EXPERIMENTAL

CELL

RESEARCH

203,

289-296 (1992)

Rat Hepatocytes in Primary Culture Synthesize and Secrete Cellular Fibronectin I. Medirinische

Klinik

und Poliklinik der Uniuersitiit Mainz, Federal Republic of Germany; *Institute Sieroterapico and tDepartment of Immunology, Scripps Clinic Research Foundation, La Jolla, California

Milan,

Italy;

By their participation in modulating many cellular processes such as cell adhesion, spreading, and migration, as well as cytoskeletal organization, fibronectins have significant part in tissue development and differentiation, wound healing, opsonization, and hemostasis. They are involved in these processes by functioning as ligands between cells and various extracellular matrix components, due to their specific binding capabilities residing in different domains of the fibronectin molecule [for reviews see 1, 21. The circulating plasma fibronectin and the extracellular matrix fibronectin are very similar in overall structure, but differ slightly in their solubility properities, their carbohydrate residues, and their mobility in SDS-polyacrylamide gels [3-51. Both the circulating plasma and the extracellular matrix proteins are dimers or multimers, respectively, of identical or very similar polypeptide chains, joined by sulfide bridges at their carboxyl termini. As proven by primary sequence analyses the polypeptides consist of three types of homologous repeats referred to as types I, II, and III [6, 71. However, on the basis of proteolytic cleavage primary sequence variability has also been suggested and could be confirmed by analyses of different cDNA clones, revealing fibronectin heterogeneity in three distinct regions, named ED-A, ED-B, and III CS. This fibronectin polymorphism is due to alternative splicing, since many different mRNA originate from a common primary precursor transcript of a single gene. The splicing patterns are tissue specific and vary in different stages of cell growth and differentiation. Alternative splicing on the fibronectin gene transcript occurs in two forms: exon skipping accounts for the presence or absence of the additional type III segments, ED-A or ED-B, and exon subdivision leads to variations in the III CS area, located between the penultimate and the final type III repeat [for review see 81. Antibodies raised against the ED-A domains show that they are present in fibroblast-derived cellular fibronectin, but not in plasma fibronectin [9, lo]. In addition, protease sites characteristic of the ED-B domain are not present in plasma fibronectin, whereas they are found

involved in cell-matrix interactions Fibronectins, and cell attachment, are glycoproteins which show a remarkable heterogeneity, due to alternative splicing. The type III-related domains, ED-A and ED-B, are present in cellular fibronectin in a variety of ratios whereas they are absent in circulating plasma fibronectin. Fibronectin synthesis by hepatocytes which are accepted as suppliers of plasma fibronectin was studied in primary cultures during a g-day culture period. Using site-specific antibodies we demonstrate that rat hepatocytes are also able to synthesize and secrete fibronectin bearing the ED-A domain from Day 3 on after inoculation. By immunocytological characterization of the hepatocyte monolayer with antibodies directed against desmin, laminin, collagen IV, ol-SM-actin, or ED-l or factor VIIIrelated antigen, contaminating mesenchymal hepatic cell-types as a source for cellular fibronectin production could be ruled out. Dexamethasone treatment caused enhanced fibronectin synthesis and cellular fibronectin was already detectable at Day 1 after plating. Elevation of cellular fibronectin synthesis after prolonged culture-terms and by dexamethasone could also be demonstrated on mRNA steady-state level, using ED-A cDNA as a probe in hybridization analysis. Dot blot hybridisation proved a prominent response of cellular fibronectin mRNA level to dexamethasone at Day 1 when dexamethasone treatment resulted in an increased contribution of ED-A-positive fibronectin tranc I%IW Academic scripts to total fibronectin mRNA level. Press,

Milanese,

Inc.

INTRODUCTION Fibronectins are high molecular weight glycoproteins found as extracellular matrix proteins in association with cell surfaces and connective tissues and as circulating plasma proteins in blood and other body fluids.

’ To whom reprint requests should be addressed at Zentmm Innere Medizin Der Universit;it Gijttingen, Robert-Koch-Strasse 40, 3400 GGttingen, FRG. Fax: /6131/222332.

289 All

Copyright 0 1992 rights of reproduction

0014-4827/92 $5.00 by Academic Press, Inc. in any form reserved.

290

ODENTHAL

in cellular fibronectin [ll]. According to data pointing out hepatocytes as the main source of plasma fibronectin, the expression of the ED-A and ED-B domains in plasma and in cellular fibronectin is consistent with the presence of these exons in fibronectin mRNA of fibroblasts and their absence in adult liver [7, 121. Thus, whereas in the majority of tissues and cell lines a mixture of fibronectin subunits is observed, bearing the extra ED-A and ED-B type III repeats in various ratios depending on cell-type and differentiation [13151, the liver, made up of hepatocytes bo about 90%, only gives rise to fibronectin transcripts where ED-A and ED-B domains are absent or undetectable. Oyama et al., however, report a different alternative splicing pattern in fetal liver and in malignant human liver tumors, where fibronectin transcripts bearing the additional ED-A domain of cellular fibronectin were found [15, 161. Among hepatic cell types, fat-storing and endothelial cells turned out to synthesize cellular fibronectin in primary culture [18, 191. In the present study we investigated whether rat hepatocytes in primary culture are able to produce fibronectin containing the ED-A segment as well. In regard to the stimulating effect of glucocorticoids on total fibronectin synthesis, resulting in a marked increase of expression in hepatoma cells [ 14,201 and in hepatocytes of different species [21-231, we examined synthesis of cellular fibronectin in rat hepatocyte monolayers subjected to dexamethasone treatment. MATERIALS

AND

METHODS

Isolation and primary culture of hepatocytes. Hepatocytes were isolated from adult female Wistar rats by collagenase perfusion according to Seglen [24]. Cells (2.5 X 105) in Dulbecco’s modification of Eagle’s medium (DMEM), supplemented with 0.05% insulin, 1% glutamine, and 10% fetal calf serum were seeded onto each well of a collagen-coated 24 well plate. After 4 h the culture medium was replaced by fresh medium containing 0.2% bovine serum albumin instead of fetal calf serum. Hormonal treatment of hepatocytes consisted of the addition of 10d7 M dexamethasone to the culture medium. Zmmunocytochemistry. For immunofluorescence staining cells were plated onto collagen-coated Lab Tek tissue-culture slides (80,000 cells/well) and fixed in cold methanol (-2O”C, 5 min) with subsequent rinsing in cold acetone on Days 1, 3, and 5 of primary culture. To detect contaminating cell types, fixed cells were exposed either to monoclonal anti-desmin (Dako, Copenhagen) or to anti-uSM-actin (Sigma), to anti-ED1 antibodies (a generous gift of C. D. Dijkstra [25]), or to polyclonal antibodies recognizing laminin, collagen IV, or factor VIII-related antigen (Camon, Wiesbaden; Dako, Copenhagen). For evaluation of contaminants 3000 cells were counted. The identification of fibronectin was performed using affinity-purified rabbit antibodies directed against human fibronectin (Calbiothem, Munich). Cellular fibronectin was identified by site-specific antibodies (Sigma) or anti-ED-A antiserum [26]. Antigen-IgG complexes were detected by fluorescein isothiocyanate-coupled sheep antirabbit or antimouse IgG (Sigma). In a double immunostaining procedure cells were first exposed to anti-human fibronectin or antilaminin antibodies, followed by tetramethyl rhodamine isothiocyanate-coupled sheep antimouse IgG antiserum. In a second step, monoclonal anti-desmin antibodies were used, followed by the exposure to the respective fluorescein isothiocyanate conjugate. The specimens

ET AL. were embedded with Fluoromount (Southern Biotechnology, Birmingham) and fluorescence of fluorescein isothiocyanate and tetramethyl rhodamine isothiocyanate was monitored under a Leitz light microscope equipped with the appropriate filters. Determination of protein synthesis in primary hepatocyte cultures. At different days after isolation, hepatocytes of primary cultures were washed three times in methionine-free DMEM medium and cell culture was continued for 2 h in the presence of [?S]methionine (9.35 MBq/2.5 X lo5 cells in 250 ~1 culture medium; sp act 37 TBq/mmol). The media were collected and cells were lysed by freezing and thawing in 1% SDS, 0.5% Na desoxycholate PBS 1X as described previously [27]. Protein synthesis and secretion was determined by measuring [35S]methionine incorporation in trichloroacetic acid insoluble proteins of cell lysates and of the culture media. For immunoprecipitation with anti-human fibronectin (Calbiochem, Munich) or with antiED-A fibronectin antibodies (27) de nouo-synthesized proteins corresponding to lo6 cpm (cell lysate) or to lo5 cpm (media) were used. IgG-antigen complexes were precipitated with Staphylococcus aureus cells according to Cullen and Schwartz 1281. Immunoprecipitates were analysed in 5% polyacrylamide gel electrophoresis following the method of Laemmli 1291. For analysis of fucose incorporation into fibronectin synthesized by hepatocytes, [“Hlfucose was added to the culture media instead of [“‘Slmethionine (5.55 MBq; sp act, 3 TBq/mmol) and cells were incubated for 4 h. After immunoprecipitation and SDS-PAGE, [3H]fucose-labeled fibronectin was visualized by autofluorography [30]. Northern blot and dot blot hybridization. Total RNA extracted by guanidine isothiocyanate [31] was purified by CsCl density centrifugation [32] and electrophoresed in 1% agarose gels under denaturating conditions 1331. Separated RNA was immobilized on nylon membranes (Amersham) according to the technique of Southern [34]. Alternatively, the denaturated, total RNA was spotted directly onto the nylon membrane by means of a 96-well dot blot apparatus. Northern and dot blots were hybridized with “P-labeled cDNA probes. The human fibronectin cDNA of plasmid pFH 111 [7] was digested with PstI and the resulting 100. and 180.bp cDNA fragments were isolated by polyacrylamide gel electrophoresis and purified by DE 52 chromafragments of 100 bp, representing tography. Labeling of the PstI-PstI part of the 11th type III repeat, on one hand, and of 180 bp containing the cDNA of the 5’-end of the ED-A domain on the other, was performed by random priming with [o-32P]dCTP. Hybridization analyses to actin, a2-macroglobulin, and albumin transcripts were carried out with pAC 269, pRL29J, and alb 2M plasmid DNA [35-371 after labeling by nick translation.

RESULTS

Characterization of Primary Immunocytochemistry

Cultures of Hepatocytes by

Primary cultures of hepatocytes were examined for the presence of mesenchymal hepatic cells at Days 1,3, and 5 after isolation. As markers for fat-storing cells, laminin [39] and desmin [40], and for myofibroblasts a-SM-actin, were used [41]. Kupffer cells were identified by immunostaining with monoclonal antibodies specific for the monocyte and macrophage antigen ED1 [25], and endothelial cells by detection of the factor VIII-related antigen. Freshly isolated hepatocytes contained only a few ED 1 and very few laminin-positive cells. At Day 1 after plating no desmin-, cy-SM-actin-, or factor VIII-related antigen-positive cells were detectable. Kupffer-cell contamination could be shown to be below 4% (Fig. 1). Because immunostaining of desmin

RAT

HEPATOCYTES

SYNTHESIZE

CELLULAR

FIBRONECTIN

291

IG. 1. Monolayers of rat hepatocytes in primary culture at Day 3 (a, c-f) and Day 5 (b) after isolation, immunostained with anti1 lodies nst desmin (a, b), a-SM-actin (c), collagen IV (d) and laminin (e). As a control (f) rabbit serum was used. Nonspecific bright Auores :cence associated with nonviable hepatocytes. The scale bar corresponds to 50 pm.

det ected less than 50% of freshly isolated fat-storing cell Is (data not shown), additional immunocytology of lan Cnin-positive cells was performed. But even by double staining of laminin and desmin only slight contamicells (below 3%) could be obnat ion with fat-storing

served in the hepatocyte monolayer at Day 3. Imm unocytological detection of collagen IV revealed scatt ;ered positivity, which was restricted to single desmin- anId/or a-SM-actin-positive cells and did not deposit as a matrix (Fig. 1). Kupffer-cell contamination did not ch ange

292

ODENTHAL

3%

3H

4- fibronectin

ET AL.

cellular fibronectin synthesized in longer-term culture into the media at Day 3 of culture. Glucocorticoids are known to increase fibronectin synthesis and secretion by hepatocytes in primary culture [21-23, 431. The stimulating effect of glucocorticoids on fibronectin production in rat hepatocytes was shown by hormonal treatment with dexamethasone (1O-7 M) of the presented, serum-free primary culture (Fig. 3). The response to dexamethasone could be observed not only in total fibronectin synthesis, but even more prominently in the synthesis and secretion of cellular fibronectin. In fact addition of dexamethasone caused secretion of cellular fibronectin in hepatocyte cultures already on Day 1 after inoculation. Assembly of Cellular and Plasma Fibronectin Cultures of Hepatocytes

136

136 days

FIG. 2. Fibronectin synthesized by hepatocytes in primary culture at Days 1,3, and 6, after [%]methionine (%l) or [3H]fucose (3H) labeling. For immunoprecipitation with anti-fibronectin antibodies IO6 cpm of the labeled cell lysate was used. Immunoprecipitates were analyzed by SDS-PAGE and autofluorography.

at Day 3 compared to the first day of primary culture. At Day 5 after isolation no significant increase in the numher of contaminating cells in primary cultures of hepatocytes had occurred.

Synthesis of Fibronectin Hepatocytes

in Primary

Cultures of

In order to examine fibronectin synthesis in primary cultures of rat hepatocytes proteins synthesized de nouo were labeled by [35S]methionine incorporation at different days of culture and immunoprecipitated with antifibronectin antibodies. The data presented in Fig. 2 show elevated fibronectin synthesis and secretion in longer-term primary cultures, which is in agreement with previous work concerning fibronectin synthesis by isolated hepatocytes in culture [23,42]. By immunoprecipitation of [3H]fucose-labeled proteins we were able to show that the synthesized fibronectin is partly fucosylated (Fig. 2). Because incorporation of fucose is a characteristic of cellular fibronectin in hamster and guinea pig [3, 191, we studied the contribution of cellular to total fibronectin synthesis by using monospecific antibodies against the additional ED-A domain of cellular fibronectin. Whereas no fibronectin form containing the additional ED-A sequence was detectable at Day 1 after seeding, a reasonable part of the newly synthesized fibronectin could be identified in the cell lysate as originating from cellular fibronectin after Day 3 of primary culture (Fig. 3). Figure 3 also indicates the secretion of

in Primary

In the well-defined monolayer of rat hepatocytes fibronectin fibers are shown to be already present at Day 1 after isolation and the fibrous patch-like distribution of fibronectin grows up to an extensive fibrous network during culture (Figs. 4a and 4b). This result corresponds to previous work [21, 23, 431. By indirect immunofluorescence experiments with monoclonal antibodies recognizing exclusively the additional ED-A domain of cellular fibronectin, however, cellular fibronectin fibers could also be clearly detected on the monolayers of hepatocytes 3 days after isolation (Fig. 4d). Because immunostaining revealed no positivity surrounding single cells but an even cellular fibronectin matrix similar to the distribution of total fibronectin, contaminating mesenchymal liver cell-types can be excluded as a source for cellular fibronectin (Fig. 4). Therefore, in primary hepatocyte culture the enhancement of extracellular fibronectin is not only due to an increased secretion of

medium

cell -

1

3

+

1313

-

l

+

total tibronectin

+

cellular fibronectin

13

FIG. 3. Synthesis and secretion (cell, medium) of total and cellular fibronectin by hepatocytes in primary cultures at Days 1 and 3. The hepatocyte cultures were exposed to lo-’ M dexamethasone (+) or not (-). For immunoprecipitation with anti-fibronectin or anticellular fibronectin antibodies, lo6 cpm of the respective [%methionine-labeled cell lysates and lo5 cpm of the respective medium were used. Immunoprecipitates were analyzed by SDS-PAGE and autofluorography.

RAT

HEPATOCYTES

SYNTHESIZE

CELLULAR

FIBRONECTIN

FIG. 4. Monolayers of hepatocytes in primary culture at day 1 (a, c, e) and 3 (b, d, f), subjected to 10m7M dexamethasone (e), after immunostaining with anti-fibronectin (a, b) or anti-cellular fibronectin antibodies (c, d, e) or with rabbit-serum as a control (f). Nonspecific bright fluorescence was associated with nonviable hepatocytes. Scale bar represents 50 Km.

plasma fibronectin, but part of the extracellular fibronectin matrix originates from the delivery of cellular fibronectin. Upon dexamethasone treatment, formation of extracellular matrix fibronectin filaments was elevated and already at Day 1 after inoculation anti-cellular fibronectin antibodies stained a distinct fibrillar cellular fibronectin network (Fig. 4e). Effect of Dexamethasone

on Steady-State

mRNA Level

The stimulating effect of dexamethasone shown in fibronectin synthesis and secretion was further examined on the mRNA steady-state level in comparison

with hepatic proteins as albumin and cu2-macroglobulin. Total RNA extracted from hepatocytes at different days in primary culture either exposed or not to hormonal treatment was analyzed by hybridization with cDNA probes, recognizing actin, albumin, and a2-macroglobulin transcripts. For determination of the steadystate mRNA level of total and cellular fibronectin, cDNA of the 11th type III repeat, common in all fibronectin forms, and cDNA of the additional ED-A domain in cellular fibronectin were used. Prolonged culture and dexamethasone treatment of rat hepatocytes caused an enhanced transcript level of both fibronectin forms (Figs. 5 and 6). Whereas in total RNA from freshly iso-

294

ODENTHAL ET AL. -

f

+

-+-+

1

3

+

cellular fibronectin

4-

total fibronectin

e

a2-M

A

m total

5

w EDA+

fibronectin tlbmnectir

days FIG. 5. Northern blot hybridization of total RNA (5 pg) obtained from freshly isolated hepatocytes (f) or from hepatocytes maintained in primary culture for 1,3, and 5 days, treated with 10-‘Mdexamethasone (+) or not (-). The Northern blot was probed with the 180.bp cDNA-fragment of the ED-A domain recognizing cellular fibronectin mRNA (cellular fibronectin), or with the 100 bp cDNA-fragment of the 11th type III repeat (total fibronectin) or with pRL29J plasmid DNA (u2-macroglobulin, (~2-M).

lated hepatocytes as well as in RNA from hepatocytes lysed at Day 1 after isolation total fibronectin-specific transcripts were detectable in small amounts cellular fibronectin mRNA could be shown to be present only in hepatocytes of primary culture at Day 1 upon dexamethasone treatment (Figs. 5 and 6). The marked response to dexamethasone in early primary culture was also proved in Lu2-macroglobulin synthesis, leading to a delayed decrease in a2-macroglobulin transcript level during primary culture. Although albumin expression is affected by hormonal treatment in a similar way, the retarded decline was not observed until Day 5 after plating. In contrast to the modulation of mRNA level of these hepatic proteins, the expression of actin in cultured hepatocytes is not significantly affected by dexamethasone (Fig. 6). By dot-blot hybridization we investigated whether the enhancement of fibronectin mRNA level in re-

t

fibronectin

+

albumin

t

a2-M

4- actin

1

davs

I

FIG. 6. Northern blot hybridization of total RNA (5 pg) extracted from hepatocyte primary cultures at Days 1 and 5 after plating and treated with 10m7M dexamethasone (+) or not (-). Hybridization was performed with plasmid DNA pFH 111 (fibronectin), alb 2M (albumin), or the plasmid pRL29J recognizing ot2-macroglobulin mRNA (otP-M) or with plasmid DNA PAC 269 (actin).

1

1.

1+

3-

3+

SMC 6

0

1

3

SMC

FIG. ‘7. Dot-blot analysis of fibronectin mRNA-level in freshly isolated hepatocytes (0) and hepatocytes during primary culture at Days 1 and 3 (1, 3) or in smooth muscle cells (SMC). For dot blot analysis, three single experiments were performed. (A) mRNA-level of fibronectin containing the ED-A domain (ED-A+) or of total fibronectin. Dexamethasone treatment of hepatocytes is indicated by (k). (B) Ratio of fibronectin mRNA containing the ED-sequence (ED-A+ FN) to total fibronectin mRNA (total FN) in freshly isolated hepatocytes (0) and at Days 1 and 3 of primary culture (1, 3) or in smooth muscle cells (SMC).

sponse to dexamethasone correlates with an altered ratio of fibronectin transcripts containing the additional ED-A sequence to the mRNA of total fibronectin. Dotblot analysis confirmed the prominent response of fibronectin mRNA level to dexamethasone during the early primary culture of hepatocytes already shown in Fig. 5. At Day 1 of primary culture dexamethasone treatment caused an enhancement of total fibronectin mRNA level of about 3 times whereas at Day 3 the stimulation declined to a factor of 1.2. As Fig. 7 shows the mRNA level of fibronectin bearing the ED-A sequence is also more affected at the very early state of primary culture when dexamethasone increased the portion of ED-Apositive fibronectin transcripts to 50% of the fibronectin mRNA population (Fig. 7B). Since at Day 3 the contribution of ED-A-positive to total fibronectin transcripts had become nearly equal in untreated and dexamethasone-treated hepatocytes (about 40% each), prolongation of primary culture led to a significant decline in the dexamethasone effect (Fig. 7B). DISCUSSION In the past 10 years hepatocytes were the main source of plasma fibronectin [5,44]. In the present report, how-

RAT

HEPATOCYTES

SYNTHESIZE

ever, using epitope-specific antibodies, distinguishing fibronectin, bearing the ED-A domain from other forms, and by fucose incorporation, we accumulated evidence that rat hepatocytes are also able to synthesize and secrete cellular fibronectin. Increased synthesis of fibronectin in the primary culture of rat hepatocytes was shown by analysis of de novo synthesis and by immunocytology, confirming previous reports [21-23,431. Since a pronounced part of fibronectin was fucosylated and Fukuda and Hakomori [3] pointed out fucose incorporation as a criterion for cellular fibronectin in hamster, we presumed that not only is the predominant plasma fibronectin produced in the monolayer of hepatocytes, but cellular fibronectin is produced as well. Although Tamkun and Hynes [5] could not identify cellular fibronectin in the culture media of rat hepatocytes, the ability of rat hepatocytes to produce cellular fibronectin still had to be clarified. Because the investigation of Tamkun and Hynes was concerned with hepatocytes cultured in the presence of fetal calf serum, which has been shown to inhibit fibronectin elaboration and assembly in rat hepatocyte monolayers [45], cellular fibronectin synthesis might have become undetectable in their primary culture system. By immunoprecipitation of newly synthesized proteins and immunocytological detection with site-specific a-cellular fibronectin antibodies, we were able to show an enhanced cellular fibronectin synthesis and secretion after a lag period of the first day in a serum-free primary culture of rat hepatocytes. Several authors describe hepatocytes as possible suppliers of biomatrix components [46-481. Following these reports, hepatocytes are able to synthesize collagen in primary culture. Maher et al., however, point out that collagen synthesis in rat hepatocyte primary culture might result from contaminating fat-storing cells [49]. Although the hepatocyte isolates obtained by the collagenase-perfusion procedure of the liver are almost free of mesenchymal cells [50], in primary hepatocyte cultures their presence by proliferation has to be considered. According to the determination of up to 10% contaminating fat-storing cells in primary cultures of hepatocytes at Day 8 after plating [49], the authors conclude that the contribution of hepatocytes to biomatrix production in vivo is only minute. By immunocytologically monitoring the hepatocyte monolayers for laminin and desmin, which can be regarded as marker proteins for fat-storing cells [39, 401, we determined no significant contamination at Day 1 and a contamination below 3% in the late primary culture. The few proliferating fatstoring cells in the hepatocyte monolayer can be excluded as a source for cellular fibronectin, because fibronectin synthesis of fat-storing cells takes place only after activation in primary cultures after Day 3 [ 181. Moreover, fibronectin synthesis in fat-storing cells is not inducible by dexamethasone [51], corresponding to

CELLULAR

FIBRONECTIN

295

the synthesis in liver epithelial cells [52], but unlike fibronectin synthesis in primary cultures of hepatocytes. Dexamethasone is implicated in the regulation of the acute-phase response and stimulates the production of fibrinogen, a2-macroglobulin, and other acute-phase reactants. Increased synthesis and formation of fibronectin upon dexamethasone treatment observed in the monolayer of rat hepatocytes (Figs. 3,4,6) is consistent with previous work [21-23,431. The dexamethasone response shown in primary cultures of rat hepatocytes confirms fibronectin as an acute-phase protein and correlates with an increased fibronectin level after an induced acute inflamation in vivo [44]. In the presented results, however, we demonstrate that in rat hepatocytes dexamethasone also causes enhanced synthesis of fibronectin bearing the ED-A domain. Recent work of Peters and co-workers presents ample evidence of irregular high levels of fibronectin containing the ED-A segment in plasma of patients with vasculitis [53] or an acute pulmonary injury [54]. Elevations in plasma content of ED-A-positive fibronectin, but not of total fibronectin, were also noted in individuals with acute vascular tissue injury associated with major trauma of sepsis syndrome. Peters et al. consider endothelial cells as a source of circulating ED-A-positive fibronectin [26,55]. According to our results hepatocytes are also supposed to be a prominent cell type, responsible for delivery of ED-A-positive fibronectin into the blood. The modulation of fibronectin synthesis by dexamethasone is due to an increase of mRNA for both fibronectin subforms (Figs. 6 and 7). By runoff transcription assays a transcriptional control by dexamethasone has been suggested [56]. This assumption could be confirmed by cloning of the 5’-genomic DNA segment of the fibronectin gene, which revealed regulatory sequences for glucocorticoidand CAMP-mediated regulation, both in human and in rat DNA [57]. While changes in alternative splicing products and reported shifts in EDA content of fibronectin were demonstrated during liver tissue development and tumorigenesis [ 16,171 as well as in fibroblasts exposed to TGF 0 stimulation [ 131 or with in vitro passage [15], Balza et al. report that in human fibroblasts the elevated accumulation of total fibronectin by dexamethasone does not correlate with a change in the ratio of ED-A positive to total fibronectin [ 131. In our investigation on hepatocyte primary cultures, however, we demonstrate a remarkable increase of ED-A content in fibronectin transcripts upon dexamethasone treatment at Day 1 after plating, whereas there was no significant alteration due to the hormonal treatment in the later state of primary culture, suggesting only an indirect effect on the splicing mechanism by the glucocorticoid dexamethasone. We thank Christine Waldmann for excellent technical assistance. This work was supported by the Deutsche Forschungsgemeinschaft (SFB 311 A7, and Ra 36215-2). G. Ramadori held a Herrmann and Lilly Schilling professorship.

296

ODENTHAL REFERENCES

ET AL.

29. Laemmli, 30.

1.

Hynes, R. O., and Yamada, 377.

2.

Yamada, K. M. (1988) in Fibronectin (Mosher, D. F., Eds.) pp. 47-121, Academic Press, San Diego. Fukuda, M., and Hakomori, S. (1979) J. Biol. Chem. 254,54515457.

3. 4. 5. 6.

7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18.

19. 20. 21. 22.

K. M. (1982) J. Cell Biol. 95, 369-

Paul, J. I., and Hynes, R. 0. (1984) J. Biol. Chem. 259, 1347713487. Tamkun, J. W., and Hynes, R. 0. (1983) J. Biol. Chem. 258, 4641-4647. Skorstengaard, K., Thorerson, H. C., Vibe-Pederson, K., Peterson, T. E., and Magnusson, S. (1982) Eur. J. Biochem. 140, 235-243. Kornblihtt, A. R., Vibe-Pedersen, K., and Baralle, F. E. (1984) EMBO J. 3, 221-226. Hynes, R. 0. (1985) Annu. Rev. Cell Biol. 1, 67-90. Vartio, T., Lattinen, L., Narvanen, D., Cutolo, M., Thornell, L. E., Zardi, L., and Virtamen, I. (1987) J. Cell Sci. 88, 419. Sekiguchi, K., and Titani, K. (1989) Biochemistry 28, 32933298. Zardi, L., Carnemolla, B., Siri, A., Petersen, T. E., Paolella, G., Sebastio, G., and Baralle, F. E. (1987) EMBO J. 6, 2337-2342. Schwarzbauer, J. E., Patel, R. S., Fonda, D., and Hynes, R. 0. (1987) EMBO J. 6,2573-2580. Balza, E., Borsi, L., Allemanni, G., and Zardi, L. (1988) FEBS Lett. 228, 4244. Barone, M. V., Henchcliffe, C., Baralle, F. E., and Paolella, G. (1989) EMBO J. 8, 1079-1085. Burke, E. M., and Danner, D. B. (1991) Biochem. Biophys. Res. Commun. 178,620-624. Oyama, F., Murata, Y., Suganuma, N., Kimura, T., Titani, K., and Sekiguchi, K. (1989) Biochemistry 28, 1428-1434. Oyama, F., Hirohashi, S., Shimosato, Y., Titani, K., and Sekigu chi, K. (1989) J. Biol. Chem. 264, 10331-10334. Ramadori, G., Knittel, T., Odenthal, M., Schwogler, S., Neubauer, K., and Meyer zum Biischenfelde, K. H. (1992) Submitted for publication. Rieder, H., Ramadori, G., Dienes, H. P., and Meyer zum Btischenfelde, K. H. (1987) Hepatoology 7, 856-864. Baumann, H., and Eldredge, D. (1982) J. Cell Biol. 95, 29-40. Amrani, D. L., Mauzy-Melitz, D., and Mosesson, M. W. (1986) Biochem. J. 238, 365-371. Marceau, N., Goyette, R., Valet, J. P., and Deschenes, J. (1980)

31.

Chirgwin, J. M., Przybyla, A. E., and McDonald, Biochemistry 18,5284-5287.

32.

Glisin,

24. 25. 26. 27. 28.

Voss, B., Allam, S., Rauterberg, J., Ullrich, K., Ginselmann, V., and van Figura, K. (1979) Biochem. Biophys. Res. Commun. SO, 1348-1354. Seglen, P. 0. (1976) in Methods in Cell Biology, Vol. 13, pp. 29-83, Academic Press, San Diego. Dijkstra, C. D., Diipp, E. A., Joling, P., and Kraal, G. (1985) Immunology 54,589-599. Peters, J. H., Ginsberg, M. H., Bohl, B. P., Sklar, L. A., and Cochrane, C. G. (1986) J. Clin. Znuest. 78, 1596-1603. Ramadori, G., Sipe, J. D., Dinarello, C. A., Mizel, S. B., and Colton, H. R. (1985) J. Exp. Med. 162, 930-942. Cullen, S. E., and Schwartz, B. D. (1976) J. Zmmunol. 117,136142.

Received February 27, 1992 Revised version received August 6, 1992

V., Crkvenjakow,

R. J. (1979)

R., and Byus, C. (1973) Biochemistry

13,2633. 33. 34. 35. 36.

37. 38. 39. 40. 41.

42. 43. 44. 45. 46. 47.

48. 49. 50. 51.

Exp. Cell Res. 125, 497-502. 23.

U. K. (1970) Nature 227, 6809685. Lasky, R. A., and Mills, A. D. (1975) Eur. J. Biochem. 56,335341.

52. 53. 54. 55. 56. 57.

Lehrbach, H., Diamond, D., Wozney, J. M., and Boedtker, H. (1977) Biochemistry 16, 4743. Southern, E. M. (1975) J. Mol. Biol. 98, 503-517. Schwartz, R. J., Haron, J. A., Rothblum, K. N., and Dugaiczyk, A. (1980) Biochemistry 19, 5883-5890. Gehring, M. R., Shiels, B. R., Northemann, W., M. H. L., D., Khan, C. C., Chain, A. C., Noonan, D. J., and Fey, G. H. (1987) J. Biol. Chem. 262, 446-454. Kioussis, D., Eiferman, F., V. D., Rijn, P., Gorin, M. B., Ingram, R. S., and Tilgham, S. M. (1981) J. Biol. Chem. 256,1960-1968. Ramadori, G., Rieder, H., Knittel, T., Dienes, H. P., and Meyer zum Biischenfelde, K. H. (1987) J. Hepatol. 4, 190-197. Maher, J. J., Friedman, S. L., Roll, F. J., and Bissel, D. M. (1988) Gastroenterology 94, 1053-1062. Yokoi, Y., Namihisa, T., Kuroda, H., Komatsu, I., Miyazaki, A., Watanabe, S., and Usui, K. (1984) Hepatology 4, 709-714. Ramadori, G., Veit, T., Schwogler, S., Dienes, H. P., Knittel, T., Rieder, H., and Meyer zum Btischenfelde, K. H. (1990) Virchows Arch. B Cell Pathol. 59, 349-357. Amrani, D. L., Falk, M. J., and Mosesson, M. W. (1985) Exp. Cell Res. 160, 171-183. Nimmer, D., G., B., Hirano, H., and Amrani, D. L. (1987) J. Biol. Chem. 262, 10369-10375. Owens, M. R., and Cimino, C. D. (1982) Blood 59, 130551309. Stamatoglou, S. C., Hughes, R. C., and Lindahl, U. (1987) J. Cell Biol. 105,2417-2425. Guzelian, P. S., Lindblad, W. J., and Diegelmann, R. F. (1984) Gastroenterology 86, 897-904. Hata, R. L., Ninomaya, Y., Sano, J., Konomi, H., Hori, H., Sunada, H., Tanaka, S., Kabuki, K., Nagai, Y., and Tsukaba, Y. (1985) J. Cell Phys. 122, 333-342. Tseng, S. C. G., Smuckler, E. A., and Stern, R. (1983) Hepatology 3,955-963. Maher, J. J., Bissel, D. M., Friedman, S. L., and Roll, F. J. (1988) J. Clin. Znuest. 82, 450-459. Bissel, D. M., Hammaker, L. E., and Meyer, U. A. (1973) J. Cell Biol. 59, 722-734. Ramadori. G., Knittel, T., Schwogler, S., Bieber, F., Rieder, H., and Meyer zum Biischenfelde, K. H. (1991) Hepatology 14,875882. Seebacher, T., Manske, M., Kornblihtt, A. R., and Bade, E. G. (1988) FEBS Lett. 239, 113-116. Peters, J. H., Maunder, R. J., Woolf, A. D., Cochrane, C. G., and Ginsberg, M. H. (1989) J. Lab. Clin. Med. 113, 586-597. Peters, J. H., Ginsberg, M. H., Case, C. M., and Cochrane, C. G. (1988) Am. Reu. Resp. Dis. 138, 167-174. Peters, J. H., Sporn, L. A., Ginsberg, M. H., and Wagner, D. D. (1990) Blood 75, 1801-1808. Tyagi, J. S., Hirano, H., Merlino, G. T., and Pastan, I. (1983) J. Biol. Chem. 258,5787-5793. Patel, R. S., Odermatt, E., Schwarzbauer, J. E., and Hynes, R. 0. (1987) EMBO J. 6, 2565-2572.