THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 262, No. 21, Issue of July 25, pp. 10369-10375,1987 0 1987 by The American Society of Biological Chemists, Inc. Printed in U.S.A.

Regulation of Plasma Fibronectin Biosynthesis by Glucocorticoids in Chick Hepatocyte Cultures*

(Received for publication, October 14, 1986, and in revised form, February 19, 1987)

David NimmerSP, Gerald BergtromS, Hideyasu Hiranoll, and David L. AmraniSII From the $Department of Biology, University of Wisconsin-Milwaukee, Milwaukee, Wisconsin 53201, the lDepartment of Biochemistry, University of Occupational and Environmental Health, Japan School of Medicine, Yahutanishi Kitakyushu 807, Jawn. and the II Devartment of Medicine. Universitv of Wisconsin Medical School, Mount Sinai Medical Center, Milwaukee, Wisconsin 53233 .

- .

Plasma fibronectin is an acute-phase reactant syn- thesized by hepatocytes. Glucocorticoids are one of the major factors implicated in controlling the hepatic acute-phase response. To study the regulatory effects of glucocorticoids on plasma fibronectin biosynthesis, a model chick hepatocyte culture system under serum- and hormone-free conditions was used. In the presence of either dexamethasone or corticosterone, secreted plasma fibronectin increased maximally to 2.8-fold above basal levels. The stimulatory effect of the hor- mones was maintained only in their continuous pres- ence, since plasma fibronectin production dropped to near basal levels within 16 h of glucocorticoid with- drawal. Pulse-chase studies indicated that pretreat- ment of cells with dexamethasone stimulated the level of secreted plasma fibronectin but had no effect on its rate of secretion. The increase in plasma fibronectin production by dexamethasone was abolished in a dose- dependent manner by the addition of progestin, an antagonist of dexamethasone known to compete specif- ically for the liver glucocorticoid receptor. Actinomy- cin D and a-amanitin, which are inhibitors of tran- scription, also blocked the early dexamethasone effect on plasma fibronectin synthesis. Slot blot hybridization of total RNA samples from dexamethasone-treated cul- tures revealed a 6-fold stimulatory rise in fibronectin mRNA during the first 6 h of treatment, which later declined and was no longer evident at 48 and 72 h. However, fibronectin mRNA levels were elevated to about the same extent in control and dexamethasone- treated cells at the later time points. During the same time period (0 to 72 h), plasma fibronectin protein levels rose and remained elevated. Evaluation of pulse- chase experiments following pretreatment with hor- mone for 48 h demonstrated that equal amounts of plasma fibronectin were translated by dexamethasone- treated and control cells, but only 42% of labeled plasma fibronectin was secreted by control cells com- pared with 93% for dexamethasone-treated cells. These findings suggest that the early phase of gluco- corticoid regulation of plasma fibronectin biosynthesis

*This research was supported in part by National Institutes of Health Program Project HL28444, National Institutes of Health Investigator Award NIH R23-AM32762 (D. L. A,), and by a Univer- sity of Wisconsin-Madison, Graduate Research Committee Award (G. B.). These results were presented in part at the 25th Annual Meeting of the American Society for Cell Biology (Nimmer, D., Bergtrom, G., Hirano, H., and Amrani, D. L. (1985) J. Cell Biol. 101, 320a). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

8 In partial fulfillment of a Masters thesis requirement.

occurs at the transcriptional level and is mediated through the specific action of the glucocorticoid recep- tor. A later phase of glucocorticoid-stimulated plasma fibronectin biosynthesis results from modulation of post-translational processing events leading to secre- tion of an increased amount of newly translated plasma fibronectin polypeptides.

Plasma fibronectin, which is synthesized by hepatocytes (1, 2), is part of a family of structurally and immunologically related glycoproteins (3-5). These proteins are synthesized by many cells in the body and are active in cellular processes such as cell adhesion, cell migration, and phagocytosis (3-5). There is a single gene for fibronectins (6-8) with the cell- associated and plasma forms arising by alternative processing of the primary fibronectin mRNA transcript (7-10). Northern blot analysis of liver RNA reveals at least three different forms of plasma fibronectin mRNA (7-9) which account for the major subunits found for mammalian plasma fibronectins (1, 3, 11, 12).

The mechanisms regulating biosynthesis of cell-associated fibronectin and plasma fibronectin are not well understood. Earlier studies of cultured cells showed that transformed and neoplastic cells have reduced quantities of cell-associated fibronectin on their surfaces and in their culture media (4). As assessed by in vitro transcription studies (13), the reduced fibronectin production by Rous sarcoma virus-transformed fibroblasts result from reduced fibronectin mRNA transcrip- tion. In contrast, SV40 transformed keratinocytes show a marked increase in overall fibronectin synthesis with much of the fibronectin being secreted into the culture medium (14). Clinical studies of individuals undergoing treatment for severe trauma, burn injury, or infection demonstrate acute decreases in plasma fibronectin levels with restoration to normal levels within 7 to 14 days after the onset of the condition (15, 16). Circulating plasma fibronectin levels also vary in experimen- tal animals during acute inflammation (15-19), strongly sug- gesting that plasma fibronectin is an acute-phase reactant.

A number of studies suggest a role for hormonal and hu- moral factors in the control of fibronectin biosynthesis. Fi- bronectin synthesis in human fibrosarcoma cells and normal fibroblasts is stimulated by glucocorticoids (20). Dexametha- sone increases the synthesis and secretion of a dysfunctional fibronectin in rat hepatoma cells (21). Plasma fibronectin levels that are decreased in patients who have had a total thyroidectomy (22), are restored after thyroxine replacement therapy (23). Our group recently showed that hepatocyte- stimulating factor and dexamethasone, either singly, or in combination, can stimulate plasma fibronectin synthesis in

10369

10370 Plasma Fibronecfin Synthesis

chick hepatocytes (19). These two factors are known to be involved in the hepatic acute-phase response (24, 25). In this study, we explored the role of glucocorticoids in regulating plasma fibronectin synthesis in a well-defined chick hepato- cytes system (2).

EXPERIMENTAL PROCEDURES’

RESULTS

We first determined the dose-response of plasma fibronec- tin synthesis to the major naturally occurring glucocorticoid in chickens, corticosterone (32) and the synthetic hormone, dexamethasone, in cultured hepatocytes (Fig. 1). The dose- response curve indicates that half-maximal stimulation of plasma fibronectin synthesis was observed at 0.25 nM dexa- methasone or 5 nM corticosterone. At -3 nM dexamethasone or -25 nM corticosterone, plasma fibronectin synthesis was maximally stimulated and produced a 2.8-fold increase over basal levels, while total secreted protein increased only 6.4% over the same period.

The stimulatory effects of glucocorticoids on cultured he- patocytes are selective, as demonstrated by the two-dimen- sional gel electrophoretic analysis of secreted proteins syn- thesized by hepatocytes treated with or without 5 nM dexa- methasone (Fig. 2). Under basal conditions, several distinct proteins, including plasma fibronectin, fibrinogen subunits, albumin, and transferrin (which were identified by compari- son to the electrophoretic position of purified standards) were clearly observed (Fig. 2A). In the presence of 5 nM dexameth- asone, a 2- to 4-fold increase in the bands corresponding to plasma fibronectin and the fibrinogen subunits as well as some other proteins was evident (Fig. 2B). In contrast, levels of albumin and transferrin were not altered significantly (Fig. 2B). The observed selective changes in hepatic secretory protein production by glucocorticoids is in agreement with results from a previous study (26).

To determine whether the dexamethasone-induced in- creases in secreted plasma fibronectin were due at least in part to stimulation of plasma fibronectin secretion time, we examined the secretion time for plasma fibronectin from control cells or cells treated with dexamethasone for 5 h. The cells were pulse-labeled with [35S]methionine for a short pe- riod (3 or 10 min) in the absence of any unlabeled methionine and subsequently chased with medium containing only unla- beled amino acids, including methionine (Fig. 3). Consistent with an earlier experiment using a 15-min pulse period (2), a pulse period of 3 (Fig. 3.4) or 10 min (Fig. 3B) demonstrated a t,,* of approximately 40 to 45 min under basal or stimulated conditions for secreted plasma fibronectin. Also, dexametha- sone-treated cells synthesized a 2.2-fold greater amount of plasma fibronectin compared to control cells, and both control and hormone-treated cells had secreted 54% to 56% of their newly made plasma fibronectin over the 3-h period (Table I). Plasma fibronectin immunoprecipitates from the 10-min pulse-chase experiment (Fig. 3B) were examined by sodium

Portions of this paper (including “Experimental Procedures” and Figs. 1-9) are presented in miniprint at the end of the paper. The abbreviations used are: PFn, plasma fibronectin; Fn, fibronectin; SDS, sodium dodecyl sulfate NP-40, Nonidet P-40; PBS, phosphate- buffered saline; kb, kilobase; ELISA, enzyme-linked immunosorbent assay. Miniprint is easily read with the aid of a standard magnifying glass. Full size photocopies are available from the Journal of Biolog- ical Chemistry, 9650 Rockville Pike, Bethesda, MD 20814. Request Document No. 86M-3553, cite the authors, and include a check or money order for $5.00 per set of photocopies. Full size photocopies are also included in the microfilm edition of the Journal that is available from Waverly Press.

TABLE I Distribution of newly translated plusma fibronectin (PFn)

Hepatocytes (3 X lo6) were maintained in the presence or absence of dexamethasone for 5 or 48 h, then pulse-labeled for 3 min with [35Slmethionine (40 pCi/plate). After the pulse period, cell lysates were isolated from control and hormone-treated cells. The remaining plates were chased for 3 h with fresh medium containing unlabeled methionine. At this time, the medium and cell lysates were isolated and total labeled plasma fibronectin determined in all samples by immunoprecipitation and radioactive counting.

% of Treatment Lysate 2 pulse- labeled

PFn cpm X io3

Control (5 b) 3-min pulse 786 & 128 (100) 3-h chase 3-h chase 445 & 138 56.6

Dexamethasone-treated (5 h)

3-min pulse 1746 f 395 (100) 3-h chase 3-h chase

Control (48 h) 946 f 220 54.2

3-min pulse 1224 f 450” (100) 3-h chase 746 f 269 61.0 3-h chase 516 It 162 42.2

Dexamethasone-treated (48 h)

3-min pulse 1438 f 566” (100) 3-h chase 115 f 16 8 3-h chase 1340 f 495 93.2 No significant difference ( p < 0.02); results are from two exper-

iments with each sample done in triplicate. Values are mean f S.D. The 5-h treated samples were obtained from the same experiment shown in Fig. 3A.

dodecyl sulfate-polyacrylamide gel electrophoresis analysis under nonreducing conditions, and showed that both dimeric and monomeric plasma fibronectin species are secreted (Fig. 4).

Dexamethasone-stimulated levels of plasma fibronectin synthesis by cultured hepatocytes are maximal by 24 h and remain elevated over 3 days in culture in the presence of the hormone (Fig. 5 ) . We determined whether the continual pres- ence of dexamethasone is necessary to sustain elevated levels of plasma fibronectin synthesis. Chick embryo hepatocytes were cultured for 24 h after which half the cells were treated with dexamethasone (zero time). In the next 24 h a 2.3-fold increase in the plasma fibronectin level over the amount synthesized under basal conditions (Fig. 5, upper solid line versus lower solid line). The stimulated level of plasma fibro- nectin was maintained over a minimum of 3 days for as long as the hormone was present. However, if medium containing dexamethasone was removed and replaced with medium with- out the hormone, a relatively rapid decrease in the plasma fibronectin level to basal levels occurred within 16 h as measured by enzyme-linked immunosorbent assay (Fig. 5, dashed tine). Since it is known that stimulation of fibrinogen synthesis by glucocorticoids rapidly declines when the hor- mone is removed (26), this suggests that the fibronectin gene, like other glucocorticoid-responsive genes, may require the continuous presence of an active glucocorticoid-receptor com- plex. This is further supported by the ability of increasing concentrations of progestin R-5020 (a competitive inhibitor of the glucocorticoid receptor) (33, 34) to block dexametha- sone stimulation of plasma fibronectin synthesis in cultured hepatocytes (Fig. 6). Maximal inhibition (82%) of dexameth- asone-stimulated plasma fibronectin synthesis occurs at 100

Plasma Fibronectin Synthesis 10371

mM progestin. Progestin alone, which binds the receptor but does not form an active complex (33,34), was not stimulatory a t 100 mM (Fig. 6).

Since the proposed mechanism of action for the active glucocorticoid-receptor complex is to increase the accessibility of promoter sites of specific genes to RNA polymerase in eukaryotes (35), we evaluated the ability of dexamethasone to stimulate plasma fibronectin levels in the presence of tran- scriptional inhibitors (Fig. 7). Hepatocytes treated with 2.5 nM dexamethasone alone showed a 2.3-fold increase in plasma fibronectin over that synthesized under basal conditions be- tween 90 and 180 min (Fig. 7, A uersus 0). In the presence of actinomycin D (36), a 40% reduction in plasma fibronectin synthesis to basal levels was seen (Fig. 7,O). The concentra- tion used is a minimally effective dose of actinomycin (1 pg/ ml) for the inhibition of chick hepatocyte protein synthesis (26). a-Amanitin, a specific inhibitor of RNA polymerase I1 (37), also prevented stimulation of plasma fibronectin synthe- sis by dexamethasone (Fig. 7, 0). We determined that 1 pg/ ml of a-amanitin was the minimally effective dose for inhi- bition of the dexamethasone stimulatory response (data not shown).

We next determined whether the stimulation of plasma fibronectin protein levels in the presence of dexamethasone correlated with an increase in the steady-state levels of fibro- nectin mRNA. Slot blots of total RNA (isolated from hepa- tocyte cultures at different time points) were probed with a :3'P-labeled 2.99-kilobase genomic DNA fragment of chicken fibronectin (6, 14) (Fig. 8). Since there is no detectable cell- associated fibronectin synthesis in chick hepatocytes (a), we assumed that the dominant form of fibronectin mRNA pres- ent in these cells encodes plasma fibronectin. We examined the change in plasma fibronectin protein levels in the presence and absence of dexamethasone, and found that the change in plasma fibronectin synthesis was roughly linear during the first 12 h, with the stimulated increase in plasma fibronectin peaking by 18 to 24 h (Fig. 9). The plasma fibronectin levels remained elevated for the remainder of the 72-h period. Slot blot hybridization analysis revealed that steady-state fibro- nectin mRNA levels increased both in dexamethasone and in untreated cells during this period (Fig. 8). However, the data indicated that a maximal 6-fold stimulatory surge in fibronec- tin mRNA accumulation occurred in the first 6 h of exposure to dexamethasone (Figs. 8A and 9). After 6 h, absolute fibro- nectin mRNA levels in dexamethasone-treated cells increased while the stimulation of fibronectin mRNA declined to only a 1.5-fold level by 24 h (Figs. 8A and 9). Since fibronectin mRNA levels continue to rise in both dexamethasone-treated and control cells, the fibronectin mRNA levels in hormone- treated and control cells were equivalent by 48 h. Thus, the fibronectin mRNA stimulatory surge peaks at 6 h, while the plasma fibronectin protein levels peak at 18-24 h. Further- more, while elevated levels of plasma fibronectin protein synthesis continue for 3 days in hormone-treated cultures, the stimulation of plasma fibronectin mRNA levels over con- trol cultures, although not the absolute fibronectin mRNA values, declines.

We investigated whether the hormone-treated cells main- tained stimulatory levels of secreted plasma fibronectin by increased translation of fibronectin mRNA or by increased stability of newly translated plasma fibronectin. Cells pre- treated with 5 nM dexamethasone for 48 h were pulse-labeled for 3 min. The amount of labeled immunoprecipitable plasma fibronectin in cell lysates at the end of the pulse period was equal to that in control cells ( p < 0.02; see Table I). The pulse-labeled plasma fibronectin in cells treated in a similar

manner was chased into the medium for 3 h and the amount of labeled plasma fibronectin in the medium and cell lysates determined. The medium from the dexamethasone-treated cells contained 93.2% of the pulse-labeled plasma fibronectin whereas only 42.2% of pulse-labeled plasma fibronectin was found in control cells (Table I). The remaining labeled plasma fibronectin was retained in the respective cells a t a time of steady-state secretion (Fig. 3; also Ref. 2).

DISCUSSION

In this report, we found that corticosterone and dexameth- asone stimulation of plasma fibronectin production in primary chick hepatocytes was similar to that observed for circulating plasma fibronectin during either experimentally induced in- flammation or infusion of glucocorticoids i n vivo (19, 38). Acute inflammation in chickens, rats, or mice results in a 2- to 3-fold increase in circulating plasma fibronectin levels within 24 to 72 h (17-19,38,39). This is preceded in chickens and rats by a rise in glucocorticoid levels to 10 to 20 nM within 24 h of the onset of the acute or chronic inflammatory response (19, 38). Infusion of corticosterone or dexametha- sone into normal rats enhances the circulating plasma fibro- nectin level 1.5- to 3-fold within 3 days (38). Although infusion of the synthetic hormone, dexamethasone into experimentally induced inflammatory rats further stimulates plasma fibro- nectin levels, corticosterone infusion has no effect (38). Our dose-response data for corticosterone and dexamethasone ef- fects on hepatocytes (Fig. 1) suggest that natural glucocorti- coids, e.g. corticosterone, are maximally stimulatory a t 10 to 20 nM. Since this corticosterone level is already reached in the experimentally induced inflammatory chicken or rat (19, 38), only the more potent synthetic glucocorticoid can cause any further increase in plasma fibronectin levels. In addition, plasma fibronectin levels maximally increased at 48 h in turpentine-treated chickens remain elevated for at least 3 days (19); at the same time, circulating glucocorticoids also remain above basal levels (19). Our finding that the stimula- tion of plasma fibronectin synthesis by glucocorticoids is dependent on their continual presence (Fig. 5 ) is consistent with physiological observations (19, 38) and is similar to the continual hormone requirement for stimulation of fibrinogen levels in chickens (26). Also, the ability of inhibitors of glucocorticoid receptors to abrogate hormone-induced in- creases in plasma fibronectin synthesis in fibroblasts (RU- 486; Ref. 20) and in hepatocytes (R-5020; Fig. 6) suggests that continual formation of active hormone-receptor complexes are involved in sustaining the response. Taken together, these observations reinforce the hypothesis that glucocorticoids re- leased during acute and chronic inflammation play a central role in the regulation of the hepatic acute-phase response and of plasma fibronectin synthesis in particular.

That glucocorticoids stimulate synthesis of cell-associated and plasma fibronectin (20, 38-44; this study) is clear, but the mechanism by which this occurs in fibroblasts may differ from that in hepatocytes. Studies on glucocorticoid regulation in normal fibroblasts suggest that the plasma fibronectin gene is regulated by a transformation-sensitive, glucocorticoid-in- dependent mechanism for basal synthesis and that the dexa- methasone effect can be primarily accounted for by glucocor- ticoid receptor action at the level of transcription (20). Recent studies with rat fibroblasts further suggest pre-translational control of collagen and fibronectin synthesis by glucocorti- coids (44). However, the degree of glucocorticoid stimulation of fibroblast fibronectin in these two studies (20,44) is differ- ent. While our results also indicate that basal plasma fibro- nectin synthesis in normal chick hepatocytes is a hormone-

10372 Plasma Fibronectin Synthesis

independent event, the mechanism of glucocorticoid stimula- tion of plasma fibronectin synthesis in chick hepatocytes appears to be more complex than a singular, direct effect of the hormone on the expression of fibronectin genes. Further- more, this stimulation appears independent of the post-pack- aging secretory process, since our data (Fig. 3) indicates that the hormones do not affect post-translational rate of plasma fibronectin secretion. Dexamethasone treatment of chick he- patocytes induced an early, rapid increase in fibronectin mRNA levels that was 6-fold above control levels. In our studies, we assumed that plasma fibronectin mRNA is the dominant form of fibronectin mRNA in our cultured chick hepatocytes. This assumption is based on the observation that the majority of the fibronectin synthesized by these cells is the same as plasma fibronectin (2) and on studies of rat and human liver RNA in which only plasma fibronectin mRNA species were detected (7-9, 11). Thus, the maximal stimulation of fibronectin mRNA preceded a peak in plasma fibronectin protein stimulation by 12 to 18 h, suggesting that these elevated fibronectin mRNA levels might be responsible for the early increases in plasma fibronectin synthesis. This is consistent with our observation that hormone-induced in- creases in plasma fibronectin are blocked in cells treated with inhibitors of transcription (a-amanitin or actinomycin D). Recent in vitro transcription "run-off" studies with isolated human fibroblast nuclei provide further support for this con- clusion, since increases in fibronectin transcription appear due in part to enhancement in the rate of fibronectin gene initiation.' However, while the amount of secreted plasma fibronectin in cells treated continuously with glucocorticoids remain elevated for at least 3 days, fibronectin mRNA in hormone-treated cells were equivalent to control levels by 48 h and declined slightly below basal levels at 72 h under the same conditions. We noted that the absolute values for fibro- nectin mRNA in both hormone-treated and untreated cells rise during this 3-day period. The reason for this increase in fibronectin mRNA levels, especially in the control cells, is currently unknown, and speculation on the continued rise is premature at this time.

The sustained stimulation of the plasma fibronectin protein levels, even as the stimulatory effect on fibronectin mRNA of hormone-treated cells declines, appears due at least in part to hormone-induced effects on a post-translational increase in the stability or packaging of new plasma fibronectin mole- cules. Analysis of cells treated for 5 h with or without dexa- methasone show that a 2.2-fold increase in newly translated plasma fibronectin was present in dexamethasone-treated cells exposed to [35S]methionine for 3 min (Fig. 3A and Table I). This is consistent with the early dexamethasone stimula- tion of fibronectin mRNA levels. In contrast, both control and dexamethasone-treated cells, pretreated for 48 h, contain equal amounts of newly translated plasma fibronectin after the pulse period, again consistent with the similar steady- state fibronectin mRNA levels in hormone-treated and con- trol cells at this time. However, after a 3-h chase period, twice as much labeled plasma fibronectin was secreted in the 48-h dexamethasone-treated cells compared to control cells (93.2% versus 42.2%). Thus, after 48 h of hormone treatment, the hepatocytes apparently synthesize equal amounts of newly translated plasma fibronectin from similar amounts of fibro- nectin mRNA. We conclude, therefore, that post-translational regulation in relatively long term (e.g. 48 h) dexamethasone- treated cells is responsible for this increased packaging and release of plasma fibronectin compared to control cells.

Our experimental results are similar to those on the stabil-

* N. Oliver, personal communication.

ity of fibrinogen synthesized in chick hepatocytes (45). In these pulse-chase experiments (45), only 30% of the labeled fibrinogen synthesized in hepatocyte cultures maintained in the absence of added hormones or serum was released into the medium. The remainder of the labeled fibrinogen disap- peared in a manner suggestive of intracellular degradation. When the cells were exposed to serum, there was an increase in fibrinogen mRNAs and fibrinogen synthesis, resulting in a large increase in secreted protein. Most significantly, there was a complete release (>95%) of fibrinogen label (45). Taken together, these results (Ref. 45; this study) suggest that when hormones or factors present in serum are added to cultured hepatocytes, they induce changes in post-translational events, i.e. intracellular degradative events, as well as increased tran- scription of fibrinogen mRNAs (45) or fibronectin mRNAs (this study), to stimulate production of these proteins.

Thus, overall regulation of plasma fibronectin production by glucocorticoids seems to occur at the transcriptional and post-translational level in the hepatocyte. Plasma fibronectin synthesis and secretory levels rise initially through greater increases in available fibronectin mRNA. Later, while fibro- nectin mRNA and plasma fibronectin synthesis are similar in hormone-treated and untreated cells, secreted plasma fibro- nectin levels remain elevated, presumably by glucocorticoid induction of factors controlling the stability or packaging of translated plasma fibronectin.

Acknowledgments-We thank Betty Perrin for secretarial assist- ance in typing this manuscript and Joyce Mitchell for the graphic illustrations. We thank Dr. Michael W. Mosesson for his critical review and helpful suggestions in the preparation of this manuscript. We also thank Mary Faculjak for her expert technical assistance.

REFERENCES 1. Tamkun, J. W., and Hynes, R. 0. (1983) J. Biol. Chem. 258 ,

2. Amrani, D. L., Falk, M. J., and Mosesson, M. W. (1985) Exp. Cell

3. Mosesson, M. W., and Amrani, D. L. (1980) Blood 56, 145-158 4. Yamada, K. M. (1983) Annu. Reu. Biochem. 5 2 , 761-799 5. Mosher, D. F. (1984) Annu. Reu. Med. 3 5 , 561-575 6. Hirano, H., Yamada, Y., Sullivan, M., de Crombrugghe, B., Pas-

tan, I., and Yamada, K. M. (1983) Proc. Natl. Acad. Sci. U. S. A.

7. Schwarzbauer, J. E., Tamkun, J. W., Lemischka, I. R., and Hynes,

8. Kornblihtt, A. R., Vibe-Pedersen, K., and Baralle, F. E. (1984)

9. Tamkun, J. W., Schwarzbauer, J. E., and Hynes, R. 0. (1984) Proc. Natl. Acad. Sci. U. S. A. 81, 5140-5144

10. Schwarzbauer, J. E., Paul, J. I., and Hynes, R. 0. (1985) Proc. Natl. Acad. Sci. U. S. A. 82, 1424-1428

11. Sekiguchi, K., Klos, A. M., Kurachi, K., Yoshitake, S., and Hakomori, S. (1986) Biochemistry 25,4936-4941

12. Amrani, D. L., Homandberg, G. A., Tooney, N. M., Wolfenstein- Todel, C., and Mosesson, M. W. (1983) Biochim. Biophys. Acta

13. Hynes, R. O., and Yamada, K. M. (1982) J. Cell Biol. 96, 369-

14. Tyagi, J. S., Hirano, H., Merlino, G. T., and Pastan, I. (1983) J.

15. Saba, T. M., Blumenstock, F. A., Weber, P., and Kaplan, J. E.

16. Richards, P. S., and Saba, T. M. (1983) Infect. Zmmun. 39,1411-

17. Owens, M. R., and Cimino, D. C. (1982) Blood 59,1305-1309 18. Stecher, V. J., Kaplan, J. E., Connolly, K., Mielens, Z., and

19. Amrani, D. L., Mauzy-Melitz, D., and Mosesson, M. W. (1986)

20. Oliver, N., Newby, R. F., Furcht, L. T., and Bourgeois, S. (1983)

21. Baumann, H., and Eldredge, D. (1982) J. Cell Biol. 95,29-40

4641-4647

Res. 160 , 171-183

8 0 , 46-50

R. 0. (1983) Cell 35,421-431

EMBO J. 3,221-226

748,308-320

377

Biol. Chem. 258,5787-5793

(1978) Ann. N. Y. Acad. Sci. 312 , 43-55

1418

Saelens, J. K. (1986) Arthritis Rheum. 29, 394-399

Biochem. J. 238,365-371

Cell 33 , 287-296

Plasma Fibronectin Synthesis 10373

22. Watzke, H., Schwarze, H. P., and Weissel, M. (1985) Thromb. (1981) J. Bid. Chem. 2 5 6 , 10503-10508

23. Graninger, W., Pirich, K., Derfler, K., and Waldhausl, W. (1965) N. E. (1983) Science 222, 1341-1343

24. Miller, L. L., and Griffin, E. E. (1975) in Biochemical Actions of Rutter, W. J. (1970) Science 2 2 2 , 1341-1343

Haemostasis 54, 249a 35. Pfahl, M., McGinnis, D., Hendricks, M., Groner, B., and Hynes,

J. Clin. Pathol. (Land. ) 38,64-67 36. Lindell, T. J., Weinberg, F., Morris, P. W., Roeder, R. G., and

Hormones (Litwack, G., ed) Val. 3, pp. 159-186, Academic 37. Penman, s., Vesco, c., and Penman, M. (1968) J. Mol. Bid. 3 4 ,

Natl. Acad. Sci. U. S. A. 75,5506-5510 27. Grieninger, G., Plant, P. W., Liang, T. J., Kalb, R. G., Amrani,

D., Mosesson, M. W., Hertzberg, K. M., and Pindyck, J. (1983) Ann. N. Y. Acad. Sci. 408,469-489

28. O'Farrell, P. H. (1975) J. Bwl. Chern. 250,4007-4021 29. Krawetz, S., and Anwar, R. (1984) Biotechniques 2 , 342-347 30. Cheley, S., and Anderson, R. (1984) A d . Biochem. 137, 15-19 31. Singh, L., and Jones, K. (1984) Nucleic Acids Res. 12,5627-5638 32. Beuving, G., and Vonder, G. M. A. (1978) Gen. Comp. Endocrinol.

33. Svec, F., and Rudis, M. (1982) J. Steroid Biochern. 16, 135-140 34. Nordeen, S. K., Lan, N. C., Showers, M. O., and Baxter, J. D.

35,153-159

Supplementary Material to

by GlUCoCO~tlCOidS in Chick Hepatocyte Cultures RegUldtlOn of Plasma Fibronectin Biosynthesis

BY

0. Nirnmer, G. Bergtrom, H . Hirano and 0. L. Amrani

aiieri,%la - L-ornlrhine-HCI was Obtalned from P.L.Biochemicals. Milwaukee, Wisconsin. Carbamyl phosphate, phenylmethylsulfonyl- fluoride IPMSFI . Kunitz' pancreatic trypsin inhibitor IAprotinxnI. a-ananltinand dithiothreltol (DTTI were rmrchdsed fcoms~omd ChenlCal CO., St. Louls, Missouri. Trypsin l j x crystallized) ;as from ICN. Cleveland, Ohio. Actinomycin D was Obtained from Behringer Mannherm. Dexamethasone and COrtICosterone were obtained from

Coloradq2yrum Co.. Denver , Colorado. !.-J5jSSlnethlonine (>llOOci/

Steraloids. 1°C.. Wllton, New Hampshlr Chick plasma was from

mmoll, I-goat anti-rabb~t IgG, 1 - PiATP 13000 Ci/mmall and 'Enllghting' Wele from New England Nuclear. N. Blllerica, MaSSaChu- setts.

B e L l u d ~ - Primary hepatocyte monolayer c u 1 t u 1 - e ~ were prepared from 16 day-old Chick embryos. as previously described 12.2gl. Hepatocyte

Willlaas E medim without hoimonal 01 serum supplement onto either suspensions yere plated at a concentration of 3 x 10 cells/ml in

16 mm I1 m l l , 3 5 mrn 12 m l l . 60 ma 15 m l l or 100 mm (10 m l l plastic culture dlshes. For most experiments. cultures were initlated and maintained In Williams E medlurn containlng 0.7 mM a r g l n l n e . At 24 h intervals, the medium was replaced with fresh medium lacking argl- nine but supplemented with 0.7 mM ornlthlne and 1 mM Carbamyl phos-

and hirvdtn ISlgind, 2000 units/mll at 15 "q/ml and 0,2unlts/ml. phate and contalnlng sodium heparin Solution (Fisher, 156 unlts/mgl

respectively. In certain experiments. Other supplements 1e.g.. hor- mone.1 yere added as described elsewhere.

40.

41.

42.

43.

44. Raghow, R., Gossage, D., and Kang, A. H. (1986) J. Biol. Chem.

45. 26 1,4677-4684

Furcht, L. T., Mosher, D. F., Wendelschafer-Crabb, G., and Foidard, J.-M. (1979) Cancer Res. 39, 2077-2083

Furcht, L. T., Mosher, D. F., Wendelschafer-Crabb, G., Wood- bridge, P. A., and Foidart, J.-M. (1979) Nature 277,393-395

Piovella, F., Giddings, J. C., Ricetti, M. M., Almasio, P., and Ascari, E. (1982) Hwmatologica 67,58-63

Marceau, N., Goyette, R., Valet, J. P., and Deschenes, J. (1980) Exp. Cell Res. 125,497-502

I D 1

Grieninger, G., Plant, P. W., and Chaisson, M. A. (1984) J. Biol. Chem. 259,14973-14978

samples was carried Out a s pceY1ously described 121. Culture One-dimensional gel electrophoresis Of mmunopreclpltated Fn

medium samples were analyzed on two-dimensional gels with 1soelec-

the second 1261. An al~quot 1 1 5 0 ~ 1 1 Of each medium sample was tric focusing in the flcst dimension and SDS-polyacrylamide gels in

reduced in 5 0 urea, 1% Nonidet P-40 and 20 mM dithmthieitol IDTTI and r u n on 1SOeleCttiC focusing gels 12 mm, 5 % acrylanlde, 5 H ulea, pH range 8.2 to 4 . 3 1 . The isoelectric focusing pH gradient was obtained by mixlng 0 . 4 4 ml of pH 3.5-10 Ampholines ILKB) with 0.2 rnl Of pH 5-7 Ampholines lLKBl/10 ml of gel SOlYtlOn. TO measure the pH gradient of the focuslng gels, gels were C u t lnto 0.5 Crn pieces mmedlately after focusing and Crushed with a glass rod i n 1.0 ml of d15tllled watei which had been boiled and subsequently coaled to l o o m temperature for pH determlnatlon. POT 2D qels isoelectric focusing gels were placed on 6% Laemml~ s l a b gels ( 5 I; U r e a . 20 m m DTTI with 1 mM DTT in the r u n n i n g buffer. After electrophoresis. gels were Stained with Coomassie brilliant blue to indlcate posltions of bovine serum alburnrn, chicken fibrinogen subunits. transferrln. and chicken PFn whlch were rnlred with the labeled medium samples prior to LSOelectrlc focusing. Hole~ular welght markers, whlch were placed at the top of the SDS gel. were stained a5 w e l l . The qels were destained, treated wlth Enhance

-70% and developed vlthin three days. (New England Nuclear]. drled. exposed to Kodak AR X-ray f l l m at

dlne-HCl buffer. pH 6.5 containing 1% N-1auroyISarCOSine. The cells Mgt2-free Hank's buffer, then Scraped ~n the presence of 6 M g u m -

were homogenlred on ice in a teflon-glass homogenizer far one mlnute o r rlth d m e c h d n l c a l homogenizer ITiSsUemlzer Tekmarl far 3 0 seconds. The homogenate was centrifuged at 10 O b 0 x q for 10 mi". at -10%. The resulting Supernatant was transfe;red to a c lean tube to which 0.04 volumes of 4 .5 I sodium acetate buffer. pH 5 .5 and 0.75 v o l u m e of absolute ethanol were added, vortexed and Stored at -2O'C overnlght. Precipitated RNA was collected by centrifugation 13320 x g for 60 mln. at 0%. dissolved in a minlmal v o l u m e I4 m l ] Of 6 1 guanldlne-HCI buffer. pH 7.0 containlng 1% sarcoslne. The

pH 7.0 and centrifuged at 37.900 rpm for 22 hours a t 20°C in sample was layered O v e r d 1 m l cushion of 5.7 H CsCl in 100 m M EDTA

st~:;;geh,".2~h,Prought to 0.3 M NaOAC and preclpltated vlth ethanol Beckman 50T1 rotor. The resulting RNA pellet was d>$sOlved I"

total ANA pellet was collected by rnicrocentrifuga- t l o n and resuspended I" 1 5 % formaldehyde. Three c u l t u r e plates typlcally produced 140 310 vg of total RNA.

Slot blot hybridization was performed essentially a5 described by Cheley and Anderson 1301. Brlefly. equal amount9 of total RNA vece brought to t h e same volume I 2 0 u l l I" 15% formaldehyde. An equal v O l U m e Of 20X SSC (saline-sodlum Citrate) was added and 1:l s e r ~ a l dilutions were made using 1:l 15% formaldehyde/2Ox SSC. samples were heated at 55°C for 15 min. and lmmedlately chilled on ice. Twenty-microliter aliquots of the samples were then applied to equilibrated nltmCellU1OSe filters mounted ~n a HybriSlot manifold

further bound to the filter by applying 1 M ammoniim acetate and filtration apparatus IBRLI. After vacuum fllfration samples were drylng at 80°C for 2 hours. Slot blots w e r e prehybridined hybridized and washed a s described by Slngh and Jones (311 excepi that the wash Solutions contamed 0.1% sodium p rophosphate. Prehy- brldlratlon was carried out for 3 hours at 4 2 4 and hybridization f o r 18-24 hour8 at 42'C. The fllters were washed 3 t z m e 5 at 6 5 O C for 15 minute5 per wash. A 2.992 Bp ECORl fragment Of the FC 4 0 chick Fn geno c Clone (6,141 was labelled u51ng a nick-translation k l t with m-"P-CTP (NEN). R-loop and restriction map analysis

et. al., unpublished datal lndlcdtes that thzs EcoRI fragment con: 16,141 and more recently. seq~encing of thls entire clone 1H.Hxrano

t a m s exons 1 and 2. Intron 1. pact Of intron 2, including 1.5 Kb of

2,992 Bp fragment has been used for F n RNA measurements in'chlck 5' untranslated region of the chick Fn 4ene. In addiizon chis

flbcoblasts 1141. Autoradiographs were exposed at -7O'C for 72 to

were determined from densltoinetrlc scans Of autoradiograms of the 120 hours before development. The r e l a t l v e changes l n Fn mRNA levels

slot blots using a Qulck Scan J r . (Helena Laboratories C0cp.l.

10374

B

1

Plasma Fibronectin Synthesis

A 0' 5' 15' 30' 60' 90' 180'

A a0 7;O 6;O 5;O 4;5 pH

. ,Fn

205Kd-

116Kd- "

116Kd- 97Kd-

66Kd-

"

45Kd-

B 8p 50 70 5i0 4;5+PH

"E-

MONOMER- 205 Kd-

97 Kd-

66 Kd-

45 Kd,

lo] 9

0- 24 48 72

c - H

A B

10 7 ,/- . "_"

"1 /+ Plasma Fibronectin Synthesis 10375

To C D C D C D C D C D F c 4 0 2h 6h 12h 18h 24h

A 1 1 3 3 1 0

9