Journal of Cell Science 101, 873-883 (1992)Printed in Great Britain © The Company of Biologists Limited 1992

873

Initiation of HeLa cell adhesion to collagen is dependent upon collagen

receptor upregulation, segregation to the basal plasma membrane,

clustering and binding to the cytoskeleton

MICHAEL L. LU*, RICHARD J. McCARRON and BRUCE S. JACOBSONt

Department of Biochemistry and the Graduate Program in Molecular and Cellular Biology, University of Massachusetts, Amherst,MA 01003, USA

•Present address: Brigham and Women's Hospital, Harvard Medical School, 75 Francis St., Boston, MA 02115, USAtAuthor for correspondence

Summary

It was recently reported that HeLa cells have three Arg-Gly-Asp-dependent collagen receptors that do notappear to be in the integrin family of extracellularmatrix receptors and bind to either type I or IV collagenor to type I gelatin. It was our goal to determine howthese receptors function in HeLa cell-substratum ad-hesion. We report here that the sequence of events bywhich the receptors mediate adhesion to collagen orgelatin is: (1) induction of cell attachment by specificcollagen receptor-substratum interactions with culturedishes covalently coated with either type I collagen orgelatin - attachment is inhibited by soluble gelatin;(2) stabilization of attachment by exocytotic upregula-tion of the receptors to the basal plasma membrane,which was demonstrated by analyzing, during celladhesion, the redistribution of the collagen receptorsamong the apical plasma membrane exposed to theculture medium, the basal plasma membrane contactingthe culture dish, and an intracellular pool of plasmamembrane vesicles; (3) the initiation of cell spreading by

receptor clustering and cytoskeletal association. Cellspreading is a threshold effect with regard to the surfaceconcentration of gelatin, indicating that collagen recep-tor clustering is a precondition to the onset of spreading.Observations consistent with this interpretation of thethreshold effect are that cells attach but spread moreslowly on a substratum that retards receptor clustering,and that collagen receptors, when viewed by immunoflu-orescence microscopy, form a punctate pattern offluorescence in the basal plasma membrane during cellspreading. It is also shown that more collagen receptorsco-isolate with nondenaturing detergent-stable cytoskel-etal preparations after the collagen receptors have beeneither clustered by antibodies or gelatin in solution, orby a collagen matrix. This indicates that clusteringdrives the receptors to bind to the cytoskeleton and is anecessary step in the transition from cell attachment tocell spreading.

Key words: cell adhesion, collagen receptors, HeLa cells.

Introduction

Adhesion of cells to an extracellular matrix (ECM) isintimately involved in cell migration, metastasis andtissue development (e.g. see Burridge et al., 1988; Buckand Horwitz, 1987; Juliano, 1987; Thompson-Pletscher,1986). An in vitro model of these normal and aberrantfunctions is the adhesion of cells to culture dishescoated with extracellular matrix components. Thegeneral events that are thought to occur duringadhesion of cells in vitro are: first, an initial attachmentphase; second, an intermediate phase where the cellsspread or migrate by forming adhesion zones betweenthe substratum and the cell; and third, a final phasewhere the cells either detach or the cell-substratumadhesion zones become more elaborate or are re-modeled by incorporating existing or newly synthesized

materials (cf. LeBaron et al., 1988; Singer et al., 1987,1988; and reviews by Buck and Horwitz, 1987; Juliano,1987; Burridge et al., 1988; Lark et al., 1985; Geiger,1983; Geiger et al., 1984; Rollins et al., 1982; Culp etal., 1986; Rees et al., 1977).

Our aim has been to focus on the function of HeLacell collagen receptors in the first two phases of celladhesion, i.e. cell attachment and cell spreading. Wehave previously shown that HeLa cells have three Arg-Gly-Asp-dependent receptors for collagen that mediatecell-substratum adhesion but do not appear to be in theintegrin family of ECM receptors (Lu et al., 1989;Beacham and Jacobson, 1990). The receptors aredesignated by their molecular mass and that of anyprevalent proteolytic fragments (102/58 kDa, 87 kDaand 38/33 kDa). These cells are less complex than mostother cells used to study cell adhesion in that they do

874 M. L. Lu et al.

not make an extracellular matrix, which is thought to beinvolved in the formation of the more complexadhesion zones referred to as focal contacts andfibronexuses (Izzard and Lochner, 1976; Singer, 1982;Singer et al., 1987; Couchman et al., 1983; Fairman andJacobson, 1983; LeBaron et al., 1988). This is animportant aspect since it is difficult to separate, in time,the formation of the complex adhesion zones from theattachment and spreading events. Therefore, sinceHeLa cells do not form complex adhesion zones, we areable to dissect the sequence of events that take placeduring attachment and spreading.

The objectives of the work reported here were todetermine (1) whether the HeLa cell collagen receptorsact cooperatively to mediate firm cell attachment; (2)whether the collagen receptors segregate to the basalplasma membrane (PM) by diffusion from the apicalPM and/or are upregulated by exocytotic membraneflow from an internal pool of PM vesicles; (3) whetherHeLa cell spreading on collagen is a cooperativeprocess, and whether such cooperativity is due toclustering of the collagen receptors; and (4) whetherclustering collagen receptors induces them to associatewith the cytoskeleton to initiate cell spreading.

Materials and methods

Cell cultureSuspension cultures of HeLa-S3 (ATCC) cells were grown tomid-log phase (3-5 x 10s cells/ml) at 37°C in a humidified 5%CO2 incubator in RPM 1640 medium (K.C. Biologicals,Lenexa, KS) supplemented with 5% calf serum (Gibco,Grand Island, NY), 0.3% NaHCO3, 100 /ig/ml dihydrostrep-tomycin, 60 /

HeLa cell-substratum adhesion 875

at 970 g yielding a pellet containing the silica-coated apicalplasma membrane fragments and a supernatant containingthe cytoplasmic fraction. The cytoplasmic fraction was thencentrifuged with a microfuge at 14,500 g for 15 min to obtain afraction containing large intracellular vesicles termed "theinternal plasma membrane." The basal plasma membraneattached to the culture dish was washed two times with coldPBS and scraped into 2% SDS containing the proteaseinhibitor cocktail. The "internal PM" fraction was separatedfrom the cell surface PM the same way as that done with theapical PM and the internal membrane fraction, except theentire cell surface for the cells in suspension was coated withcolloidal silica and the cells were disrupted in a parr pressurebomb at 600 psi. All fractions were solubilized in the sameSDS-protease inhibitor cocktail. The samples were heated to90°C for 10 min and sonicated with a Branson sonifier(Branson Ultrasonics Co., Danbury, CT) set at 35 W andsonicated 3 times for 5 s durations. The microbeads and otherinsoluble materials were then removed by microfuging thesamples for 5 min.

Indirect immunofluorescence microscopyHeLa cells in suspension or after different times of spreadingon gelatin-coated culture dishes were fixed with 3% parafor-maldehyde in PBS, pH 7.4, for 30 min on ice. Cells weregently washed three times in PBS then incubated on ice for 5min in 0.1 M glycine in PBS, pH 7.4, to block residualformaldehyde groups. After one wash with PBS, the cellswere incubated on ice for 30 min in PBS containing 1% BSAand mouse primary antibodies to the collagen receptors. Cellswere then washed with PBS containing 1% BSA anddecorated with FITC-conjugated rabbit anti-mouse secondaryantibodies for 30 min on ice, after which the cells were washedextensively with PBS/BSA. The specimen samples werecovered by one drop of 3% (w/v) rc-propylgallate in 90% (v/v)glycerol.

Gel electrophoresis and electroblottingSDS-polyacrylamide gel electrophoresis (SDS-PAGE) wasperformed using a 4% stacking gel and an 8% resolving gel inthe presence of /J-mercaptomethanol. Proteins were electro-phoretically transferred from 8% slab gels to nitrocellulosesheets as described by Towbin et al. (1979). For staining withantibodies, the nitrocellulose blots were blocked with 150 mlNaCl, 50 mM Tris, pH 7.5, containing 5% non-fat dry milk(Carnation Co., Los Angeles, CA). The decoration ofprimary antibodies with alkaline phosphatase-conjugatedsecondary antibody was completed in the same buffer. Theunbound antibody was removed by washing extensively inwash buffer (150 mM NaCl, 50 mM Tris, pH 7.5, containing1% Triton X-100, 0.5% sodium deoxycholate and 0.1% SDS)and the color was developed by incubating in 5-bromo-4-chloro-3-indoyl phosphate p-toluidin salt (Sigma) as chromo-genic enzyme marker (Leary et al., 1983).

Results

HeLa cell adhesion to collagen-coated culture dishesfollows the same characteristic spreading morphology(Fig. 1) as that seen with the adhesion of most cells invitro. Round cells attach to the substratum and spreadby continually sending out new lamellipodia andfilopodia at the edges of the cell as the spaces betweenthe old lamellipodia and filopodia fill in with proto-plasm by a process called webbing (Rajaraman et al.,

Fig. 1. HeLa cell attachment and spreading on collagen-coated culture dishes. Scanning electron micrographs, (A,B and C). Light micrographs of hematoxylin-stained cells,(D, E and F). Cells were allowed to attach for 1 min (Aand D), and spread for 15 min (B and E), or 60 min (Cand F). Bars, 5 fjm (A,B), 10 /im (C) and 20 jan (D, Eand F).

1974; Knox, 1981). HeLa cells which attached to rat tailtype I collagen or gelatin were partially spread by 15min (Fig. 1, B and E) and fully spread by 60 min (Fig. 1,C and F). The collagen or gelatin is covalently bound tothe culture dish since cell attachment is reduced andattached cells do not spread if the collagen or gelatin isonly adsorbed to the dishes. Experiments showed thatadsorbed 125I-labeled gelatin is released from theculture dish during cell adhesion (data not shown).Desorption or removal of fibronectin from focaladhesion sites during fibroblast spreading has also beendemonstrated (Avnur and Geiger, 1981; Grinnell,1986). In addition, adsorbed fibronectin is evenremoved by antibodies to the surfaces of cells used inspreading assays to determine whether a particularantigen is involved in cell adhesion (Chen et al., 1985).For these reasons all experiments on cell attachmentand spreading reported here were done with covalentlybound substrata.

Specificity of cell substratum adhesion to variousphysical forms of collagenThe rate of HeLa cell attachment (Fig. 2) and spreading(Fig. 3) on gelatin or collagen is rapid and the

876 M. L. Lu et al.

100 -

Fig. 2. Kinetics of HeLa cell attachment to culture dishescoated with gelatin or a positively charged polymer. HeLacells were metabolically labeled with [3H]thymidine for 48h, suspended in HEPES-buffered RPMI-1640 (seeMaterials and methods) and plated onto gelatin-coatedculture dishes at 37°C (open squares) or 20°C (opencircles), or at 37°C onto culture dishes coated withpolyethyleneimine (open triangles). Attached cells weresolubilized in 0.1 M NaOH and quantitated in a liquidscintillation counter. The vertical bars are the standarderror of the mean for four separate determinations.

percentage of attached cells that spread is high.Generally, 85-95% of the attached cells spread (Fig. 3;cf. Mason and Jacobson, 1985). While attachment israpid on a positively charged substratum such aspolyethylenimine (PEI), cell spreading is significantlyslowed (Figs 2 and 3). Low temperature (20°C) alsosignificantly reduced cell attachment (Fig. 2) andinhibited cell spreading (data not shown). Interestingly,while cells attached and spread at 37°C on culture dishescovalently coated with gelatin at room temperature,cells attached but did not spread at 37°C on culturedishes covalently coated with gelatin at 45°C (Fig. 3). Itis possible that the gelatin bound to the culture dishes at45°C further denatures its cell binding conformation aswould be expected of gelatin coated on dishes at roomtemperature. Gelatin molecules above their criticalmelting temperature of 41°C are in an extendedconformation and do not have a tertiary structure asdoes gelatin below its critical melting temperature.Below 41°C even gelatin has alpha helices, although thehelix is not the same as the quaternary, triple-helicalstructure of collagen (cf. Traub and Piez, 1971).

Cell attachment and spreading on gelatin are thresholdphenomenaA convenient way to ascertain if a receptor-ligandinteraction might be a cooperative reaction is todetermine whether it has a threshold or criticalconcentration below which the reaction does not take

8 0

Fig. 3. Kinetics of HeLa cell spreading at 37°C on culturedishes covalently coated at either 22°C with gelatin (opensquares) or polyethyleneimine (open triangles), or at 45°Cwith gelatin (open circles). Cells in HEPES-bufferedRPMI-1640 were plated onto all culture dishes at 37°C (seeMaterials and methods). At the times indicated thepercentage of attached cells that spread was determinedusing a phase contrast inverted Nikon microscope. Thevertical bars are the standard error of the mean for fourseparate determinations.

place (Weigel et al., 1978, 1979; Aplin and Hughes,1981; Oka and Weigel, 1986). We followed this kineticapproach with HeLa cell attachment and spreading ongelatin. Gelatin was used instead of collagen since (1)HeLa cells attach and spread equally well on collagen orgelatin (Figs 1, 2 and 3); (2) HeLa cell collagenreceptors bind to both type I collagen and gelatin, andtype IV collagen; (3) antibodies to the collagenreceptors inhibit cell spreading on gelatin or conversely,substitute for gelatin and facilitate cell spreading (Lu etal., 1989); and (4) it is easier to control the amount anddistribution of gelatin bound to the sulfonated poly-styrene culture dishes. Collagen aggregates in solutionsused to covalently couple it to the culture dishesresulting in an uneven surface distribution.

Cell attachment was followed by metabolicallylabeling the cells with [3H]thymidine. The percentageof cells attached was determined from the percent oftrichloroacetic acid-precipitable counts associated withthe culture dish. The percentage of cells attached todishes with different surface concentrations of gelatindiffers depending upon the time of incubation beforethe unbound or weakly attached cells are washed away.If the percentage of cells attached is determined after 2or 4 min incubation and plotted against the gelatinconcentration on the dish, a sigmoidal relationship isobserved (Fig. 4). After 2 min incubation, the concen-tration of gelatin at which cells begin to attach isbetween 18 and 20 ^g/cm2, clearly indicating a threshold

HeLa cell-substratum adhesion 877

CO

.28

3?

100-

10 20 0Gelatin (/jg/cm2)

10 20

Fig. 4. HeLa cell attachment at various times of incubationas a function of the amount of gelatin on the culture dish.HeLa cells were metabolically labeled with [3H]thymidinethen suspended in HEPES-buffered RPMI-1640 containing1 mg/ml BSA, plated onto culture dishes with differentconcentrations of gelatin and incubated at 37°C (seeMaterials and methods) for either 2, 4, 10 or 20 min asindicated in the upper left corner of the figure panels. Cellswere incubated in either the absence (open circles) or thepresence (filled triangles) of 1 mg/ml gelatin in theincubation medium. The vertical bars are the standarderror of the mean for four separate determinations.

effect. The sigmoidal relationship disappears withlonger times of incubation (Fig. 4, 10 and 20 min).Additionally, gelatin in solution at 1 mg/ml blocks theattachment, which is consistent with the reasoning thatspecific cell surface receptors are involved in theattachment (Lu et al., 1989).

The percentage of attached HeLa cells that spread in60 min was also found to have a threshold concentrationof gelatin below which the cells do not spread. Theconcentration is 19 to 20 jUg/cm2 of culture dish surface(Fig. 5). Interestingly, when percent cell spreading wasmeasured after 15 min, when the cells were onlypartially spread, the threshold concentration of gelatinwas still 19 to 20 jig/cm2 (data not shown). In otherwords, the cells below the threshold concentration donot even partially spread.

Segregation of collagen receptors to the basal PMduring cell spreadingThe distribution of the receptors in HeLa cells in

3

878 M. L. Lu et al.

son, we also probed the blots with antibodies to a 45kDa collagen-binding protein (45 kDa CBP) notthought to be involved in HeLa cell adhesion to gelatin(Lu et al., 1989). The relative abundance of thereceptors in the immunoblots was quantified with avideo densitometer and the results are expressed as the

Fig. 6. Immunofluorescence staining of HeLa cells withantibodies raised against collagen receptors (CR) orcollagen binding proteins (CBP). HeLa cells were fixed in3% paraformaldehyde and stained with either anti-102/58kDa CR in suspension (A) or after spreading 60 min (B),or with anti-45 kDa CBP after 60 min of spreading (seeMaterials and methods) (C). FITC-conjugated goat anti-mouse IgG was used for visualization with a Dialux 20Leitz fluorescence microscope. Bar, 20 fim.

percent of total for each receptor in the PM fractions. Inthe case of cells in suspension, the percent of total for areceptor is given only for the internal PM fraction withthe balance being in the external or cell surface fraction.For cells spreading, the percent of total is given for allthree PM fractions, i.e. the internal PM domain, andthe apical and basal PM domains that make up the cellsurface. The most obvious finding is a decrease inrelative abundance of all the receptors in the internalPM fraction during cell adhesion (Fig. 7, bars desig-nated INTERNAL). For example, 45% of the total102/58 kDa collagen receptor is found in the internalPM fraction when cells are in suspension (open bar)while only 27% for cells spread for 15 min (diagonallyhatched bar) and 22% for cells spread for 60 min(horizontally hatched bar). All the collagen receptorsincreased in the basal PM fraction during cell spreading(Fig. 7, bars designated BASAL) and tended todecrease or remain the same in the apical PM fraction(Fig. 7, bars designated APICAL). In all cases therelative abundance of the collagen receptors wasgreatest in the basal PM, which accounts for 27% of thetotal PM in cells spread for 60 min (cf. Mason andJacobson, 1985; Mason et al., 1987).

Association of collagen receptors with the cytoskeletonIt has been hypothesized that cell attachment andspreading might be mediated by the segregation ofgelatin receptors to the basal plasma membrane domainwhere they become clustered and bind to the cytoskel-eton (cf. Rees et al., 1977; Grinnell, 1978, 1980; Rubinet al., 1984; Geiger, 1983; Cody and Wicha, 1986;Rapraeger et al., 1986; Jacobson, 1988). To determinewhether clustered HeLa cell collagen receptors bind tothe cytoskeleton, cytoskeletal preparations with associ-ated PM proteins were made by extracting the cells with

70

60

50

• In suspension 102/58 kDa CRE3 spread 15 minB spread 60 min

40

30

20

100

70

60

50

40

30

20

10

0

Fig. 7. Redistribution of collagen receptorsamong the apical PM, the basal PM and aninternal pool of PM vesicles during HeLa celladhesion on gelatin. PM domains wereisolated from HeLa cells in HEPES-bufferedRPMI-1640 in suspension or after 15 and 60min of spreading on gelatin-coated culturedishes, and the relative amounts of threecollagen receptors designated 102/58 kDa CR,87 kDa CR and 38/33 kDa CR, and one 45kDa collagen-binding protein (CBP) in eachof the PM fractions was determined byimmunoblotting (see Materials and methods).The internal PM domain, the apical PMdomain and the basal PM domain aredesignated, respectively, INTERNAL,APICAL and BASAL under the series ofbars for cells in suspension (open bars), cellsspread 15 min (diagonally hatched bars) orcells spread 60 min (horizontally hatchedbars). The relative amount of each receptor isplotted as a percent of the total. The percentof total CR or CBP for cells in suspension is

given for only the internal PM domain, with the balance being in the external or cell surface PM. The percent of total forcells spreading on gelatin is given for the internal PM, apical PM and basal PM domains. The vertical lines on the bars arethe standard error of the mean for four separate experiments.

INTERNAL APICAL BASAL INTERNAL

Plasma membrane domains

APICAL BASAL

HeLa cell-substratum adhesion 879

Table 1. The percent of HeLa cell collagen receptorsassociated with the Triton X-100 insoluble cytoskeletal

fractionCollagen receptors

Treatment 102/58 kDa 87 kDa 38/33 kDa

ControlAntibodyGelatinSpread 15 minSpread 60 min

47±370±972±271±472±8

45±468±766±572±874±9

8±494±669±591±591±5

The data are an average of five experiments with the standarderror of the mean. Control, cells in suspension extracted withdetergent; Antibody, cells extracted after they were incubated insuspension with antibodies to the collagen receptors; Gelatin, cellsextracted after they were incubated with 1 mg/ml gelatin; Spread15 min, cells were allowed to attach and spread on gelatin-coatedculture dishes for 15 min before extraction with detergent; andSpread 60 min, cells were allowed to attach and spread for 60 min.See Materials and methods for conditions of Triton X-100extraction and for detection of collagen receptors in the detergent-stable cytoskeletal fraction transferred to cellulose nitrate sheets.

a nondenaturing detergent (e.g. see Flanagan andKoch, 1978; Prives et al., 1982; Woods et al., 1986;Cody and Wicha, 1986; Rapraeger et al., 1986; Pattonet al., 1989). The intent of the experiments reportedbelow was to determine the baseline level of collagenreceptor binding to a detergent-resistant cytoskeletalfraction, and to determine whether larger amounts ofreceptors co-isolate with the cytoskeletal fractionduring cell adhesion.

Nondenaturing detergent-resistant cytoskeletonswith associated collagen receptors were prepared fromHeLa cells: (1) treated in suspension with polyclonalantibodies raised against collagen receptors; or (2)treated in suspension with 1 mg/ml gelatin; or (3) afterthe cells were allowed to partially spread (15 min); or(4) after the cells were allowed to fully spread (60 min)on gelatin-coated culture dishes. The cells were thenextracted with 1% Triton X-100 in PBS on ice for 20min. After the supernatants were removed, the cellswere extracted once in PBS without the detergent. Thesupernatants and the insoluble material were solubil-ized in SDS and resolved by electrophoresis. Therelative amounts of collagen receptors in either thesupernatant or the cytoskeletal fractions were deter-mined by immunoblotting using mouse anti-collagenreceptor antibodies. The relative abundance of thereceptors in each fraction was determined by video-densitometry and the percentage in the cytoskeletalfraction is presented in Table 1. In all cases, incubatingthe cells with polyclonal anti-receptor antibodies or 1mg/ml gelatin to cluster the receptors, or allowing thecells to partially spread for 15 min or fully spread for 60min, increases the amount of receptors associated withthe cytoskeletal fraction (Table 1). The correlationbetween the increased receptor-cytoskeletal associationdue to antibody or gelatin presentation and cellspreading is consistent with the hypothesis that spread-ing is initiated by receptor-cytoskeletal binding that

comes about as a consequence of receptor clustering inthe basal PM.

Discussion

Recently, work in cell attachment has focused on theinteraction of specific cell surface receptors with specificECM components (cf. reviews by Ruoslahti et al., 1985;Buck and Horwitz, 1987; Hynes, 1987) but there is alarge body of information indicating that attachmentcan also be mediated by nonspecific physicochemicalforces (e.g. see Maroudas, 1975, 1977; Curtis andMcMurray, 1986). HeLa cell attachment is initiated byboth nonspecific substrata such as BSA and the specificsubstrata, gelatin or collagen (Figs 1 and 2). However,the cells do not spread on the nonspecific substrataalthough they will spread slowly on the positivelycharged polymer PEI (Fig. 3; Fairman and Jacobson,1983). Other cells have been shown both to attach andspread on nonspecific substrata, but unlike HeLa cellsthey are capable of making their own ECM (cf. Damskyet al., 1984; Lark et al., 1985).

Experiments to measure attachment kinetics (Fig. 4)and experiments designed to follow the segregation ofreceptors to different PM domains during cell adhesion(Fig. 7) indicate that HeLa cell attachment to gelatininvolves the upregulation of collagen receptors byexocytosis to the basal PM. The kinetic measurementsindicate that HeLa cell attachment is a threshold effectwith regard to the surface concentration of gelatin (Fig.4) and is consistent with previous suggestions (Oka andWeigel, 1986; Weigel et al., 1978) that a minimumnumber of substratum-receptor interactions must takeplace before the cells can attach. However, thereappears to be more to attachment than just a minimumnumber of initial receptor-substratum interactions,since with increased times of incubation (10 or 20 min),the threshold effect of attachment is dampened and thecells become more adherent at lower surface concen-trations of gelatin (Fig. 4). These results indicate thatthe initial contacts of the cell with gelatin become morestable over time. Consistent with this interpretation isthe observation that, as the cells are allowed to attach tothe culture dish for longer times, gelatin in solutionbecomes less effective at inhibiting cell attachment (Fig.4). The increased strength of attachment does notappear to be simply a result of the onset of spreadingbecause the decrease in inhibition of cell attachment bygelatin occurs with dishes coated with surface concen-trations of gelatin either above or below that whichinduces cell spreading (Fig. 5). The more firm attach-ment with longer incubation times at surface concen-trations of gelatin below the threshold for cell spreadingis consistent with a time-dependent increase in thenumber of collagen receptors in the basal PM.

Previous work indicates that cell attachment is mademore strong by a nonspecific substratum attachment-induced exocytosis of collagen receptors from anintracellular pool of PM. We have shown that HeLacells in suspension have 55% of their total PM protein

880 M. L. Lu et al.

in intracellular vesicles as indicated by cell surfaceiodination. Upon cell attachment to gelatin, whichinduces cell spreading, or to BSA, which does notinduce spreading, there is a stimulation of membraneefflux from an intracellular pool of PM, presumably byexocytosis. During the first 10 min of spreading theintracellular pool decreases to 35% of the total PM.This is followed by a re-endocytosis over the next 20min, so that when the cells are fully spread on gelatinthe internal pool contains 46% of the total PM. 55% ofthe PM is found in the internal membrane pool ofunspread cells attached to BSA (Mason et al., 1987).This information taken together with the results on theredistribution of collagen receptors during cell adhesion(Fig. 7) indicates that the major movement of collagenreceptors to the basal PM domain probably occurs by anonspecific substratum attachment-stimulated exocy-tosis. It cannot be concluded that the exocytosis ofreceptors from the intracellular pool of PM vesiclesoccurred only in the direction of the basal PM and notthe apical PM. HeLa cells are like many other cells inthat they constantly recycle their plasma membranebetween an internal domain and the cell surface(Mellman et al., 1980; Fishman and Cook, 1982;Widnell et al., 1982; Mason et al., 1987). It is possiblethat the attachment-stimulated upregulation of collagenreceptors during HeLa cell adhesion occurred byrandom exocytosis from the internal PM domain toboth the apical PM and the basal PM.

Kinetics of HeLa cell spreading revealed that there isa threshold concentration of gelatin on the culture dishbelow which the cells will not spread (Fig. 5). This canbe interpreted in three ways with regard to the initiationof cell spreading (cf. Weigel et al., 1978, 1979):(1) there must be a minimum number of receptor-gelatin interactions required for spreading to occur;(2) gelatin on the culture dish must be present at asufficiently high concentration where it assumes aunique conformation which only then allows the HeLacells to spread; or (3) the collagen receptors must beclustered before cell spreading can take place. It is notlikely that the number of receptors in the basal PM isthe limiting factor in cell spreading. HeLa cells left incontact with dishes coated with low levels of gelatin stilldo not spread (Fig. 4) even though they increase theirstrength of attachment, presumably by delivery ofreceptors to the basal PM (see above). It is also unlikelythat a conformational change in the substratum-boundgelatin takes place at the higher surface concentrationsof gelatin. Transmission electron microscopy demon-strated that gelatin covalently bound to the surface ofpolystyrene cell culture microcarriers at surface-satu-rating levels maintains a random appearance of over-lapping fibrous molecules and does not form anydetectable supercoiled structures or regions of higherdensity (Fairman and Jacobson, 1983). Thus, thecooperativity in cell spreading is consistent with theneed to cluster receptors.

To further understand whether cooperativity in cellspreading is indicative of receptor clustering, weemployed immunofluorescence microscopy using anti-

bodies to the receptors to determine if visible clustersare formed during cell adhesion. Antibodies to collagenreceptors that inhibit cell spreading on gelatin-coatedculture dishes and conversely, substitute for the gelatinon the dish and facilitate cell spreading (Lu et al., 1989),exhibit a diffuse fluorescence staining pattern when thecells are in suspension and a punctate pattern in thebasal PM of spread cells (Fig. 6). On the other hand,antibodies to a cell surface collagen-binding proteinthat does not inhibit spreading or substitute for gelatinas the extracellular matrix (Lu et al., 1989) give adiffuse staining pattern of fluorescence in both suspen-sion and spread cells (Fig. 6). Punctate patterns offluorescence for extracellular matrix receptors thatmediate cell-substratum adhesion have been repeatedlyshown with fibronectin receptors (cf. Chen et al., 1985;Damsky et al., 1984; Giancotti et al., 1986), collagenreceptors (Mollenhauer et al., 1984) and cell surfaceproteoglycans (Rapraeger et al., 1986). Furthermore,extensive morphological studies indicate that there is aco-localization of the punctate fluorescence of thereceptors and cytoskeletal elements (e.g. Rogalski andSinger, 1985; Chen et al., 1985).

To further explore whether clustering of collagenreceptors is essential for HeLa cell spreading, wefollowed the adhesion of the cells to a positivelycharged substrate, polyethyleneimine (PEI). The rateof spreading was significantly slowed relative to cellspreading on collagen or gelatin (see Figs 2 and 3). Therationale was that if receptor clustering is a precon-dition to cell spreading, then cell spreading will beimpeded on a substrate that has a high positive charge,which slows the rate of lateral diffusion and therefore,receptor clustering. Positive substrates have beenshown to bind so tightly to the plasma membrane thatthe lateral diffusion of Con A receptors is restrained(Patton et al., 1990). It is likely the PEI substratenonspecifically binds to the collagen receptors and notnecessarily at the "active" collagen binding site. Whileit cannot be ruled out that the receptors might not be inthe proper conformation to initiate cell spreading, itshould be noted that monoclonal antibodies that bind toEGF or insulin receptors, at sites other than thehormone binding sites, still induce receptor clusteringand endocytosis (cf. Schlessinger et al., 1983; Maron etal., 1984; Forsayeth et al., 1987). Furthermore, endo-cytosis of membrane proteins is induced by binding tohighly cationized ferritin (Simionescu et al., 1981). Theabove, taken together with the observation that col-lagen receptors exhibit a punctate pattern of fluorescencein the basal PM of spread cells, and that cell spreading is athreshold function of the gelatin concentration on theculture dish, strongly supports the hypothesis thatreceptor clustering is a requirement for cell spreading.

The observation that clustering HeLa cell collagenreceptors causes them to bind to the cytoskeleton asdetermined by nondenaturing detergent extraction(Table 1), and experiments indicating that the gelatinsubstratum clusters collagen receptors (Figs 5 and 6) areconsistent with previous work with other cells indicatingthat receptor-cytoskeletal binding must occur before

HeLa cell-substratum adhesion 881

cell spreading can take place (e.g. see Rees et al., 1977;Grinnell, 1978; Geiger, 1983). However, the mechan-ism by which the HeLa cell collagen receptors interactwith the cytoskeleton is not known. Clustering lamininreceptors (Cody and Wicha, 1986) or a transmembraneproteoglycan that is an extracellular matrix receptor(Rapraeger et al., 1986) induce them to bind to adetergent-resistant cytoskeletal fraction. Both the lam-inin receptor (Brown et al., 1983) and the proteoglycanextracellular matrix receptor (Rapraeger and Bern-field, 1982) bind directly to F-actin in vitro. Fibronectinreceptors have been shown to bind to talin in vitro(Horwitz et al., 1986).

It should be emphasized that the measurements ofreceptor-cytoskeletal binding determined by non-denaturing detergent extraction procedures do notindicate whether the receptors that were associatedwith the cytoskeleton in the absence of an extracellularligand or substratum were either unclustered or clus-tered. The data only indicate that clustered receptors,or making larger clusters from smaller ones, arecorrelated with an enhancement of receptor associationwith the cytoskeletal preparations (cf. Brandts andJacobson, 1983). If the receptors were in the form ofdimers, trimers, etc., and exhibited a weak or even astrong affinity for the cytoskeleton, clustering them intolarger oligomers would markedly increase the affinity(Brandts and Jacobson, 1983; Shiozawa et al., 1989).For example, one estimate of the enhanced affinityreveals that, depending upon the mole fraction of thereceptors in the plane of the membrane, a tetramericcluster would bind to the cytoskeleton 1012 times moreeffectively than four unclustered receptors and anoctomer 1027 times (Jacobson, 1988). It should benoted, that these increases in cytoskeletal binding uponreceptor clustering are consistent with the cooperativityseen in cell spreading versus the concentration ofgelatin on the culture dish (Fig. 5).

Is the clustering and binding of the collagen receptorsto the cytoskeleton sufficient to induce cell spreading orare other messages involved? It is tempting to speculatethat clustering might induce a second messenger whichsignals the cells to spread. It has been shown thatmutants of Chinese hamster ovary cells that do notadhere to fibronectin have an altered type I proteinkinase. Adhesion of these mutants can be induced bythe addition of cyclic AMP (Cheung and Juliano, 1985;Cheung et al., 1987). It is also known that one of thesubunits of chick fibronectin receptor has a proteinkinase C substrate-binding site in its cytoplasmicdomain (cf. Burridge et al., 1988). Interestingly,Danilov and Juliano (1989) recently reported thatphorbol ester, a protein kinase C activator, does notalter the phosphorylation state of fibronectin receptoror talin, or the number or affinity of cell surfacefibronectin receptors, and concluded that the increasein cell adhesion induced by the phorbol ester is not dueto a direct effect on the receptors. Work is currently inprogress to determine whether a second messenger isinvolved in signaling HeLa cell spreading. Preliminaryexperiments indicate that HeLa cells have a dramatic

rise in cyclic AMP and a spike of intracellular freecalcium during cell spreading; however, both the riseand the spike occur after spreading begins indicatingthat neither cyclic AMP nor Ca are the secondmessengers initiating cell spreading (Lu, Chun andJacobson, unpublished observations).

We are deeply indebted to Dr Peter Mason and DeidraGramas for the micrographs of spreading HeLa cells, and toMr. Jang-Soo Chun for providing us with his initial obser-vations on second messengers during cell adhesion. This workwas supported in part by a grant from the National Institutesof General Medical Sciences, GM 29127.

References

Aplin, J. D. and Hughes, R. C. (1981). Cell adhesion on modelsubstrata: threshold effects and receptor modulation. / . Cell Sci.59, 89-103.

Avnur, Z. and Geiger, B. (1981). The removal of extracellularfibronectin from areas of cell-substrate contact. Cell 25, 121-132.

Beacham, D. A. and Jacobson, B. S. (1990). Mg + + mediates the cell-substratum interaction of Arg-Gly-Asp dependent HeLa cellcollagen receptors. Exp. Cell Res. 189, 69-80.

Brandts, J. F. and Jacobson, B. S. (1983). A general mechanism fortransmembrane signalling based on clustering of receptors. Surv.Synth. Pathol. Res. 2, 107-114.

Brown, S. S., Mallnoff, H. L. and Wtcha, M. S. (1983). Connectin:cell surface protein that binds both laminin and actin. Proc. Nat.Acad. Sci. U.S.A. 80, 5927-5930.

Buck, C. A. and Horwitz, A. F. (1987). Cell surface receptors forextracellular matrix molecules. Annu. Rev. Cell Biol. 56, 179-205.

Burridge, K., Fath, K., Kelly, T., Nuckolls, G. and Turner, C. (1988).Focal adhesions: transmembrane junctions between theextracellular matrix and the cytoskeleton. Ann. Rev. Cell Biol. 4,487-525.

Chandrakasan, G., Torchla, D. A. and Piez, K. A. (1976).Preparation of intact monomeric collagen from rat tail tendon andskin and the structure of the nonhelical ends in solution. J. Biol.Chem. 251, 6062-6067.

Chen, W.-T., Hasegawa, E., Hasegawa, T., Weinstock, C. andYamada, K. M. (1985). Development of cell surface linkagecomplexes in cultured fibroblasts. / . Cell Biol. 100, 1103-1114.

Cheung, E. and Juliano, R. L. (1985). cAMP-induced phenotypicreversion of adhesion, aggregation, and endocytosis in adhesion-defective DHO cell variants. /. Cell. Physiol. 124, 337-343.

Cheung, E., Brown, P. J. and Julian, R. L. (1987). Altered type Iprotein kinase in adhesion defective CHO cell variants. J. Cell.Physiol. 130, 118-124.

Cody, R. and Wicha, M. (1986). Clustering of cell surface lamininenhances its association with the cytoskeleton. Exp. Cell Res. 165,107-116.

Couchman, J. R., Hook, M., Rees, D. A. and Timpl, R. (1983).Adhesion, growth, and matrix production by fibroblasts on lamininsubstrates. J. Cell Biol. 96, 177-183.

Culp, L. A., Laterra, J., Lark, M. W., Beyth, R. J. and Tobey, S. L.(1986). Heparan sulphate proteoglycan as mediator of someadhesive responses and cytoskeletal reorganization of cells onfibronectin matrices: independent versus cooperative functions.Ciba Found. Symp. 124, 158-183.

Curtis, A. S. G. and McMurray, H. (1986). Conditions for fibroblastadhesion without fibronectin. / . Cell Sci. 86, 25-33.

Damsky, C. H., Knudsen, K. A. and Buck, C. A. (1984). In TheBiology of Glycoproteins (ed. R. J. Ivastt). pp. 1-64. Plenum Press,NY.

Danilov, Y. N. and Juliano, R. L. (1989). Phorbol ester modulation ofintegrin-mediated cell adhesion: a postreceptor event. J. Cell Biol.108, 1925-1935.

Falrman, K. and Jacobson, B. S. (1983). Unique morphology of HeLacell attachment, spreading and detachment from microcarrierbeads covalently coated with a specific and a non-specificsubstratum. Tissue & Cell 15, 167-180.

882 M. L. Lu et al.

Flshman, J. B. and Cook, J. S. (1982). Recycling of surfacesialoglycoconjugates in HTC and HeLa cells. J. Bio!. Chem. 257,8122-8129.

Flanagan, J. and Koch, G. L. E. (1978). Cross-linked surface Igattaches to actin. Nature 273, 278-281.

Forsayeth, J. R., Montemurro, A., Maddux, B. A., DePirro, R. andGoldfine, I. D. (1987). Effect of monoclonal antibodies on humaninsulin receptor autophosphorylation, negative cooperativity, anddown-regulation. J. Biol. Chem. 262, 4134-4140.

Geiger, B. (1983). Membrane cytoskeleton interaction. Biochim.Biophys. Acta 737, 305-341.

Geiger, B., Avnur, Z., Krels, T. and Schlessinger, J. (1984). In Celland Muscle Motility: The Cytoskeleton (Vol. 5) (ed. J. Shay), pp.195-234. Plenum Press, New York.

Giancotti, F. G., Coraoglio, P. M. and Tarone, G. (1986). A 135,000molecular weight plasma membrane glycoprotein involved infibronectin-mediated cell adhesion. Immunofluorescencelocalization in normal and RSV-transformed fibroblasts. Exp. CellRes. 163, 47-62.

Grlnnell, F. (1978). Cellular adhesiveness and extracellular substrata.Int. Rev. Cytol. 53, 65-144.

Grlnnell, F. (1980). Fibroblast receptor for cell-substratum adhesionstudies on the interaction of baby hamster kidney cells with latexbeads coated by cold insoluble globulin (plasma fibronectin). / . CellBiol. 86, 104-112.

Grlnnell, F. (1986). Focal adhesion sites and the removal ofsubstratum-bound fibronectin. / . Cell Biol. 103, 2697-2706.

Horwitz, A., Duggan, K., Buck, C , Beckerle, M. C. and Burridge, K.(1986). Interaction of plasma membrane fibronectin receptor withtalin-a transmembrane linkage. Nature 320, 531-533.

Hynes, R. O. (1987). Integrins: a family of cell surface receptors. Cell48, 549-554.

Izzard, C. S. and Lochner, L. R. (1976). Cell-to-substrate contacts inliving fibroblasts: an interference reflexion study with an evaluationof the technique. / Cell Sci. 21, 129-159.

Jacobson, B. S. (1977). Isolation of plasma membrane fromeukaryotic cells on polylysine-coated polyacrylamide beads.Biochim. Biophys. Acta 471, 331-335.

Jacobson, B. S. (1988). Cell-substrate adhesion: induction of cellspreading and apical/basal plasma membrane polarity. Adv. CellBiol. 2, 91-118.

Jacobson, B. S. and Ryan, U. (1982). Growth and endothelial andHeLa cells on a new multipurpose microcarrier that is positive,negative or collagen coated. Tissue & Cell 14, 69-83.

Juliano, R. L. (1987). Membrane receptors for extracellular matrixmacromolecules: relationship to cell adhesion and tumormetastasis. Biochim Biophys. Acta 907, 261-278.

Knox, P. (1981). The adhesion of cells to a solid substratum. In TheBiochemistry of Cellular Regulation, Vol. IV: The Cell Surface, (ed.P. Knox), pp. 121-149. CRC Press.

Lark, M. W., Laterra, J. and Culp, L. A. (1985). Close and focalcontact adhesions of fibroblasts to a fibronectin-containing matrix.Fed. Proc. 44, 394-403.

Leary, J. J., Brigatl, D. J. and Ward, D. C. (1983). Rapid andsensitive colorimetric method for visualizing biotin-labeled DNAprobes hybridized to DNA or RNA immobilized to nitrocellulose:bio-blots. Proc. Nat. Acad. Sci. U.S.A. 80, 4045-4049.

LeBaron, R. G., Esko, J. D., Woods, A., Johansson, S. and Hook, M.(1988). Adhesion of glycosaminoglycan-deficient Chinese hamsterovary cell mutants to fibronectin substrata. / . Cell Biol. 106, 945-952.

Lu, M. L., Beacham, D. A. and Jacobson, B. S. (1989). Theidentification and characterization of collagen receptors involved inHeLa cell-substrate adhesion. J. Biol. Chem. 264, 13546-13558.

Maron, R., Jackson, R. A., Jacobs, S., Eisenbarth, G. and Kahn, C.R. (1984). Analysis of the insulin receptor by anti-receptorantibodies and flow cytometry. Proc. Nat. Acad. Sci. U.S.A. 81,7446-7450.

Maroudas, N. G. (1975). Adhesion and spreading of cells on chargedsurfaces. J. Theor. Biol. 49, 417-424.

Maroudas, N. G. (1977). Polymer aggregation and cell adhesion. / .Cell Phys. 90, 511-520.

Mason, P. W. and Jacobson, B. S. (1985). Isolation of the dorsal,

ventral and intracellular domains of HeLa cell plasma membranesfollowing adhesion to a gelatin substratum. Biochim. Biophys. Acta821, 264-276.

Mason, P. W., Lu, M. L. and Jacobson, B. S. (1987). Cell substrateadhesion-induced redistribution of proteins among the apical, basaland internal domains of the plasma membrane of HeLa cellsspreading on gelatin. J. Biol. Chem. 262, 3746-3753.

Mellman, I. S., Steinman, R. M., Unkeless, J. C. and Cohn, Z. A.(1980). Selective iodination and polypeptide composition ofpinocytic vesicles. / . Cell Biol. 86, 712-722.

Mollenhauer, J., Bee, J. A., Lizarbe, M. A. and von der Mark, K.(1984). Role of anchorin CII, a 31,000-mol-wt membrane protein,in the interaction of chondrocytes with type II collagen. / . Cell Biol.98, 1572-1579.

Oka, J. A. and Welgel, P. H. (1986). Binding and spreading ofhepatocytes on synthetic galactose culture surfaces occur as distinctand separable threshold responses. J. Cell Biol. 103, 1055-1060.

Patton, W. F., Dhanak, M. R. and Jacobson, B. S. (1989).Differential partitioning of plasma membrane proteins into thetriton X-100 insoluble cytoskeleton of Dictyostelium discoideum. J.Cell Sci. 92, 85-91.

Patton, W. F., Dhanak, M. R. and Jacobson, B. S. (1990). Analysis ofplasma membrane protein changes in Dictyostelium discoideumduring concanavalin A-induced receptor redistribution using two-dimensional gel electrophoresis. Electrophoresis 11, 79-85.

Prives, J., Fulton, A. B., Penman, S., Daniels, M. P. and Christian, C.N. (1982). Interaction of the cytoskeletal framework withacetylcholine receptor on the surface of embryonic muscle cells inculture. J. Cell Biol. 92, 231-236.

Rajaraman, R., Rounds, D. E., Yen, S. P. S. and Rembaum, A.(1974). A scanning electron microscope study of cell adhesion andspreading in vitro. Exp. Cell Res. 88, 327-339.

Rapraeger, A. and Bernfield, M. (1982). An integral membraneproteoglycan is capable of binding components of the cytoskeletonand the extracellular matrix. In Extracellular Matrix (ed. S. Hawkesand J. Wang), pp. 265-269. Academic Press, New York.

Rapraeger, A., Jalkanen, M. and Bernfleld, M. (1986). Cell surfaceproteoglycan associates with the cytoskeleton at the basolateral cellsurface of mouse mammary epithelial cells. J. Cell Biol. 1003, 2683-2696.

Rees, D. A., Lloyd, C. W. and Thorn, D. (1977). Control of grip andstick in cell adhesion through lateral relationships of membraneglycoproteins. Nature 267, 124-128.

Rogalskl, A. A. and Singer, S. J. (1985). An integral glycoproteinassociated with the membrane attachment sites of actinmicrofilaments. J. Cell Biol. 101, 785-801.

Rollins, B. J., Cathcart, M. K. and Culp, L. A. (1982). In TheGlycoconjugates, Vol. HI, pp. 289-329. Academic Press, NewYork.

Rubin, K., Borg, T. K., Holmdahl, R., Klareskog, L. and Obrink, B.(1984). Interactions of mammalian cells with collagen. Ciba Found.Symp. 108, 93-116.

Ruoslahtl, E., Hayman, E. G. and Pierschbacher, M. D. (1985).Extracellular matrices and cell adhesion. Arteriosclerosis 5, 581-594.

Schlesslnger, J., Schrelber, A. B., Llbcrmann, T. A., Lax, I., Avlvi,A. and Yarden, Y. (1983). In Cell Membranes: Methods andReviews, Vol. 1. (ed. E. Elson, W. Frazier and L. Glaser), pp. 117-149. Plenum Press, New York.

Shlozawa, J. A., Brandts, J. F. and Jacobson, B. S. (1989). Binding ofplasma membrane glycoproteins to the cytoskeleton duringpatching and capping is consistent with an entropy enhancementmodel. Biochim. Biophys. Acta 980, 361-366.

Slmlonescu, N., Slmlonescu, M. and Palade, G. E. (1981).Differentiated microdomains on the luminal surface of the capillaryendothelium. I. Preferential distribution of cationic sites. /. CellBiol. 90, 605-613.

Singer, I. I. (1982). Fibronexus formation is an early event duringfibronectin-induced restoration of more normal morphology andsubstrate adhesion patterns in transformed hamster fibroblasts. J.Cell Sci. 56, 1-20.

Singer, 1.1., Kawka, D. W., Scott, S., Mumford, R. A. and Lark, M.W. (1987). The fibronectin cell attachment sequence Arg-Gly-Asp-

HeLa cell-substratum adhesion 883

Ser promotes focal contact formation during early fibroblastattachment and spreading. J. Cell Biol. 104, 573-584.

Singer, I. I., Scott, S., Kawka, D. W., Kazazis, D. M., Gailit, J. andRuoslahti, E. (1988). Cell surface distribution of fibronectin andvitronectin receptors depends on substrate composition andextracellular matrix accumulation. J. Cell Biol. 106, 2171-2182.

Thompson-Pletscher, H. A. (1986). In Regulation of MatrixAccumulation (ed. R. P. Mecham), pp. 400-444. Academic Press,New York.

Traub, W. and Piez, K. A. (1971). The chemistry and structure ofcollagen. Adv. Prot. Chem. 25, 243-352.

Towbin, H., Staehelln, T. and Gordon, J. (1979). Electrophoretictransfer of proteins from polyarylamide gels to nitrocellulosesheets: procedure and some application. Proc. Nat. Acad. Sci.U.S.A. 76, 4350-4353.

Welgel, P. H., Schmell, E., Lee, Y. C. and Roseman, S. (1978).

Specific adhesion of rat hepatocytes to beta-galactosides linked topolyacrylamide gels. J. Biol. Chem. 153, 330-333.

Weigel, P. H., Schnaar, R. L., Kuhlenschmidt, M. S., SchmeU, E.,Lee, R. T., Lee, Y. C. and Roseman, S. (1979). Adhesion ofhepatocytes to immobilized sugar. A threshold phenomenon. J.Biol. Chem. 254, 10830-1083S.

WidneU, C. C , Schneider, Y.-J., Pierre, B., Baudhuin, P. andTrouet, A. (1982). Evidence for a continual exchange of 5'-nucleotidase between the cell surface and cytoplasmic membranesin cultured rat fibroblasts. Cell 28, 61-70.

Woods, A., Couchman, J. R., Johansson, S. and Hook, M. (1986).Adhesion and cytoskeletal organization of fibroblasts in response tofibronectin fragments. EMBO J. 5, 665-670.

(Received 5 November 1991 • Accepted 15 January 1992)