Distribution of Vitronectin in Plasma and Liver Tissue: Relationship to Chronic Liver Disease

JUNKO KOBAYASHI, SADAKO YAMADA AND HIRONAKA KAWASAKI Second Department of Internal Medicine, Tottori University Faculty of Medicine, Yonago 683, Japan

To clarify the clinical significance of vitronectin, we compared the concentration of plasma vitronectin with serum fibrous markers and liver function test values in patients with chronic liver diseases. We also evaluated the vitronectin content in the liver by means of enzyme-linked immunosorbent assay and the local- ization of vitronectin in liver tissue with enzyme immunohistochemistry. In chronic liver disease, the concentration of plasma vitronectin was significantly lower than that in healthy controls, being related to the severity of liver disease. The plasma levels of vitronectin showed no correlation to fibrous markers but a significant correlation with those of serum albumin and prothrombin time. On the other hand, the content of vitronectin in liver tissue was significantly increased in chronic liver disease compared with that in normal controls. In the normal liver, vitronectin was observed in the portal area by light microscopy. In chronic hepatitis and cirrhosis, vitronectin was found in the connective tissue around the portal and central veins and in the areas of piecemeal and focal necrosis. These findings suggested that vitronectin is deposited in injured tissue through the process of repair and fibrosis and plays an important role as an adhesive protein. Moreover, the lower levels of plasma vit- ronectin in chronic liver disease may be due to its decreased synthesis, deposition or both in injured tissue. (HEPATOLOGY 1994;20:1412-1417.)

Vitronectin is an adhesive glycoprotein in the extra- cellular matrix and plasma and has been shown to bind with collagen and heparin (1). It has important physi- ological actions in cell adhesion (2, 3), cell necrosis mediated by complement action (4, 5) and blood coagu- lation (61, but little is known about its distribution and

Received December 9, 1993; accepted May 26, 1994. Other abbreviations used in the t a t : CH, chronic hepatitis; CPH, chronic

persistent hepatitis; DAB, diaminobenzidine; HCC, hepatocellular carcinoma; PBS, phosphate-buffered saline solution; PH, prolyl hydroxylase; PIIIP, procol- lagen 111 peptide.

This work was supported by a grant-in-aid for scientific research (no. 05670554) provided by the Ministry of Education, Science and Culture of Japan.

Address reprint requests to: Junko Kobayashi. M.D., Second Department of Internal Medicine, Tottori University Faculty of Medicine, Yonago 683, Japan.

Copyright d 1994 by the American Association for the Study of Liver Diseases.

0270-9139/94 $3.00 t 0 3111159237

metabolism in the liver or its clinical significance in chronic liver disease (7, 8).

In this study, we measured plasma and liver content of vitronectin in patients with chronic liver disease and investigated the relationship of vitronectin to serum fibrous markers such as PIIIP, 7s collagen, laminin and PH and liver function tests. We also evaluated the distribution and localization of vitronectin in the liver by enzyme immunohistochemistry to clarify the signifi- cance of the vitronectin in chronic liver disease.

MATERIALS AND METHODS Patients. The study included 108 patients with chronic liver

disease - of whom 33 had CPH, 14 had CAH, 33 had cirrhosis and 28 had HCC-and 39 healthy subjects (31 men and 8 women, age range 24 to 63 yr) as controls, who visited Tottori University Hospital between April 1989 and April 1992 (Table 1). Informed consent was obtained from all the patients, and the study was performed in accordance with the Declaration of Helsinki. Diagnoses were made by means of biochemical tests, diagnostic imaging techniques and histological examinations.

Plasma Vitronectin. Blood was drawn from all subjects early in the morning after an overnight fast. We obtained plasma for measurements of vitronectin by treating blood with citrate for anticoagulation and storing it at a temperature below - 20" C until analysis. The plasma vitronectin level was measured with a vitronectin ELISA kit (Iwaki Glass Co., Ltd., Tokyo, Japan) (9). In brief, the microtiter plate was coated with 50 ~1 of antivitronectin monoclonal antibody (Ml) diluted 100-fold with PBS and then incubated overnight at 4" C. Fifty microliters of the mixture was then added to the well and incubated for 1 hr a t room temperature. After three washes with PBS in a platewasher, 100 p1 of chromogenic solution containing o-phenylenediamine was added, and the color development, was monitored at an optical density of 492 nm with a microplate reader after 30 min at room temperature. Sera were also stored at a temperature below -20" C until analysis. Serum PIIIP was measured by means of RIA (Behringwerke AG, Marburg, Germany) (lo), 7 s collagen and laminin were measured with an RIA kit (Nippon DPC Co., Tokyo, Japan, and Hoechst Co., Tokyo, Japan) (11,12) and PH was measured with a commercial EIA kit (Fuji Chemical Industries Ltd., Takaoka, Japan) (13). Laboratory parameters, including serum albumin, total bilirubin, AST, ALT, in- docyanine green clearance rate and prothrombin time were measured by means of conventional methods.

Tissue Vitronectin. Percutaneous liver biopsy was carried out with a Silverman needle in 17 patients with chronic liver disease on the day of plasma collection. Three liver specimens

1412

HEPATOLOGY Vol. 20, No. 6, 1994 KOBAYASHI, YAMADA AND KAWASAKI 1413

plasma vitronectin (PS lm l ) 600

1

500 -

400 -

8

I I I I I

normal CPH CAH LC HCC (39) (33) (14) (33) (28 1

mean +S.E.M. 260f37 220+-46 236f57 178+86 164+48

FIG. 1. Plasma concentration of vitronectin in chronic liver disease. There were no significant differences between patients with CPH and CAH and the controls. In patients with cirrhosis or HCC, the plasma vitronectin concentrations were significantly lower than that in the controls. Numbers in parentheses indicate numbers of subjects. *p c 0.01 vs. control.

TABLE 1. Clinical and laboratory findings in patients with liver disease Group CPH CAH Cirrhosis HCC

No. of cases 33 14 33 28 Age (Yr)" 53.9 2 3.8 39.6 ? 3.3 57.3 '' 2.0 65.8 ? 1.6 Sex ( M P ) 21/12 915 20/13 17/11 Albumin (gm/dl)" 4.5 2 0.3 4.4 ? 0.3 3.4 ? 0.9 3.4 r 0.5 Total bilirubin (mg/dl)" 0.68 ? 0.20 0.70 k 0.24 1.6 f 1.7 1.5 ? 1.6 ALT (U/L)" 55 ? 15 249 ? 163 67 ? 52 86 ? 99

Normal ranges: albumin, 3.5 to 5.3 gmldl; total bilirubin, 0.2 to 1.0 mg/dl; ALT, <47 U/L. "Data expressed as mean t S.E.M.

TABLE 2. Correlation between plasma vitronectin levels and laboratory data

Laboratory data Correlation coefficient

(n) p Value"

PIIIP Laminin 7s-collagen Prolyl hydroxylase Albumin Prothrombin time

-0.291 (70) -0.183 (35) -0.183 (69)

0.011 (29) 0.637 (66) 0.602 (56)

NS NS NS NS

< 0.01 < 0.01

"p Values calculated with an F test.

from normal controls were obtained by means of wedge biopsy measured by ELISA. Liver protein was measured by means of during laparotomy for gastric disease in our hospital. Lowry's method (14). Liver vitronectin content was expressed

Liver tissue sections, frozen at - 20" C , were placed in PBS. as micrograms per milligram liver protein. The specimen homogenated for 3 min, after which it was Immunohietochemical Procedure. Twenty liver tissue adjusted to a 2% suspension and centrifuged at 15,000 g for 15 specimens were fixed in 10% formalin and embedded in min at 4" C. The supernatant was diluted 20-fold with PBS paraffin. After deparaffinization of these sections with a xylene containing 1% BSA, and the concentration of vitronectin series and inactivation of endogenous peroxidase

1414 KOBAYASHI. YAMADA AND KAWASAKI HEPATOLOGY December 1994

vitronectin ( p g / mg liver protein)

0.7

0.6 - 0.5 -

0.4 -

0.1 1 0.0 I

I 1

normal (3) CH (9) LC (8) meanfS.E.M. 0.174k0.045 0.41 1k0.120 0.419f0.182

FIG. 2. Hepatic content of vitronectin in chronic liver disease. Hepatic content of vitronectin was significantly increased in CH and cirrhosis. Numbers in parentheses indicate numbers of subjects.

n = l l 0 N.S.

100 200 300 400 plasma vitronectin ( p g / ml)

FIG. 3. There was no correlation between the liver content and plasma concentration of vitronectin.

with methanol-0.3% H,O, for 15 min at room tempera- and incubated for 1 hr at 37" C. They were washed with PBS, ture, the tissue sections were washed with PBS and re- reacted with 0.025% H,O, and the chromogen DAB for 5 min acted with the primary antibody, antivitronectin mono- at 37" C and washed with PBS again, covered with glycerin clonal antibody (M2; Iwaki Glass Co.) diluted 200-fold and gelatin and examined under light microscopy (15). incubated overnight at 4" C. Then the tissue sections were Results are expressed as mean * treated with the secondary antibody, biotin-labeled rabbit S.E.M., and differences were compared statistically with antimouse IgG, IgA and IgM (Nichirei Co. Ltd., Tokyo, Japan),

Statistical AnaZysis.

Student's t test, the F test and the least-squares method.

HEPATOLOGY Vol. 20, No. 6, 1994 KOBAYASHI, YAMADA AND KAWASAKI 1415

,, .,

' A ,

JL

J

FIG. 4. Localization of vitronectin by means of light microscopy with immunohistochemical staining techniques. Vitronectin was stained brown by DAB. (A) In normal liver, vitronectin was lightly stained in the portal region (arrows) (original magnification X 40). (B) In cirrhotic liver, vitronectin was localized in the portal region (arrows) along the collagen fibers (original magnification x 20).

RESULTS The plasma vitronectin concentration was 260 2 37

p.g/ml in the control group (Fig. 1). It was 220 ? 46 p.g/ml in patients with CPH and 236 k 57 Fg/ml in patients with CAH. There were no significant differ- ences between the two groups and the controls. However, in patients with cirrhosis and HCC, the plasma vitronectin concentrations were 178 * 86 kg/ml and 164 2 48 kg/ml, respectively, significantly lower than that in the control group (p < 0.01 for both).

Table 2 shows the relationship between plasma vitronectin and liver function, as well as various markers of fibrosis in subjects with chronic liver disease and controls. Plasma vitronectin levels and all the fibrous markers (PIIIP, 7 s collagen, laminin and PH) were not correlated (Table 2). In contrast, serum albumin content (r = 0.637, p < 0.01) and prothrom- bin time (r = 0.602, p < 0.01) were correlated with plasma vitronectin. Indocyanine green clearance rate and plasma vitronectin were negatively correlated (r = -0.552, p < 0.01). Plasma vitronectin was not correlated with total bilirubin, AST, or ALT.

The content of vitronectin in liver tissue was 0.174 ? 0.045 p,g/mg liver protein (Fig. 2). In CH and cirrhosis, it was 0.411 k 0.120 p.g/mg liver protein and 0.419 f 0.182 Fg/mg liver protein, respectively. The content of vitronectin in liver tissue was significantly increasedin CH and cirrhosis (p < 0.01). Figure 3 shows the relationship between the content of vitronectin in liver tissue and plasma in three controls and eight pa- tients with chronic liver diseases. The patients with low plasma vitronectin level seem to have high hepatic vit- ronectin level. But we found no correlation between the liver content and plasma concentration of vitronectin.

The localization of vitronectin was investigated by means of light microscopy with immunohistochemical staining techniques. Vitronectin was stained brown by DAB. In normal liver tissues, vitronectin was lightly stained in the portal region (Fig. 4A).

In cirrhotic tissues, vitronectin was localized at sites of fiber deposition in the central venous region and in the portal region (Fig. 4B). Vitronectin was also detected in areas of piecemeal necrosis and focal necrosis in liver cirrhosis tissues.

1416 KOBAYASHI, YAMADA AND KAWASAKI HEPATOLOGY December 1994

DISCUSSION Vitronectin, which was first reported by Barnes et al.

(2, 31, is a plasma glycoprotein exhibiting adhesive properties, with a molecular weight of 75,000 Da. The presence of the Arg-Gly-Asp arrangement and the binding sites of collagen (16, 17) and heparin has also been reported in the vitronectin molecule. Vitronectin causes cells to adhere to the extracellular matrix by way of the vitronectin receptors on the cellular surface, thus enabling the cells to spread (18). Vitronectin also has been shown to form membrane invasive complexes by binding with the C5b-9 complement (19). Vitronectin has been reported to inhibit cell damage after binding with the C5b-9 complement (5). Vitronectin also binds with the thrombin-antithrombin I11 complex to sup- press thrombin activity (20).

In this study, we detected vitronectin in the fibrous portions of the portal and central venous regions, as well as in the areas of piecemeal and focal necrosis (Fig. 4). Kodama et al. also detected vitronectin in the portal and central venous regions, as well as in the interfibrous areas of collagen in the fibrous septae (8). Grebb et al. (21) reported that vitronectin binds with various types of collagen fibers. The findings by Clement et al. (22) indicate that the collagen detected in the portal region was mainly type I and 111. These findings suggest that vitronectin, which has binding sites for these collagens, plays a role in the development and progression of hepatic fibrosis (22, 23).

We also showed that vitronectin is significantly increased in the liver tissues of patients with chronic liver disease. It is speculated that vitronectin is rapidly secreted after its synthesis, as judged from its distri- bution pattern in the liver tissues and plasma of healthy subjects. Even in the CH group, in whom plasma vitronectin is not significantly decreased, it is increased in the tissues.

There was no significant difference in the content of vitronectin in the liver tissues between the CH and cirrhosis groups. Furthermore, our findings did not reveal any relationship between plasma vitronectin and markers of fibrosis. These findings suggest that vit- ronectin is involved in postnecrotic healing and fibrosis at a very early stage of the repair process.

Plasma vitronectin levels and tissue vitronectin content were not correlated. The vitronectin that is deposited in liver tissue may be different from that detected in plasma. Izumi et al. (24) proposed that some forms of vitronectin are conjugated with heparin, whereas others are not. The serum vitronectin did not have collagen- or heparin-binding properties, activating that by heating, urea and other glycosaminoglycans (25). In this study, plasma vitronectin levels decreased with the progression of liver disease, being significantly lower in LC and HCC. Inuzuka et al. (26) detected vitronectin in the rough endoplasmic reticulum of hepatocytes by means of electron microscopy and suggested that vit- ronectin is synthesized in the liver. Indeed, albumin and prothrombin time which are indicators of protein syn- thesis were found to be correlated with plasma vit-

ronectin. The decrease in the concentration of plasma vitronectin in chronic liver diseases is probably due to reduction in the volume of liver parenchymal cells and the consumption of plasma vitronectin due to its deposition in the extracellular matrix.

This study revealed that vitronectin content in liver tissue is significantly increased in chronic liver disease, though the level of plasma vitronectin is decreased. These findings suggest that vitronectin is deposited in liver tissues in the process of tissue repair and fibrosis in the liver.

Acknowledgment: We are grateful to Dr. Yasushi Horie for his expert technical assistance.

REFERENCES 1. Suzuki S, Pierschbacher MD, Hayman EG, Nguyen K, Ohgern Y,

Ruoslahti E. Domain structure of vitronectin. J Biol Chem

2. Barnes D, Sato G. Growth of a human mammary tumor cell line in a serum-free medium. Nature 1979;281:388-389.

3. Barnes DW, van der Bosch J, Masui H, Miyazaki K, Sato G. The culture of human tumor cells in serum-free medium. Methods Enzymol 1981;79:368-369.

4. Podack ER, Preissner KT, Muller Eberhard HJ. Inhibition of C9 polymerization within the SC5b-9 complex of complement by S-protein. Acta Pathol Microbiol Immunol Scand 1984;284:89-96.

5. Tshopp J , Masson D, Schafer S, Peitsch M, Preissner KT. The heparin binding domain of S-proteidvitronectin binds to complement components C7, C8 and C9 and perforin from cytolytic T-cells and inhibits their lytic activities. Biochemistry

1984;259:15307-15314.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

i988;27:4103-4iog. Jenne D, Stanley KK. Molecular cloning of S-protein, a link between complement, coagulation and cell substrate adhesion.

Kemkes-Matthes B, Preissner KT, Langenscheidt F, Maatthes KJ, Muller-Berghaus G. S-protein/vitronectin in chronic liver disease: correlations with serum cholinesterase, coagulation factor X and complement component C3. Eur J Haematol 1987;39:161-165. Kodama T, Gotoh M. The distribution of vitronectin in serum and liver in fibrotic liver disease [in Japanese]. Acta Hepatol Jpn

Takahashi TA, Hosoda M, Sekiguchi S, Hasegawa Y, Shimooka T. Vitronectin levels of plasma in healthy subjects measured by an enzyme-linked immunosorbent assay (ELISA) [in Japanese]. J Med Techno1 1989;33:957-960. Ueno T, Torimura T, Inuzuka S, Ishii K, Sada M, Tanigawa K. Elevation on the measurement of serum PIIIP in patients with various liver disease by a new RIA-gnost PIIIP kit using monoclonal antibodies [in Japanese]. Igaku Yakugaku 1990;24:

Yamada S, Suou T, Kawasaki H, Yoshikawa N. Clinical signifi- cance of serum 7 s collagen in various liver diseases. Clin Biochem

Foidart J-M, Berman JJ , Paglia L, Rennard S, Abe S, Pernantoni A, Martin G. Synthesis of fibrinogen laminin and several collagens by a liver derived epithelial iine. Lab Invest 1980;42:525-532. Yoshida S, Bai Y, Muragaki Y, Ooshima A, Inada S, Nagai Y, Obata K, et al. A sandwich immunoassay for prolyl 4-hydroxylase using monoclonal antibody. Clin Chim Acta 1986;160:37-46, 1986. Lowry OH, Rosenbrough NJ, Farr AL, Randall RJ. Protein measurement with the Folin phenol reagent. J Biol Chem

Guesdon JL, Ternynck T, Avrameas S. The use of avidin-biotin interaction in immunoenzymatic technique. J Histochem 1979;

Suzuki S, Oldberg A, Hayman EG, Pierschbacher MD, Ruoslahti E. Complete amino acid sequence of human vitronectin deduced from cDNA similarity of cell attachment sites in bitronectin and fibronectin. EMBO J 1985;4:2519-2524.

EMBO J 1985;4:3153-3157.

1990;31:287-293.

675-678.

1992;25:467-470.

1951;193:265-275.

27:1131-1139.

HEPATOLOGY Vol. 20, No. 6, 1994 KOBAYASHI, YAMADA AND KAWASAKI 1417

17. Izumi M, Shimooka T, Morishita N, Ii I, Hayashi M. Identification of collagen-binding domain of vitronectin using monoclonal antibodies. Cell Struct Funct 1988;13:217-225.

18. Buck CA, Horwitz AF. Cell surface receptors for extracellular molecules. Annu Rev Cell Biol 1987;3:179-205.

19. Podack ER, Kolb WP, Muller-Eberhard HJ. The SC5b-7 complex: formation, isolation, properties and subunit composition. J Im- munol 1977;119:2024-2029.

20. I11 CR, Ruoslahti E. Association of thrombin-antithrombin 111 complex with vitronectin in serum. J Biol Chem 1985;260:15610- 15615.

21. Grebb C, Hayman EG, Engvall E, Ruoslahti E. Interaction of vitronectin with collagen. J Biol Chem 1986;261: 16698-16703.

22. Clement B, Grimand JA, Campion JP , Deugneier Y, Guillouzo A.

Cell types involved in collagen and fibronectin production in normal and fibrotic human liver. HEPATOLOGY 1986;6:225-234.

23. Reilly JT, Gnash JR. Vitronectin (serum spreading factor): its localization in normal and fibrotic tissue. J Clin Pathol 1988;41:

24. Izumi M, Yamada KM, Hayashi M. Vitronectin exists in two structurally and functionally distinct forms in human plasma. Biochem Biophys Acta 1989;990:101-108.

25. Ishikawa M, Hayashi M. Activation of the collagen-binding of endogenous serum vitronectin by heating, urea and glycosami- noglycans. Biochem Biophys Acta 1992;112 1: 173- 177.

26. Inuzuka S, Ueno T, Torimura T, Tamaki S, Sakata R, Sato M, Yoshida H, et al. Vitronectin in liver disorders: biochemical and immunohistochemical studies. HEPATOLOGY 1992;15:629-636.

1269- 1272.