Molecular and Cellular Biochemistry 96: 5%67, 1990. © 1990 Kluwer Academic Publishers, Printed in the Netherlands.

Original Article

The kinetics of fibronectin synthesis and release in normal and tumor promoter-treated human lung fibroblasts Beth A. Burrous 1 and George Wolff

Department of Applied Biological Sciences, Mas'sachusetts Institute of Technology, Cambridge, MA 02139, USA; Present address: 1 U.S. Patent and Trade Mark Office, Arlington, VA 22202; 2Department of Nutritional Sciences, University of California, Berkeley, CA 94720, USA Received 28 August 1989; accepted 8 December 1989

Key words: fibronectin, tumor promoter, 12-0-tetradecanoyl phorbol-13-acetate, TPA, cell surface

Summary The present study is a detailed kinetic analysis of the synthesis, release and multimerization of fibronectin (FN) in normal and tumor promoter-treated human lung fibroblasts. Pulse/chase and surface labeling experiments were performed to follow the fate of both newly synthesized and preexisting cell-surface FN over time. The majority of FN (80%) left the intracellular compartment within one hour of synthesis. However, the rate of direct secretion was very low and after one hour, 70% of newly synthesized FN was still at the cell surface. This material was primarily dimeric. Dimeric and multimeric (very high molecular weight) FN was detectable at the cell surface and in the medium 4 hours after synthesis. Pulse-labeled FN multimer levels peaked at 12 hours and declined thereafter. After 24 hours, 85% of pulse-labeled FN had been shed into the medium and the labeled FN remaining at the cell surface was primarily multimeric. Surface labeling experiments confirmed that the majority of FN resides at the cell surface prior to release into the medium. One hour treatment with the phorbol ester tumor promoter, 12-0-tetradecanoyl phorbol-13-acetate (TPA), stimulated a nine-fold increase in release of preexisting, dimeric cell-surface FN (125I-labeled). The maj or effect of longer term TPA treatment up to nine hours was continued depletion of dimeric cell-surface FN. Increased release of cell-surface multimeric FN was also stimulated by TPA, but to a much lesser extent. Release of newly synthesized (pulse-labeled) dimeric FN was also stimulated by TPA though much less than pre-existing FN, and TPA treatment produced a small decrease in the steady-state level of multimeric FN. Thus, preexisting cell-surface FN and newly synthesized FN differ dramatically in their susceptibility to TPA treatment.

Abbreviations: FN - fibronectin, HLF - human lung fibroblasts, PMSF - phenylmethylsulfonylfluoride, TPA - 12-0-tetradecanoyl phorbol-13-acetate, DMSO - dimethylsulfoxide, MEM - minimal essential medium, ELISA - enzyme linked immunosorbent assay, PBS - phosphate-buffered saline, BSA - bovine serum albumin, SDS-PAGE- sodium dodecylsulfate-polyacrylamidegel electrophoresis, D O C - deoxycholate, D~IT- dithiotreitol

58 Introduction

Materials and methods

Fibronectin is a large extracellular glycoprotein found in the plasma and as a major constituent of extracellular matrices. In addition to being an important constituent of the extracellular matrix, FN plays roles in opsonization, blood clotting, and would healing [1]. In vitro, FN is found on the surfaces and in the medium of cultured cells. In cultured human fibroblasts, newly synthesized FN is released into the medium or deposited on the cell surface where it is disulfide-crosslinked into a multimeric form. Cell-surface FN is gradually shed into the medium [2]. There is still little known about how dimeric and multimeric FN are anchored to the cell surface. Integrins are a class of heterodimeric cell-surface proteins, some of which have been shown to be involved in binding FN to the cell surface [3]. There is evidence that integrins may form part of the transmembrane link between the cytoskeleton and the extracellular matrix. McKeown and Mosher [4] have characterized FN multimerization at the cell surface of human skin fibroblasts and have shown that multimerization receptor site(s) in FN are located in a 70,000 dalton N terminal region of the molecule. Upon oncogenic transformation, many cell types lose cell-surface FN despite continued synthesis [1]. FN plays an important role in anchoring cells to the extracellular matrix. The loss of FN from the surface by changing the adhesive properties of transformed cells, may augment metastasis. The chemical tumor promoter, 12-O-tetradecanoylphorbol-13 acetate (TPA), stimulates the loss of FN from the surface of cultured cells and the cell surface eventually becomes stripped of surface FN [5, 6]. It has been demonstrated in one cell system that the loss of FN is linked to the promotion phase of carcinogenesis [7]. We here report observations on the kinetics of FN release and multimerization in normal and TPA-treated human lung fibroblasts.

Materials

Human lung fibroblasts (HLF) were from American Type Culture Collection, (Rockville, MD); human plasma FN and ultra pure urea from Bethesda Research Laboratories, (Gaithersburg, MD); gelatin-Sepharose was purchased from Pharmacia Co., (Piscataway, NJ); TPA and DMSO from Sigma Co., St. Louis, MO; rabbit anti-human FN antibody from Cooper Biomedical Co., Malvern, PA; Enzymobeads and goat anti-rabbit IgG Sepharose from Biorad Co., Rockville Centre, NY; 35S-methionine (> 800 Ci/mmole), 125I-sodium iodide (17 Ci/mg) and En3hance from New England Nuclear, Boston, MA.

Cell culture

HLF between passages 8 and 14 were used in all experiments. They were grown in MEM which was purchased from Sigma Co., (St. Louis, MO). Medium was supplemented with penicillin (500,000 units/l), streptomycin (500,000/xg/1), nonessential amino acids, sodium pyruvate (1 mM), 0.1% lactalbumin hydrolysate and 10% fetal bovine serum, all of which were purchased from Flow Laboratories (McLean, VA). Cells were routinely split 1:4 after reaching confluency (every 4-6 days).

Measurement of total F N in medium and cell extracts

FN content in cell extracts and in medium was analyzed using an enzyme-linked-immunosorbentassay (ELISA). The method used was a modification of the method of Ruoslahti, et al. [8]. Cells were dissolved in 8 M urea/l% Triton X-100/1 mM PMSF, as described by Mosher and Vaheri [9]. Medium was clarified by centrifugation at 1000 x g for 5 min. Cell extracts and clarified medium were diluted in dilution buffer (PBS with 1 mM PMSF, 0.5% Tween-20 and 0.1% BSA) prior to assay. A standard curve using purified human plasma FN

59 was linear between 40 and 666 rig. Those sample dilutions falling within the linear range of the assay were used for determination of FN values. Values for identical samples assayed in parallel varied by less than 10%. Background was determined by performing the ELISA in wells that were incubated with dilution buffer alone.

Determination of FN-associated radioactivity The method used was that described by Zerlauth and Wolf [7]. Diluted radioactive cell extracts and medium were added to gelatin-coated plates and incubated for three hours as described for the ELISA. At this time, plates were rinsed three times with PBS, once with PBS containing 1.5M NaC1 and then three times with PBS. Bound FN was eluted with 8 M urea/l% SDS/1 mM PMSF. Plates were rinsed with water and the rinses were added to the extracts for counting. Radioactivity of 35Smethionine samples was determined by scintillation counting and radioactivity of 125I-labeled samples was determined by gamma counting. Background was calculated using diluted FN-free radioactive medium and was subtracted from sample radioactivities. Autoradiography of SDS-PAGE of gelatin-coated plate-bound material from medium and cell extracts of radioactively labeled cells showed that material eluted from the plate was FN. Multimeric and dimeric radioactive FN were observed in unreduced samples and all material migrated as monomer when samples were reduced. Samples assayed after adsorption to gelatincoated plates contained no detectable FN as quantified by ELISA, indicating that FN binding to the plates was complete.

Immunoprecipitation For immunoprecipitation following labeling with 35S or 125I, cells and medium were processed as follows: medium was removed and centrifuged at 1000 x g for 5 minutes; cells were washed three times with PBS and dissolved in cell lysis buffer

containing 2% SDS, 0.02M Tris-HC1 (pH8.3), 2 mM PMSF, 2 mM EDTA, 2 mM iodoacetic acid, 2mM N-ethyl maleimide, 0.1% BSA and 0.15M NaC1. Polyclonal rabbit anti-human FN antibody was diluted 1:3000 in PBS with 2raM PMSF and 0.1% BSA at 4txg/ml. Goat anti-rabbit immunobeads (Sepharose beads coupled to goat anti-rabbit FN IgG) were suspended in PBS with 2 mM PMSF and 0.1% BSA at 4 txg/ml. Cell extracts were centrifuged at 10,000 x g for 5 minutes. 50/xl of cell extract was incubated with 100/xl of antibody diluted 1:3000 and 1000txl of 50mM Tris (pH8.3) containing 0.5% DOC and 2 mM PMSF for one hour at 37 ° C, with shaking. Immunobeads were added and the samples were shaken at 4°C overnight. For medium samples, 200/xl of medium was used and concentrated cell lysis buffer was added so that the final concentrations of all components were equal in the cell extract and medium samples. After overnight precipitation, the precipitated samples were washed three times with the DOC buffer.

SDS-PA GE of immunoprecipitates For SDS-PAGE, washed samples were mixed with sample running buffer ( + / - DTT) and boiled for 10 minutes. Samples were centrifuged at 10,000 x g for 5 min and solubilized material was run on SDS PAGE (3% stacking gel, 5% running gel) according to the method of Laemmli [10]. Following SDSPAGE, gels were fixed overnight in 20% methanol/10% acetic acid. Gels with 35S samples were soaked in En3hance (New England Nuclear, Boston, Mass.) prior to drying and autoradiography. 125Igels were dried without the enhancement step. Exposure of gels to Kodak XOMAT XAR-5 film was done at - 80° C. Densitometry of autoradiographs was performed using a Hewlett Packard integrator. All results are expressed as relative area under the peaks. The method for distinguishing between dimeric and multimeric FN was that of McKeown-Longo and Mosher [4] briefly as follows. Radioactive, immunoprecipitated fibronectin migrated as multimeric and/or dimeric material

60

,°°I

I

80

_.1 I-0 I--

60

0 IM CD

¢I t--Z

U.i n," hi

20

0

0

2

4

6

8

I0

12

TIME ( H O U R S ) Fig. 1. Percentage distribution of [35S]methioninepulse-labeled FN in control and TPA-treated cells between 20 minutes and 12 hours. Confluent monolayers were rinsed three times with PBS. Serum-free MEM with 10% of the normal level of methionine and 100/zCi/ml of [35S]methioninewas added. Cells were incubated for 10 minutes. Labeled medium was removed, cells were washed three times with PBS and serum-free MEM with 0.4% BSA (with or without 10-TMTPA) was added. Radioactively-labeled FiNwas quantified using the gelatin affinity method described in the Methods Section. 100% = total 3sS labeled FN at time zero. Data are means from duplicate dishes assayed for each point. Error bars indicate individual data points. Open squares, control intracellular FN; open circles, control cell-surface FN; open triangles, control medium FN. Full circles, TPA-treated cell-surface FN; full triangles, TPA treated medium FN (only the 4 and 12 h time points for TPA-treated FN on the cell surface and in the medium are shown, since other time points coincide with control points; for the same reason, TPA-treated intracellular time points are not shown). w h e n samples w e r e not reduced. Multimeric F N just e n t e r e d the stacking gel and dimeric material migrated with an a p p a r e n t molecular weight of 450,000 daltons. U p o n reduction, all i m m u n o p r e cipitated multimeric and dimeric fibronectin migrated as m o n o m e r i c fibronectin with an a p p a r e n t molecular weight of 230,000 daltons showing that the material in the stacking gel was i n d e e d multimeric FN. N o o t h e r radioactive bands were detected in either the r e d u c e d or u n r e d u c e d autoradiographs.

Kinetic experimentsProcessing of cells and medium F o r all kinetic experiments, approximately 1.5 x 106 cells per dish were seeded o n t o 60 m m dishes and cultured for 2 days in M E M containing 10% fetal calf serum. A t this point, the cells were fully confluent. T h e cells were washed three times with PBS, labeled with either 35S-methionine or 125I, and washed again with PBS and/or P B I (PBS in which sodium iodide has b e e n substituted for NaCl). Serum-free M E M with 0.4% B S A was then a d d e d (with 10-7M T P A in D M S O or D M S O alone for

61 controls; D M S O concentrations did not exceed 0.2%). At various times (5 minutes to 24 hours) after the addition of serum-free medium, plates were processed as follows: medium was removed, centrifuged at 100 x g for 5 minutes, brought to 2 m M PMSF and stored in liquid nitrogen. Cell layers were rinsed three times with cold PBS; for total cellular FN, cells were scraped into PBS containing 8 M urea, 1% TritonX-100, and 1 mM PMSF (cell lysis buffer). For determination of intracellular FN, cells were incubated with 10/~g/ml trypsin in PBS for 10min; 500/~g/ml soybean trypsin inhibitor was added, cells were scraped into centrifuge tubes, centrifuged at 1000 x g for 10rain, rinsed two times with PBS and dissolved in cell lysis buffer. When surface proteins were labeled with 125I and then treated with trypsin, radioactive FN quantified by either gelatin binding or immunoprecipitation was not detectable. All cell extracts were stored in liquid nitrogen. In each experiment, duplicate plates of cells were used for total and intracellular FN determinations for each treatment.

Pulse~chase experiments After confluent cell layers were rinsed three times with PBS, 2 ml of serum-free M E M with 10% of the normal level of methionine and 100 ~Ci/ml of 35Smethionine were added for 10 minutes (pulse). For the 24 hour immunoprecipitation experiments, cells were labeled for 20min with 200/xCi/ml of 35S-methionine. Labeled medium was removed, cells were washed three times with PBS; serumfree M E M (with or without 10-7M T P A ) with three times the normal methionine level and 0.4% BSA was added. A t various times after pulse-labeling, medium and cells were processed to determine total FN content (ELISA), FN-associated radioactivity (gelatin-binding assay), or dimer and multimer content (immunoprecipitation) as described above.

Fig. 2. Autoradiograph of SDS-PAGE of immunoprecipitates of total cell extracts of [35S]methioninepulse-labeled control and TPA-treated cells between 1 and 12 hours. Samples were not reduced. Confluent monolayersof cells were pulse-labeled with [35S]methioninefor 10 minutes. Followingpulse-labeling, cells were rinsed three times with PBS; serum-free medium (with or without 10-TMTPA) and 0.4% BSA was added. At the indicated times, duplicate plates of cells were dissolved in cell lysis buffer. Equal volumes of cell extract were immunoprecipitated and immunoprecipitateswere washed. After boilingwith sample cocktail, samples were run on SDS-PAGE. Gels were fixed, enhanced with EN3HANCE, dried and exposed to X-ray film. Left four lanes are control cells, right four are TPA-treated cells. Numbers across the top of the gel indicate hours after pulse labeling. The positionsof multimer(M) and dimer (D) are indicated. Autoradiograph is from one representative experiment.

Surface iodination After confluent cell layers were rinsed, 1.5 ml of PBS with 1% beta D-glucose and 450/xCi of Na 1251 was added to all plates. The lactoperoxidase-catalyzed iodination was initiated by addition of 50 tzl of Enzymobeads (Sepharose beads with lactoperoxidase and glucose oxidase attached; the latter generates H202 with glucose, the former then cata-

62 lyzes the peroxidation of labeled I - to I2, which labels the protein). Iodination was carried out for 12 rain. Cells were washed three times with PBI and PBS following removal of beads, Na m5I, and glucose. Serum free M E M (with or without 10-7M T P A ) with 0.4% B S A was then added. Cells were processed at various time points as described above.

Results

The kinetics of pulse-labeled fibronectin Similar to previous results from our laboratory [11] we observed that over time FN was released from cultured human lung fibroblasts and accumulated in the medium. A one-hour treatment with T P A stimulated FN release to 200% of control level. After the first hour, the amount of FN released by T P A continued to be 25% above control level; after 24 h of T P A treatment, cell surface FN had decreased to 60% of controls. We asked the question: what is the origin of the early burst of released FN upon T P A treatment? One hour after a pulse label, the amount of intracellular labeled FN in untreated (control) cells

had dropped to 20% of the zero time value. However, only 10% of the labeled FN appeared in the medium and 70% was on the cell surface (Fig. 1). These results indicate that very little FN was secreted directly into the medium without residence time at the cell surface. One hour of T P A treatment did not markedly affect the release of pulse-labeled FN into the medium; thus, the initial burst of TPAstimulated FN release was not due to increased direct secretion of newly-synthesized FN from the intracellular compartment, but to FN residing at the cell surface. Release of labeled FN continued at a linear rate up to 4 hours following synthesis in control cells and T P A treatment did not markedly increase the amount or rate of release (data not shown). The rate of release of pulse-labeled FN in both control and TPA-treated cells had declined at 12 hours, but the level of pulse-labeled FN in the medium remained slightly elevated in TPA-treated cells (Fig. 1). There was a parallel, but small decrease in the pulse-labeled cell surface FN level in TPA-treated cells at 12 hours (Fig. 1). Thus, a four-hour T P A treatment exerted no marked effects on the fate of pulse labeled FN, and therefore T P A did not significantly affect the fate of newly synthesized FN during this treatment time. Longer-term T P A treat-

Table 1. Densitometryof autoradiographs of gel electrophoreses of cell-layer fibronectin after pulse-chase with pS]methionine Time

Control dimer

Control multimer

TPA treatment dimer

TPA treatment multimer

Hours Expt. I 1 4 8 12 Expt. II 12 18 24

Relative density 2.62_+ 0.51 1.06+ 0.4 1.2 + 0.l 0.7 _+ 0.1

0 0.87+ 0.3 1.1 +_ 0.1 1.21-+ 0.1

2.8 + 0.4 1.12_+ 0.3 0.98+ 0.02 0.55+ 0.1

0.06 + 0.02 0.74_+ 0.12 0.95-+ 0.2 0.8 _+ 0.11

3.44+ 0.33 1.46_+ 0.00 0.96-+ 0.2

1.2 + 0.2 1.2 _+ 0.11 0.8 -+ 0.15

1.82+ 0.16 0.84+ 0.2 0.52+ 0.02

0.93_+ 0.11 0.80-+ 0.04 0.50-+ 0.01

CelI monolayers were treated in experiment I with 100~Ci/ml [35S]methionine for 10 minutes; in experiment II, with 200/~Ci/ml [35S]methioninefor 20 minutes. After pulse-labeling, cellswere treated as described in the Methods Section (pulse-chase experiments). At the indicated times, medium was removed and cells were dissolved, FN immunoprecipitated and subsequently electrophoresed as described in the Methods section. Followingdryingand autoradiography, radioactivebands were scannedby densitometry. Each band was scanned three times in different locations. Experiment I: Values are mean + standard deviation of the densitometry representing means of data from duplicate plates. For Experiment II, values are mean + standard deviation in which three separate plates of cells were assayed for each point. Autoradiograph shown in Fig. 2.

63 ment decreased the cell surface level of pulse-labeled FN, which parallelled the decrease in total cell surface FN levels (data not shown). This suggests that the mechanism(s) responsible for TPAstimulated release of FN may vary with time or possibly the susceptibility to TPA of newly-synthesized FN may change over time. At 24 h, almost all (85%) of pulse-labeled FN had been shed from both control and TPA-treated cells;there were no significant differences between them at this time.

Effect of TPA on dimeric and multimeric fibronectin The next question we asked was: is the FN released by TPA-stimulation derived from the pool of dimeric FN residing at the cell surface, or from FN after multimerization had taken place. Following pulse labeling, samples were immunoprecipitated, subjected to SDS-PAGE and analyzed by autoradiography. In control cells, all medium and cellassociated FN was dimeric one hour after pulselabeling. After 4 hours, multimerization had taken place and both dimeric and multimeric FN were detected at the cell surface (Fig. 2 controls and Table 1) and in the medium. (Fig. 3, controls.) Four hours o f TPA treatment did not affect the levels of dimeric or multimeric FN in the cell layer (Fig. 2, TPA and Table 1) or medium FN (Fig. 3, TPA). After 12 hours, small decreases in both dirneric and multimeric pulse-labeled cell surface FN were observed (Table 1). These results parallelled those from the gelatin-binding assay (Fig. 1).

The fate of pre-existing cell-surface FN Following cell-surface labeling with 125I-iodide-lactoperoxidase, the control cell-surface level of labeled FN gradually declined and the medium level gradually increased over 12 hours when quantified by gelatin-bound radioactivity (Fig. 4). Within the first hour of TPA treatment, there was a three fold increase over control in the amount 125I-iodide-lactoperoxidase-labeled FN shed to the medium (Fig. 4). After 4 hours, the increase was two-fold. After

Fig. 3. Autoradiographof SDS-PAGE of immunoprecipitates of medium from [35S]methioninepulse-labeled control and TPA-treated cells between 1 and 12 hours. Unreduced gel. Plates were processed as in Fig. 2. Equal volumesof medium from duplicateplates were immunoprecipitatedfor each point. Left four lanes are control medium and right four lanes are mediumfrom TPA-treated cells. Numbers across the top of the gel indicate time after pulse-labeling (hours). The position of rnultimer (M) and dimer (D) are indicated. 12 hours, almost all the pre-existing FN had been shed into the medium equally in control and TPAtreated ceils (Fig. 4). At the same time, a decrease in the cell-surface level of 125I-iodide-lactoperoxidase-labeled FN was observed after 1 and 4 hours of TPA treatment. Thus, TPA treatment more strongly affected pre-existing cell-surface FN than pulse-labeled FN that was intracellular at the time that TPA treatment was initiated. Immunoprecipitation showed that l:sI-iodidelactoperoxidase-labeled FN at the cell surface was composed of both dimeric and multimeric FN (Fig. 5, control). Medium 125I-FN was also composed of dimeric and multimeric FN (Fig. 6, control). Immunoprecipitation followed by SDS-PAGE and autoradiography revealed that the majority of FN

64

400 I

O Z

,,, 3 0 0 I-< .J Q.

o_ 2 0 0 Z b.. w

Z O n." DD

I.L I H

Lo

I00

0

0

I 2

I 4

I 6

TIME

I 8

I I0

I 12

(HOURS)

Fig. 4. 125I-labeled fibronectin in the medium and cell surface of control and TPA-treated cells. Cell monolayer surface proteins were labeled with 125Ias described in the Methods section. Serum-free medium with 0.4% BSA (with or without 10-7M TPA) was added. Radioactive FN was quantified using the gelatin affinity method as described in the Methods section. Data show means from duplicate plates and error bars individual data points. Medium: open circles, control cells; full circles TPA-treated cells. Cell surface: open squares, control cells; full squares, TPA-treated cells.

released following one hour of TPA treatment was dimeric (Fig. 6, TPA). As quantified by densitometry, there was an 8-fold increase over controls in dimeric 125I-FN release into the medium after one hour of TPA treatment and after 4 hours, this increase was still 3-fold (Table 2). Decreases in labeled cell surface dimeric FN levels were evident after 4 hours of treatment (Table 3). After 9 hours of TPA treatment, the decline in cell-surface dimeric 125I-FN was still marked, with TPA-treated cell surface levels being 25% of control cell levels (Fig. 5, TPA, and Table 3). The action of TPA on multimeric FN was much less dramatic than on dimeric FN: there was a 1.25 fold increase over control in labeled multimer release after 1 hour of TPA treatment and after 4 hours, this increase was 1.5-fold. After 9 hours of TPA treatment, there was a 25% decrease in cell surface multimeric 125I-

FN. Thus, the major effect of TPA was to decrease cell surface levels of dimeric, preexisting FN.

Discussion

Using two different isotopic tracer techniques and several analytical techniques (ELISA, gelatinbound radioactivity, and immunoprecipitation followed by SDS-PAGE), we have demonstrated that the major effect of 1-4 hour treatment of human lung fibroblasts by TPA is to stimulate FN release from the cell surface. Dimeric cell-surface levels were most strongly affected. In contrast, TPA treatment had a much lesser effect on newly synthesized FN that is intracellular at the time TPA treatment was initiated. Thus a 1--4 hour TPA treatment preferentially affected pre-existing cellsurface FN. Longer-term TPA treatment of up to

65

Fig. 5. Autoradiograph of SDS-PAGE of immunoprecipitates of total cell extracts of 12SI-labeledcontrol and TPA-treated cells. Unreduced samples. Following surface iodination, cell monolayers were immunoprecipitated as described for Fig. 2. Autoradiography was performed without EN3HANCE. Left four lanes are control cells and right four lanes are TPA-treated cells. Numbers across the top of the gel indicate time after surface-labeling (hours). The position of multimer (M) and dimer (D) are marked.

24 hours decreased the cell surface levels of FN, as shown by both pulse and surface labeling. M c K e o w n - L o n g o and Mosher [12] have found that in h u m a n skin fibroblasts, exogenously added FN is first reversibly bound into a deoxycholate-soluble pool at the cell surface and after 3 hours, the FN is then incorporated irreversibly into a deoxycholateinsoluble pool, which is primarily composed of multimerized FN. These two pools may be bound to the cell in different ways. The fact that shortterm T P A treatment differentially affects newlysynthesized and pre-existing cell-surface FN differently suggests that the m a n n e r in which FN molecules are bound to the cell surface may vary with the length of time they have been at the cell surface. A n o t h e r interpretation of our results could be the following: the mechanism(s) by which T P A stimulates short and long-term FN release may dif-

Fig. 6. Autoradiograph of SDS-PAGE of immunoprecipitates of medium from 1251surface-labeled control and TPA-treated cells. Following surface iodination, medium was immunoprecipitated as described in Fig. 2. Autoradiography was performed without EN3HANCE. Left four lanes are control medium and right four lanes are medium from TPA-treated cells. Numbers across the top of the gel indicate time after surface labeling (hours). The positions of multimer (M) and dimer (D) are marked.

fer because short and long-term T P A treatment affected FN release differently. In other systems, it has been demonstrated that during the first few hours of T P A treatment, the effects of T P A are confined to rapid events at the cell surface. After several more hours, T P A begins to cause loss of binding sites for phorbol esters [13]. The major effect of T P A was to stimulate the release of dimeric FN from the cell surface into the medium. Whether the dimeric ~251-FN in medium is derived by direct release of cell surface dimer and/ or breakdown of multimeric FN could not be determined. The fact that the cell-surface level of dimeric 12~I-FN (i.e. pre-existing) is greatly decreased after nine hours of T P A treatment suggests that dimeric surface FN release is stimulated directly. At early treatment times (1-4 hours), T P A preferentially affected those FN molecules already at

66 Table 2. Densitometry of autoradiographs of gel electrophoreses of medium fibronectin after cell-surface labeling with 125I-iodidelactoperoxidase Time

Control dimer

Hours

relative density

0.5 1 4 9

0 0.12_+ 0.03 0.93_+ 0.04 2.45_+ 0.1

Control multimer

TPA treatment dimer

TPA treatment multimer

0 0.42_+ 0.07 0.48_+ 0.04 1.10_+ 0.05

0.1 _+ 0.02 1.00_+ 0.1 2.5 _+ 0.10 2.70_+ 0.2

0 0.52+ 0.08 0.72__. 0.1 0.9 + 0.101

Medium samples were processed as described under Table I, experiment II, except that FN was immunoprecipitated from the medium. Autoradiograph shown in Fig. 6.

the cell surface. Longer TPA treatment, up to nine hours, significantly decreased the levels of surfacelabeled dimeric FN, and thus had a prolonged effect on surface FN. TPA could reduce dimeric cell surface FN by directly stimulating its release and/or increasing its rate of incorporation into the multimerized FN at the cell surface. Indeed, four hours after cellsurface labeling and TPA treatment, the level of labeled multimer is higher in the TPA treated cells (Table 3). This argues for prior multimerization of the dimer. However, TPA did not significantly decrease the level of dimeric pulselabeled cell-surface FN. These differences could be accounted for by variations in the mechanism(s) of FN-cell surface binding over time and varying effects of TPA over time. It is clear that TPA had a preferential effect on

preexisting cell-surface FN. This suggests that those FN molecules at the cell surface when TPA treatment is initiated may differ in their susceptibility to TPA from those inside the cell. This may be accomplished through the effect of TPA on cellsurface FN binding proteins such as integrins [3]. Fibronectins purified from normal and TPA-treated cells did not differ in electrophoretic mobility on SDS-PAGE or in ability to support cell attachment when coated to a plastic substratum (unpublished results, Burrous and Wolf). Brown [14] recently showed that TPA treatment caused increased phosphorylation of the cell-surface FN receptor integrin, leading to increased cell adhesion. Thus, there is evidence that TPA can in some way affect integrins. Future studies using TPA as a disrupter

Table 3. Densitometry of autoradiographs of gel electrophoreses of cell layer fibronectin after cell-surface labeling with lasf-iodidelactoperoxidase Time

Control dimer

Hours

Relative density

0.5 1 4 9

0.67+ 0.08 1.12_+ 0.08 0.49_+ 0.10 0.30 _+ 0.05

Control multimer

TPA treatment dimer

TPA treatment multimer

1.14_+ 1.45_+ 1.12_+ 1.00_+

0.80_+ 0.1 0.86_+ 0.13 0.36_+ 0,14 0.07 _+ 0,02

1.24_+ 0.10 1.40_+ 0.10 1.34_+ 0.16 0.75 _+ 0.05

0.01 0.10 0.18 0.05

Cell-surface proteins of cell monolayers were labeled with 125Ias described in the Methods section. Serum free MEM with 0.4% BSA and with or without 10-7 TPA was added. At the designated times, medium was collected, ceils were washed three times with PBS and dissolved. Samples were immunoprecipitated and subsequently electrophoresed as described in the Methods section. Following drying autoradiography, radioactive bands were scanned by densitometry. Each band was scanned three times in different locations - mean and standard deviation were calculated. Data shown represent averages from duplicate plates. Autoradiograph shown in Fig, 5.

67 of integrin/FN interactions may further our understanding of FN-cell surface interactions.

Acknowledgements

7.

8.

T h e a u t h o r s g r a t e f u l l y a c k n o w l e d g e financia! supp o r t f o r this w o r k p r o v i d e d by U . S . P u b l i c H e a l t h

9.

S e r v i c e G r a n t N o . C A 13792.

References 1. Yamada KM: Cell surface interactions with extracellular materials. Ann Rev Biochem 52: 761-799, 1983 2. ChoiMY, HynesRO: Biosynthesisandprocessingoffibronectin in NIL-8 hamster cells. J Biol Chem 254: 1205012055, 1979 3. Hynes RO: Integrins: a family of cell surface receptors. Cell 48: 549-554, 1987 4. McKeown-Longo PJ, Mosher DF: Mechanism of formation of disulfide-bonded multimers of plasma fibronectin in cell layers of cultured human fibroblasts. J Biol Chem 259: 12210-12215, 1984 5. Blumberg PM, Dreidger DE, Rossow PN: Effect of phorbol ester on a transformationsensitive surface protein of chick fibroblasts. Nature 264: 446-447, 1979 6. Keski-Oja J, Shoyab M, DeLarco JE, Todaro GJ: Rapid release of fibronectin from human lung fibroblasts by bi-

10. 11.

12.

13.

14.

ologically active phorbol esters. Int J Cancer 24: 218-224, 1979 Zerlauth G, Wolf G: Kinetics of fibronectin release from fibroblasts in response to 12-0-tetradecanoyl phorbol-13 acetate and retinoic acid. Carcinogenesis 5: 863-868, 1984 Ruoslahti E, Hayman EG, Pierschbacher M, Engvall E: Fibronectin: Purification, immunochemical Properties and Biological activities. Methods in Enzymology 82: 803-831, 1982 Mosher DF, Vaheri A: Thrombin stimulates the production and release of a major surface-associated glycoprotein (fibronectin) in cultures of human fibroblasts. Exp Cell Res 112: 323-334, 1978 Laemmli UK: Cleavage of structural proteins during assembly of bacteriophage T 4. Nature 227: 680-685, 1970 Zerlauth G, Wolf G: Release of fibronectin is linked to tumor promotion: response of promotable and non-promotable clones of a mouse epidermal cell line. Carcinogenesis 6: 73-78, 1985 McKeown-Longo PJ, Mosher DF: Binding of plasma fibronectin to cell layers of human skin fibroblasts. J Cell Biol 97: 466-472, 1983 Collins MKL, Rozengurt E: Binding of phorbol esters to high-affinity sites on fibroblastic ceils elicits a mitogenic response. J Cell Physiol 112: 42-50, 1982 Brown PJ: Phorbol ester stimulation of fibronectin-mediated cell adhesion. Biochem Biophys Res Comm 155: 603607, 1988

Address for offprints: G. Wolf, Dept. of Nutritional Sciences, University of California, Berkeley CA 94720, USA

The kinetics of fibronectin synthesis and release in normal and tumor promoter-treated human lung fibroblasts.

The present study is a detailed kinetic analysis of the synthesis, release and multimerization of fibronectin (FN) in normal and tumor promoter-treate...
810KB Sizes 0 Downloads 0 Views