Stimulation of fibrinogen synthesis in cultured rat hepatocytes by fibrinogen fragment E

288

Biochimica etBiophysicaActa 844 (1985) 288 295

Elsevier BBAl1426

Stimulation of fibrinogen synthesis in cultured rat hepatocytes by fibrinogen fragment E G. Dastgir Qureshi *, Philip S. Guzelian, R. Marcus Vennart and Herbert J. Evans Department of Medicine, Biochemistry, ClinicalPathologyand the Division of Clinical Toxicology and Environmental Medicine, Medical College of Virginia, Richmond, VA 23298 (U.S.A.)

(ReceivedOctober 30th, 1984)

Key words: Fibrinogensynthesis; Fragment E; Albumin; (Rat hepatocyte)

Because of the inherent difficulties of experimentation in intact animals, we used primary monolayer cultures of non-proliferating adult rat hepatocytes to study the effects of fibrinogen degradation products on fibrinogen biosynthesis. The freshly isolated hepatocytes obtained by collagenase perfusion of the liver in situ were cultured in a chemically defined sermn-free medium. The rate of fibrinogen synthesis in control cultures was 40-50 pmol/2.5 • 10 6 cells per 24 h. Additions of 20, 60 or 100/~g of homologous stage I fibrinogen degradation products had no effect on fibrinogen synthesis. In contrast, addition of the same amounts of homologous or heterologous (human) stage III fibrinogen degradation products resulted in a concentrationdependent increase in fibrinogen biosynthesis without affecting the rate of synthesis of albumin. When purified stage III fibrinogen degradation products D and E (human) were tested in 10, 30 or 5 0 / ~ g / 3 ml medium only fragment E showed a significant increase in fibrinogen biosynthesis (1.9-, 2.8- and 5.6-fold, respectively, over the control cultures). The presence of excess fibrinogen had no effect. These results suggest that fibrinogen fragment E may be a specific stimulator of fibrinogen biosynthesis which may play an important role in maintaining normal levels of plasma fibrinogen.

Introduction It is well accepted that fibrinogen is synthesized by the liver, but the mechanism(s) which regulate its production remain unknown. Many diverse agents such as glucocorticoids [1], catecholamines, growth hormone, turpentine abscess [2], endotoxin [3] and leukocyte factor(s) [4] stimulate fibrinogen biosynthesis in animals, but it remains uncertain whether these agents proximally regulate the process. It has been suggested that plasmin-derived * To whom correspondenceshould be addressed. Abbreviation: EGTA, ethylene glycolbis(fl-aminoethylether)N, N'-tetraacetic acid.

fibrinogen degradation products may increase the rate of synthesis of fibrinogen. Early studies to demonstrate the effects of fibrinogen degradation products on fibrinogen synthesis gave conflicting results [5-9]. These studies were done in living animals, wherein it is difficult to establish a direct effect of administered fibrinogen fragments on the liver. To permit more detailed analysis of the effects of fibrinogen degradation products on biosynthesis of fibrinogen by the liver, we have turned to the system of primary monolayer cultures of parenchymal cells prepared from adult rat liver [10]. The cultures can be maintained in a chemically defined, serum-free medium for many days. Under these conditions, the cells produce fibrino-

016%4889/85/$03.30 © 1985 ElsevierSciencePublishers B.V. (BiomedicalDivision)

289 gen and express a number of liver-specific functions, many at rates in the same range as the liver in vivo [13]. Material and Methods

Materials. Male Sprague-Dawley rats were purchased from Flow Laboratories, Dublin, VA. New Zealand white rabbits from Spring Valley Farm, Howardsville, VA; Waymouth's culture medium (752/1) from Grand Island Biologicals, Grand Island, NY; goat anti-rabbit IgG polyacrylamide beads (Immunobeads®) and Affigel-10 from Bio-Rad Laboratories, Richmond, CA; monospecific antiserum to albumin from ICN Pharmaceuticals, Cleveland, OH; agarose from Marine Colloids, Inc., Rockland, MA; and human fibrinogen (Grade L) purchased from Kabi Group, Inc., Stockholm, Sweden. All other biochemicals were purchased from Sigma Chemical Company, St. Louis, MO. Monolayer cultures of rat hepatocytes. Primary cultures of non-proliferating adult rat hepatocytes were prepared from male Sprague-Dawley rats (250-275 g) by a modification [11] of the previously described method [10,12-14]. The liver was perfused in situ with a calcium-free buffer containing EGTA followed by culture medium containing 0.03% collagenase. The softened liver was excised and suspended briefly in collagenase solution. Freshly isolated hepatocytes were prepared by repeated centrifugation. Hepatocytes (2.5 • 106) were inoculated into 60 mm collagen-coated culture dishes in a final volume of 3 ml of serum-free culture medium. 85% of the cells remained attached to the culture dishes after 8 h of incubation as determined by recovery of DNA. Extensive evidence of sustained viability of our cultures has been reviewed elsewhere [13,15]. Because collagenase retained by the hepatocyte can cause excessive degradation of secreted proteins during the first 24 h in cultures, all experiments were made during the second 24 h period. There was no quantitative variation of DNA per plate between the first and the second 24 h period. Test material was sterilized by passage through a 0.45/~m Millipore filter prior to addition to the culture plates. Purification of rat and human fibrinogen. Rat fibrinogen was purified by the methods of Atencio

et al. [16] and Finlayson and Mossesson [17]. Prior to purification, plasminogen was removed by affinity chromatography [18]. The purified fibrinogen preparations were 90% clottable when tested by the method of Birkens et al. [19] and appeared homogeneous on SDS-polyacrylamide gel electrophoresis [20,21]. The starting material for the purification of human fibrinogen was commercially available fibrinogen (Grade L, Kabi, Stockholm, Sweden). Measurement of rat fibrinogen in culture medium. The concentration of fibrinogen in the culture medium was measured by a rat fibrinogen radioimmunoassay using a modification of a previously described double antibody technique [22]. The bound 125I-labeled fibrinogen was separated from the unbound using a suspension of goat anti-rabbit IgG coated polyacrylamide beads (2.5 ~g). Reproducibility of the assay as tested by repeated assays of the same sample was 90-92%. A 50% inhibition of binding was achieved by 0.9 pmol fibrinogen. The smallest amount of rat fibrinogen measured accurately by the assay was 0.01 pmol. The hepatocyte culture medium containing fibrinogen was tested in three replicate dilutions. A standard curve comprising at least seven points was included in each test assay. The culture medium incubated in the absence of hepatocytes produced no cross-reactivity in the rat fibrinogen radioimmunoassay. Fibrinogen degradation products demonstrated a concentration-dependent immunoreactivity in the assay. Therefore, in experiments involving the addition of fibrinogen degradation products to the cultures, the net fibrinogen produced was calculated as the difference in immunoreactive fibrinogen in culture medium containing fibrinogen degradation products before and after 24 h exposure to monolayer cells. This method of calculation appeared valid because the dose-response curve of the fibrinogen degradation products in the radioimmunoassay paralleled that of rat fibrinogen. Also, the maximum immunoreactivity due to fibrinogen degradation products in any one test sample constituted less than 5% of the total immunoreactive fibrinogen present in the culture medium. Measurement of rat albumin in culture medium. Rat albumin was measured by the quantitative immunoelectrophoresis method of Laurell [23]. The

290 gels (2 × 3 in.) were composed of 0.9% Agarose and 1% Dextran and contained 0.1% monospecific antiserum to rat albumin. Electrophoresis was performed at 4°C for 3 h at 10 v / c m constant voltage. The amount of rat albumin in unknown solutions was measured by reading directly from an albumin standard curve using 0.025, 0.05 and 0.10 /~g purified rat albumin.

Preparation of early (fragments X and Y) and late (fragments D and E) degradation products of rat and human fibrinogen. Plasmin degradation products of fibrinogen were prepared by the method of Takagi and Doolittle [24]. Rat fibrinogen (2 mg) in 0.7 ml of 0.05 M Tris-buffered saline (pH 7.5) was incubated at 4°C with 0.3 ml plasmin (0.045 casein units/ml). Aliquots (50/~1) were removed at 0, 5, 10, 20, 40 and 60 min and added to 20 ~1 (1 mg/ml) of soybean trypsin inhibitor to stop digestion. Samples equivalent to 5/~g of the original fibrinogen were subjected to SDS-polyacrylamide gel electrophoresis [20]. Fragments X and Y were the major products after 20 rain digestion. Complete neutralization of plasmin at the end of digestion was documented by the lack of digestion of autologous fibrinogen in the presence of an aliquot of the samples of fibrinogen degradation products. Stage III fibrinogen fragments D and E were prepared by the same method except that the digestion was carried out at 22°C for 2 h. The presence of D and E fragments was confirmed by SDS-polyacrylamide gel electrophoresis [20].

Tris-buffered saline (pH 7.5) at 4°C, identification of the protein peaks was confirmed by SDS-polyacrylamide gel electrophoresis. Fragment D (pool 1) was homogenous as judged by SDS-polyacrylamide gel electrophoresis whereas the fragment E preparation (pool 6) contained a minor contaminant of fragment D (Fig. 1).

Slab gel electrophoresis and autoradiography of immune precipitated fibrinogen. Duplicate samples (500/~1) of culture medium obtained at 0 and 24 h were dialysed for 24 h at 4°C against 0.15 M barbital saline buffer (pH 7.4). The samples were then incubated with heat-treated (56°C for 30 min) rabbit anti-rat fibrinogen serum (1 : 130 dil.) for 60 min at 37°C. The immune-complexed fibrinogen was precipitated with Staphyloccus aureus Cowan I strain at 37°C for 60 min as described by Kessler [25]. The immunoprecipitates were washed eight times in 0.05 M Tris-buffered saline, and dissolved in 50 /~1 buffer (0.0625 M Tris (pH 6.8)/2.3% SDS/10% glycerol). Insoluble material was removed by centrifugation at 15 000 × g for 2 min at 22°C. The supernatant (40 t~l) was subjected to electrophoresis for 19 h at 3.8 mA

Purification of human fibrinogen fragments D and E. Purified human fibrinogen (200 mg) in 18 ml 0.05 M Tris-buffered saline (pH 7.5) was digested for 20 h at 4°C with 2 mg (2 ml) of plasmin (0.30 casein units/mg) in the absence of Ca z+ . The reaction was terminated by adding 20/~1 (50 m g / m l ) of soybean trypsin inhibitor. The digestion mixture was dialyzed overnight at 4°C against 0.01 M potassium phosphate buffer (pH 8.6) and applied to a 1.5 x 46 cm column of DEAE Bio-Gel A maintained at 8°C. The column was washed with 250 ml of the initial buffer, and the fibrinogen fragments were eluted with a linear gradient prepared from 300 ml of 0.01 M potassium phosphate (pH 8.6) and 300 ml of 0.1 M potassium phosphate (pH 7.0). The protein peaks monitored at 280 nm were lyophilized and reconstituted in 2 ml distilled water. After dialysis against 0.05 M

Fig. 1. SDS-polyacrylamidegel electrophoresis of human fibrinogen and purified human fibrinogenfragmentsD and E. Samples containingapprox. 10 btg protein were placed on 5% polyacrylamidegels and electrophoresed for 2 h at constant current of 5 mA/gel. The gels were stainedby the procedure of Fairbanks et al. [38]. Molecular weight standards: phosphorylase A (94000), bovine serum albumin (68000), ovalbumin (43000), pepsin (36000) and myoglobin(17200).

291

constant current on a 5-15% polyacrylamide gradient slab gel containing 1.2% N,N-diallyltartardiamide and 6.3 M urea [26]. After electrophoresis, half of the gel was stained for protein with 0.5% Coomassie blue stain and the other half was impregnated with 2,5-diphenyloxazole. After drying for 90 rain on a Bio-Rad gel dryer, the gel was exposed for 3 days to a pre-flashed photographic film [27] and processed on Kodak RPXOMAT processor. To measure incorporation of radioactivity in newly synthesized fibrinogen, a part of the immunoprecipitates was dissolved in Soluene®-350 and counted in a Beckman LS 8000 counter. Results

Synthesis of rat fibrinogen and albumin by rat hepatocytes Accumulation of fibrinogen in the culture medium was undetected during the first 2 h, but thereafter it increased steadily (Fig. 2). In a typical monolayer culture, the rate of fibrinogen production was 40-50 pmol (12-15 #g)/plate per 24 h. Accumulation of albumin in the same culture was

o7 l

approx. 12-16-times greater than that of fibrinogen. We established that de novo synthesis accounted for the accumulation of fibrinogen in medium by: first, addition of cycloheximide to the culture medium blocked the accumulation of fibrinogen; second, when 24-h cultures were exposed to medium containing [3H]leucine (5 /~Ci/plate), there was a progressive increase in radioactivity in culture medium immunoprecipitated using rabbit anti-rat fibrinogen antibody. The increase in radioactivity incorporated into immunoreactive fibrinogen paralleled the accumulation of fibrinogen in the culture medium as measured by radioimmunoassay (Fig. 3). Finally, the immunoprecipitate from 24-h cultures when analysed by polyacrylamide gel electrophoresis and visualized by Coomassie blue stain and by autoradiography, exhibited the same mobility as did the purified rat and human fibrinogen. This band was absent from the immunoprecipitate of medium incubated in the absence of cells (data not shown). Furthermore, we have failed to observe a loss of immunoreactive radiolabeled fibrinogen when added to the hepatocyte cultures over a 24 h period (data not shown). These experiments suggested that the fibrinogen in the culture medium was newly synthesized.

Effect of homologous stage 1 degradation products on synthesis of fibrinogen and albumin

0.6

Addition of 20, 60 or 100 #g of homologous

05 !

I4o

-0.4

x

~ 40.

-0.3

g~

° E=

- 2 SE.

.40

W : 'LU U:" ::3 U ClO

30.

- 50

I1: Q. O. UJ

-o2

zx

=~ Q.

O.I

0

v

i

r

I

i

i

i

i

2

4

6

8

IO

12

18

::'4

HEPATOCYTE CULTURE TIME

_z

(hrs)

Fig. 2. Production of fibrinogen and albumin by rat hepatocytes. 100-#1 aliquots of culture medium were removed at the indicated times to measure fibrinogen by rat fibrinogen radioimmunoassay and albumin by quantitative immunoelectrophoresis. Each point is the mean of at least three determinations. CHM, cycloheximide (A A).

I

%

20-

' 20

~x

x

I0-

i

i

i

i

i

i

4

8

12

16

~'0

:;'4

INCUBATION TIME

( HOURS )

Fig. 3. Rate of accumulation of' fibrinogen in the culture medium pulsed for 24 h with [3H]leucine (5 #Ci). (A A) Measured by rat fibrinogen radioimmunoassay; (O O) measured by immune precipitation of [3H]labeled fibrinogen using anti-rat fibrinogen serum. See Materials and Methods for experimental details.

292

stage I fibrinogen degradation products (fragments X and Y) to hepatocyte cultures had no effect on the rate of synthesis of either fibrinogen or albumin.

Effect of homologous and heterologous stage III fibrinogen degradation products on synthesis of fibrinogen and albumin In contrast to the effect of stage I fibrinogen degradation products, the addition of 20, 60 or 100 /ag of rat stage III fibrinogen degradation products to hepatocyte cultures resulted in statistically significant increases in fibrinogen production, of 2.2-, 4.5- and 5.7-fold, respectively (Table I). Similar stimulatory effects were observed using heterologous (human) stage III fibrinogen degradation products (Table I). To determine whether the fibrinogen degradation products specifically stimulated fibrinogen biTABLE 1 EFFECT OF F I B R I N O G E N D E G R A D A T I O N PRODUCTS (FDP) ON T H E SYNTHESIS OF F I B R I N O G E N A N D ALBUMIN BY C U L T U R E D RAT HEPATOCYTES To 24-h-old cultures was added 50/~1 of 0.05 M Tris-buffered saline (pH 7.4) containing 20, 60 o'r 100/~g stage III fibrinogen degradation products or buffer alone (control). The culture plates contained 2.5-106 hepatocytes in 3 ml media (see Materials and Methods). Culture media collected before and after 24 h incubation with hepatocyte was subjected to assays for fibrinogen (by radioimmunoassay) and albumin (by quantitative immunoelectrophoresis). The values represent mean-+S.E, of three assays. P values were calculated using Student's t-test. Fibrinogen (pmol/plate per 24 h)

Albumin (nmol/plate per 24 h)

Homologous (rat) Control S t a g e l l I F D P 20/xg 60/tg 100/~g

46.09+ 0.36 97.15+ 4.5 a 200.66 + 31.8 a 252.00-+23.7 b

0.50+0.03 0.63+0.03 ~ 0.54 + 0.002 ~ 0.43-+0.01 c

Heterologous (human) Control S t a g e I I I F D P 20/~g 60/~g 100/~g

57.7 178.5 195.1 247.8

0.92 5:0.03 0.90+0.03 ~ 1.00+0.09 c 0.85+0.10 c

a p < 0.01 vs. control. b p < 0.001 vs. control. ¢ Not significant.

5:7.0 _+ 5.8 ~ 5:22.2 a 5:8.6 b

osynthesis, and not proteins in general, the rate of synthesis of albumin in the presence or absence of homologous and heterotogous stage III fibrinogen degradation products was measured. Albumin synthesis was unaffected by the addition of either homologous or heterologous stage III fibrinogen degradation products (Table I). In simultaneously incubated control cultures, the addition of soybean trypsin inhibitor in amounts equal to that contained in the mixture of stage III fibrinogen degradation products had no effect on the rate of synthesis of fibrinogen or albumin.

Effect of purified heterologous (human) fibrinogen fragments D and E on fibrinogen and albumin production To determine which component(s) of stage III fibrinogen degradation products stimulated fibrinogen biosynthesis, heterologous (human) fibrinogen fragments D and E were purified (Fig. 1) and tested separately. The addition of 10, 30 or 50 /~g purified fibrinogen fragment E resulted in 1.9-, 2.8- and 5.6-fold increases in fibrinogen synthesis, respectively, as compared to the control TABLE II EFFECT OF P U R I F I E D H U M A N F I B R I N O G E N DEG R A D A T I O N PRODUCTS D A N D E ON THE SYNTHESIS OF F I B R I N O G E N A N D ALBUMIN BY C U L T U R E D RAT HEPATOCYTES To 24-h-old cultures was added 30/~l of 0.05 M Tris-buffered saline (pH 7.4) containing 10, 30 or 5 0 / l g purified fibrinogen fragment E or D or buffer alone (control). The cell culture conditions and measurement of fibrinogen and albumin were as described under Table I.

Control 1 10/xg fragment E 30/xg fragment E 50/~g fragment E

Fibrinogen (pmol/plate per 24 h)

Albumin (nmol/plate per 24 h)

40.9 5 : 4 . 0 82.05:19.0 ¢ 164.0 5:15.0 a 230.0 5:25.0 a

0.57 5:0.02 0.55+0.03 ¢ 0.62 :t: 0.07 c 0.65 5:0.12 ~

Control 2 10 #g fragment D 30 # g fragment D 50 # g fragment D a p < 0.02 vs. control 1. b p < 0.005 vs. control 2. c Not significant.

53.3 5 : 6 . 0 65.0_+ 0.40 ¢ 61.9 -+ 6.0 c 88.3 + 12.0 b

0.70+0.007 0.81+0.02 c 0.77 -+0.01 c 0.57 5:0.007 ~

293 cultures (Table II). In contrast, addition of 50/xg purified fibrinogen fragment D produced only a small increase in fibrinogen production. Lower concentrations of fragment D were totally without effect (Table II). Synthesis of albumin by the rat hepatocytes remained unaltered by either fragment D or E.

Effect of excess homologous fibrinogen on fibrinogen synthesis If fibrinogen biosynthesis is self-regulated by its proteolytic products, we might expect excess fibrinogen to down-regulate its biosynthesis by hepatocytes. To test this hypothesis we transferred monolayers to fresh medium containing 50 or 500 /~g homologous rat fibrinogen per plate. This amount of fibrinogen added at zero time corresponded to 10- or 100-fold excess, respectively, of the total fibrinogen produced by culture plate per 24 h period. This addition did not produce significant difference in the rate of synthesis of fibrinogen in fibrinogen-treated cultures as compared with the control cultures (data not shown). Discussion

The control of synthesis and release of macromolecular secretory products by the liver remains a poorly understood but clinically important area. The liver appears to respond to diverse conditions of trauma, infection, pregnancy and malignancy by rapidly increasing plasma fibrinogen, but the factors which regulate the increase or decrease in synthesis of fibrinogen are not known. There is growing belief that fibrinogen itself or its degradation products might act as feedback regulatory molecules [5,6,8,36,37]. Earlier studies to demonstrate the effect of fibrinogen degradation products on fibrinogen synthesis, however, have yielded conflicting results [5-9]. Bocci and Pacini [5] observed a 2-fold increase in plasma fibrinogen in rabbits after injection of homologous fibrinogen degradation products [5]. Kropatkin and Izak [6] demonstrated sustained hyperfibrinogenemia after infusion of fibrinogen degradation products generated in vivo from streptokinase-treated donor rabbits. Otis and Rappaport [7] failed to demonstrate an increase in plasma fibrinogen in rabbits after infusion of homologous fibrinogen degradation

products. Recently, Kessler and Bell [8,37] showed stimulation of fibrinogen biosynthesis after infusion of homologous stage III fibrinogen degradation products into rabbits. All of these studies were done in intact animals, which posed several disadvantages for attempting to identify regulatory factors. In addition to the biological variation among different animal species, it is often difficult to distinguish primary hepatic effect of the putative regulatory factor from secondary responses of the liver to an unknown soluble cellular mediator. To avoid the complexities of the intact animals, we have used a system of primary cultures of non-proliferating adult rat hepatocytes, which allow the study of plasma protein production as well as other biological functions of the liver in vitro [10-14,28,30-33]. With careful attention to the technical details, cultures from normal liver can be reproducibly established and maintained for many days in a chemically defined media without resorting to the temporary or sustained use of serum or serum proteins. Hence, the cultures are especially well-suited for studies involving blood-derived products. In the present studies, the total fibrinogen and albumin produced over a 24 h period was approx. 30-40 pmol and 0.6-0.7 nmol/plate, respectively. These values are comparable to those reported by others after normalization for the number of hepatocytes in their cultures [28,34,35]. The present studies demonstrated a concentration-dependent stimulatory effect of stage III fibrinogen degradation products on fibrinogen production by the cultured hepatocytes. Stage I degradation products were without effect. These observations are consistent with the observations made in intact animals [9]. The stimulatory effect appeared to be specific for fibrinogen synthesis, since there was no increase in albumin production when measured simultaneously in the same cultures. The stage III fibrinogen degradation products obtained from human fibrinogen were as effective as the homologous degradation products, suggesting that the stimulatory response was independent of the source of degradation products. The effect of stage III degradation products on fibrinogen synthesis seemed to be direct rather than through a mediator [36]. Neither hepatocytes nor the degradaljon products tested were exposed

294 to any of the cellular elements of the rat blood. Further, the plasmin-cleaved fibrinogen degradation products were bacteriologically sterile and free from endotoxin. The hepatocyte culture medium was chemically defined, and was free from serum or serum products, thus excluding a protein mediator. All of these observations suggest that the effect of degradation products was direct rather than through a mediator. Of the two major stage III fibrinogen degradation products, purified fragment E appeared to cause specific stimulation of fibrinogen production. The lack of a stimulatory effect of purified fragment D in equivalent concentrations suggest that the stimulatory effect of E cannot be attributed to the slight contamination of our preparation of fragment E with fragment D, which when tested alone had minimal effect on fibrinogen synthesis (Table II). The results of previous studies related to the effect of purified degradation products on fibrinogen synthesis have been variable. Franks et al. [9] observed 2.5-fold increase in fibrinogen fractional synthetic rate after intraperitoneal injection of 2 mg fibrinogen fragment D, (M r 100 000) in rats, and no increase in fibrinogen production after injection of 1 mg fragment E. Kessler et al. [37], who measured incorporation of [75Se]selenomethionine in circulating fibrinogen in rabbits observed 4.5- and 1.5-fold increases in fibrinogen synthesis after infusion of 4.5 mg purified fibrinogen fragment D and E, respectively. Both of these studies employed purified fibrinogen fragments, and, were done in intact animals. Recently, Ritchie et al. [36] failed to observe a stimulatory effect of purified homologous fibrinogen fragment D and E on fibrinogen synthesis by rat hepatocytes in in vitro cultures, unless the fragments were first exposed to leukocytes. The authors proposed that fibrinogen degradation products initiated a release from leukocytes of a hepatocyte-stimulating factor which caused a generalized increase in several plasma proteins, including fibrinogen, in the culture media. The reason why these investigators failed to observe a direct effect of fibrinogen degradation products on fibrinogen synthesis is not clear; however, it is possible that incubation of serum which was contained in their culture medium might interfere with the recognition of the fragments by the hepatocytes.

The mechanism by which stage III fibrinogen degradation products or purified fragment E stimulated the production of fibrinogen and hence accumulation in the medium is not known. An extensive cleavage of the fibrinogen molecule appeared to be important to produce the stimulatory effects, since neither the intact fibrinogen nor minimally degradation fibrinogen (stage I fibrinogen degradation products) were found to stimulate fibrinogen production. The addition of a purified preparation of NH2-terminal disulfide knot of human fibrinogen or plasmin-cleaved small molecular weight fragments of fibrinogen isolated during the purification of fragments D and E and designated as A, B and C had no effect on the rate of synthesis of fibrinogen (unpublished data). Similarly, infusion of fibrinopeptides A and B in rabbits has not been found to increase the fibrinogen synthesis [8]. The lack of effect of added fibrinogen on the rate of fibrinogen synthesis suggests that the cultured hepatocyte recognized specific molecular characteristic of the degradation products not available on the parent molecule. This could be related to specific sites in the fibrinogen molecule which are only exposed upon molecular cleavage by an enzyme. In summary, our results provide evidence for induction of fibrinogen synthesis in cultured rat hepatocytes by fibrinogen degradation products. These studies support the concept of an autogenous feedback regulation of fibrinogen synthesis by the liver. In this regulatory mechanism, the excess of fibrinogen does not seem to cause inhibition of its synthesis.

Acknowledgements We gratefully acknowledge the technical assistance of Mrs. Margaret B. Slate and Mrs. Suzanne Fernadez. We also thank Ms. Hattie Wyche and Ms. Carolyn Sayles for the secretarial work. This work was supported by National Institute of Health (U.S.A.) grant HL-24554 and HL-24281.

References 1 Grieninger, G., Hertzberg, K.M. and Pindyck, J. (1978) Proc. Natl. Acad. Sci. USA 75, 5506-5510 2 Seligsohn,U., Rapaport,S.I. and Zivelin,A. (1973) Thromb. Diath. Haemorrh.29, 76-86

295 3 Wycoff, H.D. (1970) Proc. Soc. Exp. Biol. Med. 133, 940-943 4 Rupp, R.G. and Fuller, G.M. (1979) Exp. Cell. Res. 118, 23-30 5 Bocci, V. and Pacini, A. (1973) Thromb. Diath. Haemorrh. 29, 63-75 6 Kropatkin, M.L and Izak, G. (1968) Thromb. Diath. Haemorrh. 19, 547-555 7 Otis, P.T. and Pappaport, S.I. (1973) Proc. Soc. Exp. Biol. Med. 144, 124-129 8 Kessler, C.M. and Bell, W.R. (1979) J. Lab. Clin. Med. 93, 768-782 9 Franks, J.J., Kirsch, R.E., Frith, L.C., Purves, L.R., Franks, W.T., Franks, J.A., Mason, P. and Saunders, S.J. (1981) J. Clin. Invest. 67, 575-580 10 Bissell, D.M., Hammaker, L.E. and Meyer, U.Z. (1973) J. Cell. Biol. 59, 722-734 11 Blackard, W.G., Guzelian, P.S. and Small, M.E. (1978) Endocrinology 103, 548-553 12 Berry, M.N. and Friend, D.S. (1969) J. Cell. Biol. 43, 506-520 13 Bissell, D.M. and Guzelian, P.S. (1980) Ann. N.Y. Acad. Sci. 349, 85-98 14 Bonney, R.S., Becker, J.E., Walker, P.R. and Potter, V.R. (1974) In vitro 9, 399-413 15 Bissell, D.M. and Guzelian, P.S. (1975) In Gene Expression and Carcinogenesis in Cultured Liver (Gershenson, L.E. and Thompson, E.B., eds.), pp. 119-136, Academic Press, New York 16 Atencio, A.C., Burdick, D.C. and Reeve, E.B. (1965) J. Lab. Clin. Med. 66, 137-145 17 Finlayson, T.S. and Mosesson, M.W. (1963) Biochemistry 2, 42-46 18 Deutsch, E.E. and Mertz, E.T. (1970) Science 170, 1095-1096 19 Birken, S., Wilner, G.D. and Canfield, R.E. (1975) Thromb. Res. 7, 599-610

20 Weber, K. and Osborn, M. (1969) J. Biol. Chem. 244, 4406-4412 21 Bouma, H., III and Fuller, G.M. (1975) J. Biol. Chem. 250, 4678-4683 22 Qureshi, G.D., Butler, V.P., Jr., Harfenist, E.J., Canfield, R.E. and Nossel, H.L. (1975) Thromb. Res. 6, 357-374 23 Laurell, B.C. (1976) Anal. Biochem. 15, 45-52 24 Takagi, T. and Doolittle, R.F. (1975) Biochemistry 14, 940-946 25 Kessler, S.W. (1976) J. Immunol. 117, 1482-1490 26 Ames, G.F. (1974) J. Biol. Chem. 249, 634-644 27 Bonner, W.M. and Laskey, R.A. (1974) Eur. J. Biochem. 46, 83-88 28 Crane, L.J. and Miller, D.L. (1977) J. Cell. Biol. 72, 11-25 29 Lin, R.C. and Snodgrass, P.J. (1975) Biochem. Biophys. Res. Commun. 64, 725-734 30 Michalopoulos, G. and Pitot, H.C. (1975) Exp. Cell Res. 94, 70-78 31 Guzelian, P.S., Qureshi, G. Dastgir and Diegelmann, R.F. (1981) Coll. Res. 1, 83-93 32 Guzelian, P.S. and Diegeimann, R.F. (1979) Life Sci. 24, 513-518 33 Gurr, J.A. and Potter, V.R. (1980) Ann. N.Y. Acad. Sci. 349, 57-66 34 Jeejeeghoy, K.N., Bruce-Robertson, A., Sodtke, V. and Foley, M. (1970) Biochem. J. 119, 243-249 35 Crane, L.J. and Miller, D.L. (1974) Biochem. Biophys. Res. Commun. 60, 1269-1277 36 Ritchie, D.G., Levy, B.A., Adams, M.A. and Fuller, G.M. (1982) Proc. Natl. Acad. Sci. USA 79, 1530-1534 37 Kessler, C.M., Bell, W.R. and Townsend, R.R. (1978) Blood 52, Suppl. 1, 186 38 Fairbanks, G., Steck, T.L. and Wallach, D.F. (1971) Biochemistry 10, 2606-2617