Retinoic acid enhances adhesiveness, laminin and integrin ?1 synthesis, and retinoic acid receptor expression in F9 teratocarcinoma cells

JOURNAL OF CELLULAR PHYSIOLOGY 159:263-273 (1994)

Retinoic Acid Enhances Adhesiveness, Laminin and lntegrin pl Synthesis, and Retinoic Acid Receptor Expression in F9 Teratocarcinoma Cells SHARON A. ROSS, RICHARD A. AHRENS,

AND

L U l G l M. DELUCA*

Differentiation Control Section, Laboratory of Cellular Carcinogenesis and Tumor Promotion, National Cancer Institute, National lnstitutes of Health, Bethesda, Maryland 20892 (S.A.R., L. M.D.L.) and Nutritional Sciences Program, University of Maryland, College Park, Maryland 20742 (S.A.R., R.A.A.) The teratocarcinoma-derived F9 cells respond to retinoic acid (RA) and RA plus dibutyrylcyclic adenosine monophosphate (dcAMP) by differentiating into endoderm cells, which elaborate a laminin and type IV collagen-rich matrix. W e found that the induction of differentiation is accompanied by a small but consistent increase in cell adhesiveness to a variety of substrates, including laminin. Therefore we investigated biochemical mechanisms involved in this phenomenon. Endoglycosidase treatment showed that laminin contains complex and hybrid oligosaccharide structures. RA enhanced general biosynthesis of laminin without a specific increase in galactose incorporation: this sugar was mainly in polylactosamine structures in the A chain of laminin and as terminal galactose a 1,3 galactose in the B chain. Laminin receptor analysis showed that RA decreased laminin binding protein-37 (LBP-37) but increased the amount of p l integrin, suggesting the involvement of f3l integrin in the attachment process. Northern blot analysis showed increased expression of retinoid receptors within hours of RA exposure. These studies demonstrate that RA increases cell to substrate interactions by increasing the biosynthesis of laminin and pl integrin. These effects are most likely subsequent to the RA-induced biosynthesis of the retinoid receptors. 01994 Wiley-Liss, Inc."

Retinoic acid (RA)has been shown to be necessary for normal embryonal development (Thompson et al., 1969); in addition it induces differentiation of the embryonal carcinoma F9 cell line to primitive endoderm (Strickland et al., 1980; Strickland and Mahdavi, 1978). Extraembryonic endoderm comprises the primitive, parietal, and visceral phenotypes with visceral endoderm formed by aggregation of primitive endoderm (Hogan et al., 1981)and parietal endoderm generated after treatment with RA and cyclic adenosine monophosphate (CAMP)(Strickland et al., 1980). Differentiation toward primitive endoderm by RA and toward parietal endoderm by RA plus dibutyrylcyclic adenosine monophosphate (dcAMP) results in the production of the extracellular matrix (ECM) components laminin, type IV collagen, entactin, and proteoglycans (Strickland et al., 1980; Carlin et al., 1983; Cooper et al., 1983; Marotti et al., 1985; Wang and Gudas, 1990). Prior to the increase in ECM proteins, RA increases the expression of several transcription factors, including RARp (Zelent et al., 1989; Hu and Gudas, 1990;Martin et al., 1990) and several hox genes known to be involved in embryogenesis (Gudas, 1991). It is thought that the basement membrane is involved in the migration and segregation of embryonic cells so that they can differentiate into specific tissues 0 1994 WILEY-LISS?INC. **This article is a US Government work and, a s such, is in the public domain in the United States of America.

(Hogan and Tilly, 1981; Gardner, 1983; Grabel and Watts, 1987).The enhanced synthesis of ECM, after RA treatment of F9 cells, suggests that there is a n attachment function for ECM during differentiation in this system. Therefore, we investigated the effect of RA on cell attachment and the biochemical mechanism underlying this effect.

MATERIALS AND METHODS Cell culture F9 teratocarcinoma cells (ATCC)were maintained in Dulbecco's modified Eagle's medium (DMEM) with 4.5 @liter glucose (Gibco Laboratories) containing 15% heat inactivated fetal calf serum (JRH Biosciences) and 1%antibiotic-antimycotic solution (Gibco) and grown on culture plates coated with 0.1% gelatin (Biorad). Cells were seeded in 100 mm tissue culture plates a t 3 or 1 x lo5 cells per plate. Culture media were changed 1day after seeding with one of the following four treatment media: 0.1% dimethylsulfoxide (DMSO) (ATCC), lop3 M N6,2'-0-dibutyryladenosine3'5'-cyclic mono-

Received August 17,1993; accepted November 17,1993. "To whom reprint requests/correspondence should be addressed.

264

ROSS ET AL.

phosphate (dcAMP) (Sigma Chemical Co.), M RA (Sigma Chemical Co.) in 0.1%DMSO, or lop3M dcAMP plus M RA in 0.1% DMSO. Treatment media were changed daily. Except where noted, cells were used in assays after 72 hours of treatment because the largest differences were observed at this time. In Northern blot analysis, cells were treated for 0,4,or 8 hours, at which time mRNA was isolated.

Cell labeling After 72 hours of treatment, as described above, the treatment media were aspirated from the cells and replaced with either methionine-free DMEM with 80 pCi/ml of 35S-methionine (1,050 Ci/mM specific activity [SPA]), DMEM with 20 pCi/ml of galact0se-[6-~Hl (25 Ci/mM SpA), or methionine-free DMEM with 100 yCi/ml of galacto~e-[6-~H] (25 Ci/mM SPA) plus 40 pCi/ml of 5S-methionine (1,115 Ci/mM). All radiochemicals were from American Radiolabeled Chemicals, Inc. These media were supplemented with 2 mM glutamine and 1% bovine serum albumin (BSA). In each experiment, cells were incubated in radiolabeled media for 18 hours. After incubation, media were collected and proteins were precipitated with 10% trichloroacetic acid on ice (TCA) (1:1.4, voUvo1, TCNsample) (Wang and Gudas, 1984). The precipitated proteins were washed twice using cold ether and the protein pellet was resuspended in lysing buffer (0.2 M phosphate buffer, pH 6.5; 1% Triton X-100, 0.1% sodium azide; 0.5%deoxycholic acid sodium salt; 0.1 M sodium chloride plus the following protease inhibitors: 1 mM phenylmethylsulfonyl fluoride [PMSF], 1 pg/ml leupeptin, 5 pg/ml aprotinin, 10 pglml benzamine hydrochloride, 1 pg/ml pepstatin, 5 pg/ml antipain, and 5 pg/ml chymostatin). Protein determinations were performed using the bicinchoninic acid (BCA) protein assay, Pierce (Smith et al., 1985). Radioactivity (counts per minute [CPMI) in each sample was determined using a Beckman Liquid Scintillation Counter (Model LS250). In the double label study, CPM were corrected for the spillover of 35S-radioactivity into the 3H channel. Cell lysates from metabolic labeling studies were also obtained as described under endoglycosidase reaction. In addition, proteins from whole cell lysates and media were analyzed by autoradiographic analysis following electrophoresis.

4°C. Samples were microfuged at 4"C, pellets were washed four times with phosphate buffer (0.2 M Na3P04,pH 6.5; 0.1% Triton-X 100; 0.1% sodium azide; 0.5% deoxycholic acid sodium salt; 0.1 M sodium chloride), and once with high salt buffer (10 mM Tris, pH 7.2; 1% Triton-X 100; 1 M sodium chloride) (Bjeurum and Aeegaard, 1988).

Electrophoresis Pellets were resuspended in 40 p1 of sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDSPAGE) loading buffer, boiled for 3 minutes, microfuged, and a n aliquot of the supernatant was added onto gel lanes for SDS-PAGE and another aliquot was counted in a liquid scintillation counter. In some experiments, total cell lysate or media protein samples without immunoprecipitation were analyzed by SDS-PAGE analysis, gels were transferred to nitrocellulose, and autoradiography was performed by exposing to X-ray film for various times a t -70°C. For visualization of the protein bands in the 3H-galactose labeled samples, gels were stained with Coomassie blue, destained, and immersed in AmplifyTM(Amersham) for 30 minutes. Gels were dried under vacuum and placed in film cassettes for exposure to Hyperfilm'" (Amersham).

Attachment assay Laminin (Collaborative Res.), fibronectin (Collaborative Res.), or gelatin (8 pg/lOO p1 of phosphate buffered saline [PBS], Mg++ and C a + + free) were incubated in wells of a 96-well tissue culture plate (Costar, no. 3596) for 2 hours at 37°C and the remaining EC liquid solution was removed. BSA (Fluka) (0.1% in PBS) was added to each well, to cover nonspecific binding sites, for 10 minutes at 37°C. The wells were washed twice with PNS following the 10-minute incubation period with BSA, using a Transtar-96 plate washer (Costar). Cells, cultured in the various treatment media for 72 hours, were rinsed with PBS and 0.1 ml of 0.25% trypsin-EDTA (5 x lo-* M) was added to allow detachment from the culture dish at room temperature. Cells were resuspended in medium with serum and counted using a Coulter counter. Cells were resuspended in medium without serum for the attachment assay. Cells (2.5 X 104/100 p.1, unless noted) were added to wells precoated with attachment substrates and incuImmunoprecipitation The protein concentration of each radiolabeled sam- bated at 37°C for 2 hours, unless otherwise noted. Unatple was adjusted to 1 pg/ml using lysing buffer. Sam- tached cells were removed and wells rinsed once with ples were precleared of nonspecific binding with pro- PBS. The fluorogenic substrate 4-methyl-umbelliferyl tein A sepharose, antibody (5 pl) was added, e.g., anti- heptanoate (MUH) (Dotsika and Sanderson, 1987; Stadmouse laminin rabbit IgG (Collaborative Res., Inc.), ler e t al., 1989) was added to each well (10 pg/lOO pl) and incubated at 37°C for 60 minutes on a multipurpose and after a 30-minute incubation (37°C) fluorescence rotator. For laminin binding protein (LBP)-37 analysis was measured using a Dynatech Instruments Microwe used antisera raised in rabbit against a 17-mer syn- fluor Reader. We selected a MUH concentration of 10 thetic peptide from the N-terminal region of LBP-37 pgi100 p1 for attachment assays because this concen(residues 2-1) which was conjugated to keyhole lim- tration provided a low background and was in the linpet hemocyanin (from Dr. Yoshihiko Yamada, National ear fluorescence range for the Dynatech Microfluor Institute of Dental Research, NIH). For p l integrin Reader. Using this dye concentration, fluorescence was analysis we used polyclonal antisera raised in rabbit not altered when substrates alone were added to wells. against a synthetic peptide from the C-terminal cytoStatistical analysis plasmic domain of the human p l integrin subunit Data were analyzed by one-way analysis of variance (Chemicon). Protein A sepharose, 200 p J (30 mg/ml solution), was added to each sample rotated overnight at (ANOVA) and mean comparisons were conducted with

RETINOIC ACID ENHANCES ADHESIVENESS OF F9 CELLS

265

Fig. 1. RA plus dcAMP enhance laminin secretion in the medium of F9 teratocarcinoma cells. Immunoprecipitation of laminin from 35Smethionine-labeled media proteins (200pg). Cells were treated for 72 hours with DMSO (0.1%),dcAMP MI, RA M in 0.1% DMSO), or RA plus dcAMP. Lane 1: labeled molecular weight (MW)

standards (Amersham); lane 2 antibody control, no sample; lane 3 DMSO plus antibody; lane 4 DMSO alone; lane 5 dcAMP plus antibody; lane 6: dcAMP alone; lane 7: RA plus antibody; lane 8: RA alone; lane 9 RA+dcAMP plus antibody; lane 10 RA+dcAMP alone. Samples were processed as described under Materials and Methods.

the Bonferroni test (Steel and Torrie, 1980). Values plotted in Figure 3C, were analyzed by the nonparametric ANOVA test and Dunn’s multiple comparisons test (Gill, 1993). This method was used because standard deviations were significantly different among different ratios. The Graphy Pad Instat program (ISI, Philadelphia, PA) was used for calculation.

(7.5% polyacrylamide) (Laemmli, 1970). Proteins were transferred to nitrocellulose (Towbin et al., 1979) in 25 mM Tris base, 192 mM glycine, 20% (voYvo1)methanol, and 0.01%SDS, pH 8.3, at 50 V overnight. Western blot analysis was performed using rabbit anti-laminin antibody (1:200 dilution) (Collaborative Res., Inc.). The membranes were blocked with 3% gelatin/l% BSA in Tris-buffered saline (TBS)before incubation for 2 hours with anti-laminin antibody. Immunoblots were washed in Tween-20 (Biorad) plus TBS (TTBS) and incubated with a biotinylated secondary antibody, anti-rabbit IgG goat (Boehringer Mannheim) for 45 minutes, washed twice with TTBS, and incubated with strepavidin alkaline phosphatase conjugate (Boehringer Mannheim) for 30 minutes. Following two washes in TTBS and one wash in TBS, the blots were developed using an alkaline phosphatase conjugate substrate kit (Biorad).

Endoglycosidase reaction Cell lysates were prepared after 72 hours of treatment, media removed, and cells washed with PBS. Cells were lysed in lysing buffer. The lysate was centrifuged a t 40 Krpm for 45 minutes at 4°C and the supernatant was saved for analysis. Cell lysate protein, 10 pg, from each treatment group was denatured by boiling in 10 pl of 1%SDS for 2 minutes. Sodium phosphate buffer (90 p1) (20 mM, pH 7.2, for endoglycosidase F/Nglycosidase F [ENDO-F; Boehringer Mannheim) or 0.2 M sodium phosphate buffer, pH 6.5, for endoglycosidase D (ENDO-D) and endoglycosidase H (ENDO-H) reactions) was added and samples boiled for 2 additional minutes (Olden et al., 1980). After cooling, 0.4 units of ENDO-F (or 10 mU of ENDO-D or 5 m u of ENDO-H) was added to each sample. Samples were incubated for enzymatic cleavage at 37°C for 20 hours and the enzymatic reaction was stopped by SDS-PAGE loading buffer (22.5 mM Tris, 12 mM Na3P0,, pH 6.8,6%SDS, 3% f3-mercaptoethanol, 30% glycerol, 0.01% bromopheno1 blue [Laemmli, 1970; Lotan and Irimura, 19871) and placed in a boiling water bath for 3 minutes. Endoglycosidase (F,D, or H) treated (10 pg) and untreated cell lysate (10 pg) were separated using SDS-PAGE

Lectin staining Cell lysate and media protein preparations were prepared as described above. Proteins were separated by electrophoresis and transferred to nitrocellulose. Lectin staining was performed with the Vectastain ABC Kit (Vector Laboratories) and described elsewhere (Cai et al., 1991). Northern blot analysis Total RNA was extracted from cells using the RNAzol procedure (CinndBiotecx Laboratories) (Chomczynski and Sacchi, 1987). Inhibit-ace (5 Prime> 3 Prime, Inc.) was used t o reduce RNAase activity. Following photometric measurement of RNA a t 260 nm, poly

ROSS ET AL.

266

5000

A

oDMSO

-

H RA

4000

3000

L T

2000

1000

0

L 1.25

2.5

10

5

Cell Number ( x

I

20

lo4)

3000 DMSO RA

b

b

a

I

2000

T

1000

T

BI

a

1

T

a

I

0

I

PL

GL

LM

FN

Attachment Substrates

Fig. 2. RA increases cell attachment t o various substrates. A: Relation of fluorescence to cell number was established for each treatment group. Six wells each of 10 x lo4, 5 x lo4, 2.5 x lo4, and 1.25 x lo4 cells were added to wells with MUH to determine the ratio of cell number to dye uptake for each treatment group a t 30 minutes. B: Six

wells per treatment group were used. PL = plastic; GL = gelatin; LM = laminin; FN = fibronectin. Means are significantly different within each substrate group (P< 0.051, as indicated by letters “a” and “b.”

RETINOIC ACID ENHANCES ADHESIVENESS OF F9 CELLS

A

.-v)

CI

5

1500

g

1250

5

1000

?!?

750

0

2

i 2 .-E

-9

b

500

I

+RA

250

0

a

Hour of Incubation R v)

c .-

5

g 5: $ ! 0

Y

3

500-

i i E c. .-

250-

*

2 a-

I

I 1750150012501000750-

01

2

3

4

5

Hour of Incubation

C 1

1

2

3

4

8

Hour of Incubation Fig. 3. RA increases the initial rate of F9 cell attachment to gelatin. A: RA-treated cells. B: Control cells. Different letters indicate significant (P < 0.05) differences in means. C: Ratio of values in A over values in B. Means for RA and control cells are significantly different (P < 0.05)except for the 8-hourvalues. Ratios of the means are plotted and different letters indicate significant differences (P < 0.05) among the ratios. Six wells per treatment group were used. Cells were treated for 48hours and used in attachment assays at 5 x lo4cells per well. Data are from t w o experiments (1,2,3,and 4 hours of attachmentand 2,4, and 8 hours ofattachment).Proportionsof attached cells (e.g.,7%for control vs. 17%for RA-treated in the first hour) is calculated from fluommetric readings for each sample (e.g., 297 for control and 685 for RA-treated over a total of 3999, at 1 hour). S i x wells per treatment group were used.

267

(A+ImRNA was isolated using oligo(deoxythymidine) cellulose (Pharmacia) column (Davis et al., 1986). Poly(A+) mRNA was eluted with 10 mM Tris, pH 7.4,l mM EDTA, and 0.05% SDS. Poly (A+)mRNA was concentrated using l/lOth vol of 3 M sodium acetate and 2.5 vol of 100%ethanol. The samples, after drying, were resuspended in dH,O. Electrophoresis, transfer to nitrocellulose, and Northern hybridization were performed as described previously (Sambrook et al., 1989). The cDNA probes for RARa and RARy were obtained from Dr. Pierre Chambon (Laboratoire de Genetique Moleculaire des Eucaryotes du CNRS, L'Inserm, Strasbourg, France). The complementary DNA was digested with the appropriate restriction enzymes to yield sequences specific for the D-F regions of the receptors. The cDNA sequences were then separated by gel electrophoresis and eluted from DEAE-nitrocellulose membrane (Schleicher and Schuell, Keene, NH) prior to hybridization. A 1,135 bp RARa cDNA was obtained after endonuclease digestion with Sacl and EcoRI. The RARy (990 bp) was obtained after endonuclease digestion with Sacl and EcoRI. Probes were radiolabeled by the double random ~riming-~'P label procedure (Loftstrand Labs Unlimited, Gaithersburg, MD). One hundred twenty x lo6 DPM of probe/lO ml of buffer was used in hybridization.

RESULTS RA enhances laminin synthesis in RA-differentiated F9 cells Immunoprecipitation of 35S-laminin from 35S-methionine labeled media protein and cell lysate preparations (data not shown) showed that laminin biosynthesis was stimulated after 72 hours of treatment in both RA and RA plus dcAMP media. Figure 1 shows that both RA and RA plus dcAMP enhanced laminin secretion in the media of F9 cell cultures (lanes 7 and 9). These results support the concept that true differentiation to primitive and parietal endoderm is occurring in our system.

RA enhances F9 cell attachment Relation of uptake of the fluorescent dye MUH to cell number was established for each treatment group (Fig. 2A). We found that the dye uptake was proportional to the cell number and that the same proportionality was seen for both RA and control cells over a range from 1.25 x lo4 to 20 x lo4. Figure 2B shows the percent of control cells that attach to plastic (21%), gelatin (22%), laminin (42%), and fibronectin (48%) in a 2-hour attachment assay. Control cells displayed the greatest cell attachment for laminin and fibronectin but they also attached to plastic and gelatin. RA-treated cells displayed an increased attachment to all substrates (Fig. 2B) in the following order: laminin (57%),fibronectin (56%),plastic (32%),and gelatin (30%).Treatment of F9 cells with RA plus dcAMP caused a small but reproducible increase in attachment to laminin (from 3 4 4 0 % ) as well as to plastic (from 254%). We next compared initial and maximal rates of cell attachment to gelatin and plastic. RA stimulated the initial rate of cell attachment to gelatin from 7% (Fig. 3B) to 17% in the first hour (Fig. 3A). Control cells

268

ROSS ET AL.

Fig. 4. Laminin contains N-linked sugar chains. ENDO-F-treated (10 pg protein per lane) and untreated whole cell lysates (20 kgprotein per lane) were analyzed by Western blot, using rabbit anti-mouse laminin IgG. Cells were treated with DMSO, dcAMP, RA,and RA plus dcAMP for 72 hours, See Materials and Methods for details.

showed a cell attachment increase over a period of 8 hours, eventually reaching the level of attachment (26%) displayed by the RA-treated cells (Fig. 3B,C). When the rate of cell attachment to plastic was measured, RA-treated cells showed enhanced attachment to plastic at all time points (from 2 9 4 6 %at 1 hour, from 33-56% at 2 hours, from 3 6 4 9 % at 3 hours, and from 2 6 4 6 %at 4 hours) (data not shown). Laminin from RA-differentiated F9 cells contain complex a n d hybrid oligosaccharide chains Treatment with ENDO-F caused a shift in the bands consistent with a reduction in the sizes of the A (400 Kd) and B (200 Kd) chains of laminin as judged by SDS-PAGE (Fig. 4). This demonstrates that laminin from RA-differentiated cultures contains N-linked oligosaccharide chains. Furthermore, Western blot staining with the griffonia simplicifolia lectin-1 (GSL-1)lectin, specific for galactose a 1,3 galactose residues, was detected in the RA cultures for the 400 and 200 Kd bands (data not shown), but in the presence of ENDO-F the GSL-1 staining was lost. This suggests that RAdifferentiated F9 cell glycoconjugates contain terminal galactose a 1,3 galactose residues. ENDO-H treatment (specific for hybrid and high mannose structures) shows lower molecular weight bands for the laminin chains (Fig. 5). ENDO-D treatment (with specificity for high mannose oligosaccharides),however, did not affect the mobility of the 400 (A)or 200 (B) Kd laminin chains

(not shown). Therefore, we conclude that laminin contains mainly complex and hybrid oligosaccharide structures. Galactose incorporation into laminin oligosaccharide chains To determine the extent of alactosylation of RAdifferentiated F9 cells we used 8H-galactose in a metabolic labeling procedure. Analysis of the sugar labeled media protein by SDS-PAGE and subsequent autoradiography revealed enhanced labeling of the 200 and 400 Kd bands in RA plus dcAMP cultures (Fig. 6). Using the total CPM from immunoprecipitated laminin (200 pg of starting protein) we could determine that RA increased the total labeling from 3H-galactose of laminin from cell lysate (data not shown) and media extracts. In the media, the increase by RA was from 13-28 fmol (2.2 fold) and with RA plus dcAMP the increase was from 17-72 fmol (4.2 fold) of galactose. We were interested in determining if the enhanced labeling from 3H-galactose in RA-differentiated cells was due to enhanced glycosylation of the laminin molecules or to overall increase in laminin biosynthesis. For this ur pose, we labeled cells with both 3H-galactose and'%: methionine to determine the ratio of sugar to amino acid incorporated into laminin. Upon immunoprecipitation of laminin from our labeled extracts we found that RA caused a parallel increase in the incorporation of both labels (Table 1). We also show that the fold

RETINOIC ACID ENHANCES ADHESIVENESS OF F9 CELLS

OMSO

dcAMP

Rh

269

RA+dcAMP

Fig. 5. Laminin contains hybrid type oligosaccharidestructures.ENDO-H-treatedand untreated whole cell lysates were analyzed by Western blot (10 kg protein per lane), using rabbit anti-mouse laminin IgG. Cells were treated with DMSO, dcAMP, RA, or RA plus dcAMP for 72 hours. See Materials and Methods for details.

increase by RA plus dcAMP was similar for both labels. Therefore, we conclude that RA increases the biosynthesis of the entire laminin molecule. Lectin staining of RA-differentiated laminin We were also interested in the types of sugar chains containing galactose in laminin. For this purpose we used biotinylated lectins in Western blots. We found that soy bean agglutinin (SBA; specificity for terminal 01 or p-N-acetylgalactosamine or galactose), Duturu strumonium lectin (DSL; recognition of the polylactosamine structure) (Fig. 7), and Lycopersicon esculenturn lectin (LEL; specific for the polylactosamine structure) (Table 2) showed enhanced staining for the 200 and 400 Kd bands in RA plus dcAMP media. Interestingly, these lectins showed greater staining for the 400 Kd band suggesting that polylactosamine chains are mainly on the A chain of laminin. Ricznus cornmunis agglutinin (RCA; specific for galactose) and GSL-1 (which recognizes the galactose (Y 1,3galactose residue) showed enhanced staining for 200 and 400 Kd bands in RA and RA plus dcAMP cultures mainly on the 200 Kd band (Fig. 8), suggesting that the B chain of laminin contains mainly the galactose 01 1,3 galactose structure.

RA treatment (Fig. 10) three fold as shown by densitometric scanning. RA receptors Northern Blot analysis shows an increase in RARa and RARy in the RA-treated cultures within a few hours of treatment (data not shown).

DISCUSSION During embryogenesis the basement membrane is involved in migration and segregation of embryonic cells before specialization into specific tissues (Hogan and Tilly, 1981; Gardner, 1983; Grabel and Watts, 1987). ECM proteins permit precursor cells to migrate from primitive endoderm onto the trophoblast and become parietal endoderm, whereas the stationary cells become visceral endoderm (Hogan et al., 1983). Greenburg and Hay (1986)have shown that culturing embryonic avian lens epithelial cells in a suspension system containing collagen type 1 caused these cells, which normally produce &crystallin, to convert to mesenchyma1 cells, which produce collagen type 1. Because of this proposed role for ECM components in embryogenesis and the enhanced biosynthesis of these components during RA-induced differentiation of F9 cells we studLaminin receptors ied the effect of RA on the ability of F9 cells to attach to RA treatment of F9 cells had opposing effects on the various ECM substrates. We observed a consistent, small increase in attachamounts of the LBP-37 and the p l integrin subunit. Figure 9 shows that immunoprecipitated 35S-methio- ment of F9 embryonal carcinoma cells to plastic as well nine-labeled LBP-37 from cell lysate was decreased fol- as to gelatin, laminin, and fibronectin substrates after lowing RA as well as RA and dcAMP. In contrast im- RA. Our finding that control F9 cells displayed a munoprecipited p l integrin subunit was increased by greater attachment to laminin (42%) and fibronectin

ROSS ET AL.

270 MEDIA

Fig. 6. RA and RA plus dcAMP increase incorporation of 3H-galactose. Autoradiography of gels after electrophoresis of 3H-galactoselabeled media protein extracts (20 pg protein per lane). DMSO, dcAMP, RA, and RA plus dcAMP cells were treated for 72 hours, as described in Materials and Methods, and subsequently labeled for 18 hours with 3H-galactosein media.

TABLE 1. RA causes a parallel increase in '%methionine and 3H-galactose incorporation into laminin' Sample

35S-methionine

3H-salactose

CPM per immunoprecipitated laminin (200 Fg of starting protein) DMSO ND2 36 dcAMP 376 312 RA 896 512 RA t dcAMP 7,624 5,300 Fold increase in label incorporation RA dcAMP/DMSO >147 147 RA + dcAMPldcAMP 20 17 RA + dcAMPIRA 9 10

+

'Two CPM measurementsper treatment group. 2Nondeteetable.

(48%)compared to plastic (21%)or gelatin (22%) is in agreement with a previous report (Tienari et al., 1989). We also found that optimal cell adhesion of control cells on gelatin was achieved at 8 hours whereas RA-treated F9 cells reached it a t 2 hours. This finding is consistent with the induction of a qualitative change on the surface of RA-treated cells. RA-treated cells showed a 1.6 fold increased attachment to plastic during the entire

4-hour period of the attachment assay, consistent with an increased amount of cell adhesion molecules at the cell surface. Tienari et al. (1989) found that F9 cells induced to differentiate into visceral or parietal endoderm showed a reduced adherence to laminin, while attachment to fibronectin remained unchanged. These workers used 5 x lo-' M RA to differentiate cells within 4-5 days, instead of our 1 x lop7 M RA treatment for 72 hours, and caused cells to differentiate as measured by an increase in laminin production. Further differences include the method of coating plates with attachment substrates as well as the actual attachment assay and the measurement of cell numbers. Other work has described the involvement of retinoids in the increased cell-to-substratum adhesion of fibroblast cell lines (Adamo et al., 1978, 1979, 1983; Sasak et al., 1980; De Luca et al., 1980; Kato and De Luca, 1987; Cai et al., 1991;Jetten et al., 1979;Bertram et al., 1981; Bertram, 1980) as well as their effects on the synthesis of ECM proteins (Wang and Gudas, 1983; Bernard et al., 1984). A suggested mechanism in cell adhesion is that the carbohydrates of laminin may be interacting with a carbohydrate binding protein or lectin on the surface of the cells. For example, it has been shown by Lotan et al. (1989,1991)that RA-induced differentiation of F9 cells results in enhanced expression of a 14.5 Kd lectin, thought to be a LBP. Another possibility is that RA may upregulate the synthesis of cell surface integrin receptors, as indicated by recent work (Dedhar et al., 1991; Rossino et al., 1991). We report that laminin from RA-differentiated F9 cells was found to contain N-linked oligosaccharides, primarily of the complex and hybrid types. These findings are in agreement with structural analysis of laminin carbohydrates of other cell systems (e.g., Engelbreth-Holm-Swarm chondrosarcoma) (Arumugham et al., 1986; Knibbs et al., 1989; Rao et al., 1983; Martin and Timpl, 1987). Our results also show that a parallel increase in both laminin protein and carbohydrate labeling occurs in RA-differentiated cells. Interestingly, the A and B chains of laminin were the predominant protein from 3H-galactoselabeling of these cells (Fig. 6). Lectin binding specificity (Table 2) shows that the A chain mostly contains a galactose in polylactosamine structures (DSL and LEL) and the B chain in galactose ci 1,3 galactose (GSL-1and RCA) type of structure. Table 2 summarizes these findings and the binding of nine different lectins to the A or B chains of laminin. Five (RCA, SBA, GSL-1, DSL, and LEL) of these lectins, which recognize terminal galactose, showed enhanced staining in RA-treated cells. Staining with other lectins, peanut agglutinin (PNA; sugar specificity galactose p 1,3 N-acetylgalactosamine) and Maackia amurensis lectin I (MAL-I;sugar specificity galactose p 1,3 N-acetylglucosaminef,which recognize sugars common in 0-linked oligosaccharides, was minimal and did not show enhancement due t o RA treatment. Recently, two groups have reported a correlation between RA-induced differentiation and upregulation of integrin subunits. In one study the upregulation of the avpl integrin by RA in P19 murine embryonal carci-

RETINOIC ACID ENHANCES ADHESIVENESS OF F9 CELLS

271

Fig. 7. RA plus dcAMP enhances SBA and DSL lectin staining in 200 and 400 Kd bands. Western blot analysis of media protein from F9 cells treated with DMSO, dcAMP, RA, or RA plus dcAMF’ for 72 hours (30 pg protein per lane). Lectins were used to probe for specific sugar residues.

TABLE 2. Lectin binding to laminin

Latin RCA SBA GSL-I PNA MAL-I

MAL-II DSL LEL PHA-E: Phnseolw vulgaris

Suear suecificitv Terminal galactose Terminal LI or P-N-acetylgalactosamineor galactose Galactose a 1,3 galactose Galactose p 1,3 N-acetylgalactosamine Galactose p 1,3 N-acetylglucosamine Sialic acid in a 2,3 linkage Polylactosamine structures Polylactosamine structures Bisecting N-acetylglucosamine

Effect of RA and dcAMP on lectin binding to laminin subunits 400 Kd (A) 200 Kd (B)

tl t

tt

tt

t

tt

-

-

-

-

-

tt

t

tt

t

t

t

agglutinin-erythroagglutinin lArrows pointing up indicate either a moderate ( t )or large ( 7 effect.

noma cells was correlated with the induction of neuronal differentiation (Dedhar et al., 1991). In the other study, an increased expression of the alp1 integrin was associated with increased neurite outgrowth in response to laminin (Rossino et al., 1991). Cell attachment was not measured in either study. Our studies on laminin receptors show a diverse effect of RA on the expression of LBP-37 and Pl integrin with a decrease in

t 1 stimulationhy RA plus dcAMP.The -.symbol indicates lack of

the former and an increase in the latter. This is consistent with the 91 integrin being specifically involved in the observed increased attachment of RA-treated cells. The observed upregulation of RARa and RARy expression and previous work on RARP expression (Zelent et al., 1989; Hu and Gudas, 1990; Martin et al., 1990) are consistent with the early action of RA on these receptors, followed by increased laminin and pl

272

ROSS ET AL.

> -= rn

250

C

n 200

DMSO

dcAMP

RA

Treatment Fig. 8. RA plus dcAMP enhances GSL-1 and RCA lectin staining in 200 and 400 Kd bands. Western blot analysis of media protein from F9 cells treated with DMSO, dcAMP, RA, or RA plus d c M for 72 hours (30 pg protein per lane). Lectins were used to probe for sugar residues.

Fig. 10. RA enhances p l integrin subunit. Anti-pl integrin antibody (+Ab) was used to immunoprecipitate the extract (200 pg protein was used). Bands on Western blots were quantified by densitometry with a Molecular Dynamics Personal Densitometer with MD ImageQuant Software 3.22.

integrin synthesis, finally resulting in increased cell attachment.

LITERATURE CITED

Fig. 9. RA decreases the LBP-37 in F9 cells. Anti-LBP-37 antibody was used to immunoprecipitate the LBP-37 from F9 cell extracts (200 pg of protein). Cells were treated with DMSO, dcAMP, RA, and RA plus dcAMP for 72 hours. Western blot analysis of the immunoprecipitates was performed. Arrows indicate samples processed without antibody.

Adamo, S., Akalovksy, I., and De Luca, L.M. (1978)Retinoid-induced adhesion of spontaneously-transformed mouse fibroblasts (Balbic 3T12-3 cells). Proc. Am. Assoc. Cancer Res., 19.27 (abs.) (Abstract). Adamo, S., De Luca, L.M., Akalovsky, I., and Bhat, P.V. (1979) Retinoid-induced adhesion in cultured transformed mouse fibroblasts. J . Natl. Cancer Inst., 62r1473-1478. Adamo, S., Sasak, W., Dion, L.D., and De Luca, L.M. (1983) Studies on the mechanism of retinoid-induced adhesion of spontaneously transformed mouse fibroblasts. Acta Vitaminol. Enzymol., 5.3-10. Arumugham, R.G., Hsieh, T.C., Tamer, M.L., and Laine, R.A. (1986) Structures of the asparagine-linked sugar chains of laminin. Biochim. Biophys. Acta, 883:112-126. Bernard, B.A., De Luca, L.M., Hassell, J.R., Yamada, K.M., and Olden, K. (1984)Retinoic acid alters the proportion of high mannose to complex type oligosaccharides on fibronectin secreted by cultured chondrocytes.J . Biol. Chem., 259:5310-5315. Bertram, J.S. (1980) Structure-activity relationships among various retinoids and their ability to inhibit neoplastic transformation and to increase cell adhesion in the C3HilOT112 CL8 cell line. Cancer Res., 40t31413146. Bertram, J.S., Mordan. L.J., Blair. S.J.. and Hui. S. (1981) Effects of retinoids on neoplastic transformation, cell adhesion, and membrane topography of cultured 10T112 cells. Ann. N. Y. Acad. Sci., 359:21&236. Bjeurum, O.J., and Aeegaard, N.H.H. (1988)CRC Handbook of Immunoblotting of Proteins. CRC Press Inc, Boca Raton, FL. Cai, D., Webber, M., and De Luca, L.M. (1991) Retinoids enhance lectin binding to gp130, a glycoprotein of NIH-3T3 cells: Correlation with cell growth and adhesion. Exp. Cell Res., 192:366372. Carlin, B.E., Durkin, M.E., Bender, B., Jaffe, R., and Chung, A.E.

RETINOIC ACID ENHANCES ADHESIVENESS OF F9 CELLS (1983) Synthesis of laminin and entactin by F9 cells induced with retinoic acid and dibutyryl cyclic AMP. J. Biol. Chem., 258~772% 7737. Chomczynski, P., and Sacchi, N. (1987) Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroformextraction. Anal. Biochem., 162t156159. Cooper, A.R., Taylor, A., and Hogan, B.L. (1983) Changes in the rate of laminin and entactin synthesis in F9 embryonal carcinoma cells treated with retinoic acid and cyclic AMP. Dev. Biol., 99.510-516. Davis, L.G., Dibner, M.D., and Battery, J.F. (1986) Basic Methods in Molecular Biology. Elsevier Science Publ. Co. Inc., New York. Dedhar, S., Robertson, K., and Gray, V. (1991)Induction ofexpression of the alpha v beta 1 and alpha v beta 3 integrin heterodimers during retinoic acid-induced neuronal differentiation of murine embryonal carcinoma cells. J. Biol. Chem., 266r21846-21852. DeLuca, L.M., Sasak, W., Adamo, S.,Bhat, P.V., Akalovsky, I., Silverm a d o n e s , C.S., and Maestri, N. (1980) Retinoid metabolism and mode of action. Environ. Health Perspect., 35:147-152. Dotsika, E.N., and Sanderson, C.J. (1987) A fluorometric assay for determining cell growth in lymphocyte proliferation and lymphokine assays. J. Imrnunol. Methods, 105:55-62. Gardner, R.L. (1983) Origin and differentiation of extraembryonic tissues in the mouse. Int. Rev. Exp. Pathol., 24:63-133. Gill, J.L. (1993) Design and Analysis of Experiments in the Animal and Medicinal Sciences. Iowa State University Press, Ames, Iowa, Vol. 1. Grabel, L.B., and Watt, T.D. (1987) The role of extracellular matrix in the migration and differentiation of parietal endoderm from teratocarcinoma embryoid bodies. J . Cell Biol., 105r441.448. Greenburg, G., and Hay, E.D. (1986) Cytodifferentiation and tissue phenotype change during transformation of embryonic lens epithelium to mesenchyme-like cells in vitro. Dev. Biol., 115:363379. Gudas, L.J. (1991) Retinoic acid and teratocarcinoma. Semin. Dev. Biol., 2:171-179. Hogan, B.L., and Tilly, R. (1981) Cell interactions and endoderm differentiation in cultured mouse embryos. J . Embryol. Exp. Morphol., 62:379-394. Hogan, B.L.M., Taylor, A,, and Adamson, E. (1981) Cell interactions modulate embryonal carcinoma cell differentiation into parietal or visceral endoderm. Nature, 291:235-237. Hogan, B.L.M., Barlow, D.P., and Tilly, R. (1983) F9 teratocarcinoma cells as a model for the differentiation of parietal and visceral endoderm in the mouse embryo. Cancer Surv., 2:115-140. Hu, L., and Gudas, L.J. (1990) Cyclic AMP analogs and retinoic acid influence the expression of retinoic acid receptor alpha, beta, and gamma mRNAs in F9 teratocarcinoma cells. Mol. Cell. Biol., 10: 391-396. Jetten, A.J., Jetten, M.E., Shapiro, S.S., and Poon, J.P. (1979)Characterization of the action of retinoids on mouse fibroblast cell lines. Exp. Cell Res., 119:289-299. Kato, S., and De Luca, L.M. (1987) Retinoic acid modulates attachment of mouse fibroblasts to laminin substrates. Exp. Cell Res., 173:450--162. Knibbs, R.N., Perini, F., and Goldstein, I.J. (1989) Structure of the major concanavalin A reactive oligosaccharides of the extracellular matrix component laminin. Biochemistry, 28:63794392. Laemmli, U.K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T,. Nature, 277:680-685. Lotan, R., and Irimura, T. (1987) Enhanced glycosylation of a melanoma cell surface glycoprotein by retinoic acid: Carbohydrate chain analysis by lectin binding. Cancer Biochem. Biophys., 9:211-221. Lotan, R., Lotan, D., and Carralero, D.M. (1989)Modulation of galactoside-binding lectins in tumor cells by differentiation-inducing agents. Cancer Lett., 48:11&122. Lotan, R., Lotan, D., and Amos, B. (1991)Differentiation of F9 embryonal carcinoma (EC) cells is accompanied by modulation of lactoside-and laminin binding lectins. Proc. Am. Assoc. Cancer Res., 32:13(Abstract).

273

Marotti, K.R., Brown, G.D., and Strickland, S. (1985)Two-stage hormonal control of type IV collagen mRNA levels during differentiation of F9 teratocarcinoma cells. Dev. Biol., 108:2641. Martin, G.R., and Timpl, R. (1987)Laminin and other basement membrane components. Annu. Rev. Cell Biol., 357-85. Martin, C.A., Ziegler, L.M., and Napoli, J.L. (1990) Retinoic acid, dibutyryl-CAMP, and differentiation affect the expression of retinoic acid receptors in F9 cells. Proc. Natl. Acad. Sei. USA, 87r48044808. Olden, K., Hunter, V.A., and Yamada, K.M. (1980) Biosynthetic processing of the oligosaccharide chains of cellular fibronectin. Biochim. Biophys. Acta, 632:408-416. Rao, C.N., Goldstein, I.J., and Liotta, L.A. (1983) Lectin-binding domains on laminin. Arch. Biochem. Biophys., 227r118-124. Rossino, P., Defilippi, P., Silengo, L., and Tarone, G. (1991)Up-regulation of the integrin cx l/p 1in human neuroblastoma cells differentiated by retinoic acid Correlation with increased neurite outgrowth response to laminin. Cell Regul., 2:1021-1033. Sambrook, J., Fritsch, E.F., and Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press, New York, 2nd Ed. Sasak, W., De Luca, L.M., Dion, L.D., and SilvermanJones, C.S. (1980) Effect of retinoic acid on cell surface glycopeptidesof cultured spontaneously transformed mouse fibroblasts (Balb/c 31'12-3 cells). Cancer Res., 40:194&1949. Smith, P.K., Krohn, R.I., Hermanson, G.T., Mallia, A.K., Gartner, F.H., Provenzano, M.D., Fujimoto, E.K., Goeke, N.M., Olson, B.J., and Klenk. D.C. (1985) Measurement of orotein using bicinchoninic acid [published erratum appears in Anal Biochem 1987 May 15;163(1):2791. Anal. Biochem., 150r76-85. Stadler, R., Detmar, M., Stephanek, K., Bangemann, C., and Orfanos. C.E. (1989) A rapid fluorometric assay Tor the determination of keratinocyte proliferation in vitro. J . Invest. Dermatol., 93.532534. Steel, R.G.D., and Torrie, J.H. (1980) Principles and Procedures of Statistics: A biometrical Approach. McGraw-Hill, Inc, New York, 2nd Ed. Strickland, S., and Mahdavi, V. (1978) The induction of differentiation in teratocarcinoma stem cells by retinoic acid. Cell, 15:393403. Strickland, S., Smith, K.K., and Marotti, K.R. (1980) Hormonal induction of differentiation in teratocarcinoma stem cells: Generation of parietal endoderm by retinoic acid and dibutyryl CAMP. Cell, 21: 347-355. Thompson, J.N., Howell, J.M., Pitt, G.A.J., and McLaughlin, C.I. (1969) The biological activity of retinoic acid in the domestic fowl and the effects of vitamin A deficiency on the chick embryo. Br. J , Nutr., 23:471490. Tienari. J., Lehtonen. E., Vartio. T., and Virtanen. I. (1989) Embrvonal carcinoma cells adhere preferentially to fibronectin and lamikin but their endodermal differentiation leads to a reduced adherence to laminin. Exp. Cell Res., 182:2642. Towbin, H., Staehelin, T., and Gordon, J . (1979)Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: Procedure and some applications. Proc. Natl. Acad. Sci. USA, 76: 4350-4354. Wang, S.Y., and Gudas, L.J. (1983)Isolation of cDNA clones specific for collagen IV and laminin from mouse teratocarcinoma cells. Proc. Natl. Acad. Sci. USA, 80.5880-5884. Wang, S.Y., and Gudas, L.J. (1984) Selection and characterization of F9 teratocarcinoma stem cell mutants with altered responses to retinoic acid. J Biol. Chem., 2595899-5906. Wang, S.Y.,and Gudas, L.J. (1990) A retinoic acid-inducible mRNA from F9 teratocarcinoma cells encodes a novel protease inhibitor homologue. J. Biol. Chem., 265r1581tL15822. Zelent, A., Krust, A,, Petkovich, M., Kastner, P., and Charnbon, P. (1989)Clonmg of murine alpha and beta retinoic acid receptors and a novel receptor gamma predominantly expressed in skin. Nature, 339r71P717.