhesion and spreading of modified cells on avidin-coated substrates were studied and compared to those of non-modified cells. Both types of membrane alteration were shown to induce specific (biotin-dependent) interaction with immobilized avidin with resultant cell spreading. Spread cells attained cpithelioid-like morphology with the formation of wide thin lamellae, focal contacts with substrate, and circular actin bundles. The process of spreading was shown to be energy-dependent: it could be blocked by metabolic inhibitors and by low temperature. Formation of extended lamellae was prevented by preincubation of cells in the presence of cytochalasin B. The effects of metabolic poisons, low temperature, and microfilament-disruptive drugs were reversible and after the restoration of physiological conditions the cells resumed the spreading process. lmmunoprecipitation of biotinylated cell lysates with antiserum to cytoplasmic domain of P,-integrin subunit revealed a major 110 kD avidin-binding component. We conclude that lack of spreading of ascites carcinoma cells may be explained by the lack of functionally active adhesion- and spreading-competent cell-surface receptors, but may not be attributed to the defects in intracellular function or organization. lntracellular machinery of cell spreading is preserved in these ascites cells and could be turned on by cell attachment to the substrate via artificial adhesive site incorporated into plasma membrane.
Numerous intercellular processes and cell functions are dependent on a complex process of cell-substrate interaction, known as adhesion. Attachment and spreading of normal eucariotic cells onto solid substrate result in the initiation of mRNA translation (Benecke et al., 1978; Farmer et al., 1978) recovery of mRNA synthesis and cell-shape controlled expression of cytoskeletal genes (Benecke et al., 1980; Ben-Ze’evet al., 1980; 1984),alteration of fluidity of plasma membrane (Nakache et al., 1985), and activation of Na+/H+antiporters with the concomitant increase of intercellular pH (Margolis et al., 1988; Schwartz et al., 1990). As a result, cellular proliferation (Folkman and Greenspan, 1975; Folkman and Moscona, 1978) and migratory activity (Vlodavskyet al., 1980)are potentiated. Therefore, it is reasonable to consider solid substrate as a complex “growth/activation factor” for many different cells. Various transformed cells exhibit partial or complete release of cell proliferation from anchorage dependence. This is accompanied by the reduced ability of transformed cells to spread on a solid substrate. Ascites cells may serve as an extreme example of transformed 0 1991 WILEY-LISS, INC.
cells which have lost completely the anchorage dependent proliferation. For example, Ehrlich ascites carcinoma cells do not adhere tightly to either clean or t o serum-coated glass. The inability of these cells to adhere and to spread on a solid substrate may be explained in two different ways. On the one hand, transformation may affect the level of expression of integrin subunits involved in cell-substrate adhesion and spreading (Piantefaber and Hynes, 1989). On the other hand, cytoskeletal genes may be expressed in an aberrant way as a result of transformation. In the latter Received June 21,1990;accepted November 15, 1990. *To whom correspondence should be addressed. Alexei A. Bogdanov is now a t Webster Center for Biological Sciences Amherst College, Amherst MA 01002. Abbreviations used: BAH-PE, N-[(6-biotinylamido)hexanoyll phosphatidylethanolamine dioleoyl; HBS, Hepes buffered saline; HBSS, Hanks’ balanced salts saline; NHS-LC-B, N-hydroxysulfosuccinimidyl(6-biotiny1amido)hexanoate; NHS-SS-B, N-hydroxysulfosuccinimidyl 2-(biotinamido~ethyl-1,3-dithiopropionate; SEM, scanning electron microscopy.
ADHESION OF ASCITES CELLS
case the assembly of normal cytoskeleton necessary for cell spreading is altered (Vasiliev, 1985). Several model systems designed for controlled alteration of cell adhesion and spreading have been reported (Folkman and Moscona, 1978; Schnaar, 1984). For example, coating of plastic with increasing amounts of poly(2-hydroxyethy1)methacrylic acid results in progressive decrease of spreading of cultured cells (Folkman and Moscona, 1978). Several artificial polyacrylamide supports containing charged groups (Brandley et al., 1987a) or carbohydrates (Brandley et al., 1987b; Schnaar et al., 1978)were shown to increase hepatocyte and lymphocyte adhesion. It was demonstrated that plastic substrate covered with polyclonal (Grinnell and Hays, 1978) or monoclonal antibodies to cell surface proteins (Grinnell et al., 1988)promote attachment and spreading of BHK cells. However, none of the substrates reported so far were used for modulation of attachment and spreading of ascites cells. We also are not aware of any attempts to modify the surface of these cells to facilitate adhesion. Here we demonstrate, that modification of plasma membrane of Ehrlich ascites carcinoma cells with adhesion-promoting molecules not only facilitates the adhesion to a model substrate, but also initiates normal cell spreading, which involves rearrangement of the cytoskeleton, organization of focal contacts, and attainment of epithelioid morphology. MATERIALS AND METHODS Model substrates Avidin-coated glass coverslips were prepared as follows: coverslips were treated by 4% HF in 50% HNO, for several seconds, washed in water, and dehydrated in acetone. A solution of 1,l'-carbonyldiimidazole in dimethylformamide (100 mg/ml) was applied to coverslips and reaction proceeded for 1 h in a inert atmosphere with stirring. After washing with cold water coverslips were treated overnight with a solution of egg white avidin (Sigma, 2 mg/ml) in 15 mM sodium tetraborate, pH 8.0. Coverslips were washed extensively in 10 mM Hepes, 0.15 M NaC1, pH 7.5 (HBS) and used the same day. Alternatively, avidin-coated coverslips were prepared by avidin attachment to the biotinylated glass. Glass coverslips were treated with 7% 3-(triethoxysily1)-propylamine (Merck) in benzene (1.5 h at SOT) and washed with n-hexane. The coverslips were incubated in a solution of N-hydroxysulfosuccinimidyl(6biotinylamido) hexanoate (NHS-LC-B, Pierce) (1 mgi ml, in 20 mM Hepes, 0.15 M NaC1, pH 7.5) for 2 h at 20°C. After extensive washing with water coverslips were incubated in avidin solution (1.5 mg/ml in Hanks' balanced salts solution (HBSS). Other types of substrates were prepared by immobilization on carbonyldiimidazole-activatedglass of bovine serum albumin (fraction V, Sigma), 10% fetal calf serum (Gibco), or poly-L-lysine (100 kD, Sigma), respectively. Biotinylation of cells Ehrlich ascites carcinoma cells, propagated in C3HA mice, were washed from ascites fluid with HBSS and suspended at concentration of 1 x lo7 celldm1 in 20 mM
183
Hepes, 0.15 M NaC1, and 7 mM glucose (pH 7.4). Suspended cells (0.5 ml) were mixed with the equal volume of 1 mgiml NHS-LC-B or 1 mgiml N-hydroxysulfosuccinimidyl 2-(biotinamido)ethyl-1,3-dithiopropionate (NHS-SS-B, Pierce) in HBS and incubated for 20 min at 4°C. Five to six washings of modified cells were performed by sedimentation through 30% FicollPlaque (Pharmacia) solution in HBSS (2 min at 8OOg). Cells were resuspended in HBSS and used in further experiments. Alternatively biotinylated lipid N-[(G-biotiny1amido)hexanoyl] phosphatidylethanolamine dioleoyl (BAHPE, prepared as in (Bayer et al., 1979) was introduced into plasma membranes of cells. Then 10 p,l of BAH-PE solution (1 mgiml in ethanol) were added to 1 ml of HBSS (Ca, Mg free, Gibco), containing 5 x lo5 cells. After incubation for 30 min at 4"C, cells were washed 2 times by centrifugation through 30% Ficoll-Paque in HBSS (2 min at 8OOg).Washed cells were susgended in serum-free MEM without phenol red at 4 x 10 cellslml. Incubation conditions In different experiments 3-5 x lo5 cellsiml were added in 50 p,l aliquots t o coverslips (8x8 mm) and incubated for 15-180 min at 37°C in humid chambers. In some ex eriments sodium azide (1.3 mg/ml in glucose-free H SS) or cytochalasin B (10 pgiml), respectively, were added in the incubation mixture. All the additives were present in washing solutions throughout the experiments. In control experiments cells were incubated with substrates coated with BSA or with avidin-coated substrates preincubated with free biotin (1mgiml). In some experiments cells were incubated at 4°C. Cells modified with NHS-SS-B were treated with 2 mM cysteine in MEM, pH 7.6, at different stages of incubation. After 4 washings with HBSS, cells were examined by light, phase-contrast, interference-reflection microscopies either live or fixed with 4% formaldehyde in PBS. Microscopy For fluorescent microscopy biotinylated cells were stained with FITC-avidin (10 pg/ml prepared as in Heggeness and Ash (1977). To visualize polymerized actin, formaldehyde-fixed cells were extracted with 1% Triton X-100 in PBS, fixed with 2% formaldehyde, and stained with TRITC-labeled phalloidin (2 pg/ml). Fluorescence and interference-reflection images were observed using an Opton Photomicroscope I11 equipped with HBO mercury lamp and standard glass filters. Scanning electron microscopy (SEM) was performed on a PSEM 500 (Phillips) instrument operated at 35 kV. Samples were fixed with 2.5% glutaraldehyde in PBS, washed, dehydrated in acetone series, dried at C02 critical point, and shadowed with platinum. Preparation of cell lysates, electrophoresis, a n d blotting A total of 1 x lo7 biotinylated or control cells were lysed in 1%Nonidet NP-40,lO mM Tris, 1mM EDTA, 1 mM EGTA, 50 pg/ml pancreatic trypsin inhibitor, 2 mM PMSF (pH 7.5) for 5 min on ice and passed several times through a blue pipette tip (Gilson). Nuclei were separated at 1,0008-for 5 min. Lysate was treated with
i
184
BOGDANOV ET AL.
A
B
2 209 7*
60-
45-
Fig. 1. Staining of biotinylated ascites carcinoma cells with FITCavidin in suspension. Suspended cells were stained with FITC-avidin (20 Fgirnl) for 5 min at 4°C. A Fluorescent. B Phase-contrast images. Bar = 10 pm.
5% trichloroacetic acid, sediment was collected by centrifugation and redissolved in 0.5% SDS in water. Protein concentration was determined using BCA protein assay reagent (Pierce) according to manufacturer's recommendations. SDS-polyacrylamide gel electrophoresis was run in 7.5-15% gradient gels or in 10% gels using Laemmli discontinuous buffer system (Laemmli, 1970). Gels were electroblotted on nitrocellulose sheets (0.2 p,m, Schliecher and Schull) and biotinylated proteins were visualized by ABC-kit (Vector).Alternatively, radioiodinated avidin was used for autoradiographical visualization as described below. Immunoprecipitation A total of 5-106 x lo7 cells were lysed and lysates obtained were immunoprecipitated with the use of antiserum to pl-integrin subunit (a generous gift of Dr. R.O. Hynes) as described in Marcantonio and Hynes (1988). Proteins eluted from protein A-Sepharose were separated by electrophoresis and electroblotted as described before. Nitrocellulose membranes were blocked
18-
Fig. 2. Visualization of avidin-binding proteins in modified ascites carcinoma cells. A Cells were lysed and proteins were separated by SDS electrophoresis in 10%' polyacrylamide gels. Gels were electroblotted on nitrocellulose. Biotinylated proteins were detected by incubation with ['251]-labeledavidin and visualized by autoradiography. B: Cell lysates were immunoprecipitated with anti+, antiserum (5 pgiml) and protein A-Sepharose.Eluted proteins were separated by SDS-electrophoresisin 10%polyacrylamide gels, electroblotted, incubated with [12511-avidin,and subjected to autoradiography. To the left indicated are the positions of molecular weight markers (kD). Details are given in Materials and methods.
by 3% gelatin, 1%BSA in PBS (pH 7.5) overnight, washed, and incubated with [1251]-labeledavidin (labeling was performed as described in Markwell and Fox (1978), 2 ~ 1 cpmiml 0 ~ in 0.1% Tween-20, 1% BSA in PBS (pH 7.5). Extensively washed and dried membranes were exposed to XAR-5 film (Kodak) without intensifying screen for 3-6 days. Determination of biotin in cell lysates Biotin content in cell lysates was determined by colorimetric method (Green, 1965),exactly as described in Pierce manual to NHC-LC-B (No. 21335).
ADHESION OF ASCITES CELLS
185
Fig. 3. Interaction of ascites cells with non-derivatizedglass substrate. A: Phase contrast image of cells adsorbed on glass, bar= 10 km. B SEM micrograph of attached cell; bar= 1 pm.
RESULTS Ehrlich ascites carcinoma cells were modified by water-soluble membrane-impermeant activated biotin derivative (NHS-LC-B). Biotin residues on the cell surface were detected by staining of modified cells with FITC-labeled avidin. All biotinylated cells exhibited fluorescent staining (Fig. 1). In the presence of free biotin no fluorescent avidin was bound to cells. Biotinylated proteins were identified by radioiodinated avidin or avidin-peroxidase on blots of total cell lysates, separated by SDS-electrophoresis. One ma’or 110 kD band and five minor ones with molecu ar masses of approximately 80, 66, 62, 42, and 28 kD, respectively, were visualized (Fig. 2A). We performed immunoprecipitation of cell lysates with rabbit antiserum t o p,-subunit of integrin (Markantonio and Hynes, 1988). Antibody-bound biotinylated proteins were detected by autoradiography using radioiodinated avidin as a probe. We found that the major 110 kD band was at least partially composed of biotinylated integrin subunits. Though cell lysis and immunoprecipitation were carried out in the presence of divalent cations, no other biotinylated integrin subunits of higher molecular weight were found to associate with isolated subunits (Fig. 2B). To examine the amount of biotin residues introduced by covalent modification, cell lysates were treated with pronase and protein fragments were tested for 4’-hydroxyazobenzene-2-carboxylic acid displacement from the active center of avidin. In most experiments approximately 160 pg of biotin were coupled per pg of total cell protein. Viability of modified cells exceeded 99% as determined by trypan-blue exclusion test. Morphology of cells was studied by phase-contrast and scanning electron microscopy. Non-modified carcinoma cells were washed to remove ascites fluid by HBSS and cell suspension was incubated over a glass coverslip at 37°C for up to 180 min. A t the end of incubation period cells were visualized on a substrate. However, after washings only few cells remained on the glass. In phase-contrast micrographs these cells appear to be
!
spherical (Fig. 3A). SEM examination of the same specimen showed that those cells which remain on the substrate are anchored by cytoplasmic protrusions (Fig. 3B). If the suspension of washed cells was incubated with the serum-covered coverslip, the number of cells adhering to glass was even lower, than in the case of clean (acid-washed) glass. Similar results were obtained when a clean glass coverslip was used as substrate and serum was added to the incubation mixture. In contrast to results of experiments carried out with non-modified Ehrlich carcinoma cells, the number of adhered cells significantly increased if biotinylated cells were allowed to attach to avidin-coated glass. Shortly after the round suspended cells start to attach to the substrate they begin to spread: in 30 min at 37°C thin transparent lamellae were formed. In 1.5-2 h after the incubation had been started, 80-90% of attached cells become spread. The morphology of cells varied: some cells exhibited spherical central (perinuclear) areas and flat circular lamella. In some cells perinuclear area was more spread and lamella was narrow (Fig. 4A), Occasional elongated (polarized) cells were also present. Usually they were restricted from both sides by adjacent cells. Some preparations contained cells of several different morphologies. Examination of cells spread on avidin-coated glass by SEM revealed wide smooth thin circular lamellae and numerous microvilli in the central region of cells with wide thin lamellae (Fig. 4B). The latter were smooth and free of microvilli. The modified cells attached to the avidin-coated coverslips remain spread in HBSS another 3-5 h. If the solution was changed for the serum-supplemented media, cells proliferate remaining attached and spread for at least another 36 h. Several kinds of control experiments have been performed to prove the actual involvement of biotin-avidin interaction in cell spreading as described below. In corresponding control experiments biotinylated cells were incubated either with non-modified glass, or with BSA-covered glass. In 1-3 h of incubation at 37°C cells attached to substrate, but did not spread. Such
186
BOGDANOV ET AL.
Fig. 4. Spreading of biotinylated cells on avidin-coated glass coverslips. Cells (5 x lo6 cellsiml) were treated with 1mg/ml NHS-LC-B in 20 mM Hepes, 0.15 M NaC1,7 mM glucose (pH 7.4) for 20 min at 4"C, washed, and incubated over avidin immobilized on glass coverslip. A Phase contrast micrography;bar= 10 pm.B SEM micrograph of attached cells; bar= 1 pm.
cells remained spherical if avidin-coated coverslips were preincubated with the excess of free biotin before the addition of cells (Fig. 5). In both cases cells remained round-shaped after prolonged incubation, with only a few percent (1-3% of the total cell number) weakly spread. Non-modified cells when applied to avidin-coated coverslips remained spherical for 1-3 h of incubation period. On the polylysine-coated glass a fraction of non-modified cells developed small fan-shaped lamellae. However, perinuclear area of cells always remained spherical and these cells have never achieved epithelioid shape. A total of 95-97% of non-modified cells adhered to avidin-coated glass did not spread at all. The same is true for biotinylated ones incubated over immobilized polylysine. To find out, whether biotin residues on cell surface directly keep the cells spread on avidinylated substrate, cells were modified with NHS-SS-B. This reagent contains disulfide bridge, which can be easily cleaved with thiol-containing membrane-impermeant agents in physiological conditions. When modified cells attached to avidin-coated glass (in 10 min of incubation), cysteine was added to reduce disulfide bonds. None of adherent cells were spread after the addition of cysteine. The amount of adherent cells was noticeably reduced. In control experiment modified cells, if allowed to spread on avidin-coated glass, attained epithelioid morphology in spite of the presence of cysteine. The involvement of biotinylated plasma membrane com onents in the formation of contacts with immobilize avidin have been proved by staining with FITCavidin. Modified cells were intensively stained in suspension or shortly after attachment to glass (Fig. 1). However, after spreading they do not bind avidin any more and remain non-stained with fluorescent avidin.
B
Fig. 5. Attachment of biotinylated cells to avidin-coated glass, preincubated in the presence of free biotin (1 mg/ml). Bar= 10 pm.
Spreading of biotinylated cells on immobilized avidin at 4°C or in the presence of sodium azide or cytochalasin B was tested. After incubation for 1-2 h at 4°C the cells remained spherical. The same was true when azide was present for 50-60 min. The inhibition of spreading was reversible: 1 h after attached cells were transferred to 37°C they began to flatten. The same was true when NaN,containing solution was replaced for HBSS. In this case SEM revealed ruffles and well-developed lamellae on the edges of cells in 1.5 h after the restoration of physiological conditions. The addition of cytochalasin B to suspended biotinylated cells inhibited the formation of wide lamellae after
ADHESION OF ASCITES CELLS
187
Fig. 6. Effect of cytochalasin B on spreading of biotinylated cells on avidin. A, C, D: Phase-contrast images. B: SEM image. A: Attachment of cells to avidin in the presence of 10 pgiml of cytochalasin B; bar= 10 km. B: SEM image of cell presented in A; bar= 1 pm. C: Cells pretreated with cytochalasin B and then allowed to spread in drugfree media; bar= 10 pm.D: Effect of cytochalasin B on biotin-modified cells spread on avidin-coated coverslip. Cytochalasin B was added after cells were spread.
their attachment to immobilized avidin. Instead only long thin spikes were formed for 50-60 min of incubation at 37°C (Fig. 6A,B). When cytochalasin-treated cells were transferred to drug-free media numerous small lamellae were detected around the long protrusions; the central area of many cells was flattened (Fig. 6C). Treatment with cytochalasin B also affected already spread cells. One hour after addition of drug, cellular lamellae retracted and cells became arborized (Fig. 6D). Biotinylated cells spread on avidin were studied by interference reflection microscopy. The majority of cells had a ring of focal (black) contacts with substrate alongside the external edge. Focal contacts were located in lamellar regions of cells (Fig. 7A,B). To study the morphology of the actin cortex of the spread ascites cells, fixed cells after Triton extraction were stained by fluorescent labeled phalloidin. A cir-
cular microfilament bundle was visualized alongside the margin of many spread cells (Fig. 8). An alternative way of plasma membrane modification with biotin residues was tested. Suspended ascites carcinoma cells were incubated with biotinylated lipid (BAH-PE) at 4°C. Staining with FITC-avidin revealed biotinylated lipid uniformly incorporated andlor attached to plasma membranes (Fig. 9A). Suspended cells were washed and tested for the ability to spread on immobilized avidin. In 10-15 min at 37°C cells attached to substrate and developed well-defined epithelioid morphology (Fig. 9B). In the case of biotin-PE-mediated adhesion the cells remain less spread than when the total cell surface was biotinylated. DISCUSSION Mouse ascites carcinoma cells proliferate in suspension and attach only weakly to glass and cultural
188
BOGDANOV ET AL.
Fig. 7. Formation of focal contacts with immobilized avidin. A Interference-reflection image. B: Phase-contrast image of the same field. Bar= 5 pm.
Fig. 8. Visualization of polymerized actin in spread ascites cells. Cells were extracted with Triton X-100, fixed with 2 6 formaldehyde and stained with TRITC-phalloidin (2 pgiml).
plastic. In contrast to normal epithelia from which these cells have been derived 70 years ago they are unable to spread on these surfaces. The reason for the lack of the features which are typical for normal phenotype is not clear. The main question to be answered is whether ascites carcinoma cells lack functionally active receptors specific for spreading-competent molecules, or the defects in spreading are caused by some alterations of the intracellular machinery. We attempted to discriminate between these two possibilities by modification of cell membrane with low-molecular weight molecules which should promote cell adhesion to specially prepared model substrate. In our experiments we used glass coated with immobilized avidin as a substrate for cells, subjected to mild modi-
fication with either membrane-impermeant activated biotin analogs or with biotinylated lipid, artificially inserted into plasma membrane. Both types of plasma membrane modification dramatically increased the number of cells attached to avidin-coated substrate. This is a trivial result which can be predicted taking into account high-affinity binding of biotin molecules with avidin (K, of avidin-biotin complex have been reported as M) (Green, 1963).What is non-trivial is that cell surface modification also promoted cell spreading with the development of epithelioid morphology. Involvement of avidin-biotin interactions in ascites cell spreading was proved in control experiments, where non-modified cells were used or when biotinbinding sites of immobilized avidin were blocked with the excess of free biotin. Thus, the adhesion and spreading of modified ascites cells is not determined solely by electrostatic interactions due to the difference in charge between cationic avidin (PI=10.5; Green, 1965) and cells, bearing a net negative charge. This conclusion was confirmed by experiments with positively charged poly-L-lysine-coatedglass. The majority of cells attached to this substrate were spherical and only some of them revealed cytoplasmic processes. However, even these cells have never developed epithelioid mor hology. When kotin was linked to plasma membrane components with thiol-cleavable linker, spreading process could be blocked at the early stages of development by membrane-impermeant cysteine. Thus, biotin, either coupled to plasma membrane components (mainly t o 110-kD protein(s) or incorporated into the membrane via the lipid anchor, can serve as an “artificial receptor,” which stimulates changes in cell behavior with respect to the model adhesive substrate. The modified cells remain alive as tested with dyeexclusion test. They also remain alive after the spread-
ADHESION OF ASCITES CELLS
189
Fig. 9. Ascites cells modified by preincubation with biotinylated lipid. A: Fluorescent micrograph of suspended cells, stained with FITC-avidin (20 pgirnl). B: Spreading of modified cells on avidin, a phase-contrast image. Bar= 10 km.
ing for another few hours if they remain in salt solution (that is approximately the period other cultures can survive in similar conditions). If the solution is changed for a serum-supplemented culture media the cells remain spread on the substrate and look morphologically normal for at least another 36 hours. However, at that time we did not check whether the cells are still bound to the substrate by biotin-avidin linkage. The very process of cell spreading is also evidence that cells are alive. However, to exclude the possibility of non-physiological spreading (e.g., energy-independent flattening of cells without the involvement of intracellular structures), we incubated modified cells with immobilized avidin in the presence of inhibitor of respiration (sodium azide), a microfilament-disruptive drug (cytochalasin B) or at low temperature. All these experiments have demonstrated the requirement of energy and microfilament reorganization in biotinmodulated cell spreading. The effects of sodium azide, cytochalasin B, and low temperature were reversible, since cells resumed with spreading when transferred to physiological medium or if warmed to 37°C. By staining of filamentous actin in spread cells were have demonstrated circular organization of microfilament bundles characteristic of normal epithelial cells (Domnina et al., 1985).The process of cell spreading is hypothesized to be governed by cytoskeletal rearrangements caused by the interactions of cell-surface receptors (integrins) (Hynes, 1987) with matrix components. These receptors supposedly interact with the cytoskeleton via cytoplasmic C-terminal domain (Chen et al., 1985; Marcantonio and Hynes, 1988). Alternatively, they are able to form a linkage with cytoskeletal elements (including actin bundles) via transmembrane proteins such as talin (Horwitz et al., 1986). In the case of ascites cells the lack of cell spreading may be due t o the defects in integrin structure or in the organization of its subunits. Related phenomenon was observed for some transformed cells, which exhibit reduced ability to spread if compared to their normal counterparts (Piantefaber
and Hynes, 1989). We found that antibodies to cytoplasmic domain of PI integrin subunit (Marcantonio and Hynes, 1988) precipitate a biotinylated 110 Kd protein. Thus, we may hypothesize that biotinylation of cell surface proteins leads to modification of P1-subunits, which initiates spreading when bound to avidincoated substrate. However, we do not exclude that biotinylated proteins other than integrin subunit may provide some intracellular signal which activate adhesion of modified ascites cells since there are at least five major bands visualized with '"I-labeled avidin (Fig. 2A). Evidence showing that attachment and spreading of fibroblasts may be initiated without the involvement of integrins (e.g. fibronectin receptors) has been recently presented in (Curtis, 1987; Grinnell et al., 1988). Moreover, our results apparently do not contradict with mechanochemical explanation of processes underlying cell adhesion and motility (reviewed in Oster, 1989). From the above described experiments one can conclude that the role played by integrins or other adhesion-competent proteins may be simulated by a simpler molecule, that is biotinylated PE. In the latter case cells were modified with biotinylated lipid introduced into plasma membrane. Although these cells remain less spread than those subjected to biotinylation, they also flattened and attained epithelioid morphology (Fig. 9B). It seems unlikely that in this case plasma membrane proteins could be anchored to the substrate directly and thus facilitate the formation of polymerized actin bundles. Nevertheless, on the basis of experiments described we could assume that the driving force of spreading is high-affinity interaction of biotinylated plasma membrane components with avidin. The assembly of cytoskeleton is presumably a secondorder process initiated by cell attachment to the substrate and is the prerequisite for lamellae to be formed. It should be noted, in conclusion, that the results obtained indicate that loss of substrate dependence by ascites carcinoma cells is concomitant with the loss of functionally active adhesion- and spreading-competent receptors. The intercellular processes responsible for
190
BOGDAN(>V ET AL.
adhesion and for the attainment of normal epithelioid morphology are definitely not altered in the course of tumor progression of these cells.
LITERATURE CITED Bayer, E.A., Rivnay, B., and Skutelsky, E. (1979) On the mode of liposome-cell interactions. Biotin-conjugated lipids as ultrastructural probes. Biochim. Biophys. Acta, 550:464-473. Benecke, B.J., Ben-Ze’ev, A., and Penman, S. (1978) The control of mRNA production, translation and turnover in suspended and reattached anchorage-dependent fibroblasts. Cell, 14:931-939. Benecke, B.J., Ben-Ze’ev, A,, and Penman, S. (1980)The regulation of RNA metabolism in suspended and reattached anchorage-dependent 3T6 fibroblasts. J. Cell. Physiol., 103:247-254. Ben-Ze’ev,A. (1984) Differential control of cytokeratins and vimentin synthesis by cell-cell contact and cell spreading in cultured epithelial cells. J. Cell Biol., 99:1424-1433. Ben-Ze’ev,A., Farmer, S.R., and Penman, S. (1980) Protein synthesis requires cell-surface contact while nuclear events respond to cell shape in anchorage-dependent fibroblasts. Cell, 21:365-372. Brandley, B.K., Weisz, O.A., and Schnaar, R.L. (1987a) Cell attachment and long-term growth on derivatizable polyacrylamide surfaces. J. Biol. Chem., 262:6431-6437. Brandley, B.K., Ross, T.S., and Schnaar, R.L. (1987b) Multiple carbohydrate receptors on lymphocytes revealed by adhesion to immobilized polysaccharides. J. Cell Biol., 105:991-997. Chen, W.T., Hasegawa, T., Hasegawa, C., Weinstock, C., andYamada, K.M. (1985) Development of cell surface linkage complexes in cultivated fibroblasts. J. Cell Biol., 100:1103-1114. Curtis, A. (1987) Cell activation and adhesion. J. Cell Sci., 87:60% 611. Domnina, L.V., Rovensky, Yu.A., Vasiliev, Yu.M., and Gelfand, I.M. (1985)Effects of microtubule-destroying drugs on the spreading and shape of cultured epithelial cells. J. Cell Sci., 74:267-282. Farmer, S.R., Ben-Ze’ev,A., and Penman, S. (1978) Altered translatability of messenger RNA from suspended anchorage-dependent fibroblasts: reversal upon cell attachment to a surface. Cell, 15:627637. Folkman, J., and Greenspan, H.P. (1975) Influence of geometry on control of cell growth. Biochim. Biophys. Acta, 41 7:211-245. Folkman, J., and Moscona, A. (1978) Role of cell shape in growth control. Nature, 273:34L349. Green, N.M. (1963) The use of biotin-14Cfor kinetic studies and for assay. Biochem. J., 89585-591. Green, N.M. (1965) A spectrophotometric assay for avidin and biotin based on binding dyes by avidin. Biochem. J., 94:23c, 24c.
Grinnell, F., and Hays, D. (1978)Induction of active cell spreading by substratum adsorbed ligands directed against the cell surface. Exp. Cell Res., 116:175-190. Grinnell, F., Ho, C.H., and Tuan, T.L. (1988) Cell adhesion and phagocytosis promoted by monoclonal antibodies not directed against fibronectin receptors. J. Cell Sci., 90:201-214. Heggeness, M.H., and Ash, J.F. (1977) Use of the avidin-biotin comolex = ~ - - - - for the localization of actin and mvosin with fluorescent microscopy. J. Cell Biol., 73:783-788. Horwitz, A,, Duggan, K., Buck, C., Beckerle, M.C., and Burridge, K. (1986) .~,Interaction of olasma membrane fibronectin receDtor with talin: a transmembraie linkage. Nature, 320531-533. Hynes, R.O. (1987)Integrins: a family of cell surface receptors. Cell, 48.549-554. Laemmli, U.K. (1970) Cleavage of structural groteins during assembly of the head of bacteriophage T4. Nature, 227:680-685. Marcantonio, E.E., and Hynes, R.O. (1988) Antibodies to the conserved cytoplasmatic domain of the integrin PI subunit react with proteins in vertebrates, invertebrates and fungi. J. Cell Biol., 106:1765-1772. Margolis, L.B., Rozovskaya,I.A.,and Cragoe, E.J. (1988)Intercellular pH and cell adhesion to solid substrate. FEBS Lett., 234:449, 450. Markwell, M.A.K., and Fox, C.F. (1978)Surface-specificiodination of membrane proteins of viruses and eucariotic cells using 1,3,4,6tetrachloro-3u-6a-diphenylglycoluril. Biochemistry, 1 7:4807-4817. Nakache, M., Schrieber, A.B., Gaub, H., and McConnell, H.M. (1985) Heterogeneity of membrane phospholipid mobility in endothelial cells depends on cell substrate. Nature, 317:75-77. Oster, G. (1989) Cell motility and tissue morphogenesis. In: Cell Shape: Determinants, Regulation, and Regulatory Role. W.F. Stein and F. Bronner, eds. Academic Press, San Diego, pp. 33-61. Piantefaber, L.C., and Hynes, R.O. (1989) Changes in integrin receptors on oncogenically transformed cells. Cell, 56:281-290. Schnaar, R.L. (1984)Immobilized glycoconjugatesfor cell recognition studies. Anal. Biochem., 143:l-13. Schnaar, R.L., Weigel, P.H., Kuhlenschmidt, M.S., Lee, Y.C., and Roseman, S. (1978)Adhesion of chicken hepatocytes to polyacrylamide gels derivatized with N-acetylglucosamine. J. Biol. Chem., 253:7940-7951. Schwartz, M.A., Cragoe, E.J., Jr., and Lechene, C.P. (1990) pH regulation in spread cells and round cells. J. Biol. Chem., 265:13271332. Vasiliev. Y.M. 119851 Soreadine of transformed and non-transformed cells. Biochim. Biophis. Act; 780:21-65. Vlodavskv. I., Lui. G.M., and Gospodarowicz,D. (19801 MorDholom. amearance. erowth behavior and mimatorv activitv of human t;mor cells miintained on extracellula~matr~x ver8u8 plastic. Cell, 19:607-617. ~~
~~~
~
