British /ournu/ o/Harmutology, 1991. 78, 229-235

AWNIS

OOO71O4891O0129N

Expression of coagulation factor V gene by normal adult human hepatocytes in primary culture M A R L E N EMAZZORANA.GEORGESBAFFET,B E R N A R DK N E I P , BERNARDLAUNOIS* A N D CHRI S T I A N E G U G U E N-G u I L L O U Z O ZNSERM U 4 9 , Unite de Recherches Hepatologiques. H6pital Pon tchaillou, Rennes, and * Clinique Chirurgicale B. Hipital de Pontchaillou, Rennes, France Received 16 November 1 9 9 0 ; accepted for publication 28 January 1991

Summary. Normal human adult hepatocytes were examined for their ability to synthesize and secrete factor V using primary culture. The culture medium contained both factor V and factor Va as determined by bioassay and activation experiments. Immunoprecipitation of newly synthesized labelled factor V showed the presence of both native factor V (m.w. 3 30 000) and two fragments of respective molecular weight 300 000 and 265 000. Northern blot analysis revealed the presence of a single 7 kb factor V mRNA in cultured human hepatocytes as in liver biopsies, together with fibrinogen and albumin transcripts. Relative levels of

factor V, fibrinogen p and albumin mRNAs differed when the cells cultured. suggesting that expression of the three corresponding genes might in part be independently regulated. Furthermore. addition of glucocorticoids enhanced factor V and fibrinogen fi mRNA levels 1.6- and 5-fold respectively. but did not significantly increase that of albumin. These results provide evidence that human hepatocytes actively participate in the synthesis of plasma factor V and constitute a valuable model to study the common and specific regulations involved in the control of the expression of this gene in human liver.

Coagulation factor V is a high molecular weight glycoprotein (m.w. 3 3 0 0 0 0 ) present in the plasma and the a granules of the platelets (Nesheim et a/. 1981: Mann et a / , 1986). Factor V itself is a poorly active co-factor. However, during coagulation, thrombin converts factor V to its very active derivative (factor Va) by limited proteolysis (Susuki et a/. 1982: Nesheim et d. 1984). Factor Va is in fact a noncovalent complex between two peptides resulting from this proteolysis stabilized by a single C a t + ion. Activated factor V is involved in the prothrombinase complex (Jackson & Nemerson. 1980: Mann et a / . 1981 , 1987) as an essential non-enzymatic co-factor in the optimum conversion of prothrombin to thrombin by factor Xa on the surface of platelets or other cell membranes and more generally on phospholipid surfaces (Kane et a/. 1980: Rodgers & Shuman. 1983: Tracy et a/. 1983: Van de Waart et a/. 198 3 ) . Many clinical investigations have suggested that the liver plays a key role in the synthesis of factor V. low levels of which have been found in patients with chronic or acute liver diseases (Lechner et al, 1977: Joist, 1987). In these situations factor V is an accurate indicator of the extent of liver damage.

A decrease in factor V activity is also observed during experimentally induced hepatic necrosis (Cornillon et a / . 1985). Although a n increasing activity was found in organ perfusate (Workman & Lundblad, 1977: Olson et a/. 1966), the hepatic cell type responsible for the synthesis of circulating plasma factor V was not clearly demonstrated. The participation of hepatocytes in this synthesis has been suspected from studies on HepGL cells (Wilson et a/. 1984) and recently we have shown the participation of normal rat hepatocytes in the synthesis of plasma factor V and its active secretion into the culture medium (Mazzorana et a/, 1989). However, up to now, the regulation of the factor V gene expression remains unknown. Furthermore, it is difficult to extend the results obtained in animals to humans. For these reasons, primary cultures of normal human hepatocytes may represent a unique tool for analysing such mechanisms in man. Indeed, they mimic well the in vivo situation and have the great advantage of being more stable than their rodent counterparts, allowing preservation of the differentiated state for a longer time (Ratanasavanh e l a/. 1986: Guillouzo. 1986). In this report we have tested the reliability of our culture model using factor V bioassay. immunoprecipitation experiments and Northern blot analysis of the amounts of specific transcripts to estimate the ability of human hepatocytes to

Correspondence: Dr Marlene Mamorana. INSERM U49. Unite de Recherches Hepatologiques, Hdpital Pontchaillou. 3 5 0 3 3 Rennes Cedex. France.

229

2 30

Marlkne Mazzorana et a1

synthesize factor V. Factor V mRNA levels in cultured cells were compared to those of fibrinogen p and albumin in order to correlate them with the cell differentiation state and with the synthesis of another coagulation protein (fibrinogen). Moreover, since it has been shown that glucocorticoids enhanced fibrinogen mRNA levels in suspended and cultured rat hepatocytes (Princen et al, 1984) we have analysed the effect of hydrocortisone hemisuccinate on the human factor V. fibrinogen /Iand albumin mRNA contents in cultured human hepatocytes. MATERIALS AND METHODS Probes. Hybridization probes for mRNA analysis were as follows: human factor V cDNA probe (The Genetics Institute, Cambridge, Mass., U.S.A.) (Jenny et ul, 1987), cDNA probe encoding for the BB chain of human fibrinogen (Uzan et al. 1984) and human albumin cDNA (De Souza et al, 1984). Reagents. Tissue culture media and fetal calf serum were obtained from Gibco. Grand Island, N.Y.. U.S.A. Purified human factor V, rabbit anti human factor V antiserum, human plasma artificially depleted of factor V (hemoreactif Pc), human thrombin, human activated protein C, and freeze-dried calcium thromboplastin (neoplastin-Ca++) were purchased from Diagnostica Stago. Asnihes. France. Pepstatin, aprotinin and collagenase were from Boehringer-Mannheim. Germany. Russel viper venom factor V activating enzyme (RVV-VAE) and insulin were from Sigma Chemical Company, St Louis,U.S.A. Act-Ultrogel AcA22 was from IBFLKB. Rockville. U.S.A. Lyophilized hydrocortisone hemisuccinate was from Roussel-Uclaf, Romainville. France. "Smethionine was purchased from Amersham. U.K. All other chemicals were reagent grade or better. Cell isolation and culture. Normal human livers were obtained from three young male and one female adult kidney transplantation donors (respectively 11, 16, 1 7 and 41 years old), the cause of the death being traffic accidents. All experimental procedures were done in compliance with French laws and regulations and were approved by the French National Ethics Committee. Medical treatments and assistance during coma were as usual and without known effects on coagulation process. Hepatocytes were isolated by perfusion of the left hepatic lobe with a collagenase solution by using selective cannulation of the portal vein according to the method developed in our laboratory (Guguen-Guillouzo et al, 1982). Cell viability was estimated by trypan blue exclusion and found to range between 70% and 85%.The cells were seeded at a density of either 3 x lohviable hepatocytes per 2 5 cm2 flask in 3 ml of medium for coagulation studies or 10' cells per 8 0 cmz flask in 10 ml of medium for mRNA isolation. Medium was composed of 2 5% medium 199 and 75%minimum essential medium containing 5 pg/ml bovine insulin, 0.1% bovine serum albumin, penicillin (7.5 U/ml), streptomycin (50 pg/ ml), kanamycin (50 pg/ml) and supplemented with 10%fetal calf serum (FCS) during the first 4 h of culture. Then, the medium was added with 7 x lo5M hydrocortisone hemisuccinate and FCS was omitted. Medium was renewed daily thereafter.

For the studies of the effects of hydrocortisone hemisuccinate on factor V, fibrinogen J and albumin mRNA contents, cells were seeded in the medium previously described but without the hormone. As in control cultures, FCS was suppressed after cell spreading. Factor V bioassay. Serum-free culture medium (500 pl) was collected every day after 2 4 h incubation with the cells, and centrifuged (50 g. 5 min, 4 O C ) . The supernatant was stored at -80°C. Factor V clotting activity was measured after activation of the samples with a specific factor V activator from Russell viper venom (RVV-VAE) (Hanahan et al. 1972; Dahlback, 1988). 50 pl of culture medium were mixed with 50 p1 of the enzyme solution at a concentration of 3 ' 3 IU/ml. After a 2 0 min incubation at 37OC, factor V coagulant activity was monitored by a one-stage assay (Colman & Weinberg. 1976) using 100 pI of human plasma artificially depleted of factor V. One unit of factor V is defined as the activity contained in 1 ml of normal human plasma. Clotting time was measured on a fibrometer coagulation timer (KC4 coagulometer, Amelung. Germany). A standard curve was obtained with dilutions of a pool of fresh human plasmas. Control assays were performed by incubating the samples in the same conditions with 4%bovine serum albumin solution prepared in Michaelis buffer. The activation quotient (AQ) was the ratio between the clotting activities in activated ( A ) and control (C) assays: AQ =

mU factor V (A) mU factor V (C)

Protein assay. After a brief wash in 0 . 1 M phosphatebuffered saline (PBS) pH 7.4. cells were suspended in the same buffer by scraping with a rubber spatula and centrifuged at 50 g for 5 min at 4°C. Each cell pellet was then disrupted by sonication in 100 pl of PBS and its protein content was estimated (Protein assay, Bio-Rad Laboratories, Munich, Germany) using a standard curve constructed with dilutions of bovine serum albumin. Immunoprecipitation offactor V in culture media. 4 d after seeding, the cells were incubated for 24 h in 3 ml of the serum-free medium previously described, deprived of cold methionine and containing 2.59 mBq 35S-methionine/mlof culture medium. Then, media were harvested and centrifuged (50 g, 5 min. 4OC). Pepstatin was added at a final concentration of 0 . 7 pg/ml to inhibit acid proteases. 6 ml of medium were treated either with human thrombin (1.3 NIH units/ml) or with activated protein C (APC) ( 7 0 ng/ ml). Half of each was treated with rabbit anti factor V antiserum whereas the other half was added with rabbit control serum. In the samples treated with APC. phospholipids and calcium necessary for the enzyme activity were coming from 100 pl of freeze-dried calcium thromboplastin from fresh cerebral tissue (Neoplastin). After a 40 min incubation at 3 7 O C . the samples were submitted to immunoprecipitation. A competitive inhibition of the reaction between the antibody and the newly synthesized 35S-labelled human factor V was performed by adding an excess of exogenous unlabelled human factor V prior to immunoprecipitation. 2 0 pg of purified human factor V were added to 6 ml of

Factor V in Cultured Human Hepatocytes

231

Table 1. Coagulant activity of factor V secreted in the culture medium with and without RVVVAE treatment (mU/nil/24 h )

Donor 2

Donor 1 Age of culture (d)

+End

-Enz

2

17.6

3

20.6 19.1

4 5

Donor 4

+En2

-En2

+Enz

-En2

+En2

-End

9.4

34.0

9.5 7.2 6.2

21.0 27.0 23.0 2.8 nd nd

24.4 10.2 8.0 7.2 2.0 nd nd

8.8 6.4 8.4

5.2 3.5

22.0 12.4

9.9 nd nd nd

nd nd nd nd

27.0 22.0 26.3 18.7 15.3 ND ND

7

15.2 15.8 5.0

4.7

X

4.4

3.5

6

Donor 3

h.6

3.6

14.6

15.6 14.0 ND ND

nd: below threshold of sensitivity. ND: not determined. 3 x 10"cells were seeded in 2 5 cm' flask in 3 ml of serum-freemedium. Culture medium ( 2 4 h of incubation with the cells) was daily removed and frozen at - 80'C. It was then assayed for factor V activity without or with RVV-VAE treatment (final enzymatic concentration 1.3 1U/ ml):donor 1 ( 1 7 years old);donor 2 ( 1 6 years old): donor 3 ( 1 1 years old);donor 4 (41 years old).

enzymatically untreated medium. After a 1 0 min incubation at room temperature, immunoprecipitation of factor V was performed using activated Ultrogel. Rabbit anti human factor V antiserum and control rabbit serum were diluted respectively 1/10e and 1/50e in the suspension of the activated beads. After overnight incubation at 4°C with gentle shaking, the Ultrogel was centrifuged and washed three times with 0.1 M NaH2P04/Na2HP04 buffer pH 7.4. The remaining activated sites were then saturated with 1 M Tris/glycine buffer pH 7.4 for 8 h at 4°C with shaking. After equilibration of the gel in 0.1 M NaHLPO4/Na2HPO4 buffer containing 0.15 M NaCI. 3 ml of culture medium were treated (overnight at 4°C with gentle shaking) with 3 0 0 pI of Ultrogel either bound to rabbit anti human factor V antiserum or control rabbit serum. After mild centrifugation (50 9. 1 min, room temperature), the beads were successively washed with 0.1 M NaH,P04/ Na2HP04pH 7.4 buffer, the same buffer containing 1 M KCI, 10 TIU (trypsin inhibitor unit) per ml aprotinin and the same buffer containing 0 . 1 M NaCI. The bound antigen was eluted from the bead pellets with 2 volumes of a solution of 0.1% SDS. 10%CHICOOH for 1 h at 4°C with gentle shaking and the Ultrogel was discarded after centrifugation. Eluted supernatants were then frozen at -80°C and freeze-dried overnight. Elertrophoresis and autoradiography. After complete dissolution of the lyophilized ['%I samples for 5 min at 100°C in 50 pl of electrophoresis sample buffer (3%, SDS. 5% 2-mercaptoethanol, 30% glycerol, 0.03% bromophenol blue), 2 pl of samples were counted on a scintillation counter. Equivalent counts were added to each well of a discontinuous sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDSPAGE) 4-1074 resolving gradient gel with a 3% acrylamide stacking gel. High molecular weight (HMW) calibration kit proteins (Bio-Rad Laboratories, Germany) and 1 p g of purified human factor V were also deposited as controls. Electrophoresis was carried out for 18 h at 30 V. Following electrophoresis the gel was cut: HMW standard proteins and

human purified factor V were treated for silver staining and the remaining gel was fixed in 10%acetic acid/2 5% ethanol for 1 h and then treated with a fluorographic reagent (Amplify, Amersham, U.K.) for 1 5 min. The fixed gel was dried and exposed with Kodak XAR-5 X-ray film at - 80°C. RNA isolation and Northern blot analpis. At different times of culture, total RNAs were isolated from cultured cells using the guanidine thiocyanate method (Chirgwin et al, 1979) modified by Raymondjean et al (1983). 5 or 1 0 p g of total RNA were separated on a 1.5% agarose-formaldehyde gel as previously described by Maniatis et a/ (1989) and subsequently transferred to a nylon membrane (Hybond N. Amersham. U.K.) as described (Thomas, 1980). l'P nicktranslated human albumin and fibrinogen cDNAs (Rigby et al. 1982) as well as human factor V cDNA insert (isolated after digestion with the appropriate restriction enzyme, Sal1. and labelled by multipriming) (Feinberg & Vogelstein, 198 3 ) were then used to probe the Northern blots (Wahl et ul. 1979). RESULTS

Factor V coagulant activity Factor V coagulant activity was found in the culture medium of human hepatocytes throughout the culture period. Measurement of factor V clotting activity was performed with and without RVV-VAE treatment (Table I). The enzymatic treatment of the samples allowed us to quantify optimum factor V activity. Results showed that factor V was actively secreted in the culture medium from the onset of culture to day 6 approximately. Later on, a quick and significant decrease of factor V activity was observed. Optimum factor V coagulant activity varied from 10 to 3 5 mU/ml of culture medium/24 h depending on the donors. The comparison between the values obtained in control samples and RVV-VAE treated ones revealed in all cases the presence of some still activable molecules in the medium

232

Marlbne Mazzorana et a1 --t

D1

U D2

kD

bi b2>

b3-

b4 ,200

- 116 - 92

c1-

c2,

0

2

4

6

1

8

AGE OF CULTURE (DAYS)

Fig 1. Activation quotient (AQ)of factor V present in culture medium of normal adult human hepatocytes from the four different donors, D1,D2. D3.W.AQ=mU factor V (A)/mU factor V(C). (A) Factor V activity after treatment with RVV-VAE (C) factor V activity in untreated culture medium used as control. Factor V coagulant activities are expressed in mU/ml of culture medium/24 h.

throughout the culture period although maximum activation did not exceed 3.5-fold (Fig l),implying that numerous factor V molecules were already activated presumably as a consequence of some proteolytic activation in the culture medium. The magnitude of RVV-VAE activation vaned with time and from one donor to the other. Maximum activation quotient was obtained between day 4 and day 6 of culture indicating at this time the presence of numerous activable factor V molecules in culture medium. lmmunoprecipitation oj newly synthesizedlactor V Since the amount of activable factor V molecules secreted in the culture medium was optimum after 5 d of culture (Fig 1). therefore the immunoprecipitation experiments were carried out at this time. As shown in Fig 2. A* and A l . two proteins (bl and b2) which migrated in a very close manner were seen at the top of the gel. This doublet of proteins was particularly visible in Fig 2. A* in which we deposited only 50%ofthe cpm laid down in each other well. Only the lower band of the doublet (b2. m.w. 330000) was specific of human factor V since it was absent in the samples treated with a control rabbit serum (Fig 2, A2). Concurrently, this protein comigrated with purified human factor V detected by silver staining (results not shown). Two other products (b3 and b4) were specifically immunoprecipitated by the antiserum with apparent molecular weights of 300 000 and 265 000 respectively. To identify further the 330000 M, band (b2) as native factor V and the 300 000 M, (b3) and 265 000 M, (b4) bands as factor V fragments, the culture medium was treated

2

y*Ly

A'

A

1

2

yyl

B

C

Fig 2. Immunoprecipitation of factor V from 35S-methioninelabelled culture media of normal adult human hepatocytes. After 4 d of culture, cells (seeded at the onset of culture at 3 x 10' cells/25 cm' flask) were incubated for 24 h in 3 ml of methionine-free medium supplemented with 2.59 mBq 35S-methionine/mL Factor V was immunoprecipitated from 3 ml samples and treated as described in Methods. A*: Factor V antiserum immunoprecipitate of enrymatically untreated medium (50% of the cpm deposited in each other well). A: enzymatically untreated medium: B medium treated with thrombin (1.3 NIH units/ml): C medium incubated with exogenous unlabelled puriRed human factor V (3.3 pg/ml): 1: factor V antiserum immunoprecipitate: 2: control serum immunoprecipitate. On the right of the Rgure the molecular weights of the calibration proteins are indicated as references: Myosin (200000): Galactosidase ( 1 16 000);Phosphorylase B (92 000). Apparent molecular weights of immunoprecipitated products: b l ( 3 5 0 000): b2 (330000):b3(300000);b4(265000):cl( 1 15000):c2 ( 1 10000).

prior to immunoprecipitation with two specific enzymes: thrombin and activated protein C (APC). Indeed. both enzymes have been reported to cleave native factor V molecule (Esmon et a/. 1980; Susuki et a/. 1982, 1983: Nesheim et a/. 1984) although it is generally admitted that factor V is less sensitive to APC proteolysis than factor Va. The thrombin treatment (Fig 2, B) as well as the addition of exogenous unlabelled human factor V to the immunoprecipitation mixture (Fig 2. C) caused the three bands to disappear. Similar results were obtained after APC treatment (results not shown). These data demonstrated that the 3 30 000 M,band corresponded to native factor V and that the 300000 and 265 000 M, bands were factor V fragments presumably resulting from aspecific proteolysis. Two cleavage products (cl and c2) of respective molecular weight 115 000 and 110 000 were clearly observed in the samples treated by

Factor V in Cultured Human Hepatocytes

- 7 K b @

-

2.2 1.9

@

BTo2 3 4 5 Pig 3. Northern blot analysis of factor V. librinogen and albumin rnRNAs in normal adult human hepatocytesduring 5 d ofculture. 10 p g of total RNA were deposited. Northern blot was hybridized with human factor V cDNA probe (A), human fibrinogen b cDNA probe (B). human albumin cDNA probe (C). I% liver biopsy: To: freshly isolated hepatocytes. Arrows indicate the mRNA shes.

thrombin (Fig 2, B); however, they were immunoprecipitated both by the antiserum (Fig 2 , B 1 ) and the control serum (Fig 2, B2). The expected final cleavage products resulting from thrombin activation of native factor V (heavy and light chains of factor Va of respective molecular weight 9 4 0 0 0 and 7 4 0 0 0 ) (Tracy, 1988) were not observable in our experimental conditions probably due to the small quantities of native factor V present in our samples and represented by the thin band b2.

R N A blot analysis of factor V m R N A in cultured human hepatocytes: comparison with fibrinogen b and albumin mRNAs As expected (Jenny et al, 1987). hybridization with fulllength cDNA factor V probe revealed an abundant 7 kb factor V mRNA. For each donor factor V mRNA appeared as a single band present in liver biopsies, freshly isolated cells and cultured cells (Fig 3. A). The level of factor V mRNA was stable in culture decreasing slightly only after 5 d. In all cases studied factor V mRNA level was lower in liver biopsies and freshly isolated hepatocytes than in cultured cells. As described in the literature (Huber et a/. 1987). two mRNAs of respective size 2.2 kb and 1.9 kb were detected by the cDNA fibrinogen p probe on the autoradiogram of the Northern blot (Fig 3. B). Geqeration of two mRNAs of different sizes could be due to the presence of two polyadenylation sites (Chung ef a / , 1983) on fibrinogen p gene. separated by 260 bp, although alternative splicing events could not be completely ruled out. High levels of fibrinogen p mRNAs were found in the liver biopsies and in freshly isolated cells and gradually decreased up to day 3 of culture. Later on, they remained stable all along the culture period. Large amounts of albumin mRNA (2 kb) were found in both the liver biopsies and freshly isolated cells (Fig 3 . C). These important levels of albumin mRNA were well maintained during culture. The presence of factor V. fibrinogen fl and albumin mRNAs indicated that in our culture conditions human hepatocytes remained capable of expressing specific liver functions for several days. Eflerts olglucororticoids on the levels ojjactor V,fibrinogen /Iand albumin mRNAs Hydrocortisone hemisuccinate exerted very different effects on the levels of factor V, fibrinogen p and albumin mRNAs in the cells from a same donor. A 4 8 h exposure of cultured M hormone induced a n human hepatocytes to 7 x increase of factor V mRNA (Fig 4, A), together with

Alb

To

H'H-

233

FV

Fg

Fig 4. ElTects of hydrocortisone hemisuccinate on the levels of and albumin mRNAs in donor 4.RNA was factor V. fibrinogen /l isolated after 48 h of culture for Northern blot analysis. Hybridization was performed with factor V cDNA probe (A), fibrinogen p cDNA probe (B)and albumin cDNA probe (C). (D) Densitometric analysis of the Northern blots. 1 0 p g of total RNA were deposited in A and B. and 5 p g in C. H-:cells were seeded in a FCS-free medium without hydrocortisone hemisuccinate: H + : cells were seeded in a FCS-free medium to which hydrocortisone hemisuccinate ( 7 x 1 W 5 M) was added. To: freshly isolated hepatocytes. In (D) shaded areas: H + : white areas: H-.

234

Marlene Mazzorana et al

fibrinogen p mRNAs (Fig 4. B). but failed to produce such an increase of albumin mRNA (Fig 4. C). The densitometric analysis of the Northern blots prepared from two different donors (donors 2 and 4) and tested for the three functions studied revealed an important increase of fibrinogen j? mRNAs (means of 4.2- and 4.8-fold increase for donor 2 and donor 4 respectively)and a moderate one for factor V mRNA (means of 1.6- and 1.9-fold increase for donor 2 and donor 4 respectively). Fig 4(D) shows an example of the values obtained for donor 4. DISCUSSION This study shows that the synthesis of activable factor V is maintained for approximately 1 week in human hepatocyte primary cultures. These data are different from those obtained with rat cells in which the synthesis of factor V was undetectable after 4 d of culture and the protein unactivable. These results confirmed the well-known greater survival and stability of human normal adult hepatocytes in primary culture (Guilouzo. 1986) resulting in the maintenance of numerous specific functions such as cytochrome P450 (Ratanasavanh et al. 1986: Guillouzo et al, 1985) and in less numerous dying cells implying lower amounts of released proteases able to activate factor V. The synthesis of factor V in cultured human hepatocytes could reach a maximum activity of 10-35 mU/ml of culture medium/24 h. After 5 d of culture, progressive disappearance of factor V activity was probably accelerated by increasing amounts of proteases degrading factor V and released during cell alteration occurring before the death of an increasing number of cells. Variations of factor V clotting activities were observed from one donor to the other, while no significant differences were detected at the mRNA level. This could reflect, at least in part, the loss of the post-translational regulation processes usually controlling factor V plasma level in vivo. The immunoprecipitation revealed the presence of a 3 30 000 M, protein clearly identified as native factor V after treatment with thrombin, activated protein C and exogenously added unlabelled human factor V. Two other labelled products were immunoprecipitated (m.w. 300 000 and 265 000). We may assume that they correspond to factor V fragments resulting from factor V degradation since the treatments mentioned above caused their disappearance. Northern blot analysis provided characterization of factor V mRNA in isolated normal human hepatocytes. In the cultured cells as well as in liver biopsies we found only a single 7 kb factor V mRNA but failed to detect the thin upper band evidenced by Jenny et a1 (1987). Our results confirm those obtained by Kane et a1 (1987) on the transformed HepGLcells. The level of factor V mRNA was stable during the culture time (as that of albumin mRNA)with a weakdecrease observed only after 5 d of culture, correlating well the results obtained for factor V coagulant activity. In 63% of the cases studied in our laboratory, low levels of factor V mRNA appeared in biopsies and freshly isolated cells comparatively to those obtained in cultured cells. These data might suggest either a change in the factor V gene expression in our culture conditions or a very rapid turnover rate of the factor V mRNA

which would be responsible for the decrease of mRNA levels during the organ sampling. The reproducible high levels of fibrinogen mRNAs observed in liver biopsies, freshly isolated hepatocytes. and also during the two first days of culture might result from the fact that this protein (factor I of the coagulation cascade) is an acute phase protein which can be induced by a wide variety of stresses and injuries. In our culture conditions, the presence of hydrocortisone hemisuccinate in the culture medium resulted in an important increase of fibrinogen /3 mRNAs, as expected from a previous report (Princen et al. 1984),a moderate increase of factor V mRNA and no significant influence on albumin mRNA level probably due to the stability of the albumin mRNA and its large pool size in the hepatocytes (Nawa et al, 1986). Our data showed that high concentrations of glucocorticoids exert quantitatively different influence on fibrinogen and factor V gene expression suggesting that these genes are submitted to partially independent molecular regulations probably reflecting the participation of fibrinogen but not factor V, to the acute phase reaction. ACKNOWLEDGMENTS We would like to thank Dr A. Guillouzo and Dr D. Webb for their critical reading of the manuscript and Professer J . Y. Le Gall for his help in measurement of clotting factor V activities. We would like also to thank M. Boisnard-Rissel for her assistance in preparing photographic figures and A. Vannier for typing the manuscript. This research was supported by the Institut National de la Santk et de la Recherche Medicale. REFERENCES Chirgwin, J.M.. Przybyla. A.E.. MacDonald, R.J. & Rutter. W.J. ( 1 979) Isolation of biologically active ribonucleic acid from sources enriched in ribonuclease. Biochemistrg. 18, 5294-5299. Chung. D.W.. Que. B.G.. Riscon. M.W.. Mace, M. & Davie. E.W. ( 1 983) Characterization of complementary deoxyribonucleic acid and genomic deoxyribonucleic acid for the /lchain of human fibrinogen. Riochemistry. 22, 3244-3250. Colman. R.W. & Weinberg. R.M. (1976) Factor V. In: Methods i n Enzymology (ed. by L. Lorand). Academic Press, New York. Cornillon. B.. Paul, 1.. Belleville. J.. Aurousseau, A.M.,Clendinnen. A. & Eloy. R. ( 1 985) Experimental DMNA-induced hepatic necrosis: early course of hemostatic disorders in the rat. Comparative Biochemistry and Physiology. 80C, 277-284. Dahlback. B. (1988) A new model for coagulation factor V suggcsting a unique mechanism of activation. Scandirmvian /ourrial o/ Clinical and Laboratory Medicine. 48, S 19 1 , 47-6 1. De Souza, S.L..Frain, M.. Mornet. E.. Sala-Trepat. J.M. & Lucotte. G. (1984) Polymorphisms of human albumin gene after DNA restriction by Hae I11 endonuclease. Human Genetics. 67, 48-5 1. Esmon. C.. Comp, P. & Walker. F. (1980) Functions for protein C. In: Vitamin K Metabolism and Vitamin K-Dependent Protriris (ed. by J. Suttie). University Park Press, Baltimore. Feinberg. A.P. & Vogelstein. B. (1983) A technique for radiolabeling DNA restriction endonuclease fragments to high specific activity. Annals of Biochemistry. 132, 6-1 3. Guguen-Guillouzo. C.. Campion. 1.-P..Brissot. P.. Glaise. I).. Launois. B.. Bourel. M. & Guillouzo. A. (1982) High yield preparation of

Factor V in Cultured Human Uepatocytes isolated human adult hepatocytes by enzymatic perfusion of the liver. Cell Biology International Reports. 6. 62 5-628. Guillouzo. A. ( 1 986) Use of isolated and cultured hepatocytes for xenobiotic metabolism and cytotoxicity studies. In: Isolated arid Cultured Hepatocytes (ed. by A. Guillouao and C. Guguen-Guillouzo). INSERM/John Libbey Eurotext Ltd. Paris. Cuilloum. A., Beaune. B.. Gascoin. M.N.. Begue. J.-M., Campion. J.P., Guengerich. F.P. & Guguen-Guillouzo. C. (1985) Maintenance of cytochrome P-450 in cultured adult human hepatocytes. Biochemical Pharmacology. 34. 2991 -299 5. Hanahan. D.J.. Rolfs. M.R. & Day. W.C. (1972) Observations on the factor V activator present in Russell's viper venom and its action on factor V. Biochimica et Biophysica Acta. 286, 205-2 11. Huber. P.. Dalmon. J.. Courtois. G . . Laurent. M.. Assouline. M. & Marguerie. G. ( 198 7) Characterization of the 5'4anking region for the human fibrinogen fi gene. Nuclic Acids Research. 15, 161 51625.

Jackson, C.M. & Nemerson, Y. ( 1 980) Blood coagulation. Annual Review of Biochemistry. 49, 765-795. Jenny. R.J.. Pittman. D.D.. Toole. 1.1.. Kriz. R.W.. Aldape. R.A.. Hewick. R.M.. Kaufman. R.J. & Mann. K.G. ( 1 987) CompletecDNA and derived amino acid sequence of human factor V. Proceedings of the National Academy of Sciences ofthe United States of America. 84, 4846-4850.

Joist. J.H. (1987) Hemostatic abnormalities in liver disease. In: Hernostasis and Thrombosis. Basic Principles and Clinical Practice (ed. by R. W. Colman. J . Hirsh. V. J. Marder and E. W. Salzman). J . B. Lippincott Company, Philadelphia. Kane. W.H.. Lindhout. M.J.. Jackson. C.M. & Majerus. P.W. (1980) Factor Va dependent binding of factor Xa to human platelets. Journal of Biological Chemistry. 255. 1 1 70-1 174. Kane. W.H.. Ichinose. A,. Hagen. F.S. & Davie. E.W. ( 198 7) Cloning of cDNAS coding for the chain region and connecting region of human factor V. a blood coagulation factor with four types of internal repeats. Biochemistry. 26. 6 508-65 14. Lechner. K.. Nissner. H. & Thaler. E. ( I 977) Coagulation abnormalities in liver disease. Seminars in Thrombosis arid Hemostasis. 4, 4056.

Maniatis. T.. Fritsch, E.F. & Sambrook. 1. ( 1 9 x 9 ) Electrophoresis of RNA through gels containing formaldehyde. In: Molerular Cloning: A /ahorator!j mcinual (ed. by T. Maniatis. E. F. Fritsch and I. Sambrook). 2nd edn. Vol. 1. Cold Spring Harbor Laboratory Press. Mann. KG., Nesheim. M.E.. Hibbard. L.S. &Tracy. P.B. ( 1 9 8 1 ) T h e role of factor V in the assembly of the prothrombinase complex. Annals ofthe New York Academy of Sciences. 370, 378-388. Mann. K.C., Nesheim. M.E. & Tracy. P.B. ( 1986) Nonensymatic cofactors: factor V. In: Blood Coagulation (ed. by R. F. A. Zwaal and H. C. Hemckerl. Elsevier North Holland, Amsterdam. Mann. K.G.. Tracy, P.H.. Krishnaswamy. S., Jenny. R.J.. Odegaard. B.H. (I: Nesheim. M.E. (19x7) Platelets and coagulation. In: Thronibosis: Xlth Congress on Haemstasis. I.euven (ed. by M. Verstraete. J. Verrnylen, H. R. Lijnen and J . Arnout). International Society on Thrombosis and Haemostasis and I m v e n University Press. Mazzorana. M.. Cornillon. B.. Baffet. G.,Hubert. N.. Belleville.J . , Eloy. R. (I: Guguen-Guillouzo. C. ( 1 9 8 9 ) Hiosynthesis of factor V by normal adult rat hepatocytes. Thrombosis Resmrch. 54, 655-675. Nawa. K.. Nakamura, T.. Kumatori. A,. Noda. C. & Ichihara. A. ( 19801 (;lucocorticoid-dependentexpression of the albumin gene in adult rat hepatocytrs. journd of Biologiml Chemistry. 261, 1 6x8 3-1 68x8. Nesheim. M.K.. Katzmann. J.A.. Tracy. P.H. (I: Mann. K.G. (1981) Factor V. Mt~tliodsin Enzpnolog!/, 80, 249-274.

23 5

Neisheim, M.E.. Foster, W.B.. Hewick. R. & Mann. K.G. (1984) Characterization of factor V activation intermediates. journal of Biological Chemistry, 259, 3187-3196. Olson, J.P., Miller, L.L. & Troup. S.B. (1966) Synthesis of clotting factors by the isolated perfused rat liver. journal of Clinical Investigation, 45, 690-701. Princen. H.M.G.. Moshage. H.J.. de Haard. H.J.W.. van Gemert. P.J.L. & Yap, S.H. (1984) The influence of glucocorticoids on the fibrinogen messenger RNA content of rat liver in vivo and in hepatocyte suspension culture. Biochemical journal. 220, 63 1 637.

Ratanasavanh. D.. Beaune. P.. BafTet. G., Rissel. M.. Kremers. P.. Guengerich. F.P. & Guillouzo. A. (1986) Immunocytochemical evidence for the maintenance of cytochrome P-450 isozymes. NADPH cytochrome c reductase, and epoxide hydrolase in pure and mixed primary cultures of adult human hepatocytes. Iournalof Histochemistry and Cytorhemistry, 34, 527-533. Raymondjean, M.. Kneip. B. & Schapira. G. (1983) Preparation and characterization of mRNAs from rat heart muscle. Biochimie. 65, 65-70.

Rigby. P.W.. Levy. B.A.. Adams, M.A. & Fuller. G.M. (1982) Labeling deoxyribonucleic acid to high specific activity in vitro by nick translation with DNA polymerase 1. journal of Molecular Biology. 113,237-251.

Rodgers. G.M. & Shuman. M.A. ( 1 98 3 ) Prothrombin is activated on vascular endothelial cells by factor Xa. Proceedings of the National Academy ofsciences ofthe United States ofAmerica. 80. 7001-7005. Susuki. K.. Dahlback. B. & Stenflo. J. (1982) Thrombin catalysed activation of human coagulation factor V. journal of Biological Chemistry. 257, 6556-6564. Suzuki, K.. Stenflo. J . , Dahlback. B. & Teodorsson. B. (1983) Inactivation of human coagulation factor V by activated protein C. journal of Biological Chemistry. 258, 1914-1 920. Thomas, P.S. (1980) Hybridization of denatured RNA and small DNA fragments transfered to nitrocellulose. Proceedings of the National Academy of Sciences of the United States of America. 77. 5201-5205.

Tracy, P.B. (1988) Regulation of thrombin generation at cell surfaces. Seminars in Thrombosis and Hernostasis. 14, 227-2 33. Tracy, P.B.. Rohrbach. M.S. & Mann. K.G. (19x3) Functional prothrombinase assembly on isolated monocytes and lymphocytes. journal of Biological Chemistry. 258. 7264-7267. Uzan. G.. Courtois. G.. Besmond. C.. Frain. M.. Sala-Trepat. J.M.. Kahn. A. & Marguerie. G.A. ( 1984) Analysis of fibrinogen genes in patients with congenital afibrinogenemia. Biochemical and Biophgsical Research Communications. 120. 3 76- 38 3. Van de Waart. P.. Bruls. H.. Hemcker. H.C. & Lindhout. T. (1983) Interaction of bovine blood clotting factor V and its subunits with phospholipid vesicles. Biocheniistrg. 22. 2427-24 32. Wahl, G.M.. Stem, M. & Stark, G.R. (1979) Efficient transfer of large DNA fragments from agarose gels to diarobenzyloxymethyl-paper and rapid hybridization by using dextran sulfate. Proceedings ofthe National Academy of Sciences of the United Statcs of America. 76, 3683-3687.

Workman. E.F. & Lundblad. R.L. (1977). The role of liver in biosynthesis of the non-vitamin K-dependent clotting factors. Seminars in Thrombosis and Hemostasis. 1 1, 1 5-28. Wilson, D.B.. Salem. H.H.. Mruk. J.S.. Maruyama. 1.6; Majerus. P.W. ( 1 984) Hiosynthesis of coagulation factor V by a human hepatocellular carcinoma cell line. Iournal of Clinical Investigation. 73. 654-658.