Fibronectin: a possible factor promoting cholesterol monohydrate crystallization in bile

44

Biochimica el Biophysica Acta, 1086 ( 1991 ) 44-48 © 1991 Elsevier Science Publishers B.V. All rights reserved 00115-27611/91/$03.50 A DONIS 00052760910112738

BBAL1P 53732

Fibronectin: a possible factor promoting cholesterol monohydrate crystallization in bile Kazuo Chijiiwa, Akitoshi Kosa, Toru Yamasaki, Kazuo Shimada, Hirokazu Noshiro and Fumio Nakayama Deparlment of Surgery L Kyushu Unicersity Faculty of Medicine, Fukuoka (Japan) (Received 22 October 1990) (Revised manuscript received 5 June 1991)

Key words: Fibronectin; Nucleation time; Vesicle; Cholesterol gallstone

To examine the hypothesis that fibronectin physiologically present in bile might be a possible nucleating factor, the concentrations of fibronectin in gallbladder bile were determined and its induced effect on nucleation time and on the form of vesicle were examined in bile-model and human gallbladder bile. The gallbladder bile samples taken from patients with cholesterol gallstone had a significantly higher concentration of fibronectin and the faster nucleation time than the control. However, no significant correlation was found between nucleation time and endogenous fibronectin concentration. The addition of 0.5, 1.2, 1 0 / z g / m l of fibronectin into two kinds of bile-model significantly shortened the nucleation time in a dose-related manner. Nucleation time was significantly shortened by the addition of I p g / m l exogenous fibro~ectin into abnormal bile while such an effect was absent in the control. The addition of fibronectin increased the size of vesicles observed by the electron microscope. The results suggest that fibronectin physiologically present in bile may be one of the possible nucleating factors.

Introduction

Biliary cholesterol is mainly transported in micelles and in vesicles [1-6]. The latter plays an important role in the formation of cholesterol monohydrate crystals. Previous studies indicated that cholesterol monohydrate crystallization occurs from phospholipidcholesterol vesicles depending upon vesicular cholesterol saturation [6-8] and vesicular aggregation precedes the cholesterol monohydrate crystallization [7,9]. It is also noteworthy that the population of vesicles is heterogeneous in size and larger vesicles are cholesterol rich [10]. Since biliary cholesterol is frequently saturated in healthy man and the nucleation time can distinguish abnormal bile from control bile [11,12], the factors affecting cholesterol nucleation are of considerable interest. Among these factors, the importance of anti- and pronucleating protein have been recently demonstrated [ 13- L6]. Fibronectin, a disulfide bonded dimer of molecular weight 440 kDa, is an a2-surface binding glycoprotein and present in soluble form in plasma, urine and in ~_orrespondcnce: K. Chijiiwa, Department of Surgery i, Kyushu L livcrsity Faculty of Medicine. Fukuoka 812, Japan.

bile. Fibronectin is synthesized by hepatocytes and sinusoidai endothelial cell [17,18] and excreted in bile [19]. Much interest is the phospholipid vesicular aggregation by the presence of fibronectin [20]. Aggregated and fused vesicles may form large vesicles which are prone to nucleate [10]. Based on the fact that fibronectin induces the formation of aggregated vesicles and vesicular aggregation precedes cholesterol monohydrate crystallization [9,20], we have examined the hypothesis that fibronectin might be a possible nucleating factor. Accordingly, we first determined the concentrations of endogenous fibronectin in gallbladder bile with and without cholesterol gallstone. Then, the induced effects of fibronectin on vesicular form and on cholesterol crystal appearance time were studied using bile-model. Finally, exogenous fibronectin was added into gallbladder bile whether fibronectin shortens the cholesterol crystal appearance time. Materials and Methods

Chemicals Sodium taurocholate and phospholipid (Egg York lecithin Type XI-E) were purchased from Sigma Chemical (St. Louis, MO). Lecithin was more than 99% pure

45 by thin-layer chromatography in chloroform/methanol/water/ao=tic acid ( 6 5 : 2 5 : 4 : i). Cholesterol was obtained from Sigma and recrystallized three times from hot ethanol and dried in a vacuum dessicator. The purities of these compounds were greater than 99% by thin-la~er chromatography and gas-liquid chromatography. Fibronectin (human plasma) was obtained from the Biomedical Products Division (Two Oak Park, Bedford, M A ) a n d Green Cross (Osaka, Japan). Fibronectin was also purified from human p!asma as described [21] and its single band was assessed by sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SD~-PAGE). i~_~l-monoclonal fibronectin antibody was synthesized using ascites taken from immunizated mice followed by Protein A agarose column. lodination, elec:rophoresis and immunoblotting were carried out as described [21], Solvents were distilled prior to use. All other reagents used were of analytical grade.

Bile-model and human gallbladder bile Bile-model was prepared basically as described 1"7,22]. Necessary amounts of lecithin and cholesterol were placed in a vial with a cap. After the organic solvent was completely evaporated under a stream of nitrogen, the necq:ssary amount of taurocholate and the required volume of Hepes saline (N-2-hydroxyethylpiperazine-N-2-ethanesulfonic acid-NaCi t0 raM: 140 raM, pH 7.5) wt;re added. The bile-model samples were shaken at 5 5 ° C until microscopically isotropic. The final compositions of two kinds of bile-model were; tauroeholate 104.4 mM, lecithin 26.3 raM, cholesterol 9.0 raM, cholesterol saturation index 1.04 or taurocholate 103.9 raM, lecithin 26.2 raM, cholesterol 10.0 mM, cholesterol saturation index 1.16. Lecithin/bile salt molar ratio was 0.25 and total lipid concentration was adjusted to 8 g / d l . Human gallbladder bile samples were completely needle aspirated ~u avoid the stratification from patients with cholesterol gallstone disease and from the control patients underwent surgery for gastric cancer. The absence of gallstones was verified by preoperative examinations and intraoperative palpation, and informed consent was obtained from all patients. Care was taken to avoid the contamination with blood. The cholesterol gallstone was identified by visual inspection and chemical analysis as reported from our laboratory [23]. Cholesterol content was 75 ± 19.8% of total dry weight. Gallbladder bile samples titus obtained were kept at 37 °C in iht; dark jar and bJought to the laboratory as soon as possible. The freshly prepared bile-model and human gallbladder bile samples obtained were immediately filtered through the 0.22/.~m filter (MiUipore, Bedford, MA) at 37" C, then flushed with nitrogen and sealed in the brown tube and subsequently used for the following studies. A part of bile samples were stored at

- 2 [ ) ° C for the anatysis of biliar-i lipids artd fibronectin. The storage of gal!bladder bile at - 2 0 ° C within a month had no effect on fibronectin values. AII patients had normal liver function.

Chemical analysis Total bile acid concentration was determined by 3a-hydroxystcroid dehydrogenase method [24]. Phospholipid was determined colorimetricaily [25] and cholesterol was quantified by gas-liquid chromatography as described previously [26,27]. Cholesterol saturation index v as calculated from tables provided by Carey [28]. Chemical analysis of gallstones was carried out as reported from our laboratory [23]. The concentration of fibronectin in bile samples were determined by the immunoradiometric assay system using "--Slmonoclonal fibronectin antibody. 20 .u.1 of 1-30-times diluted bile samples was dissolved in buffer and applied on monoclonal antibody coated beads. They were incubated for 3 h and washed with deionized water and then reacted with L~51-monoclonal antibody and counted (Fig. 1). The rehability of the method was confirmed by the recovery experiment (fibroncctin recovery in bile: 95 + 13.3%). The detection limit was 0.02 /.tg/ml and the coefficient of variation was less than 5%.

Fibronectbz addition experiments Induced effects of fibronectin addition on cholesterol crystal appearance time (nucleation t i m e ) w e r e examined in bile-model and native bile. 40 tL! of various concentrations of fibronectin (1 mg/ml, 0.12 m g / m l and 0.05 m g / m l in distilled water) was added into 4 ml of bile-model samples to yield the concentrations of fibronectin 0.5, 1.2 and 10 p.g/ml of bile. These fibronectin concentrations were selected beAssay procedure Monoc~onal ant~t~ody

{OAL -DF117} co,~ted beads

Sample or standards

1 2 0 ~j=

Bulfer (50 m M phosphate buffer. C 0 1 5 M I 0 0 5 *10 N a N 3 , O1 -/oBSA, pH 5 5 )

NoCI

2 C 0 pl

Incubated at r o o m ter~,p ~ j n }

~Vashed with delontzed w a t e r

(3 times))

* monoclonal ant,body ( O A L - pF 11: } , 0 0 0 O O c p m / 2 0 0 )Jl used buffer; 5 0 r a m phosphate buffer, C 0 0 5 =/o N o N a I 01 "to BSA Inc,~bated at r o o m Washed

wltt~ de~on, zed water

1 Translerred

ternp

(3 h }

{3 times)

to clean tube

Coun ted

Fig. I, Analytical procedure to determine the concentration of fi-

broneclin in bile by the immunoradiometric assay meth,)d.

46 TABLE l Chemical ct,mposition tend nt¢ch'atio~t tbne of gallbladder bile TBA (mM)

PL (raM)

CH (raM)

CSI

FN (#g/ml)

Oallslone (n -- 7)

132 +69.3

38± 12.3

13.8 +-6 . ( )

1.17+_1}.34,1

1.2+_0.49 h

Contrc¢ {n -- 6)

1511_+84.1

48 +_23.7

13.7 + 7.8

0.85 +_0.21}

1}.5_+(}.! 2

NT (days) 3.3+_2.87 :' 19.0 _+3.58

~ P < 11.111. ~' P < i).02 compared with control. Ct I. cholesterol: CSI, cholesterol silturi=tion index: TBA, total bile acid: PL. phospholipid; FN. fibronectin: NT, nucleation time,

cause human gallbladder bile samples with and without cholesterol gallstone c o n t a i n e d 1.2 / z g / m l and 0.5 # g / m l of fibronectin, respectively. 4 / x g of fibronectin was a d d e d into 4 ml of gallbladder bile. 40 #1 of distilled water was a d d e d into the bile-model and human gallbladder bile as a control. T h e addition of albumin (1 m g / m l ) had no significant effect on the nucleation time w h e n c o m p a r e d to the control. T h e samples were mixed thoroughly, flushed with nitrogen and then stood in an incubator at 37 ° C without shaking. T h e cholesterol m o n o h y d r a t e crystal a p p e a r a n c e

(nucleation) time was m o n i t o r e d using a polarized microscope (Nikon XTM, Japan). Nucleation time was examined every 3 - 4 h for the first 2 days and twice a day t h e r e a f t e r for a 21-da~s observation period. W h e n cholesterol crystal did not a p p e a r within 21 days, nucleation time was expressed as 21 day. Electron microscopic s t u d y T h e difference in the form of vesicles by the addition of fibroneetin was e x a m i n e d using bile-model samples. For freeze fracturing, samples were m o u n t e d o n t o

Fig. 2. Electron microgn,ph by the deep-etching freeze-fracture method showing difference in the appearance of vesicles in bile-model (CS! 1.04) without (a) and with (b) i #g/ml fibronectin at the third day after initiating the experiment. Vesiclesare larger and fused by the presence of fibronectin (Fig. 4b). while small vesicles (50 nm) are seen in Fig. 4a. The bar indicates 250 rim.

47 the specimen stage and rapidly frozen in propane followed by liquid nitrogen. Frozen specimens were freeze-fractured in the JFD-700u freeze-fracture apparatus (JEOL, Tokyo, Japan) and deep-etched in the upside-down position for less than 7 min at 178 K and replicated with platinum as described by Shibata et 31. [29]. After coating with carbon, the replicas were washed with distilled water, recovered on 300-mesh copper grids and examined in a JEOL JEM-100CX electron microscope operated at !00 kV.

E~f@¢.~ Of f J b r o n e c % n o n c ~ o l e s ~ e f o i nuc I~_<~tlon -';~-no In m o d e l ~ l e

2001 ~3 0

T

'! x

150-

,I

I

Statistics Values are means + S.D. of at least six separate experiments. Effects of various concentrations of fibronectin on nucleation time in bile-model were ana:ysed by two-way analysis of variance (ANOVA).

.o IOC L

o . 005 I-

Z

F o, 005-~-t

Results

Analysis of native bile The concentrations of total bile acids, phospholipid, cholesterol, cholesterol saturation index and nucleation time of gallbladder bile are shown in Table I, Significantly higher cholesterol saturation index and faster nucleation time were observed in the cholesterol gallstone group than those in the control (P < 0.01). The concentration of fibronectin was significantly higher in the cholesterol gallstone group than the control (1.2 + 0.49 vs 0.5 + 0.12 g g / m l , P < 0.02) (Table !). However, no significant correlation was found between the concentration of endogenous fibronectin and the nucleation time ( r = - 0 . 4 2 ) . No significant correlation was also observed between cholesterol saturation index and nucleation time.

Fibronectin addition experiments Electron micrographs (Fig. 2) show that fibronectin induced the formation of larger and clustered vesicles in bile-models, as compared to those present in the absence of fibronectin. The effect of fibronectin on cholesterol monohydrate crystal appearance time (nucleation time) in bile-model is shown in Fig. 3. The nucleation time was shortened with fibronectin in a dose-dependent manner. At 0.5 # g / m l fibronectin this effect was significant both in bile-model having a cholesterol saturation index 1.04 (P < 0.001) as well as in a model whose cholesterol saturation index was 1.16 ( P < 0 . 0 5 ) compared to the corresponding control bile-model. In both bile-model, the addition of 1.2 # g / m l fibroneetin progressively shortened the nucleation time significantly ( P < 0 . 0 5 ) compared to the corresponding nucleation time at the concentration of 0.5 # g / m l . In both bile-model, the addition of 1.2 ~ g / m l fibronectin significantly shortened the nucleation time ( P < 0.001) compared to the control.

,,,,4,,,

Fibronectin

,

i

( ~g/m~ )

Fig. 3. Effectof fibronectin on nucleation time of artificial bile-model. Closed and open circles represent the bile-model with cholesterol saturation index 1.04and 1.16, respectively.

The addition of 1 ~ g / m l of exogenous fibronectin significantly shortened cholesterol monohydrate crystal appearance time in abnormal bile from 3.30 + 2.87 to 1.90 + 1.98 days (P < 0.1)5, n = 7). Thus, 0.6-times shortened nucleation time was observed by the addition of 1 tzg/ml of exogenous fibronectin. In contrast to abnormal bile, no significant effect was observed in the control bile [,y the addition of exogenous fibronectin (19.0 +_ 3.63 vs 18.7 + 3.67 days). Discussion

This is the first report suggesting that fibronectin present in bile might be a possible nucleation promoting factor. The current findings are that fibronectin indeed exist in human bile, induces the formation of clusters of large vesicles and promotes cholesterol monohydrate crystal appearance time in both bilemodel and abnormal human gallbladder bile at its physiological dose. These results support the working hypothesis that fibronectin is a possible pronucleating factor. Recent studies also demonstrated that the unidentified cholesterol nucleation promoting factor increase~ the size of the vesicles which results in the rapid nucleation time [30,31]. Based en the observations that there are significantly higher concentration of fibronectin and the faster nucleation time in abnormal bile than those in the

48 control, one would expect a significant correlation between nucleation time and endogenous fibronectin level in native bile. However, this was not observed in the p r e s e n t study. Similarly, a l t h o u g h m u c o u s g l y c o p r o t e i n is k n o w n to be a n u c l e a t i n g f a c t o r [15,16], t h e r e was also the lack o f c o r r e l a t i o n b e t w e e n t h e n u c l e a t i o n t i m e a n d t h e e n d o g e n o u s biliary m u c o u s g l y c o p r o t e i n c o n c e n t r a t i o n [32]. In c o n t r a s t to b i l e - m o d e l or a b n o r mal bile, f i b r o n e c t i n h a d n o e f f e c t on t h e n u c l e a t i o n t i m c o f c o n t r o l bile. T h i s c a n be e x p l a i n e d e i t h e r by t h e p r e s e n c e o f small a m o u n t s o f vesicles [33,34] o r by t h e p r e s e n c e o f a n t i n u c l e a t i n g f a c t o r in c o n t r o l bile

[35]. In conclusion, the results suggest that fibronectin in bile may be a pronucleating agent. The lack of the correlation between the nucleation time and endogeqous fibronectin and the lack of promoting activity of fibronectin in control bile might be ascribed to the overall balance between nucleation promoting and inhibiting activity [I4,35].

Acknowledgement The authors appreciate Professor Y. Shibata, Department of Anatomy, Kyushu University Faculty of Medicine for his kind preparation of freeze fracture replicas. They also appreciate colleagues, Research and Development Division, Otsuka Assay Laboratories, Tokushima, Japan for their kind help of the purification and determination of fibronectin. The authors appreciate Ms Chieko Tokunaga to prepare the manuscript.

References I Patlinson. N.R. (t985) FEBS Lett. 181,339-342. 2 Somjen, G.J. and Gilat, T. (1983) FEBS Lett. 156, 265-268. 3 Patti~lson, N.R. and Chapman, B.A. (1986) Gastroenterology 91, 6q7-702. 4 Somjen. G.J., Marikovsky, Y., Lelkes. P. and Gilal, T. (1986) Biochim. Biophys. Acta 879, 14-21. 5 Somjen, G.J. and Gilat, T. (1985) J. Lipid Res. 26, 699-704. 6 Lee, S,P., Park, H.Z., Madani, H. and Kaler, E.W. (1987) Am. j. Physiol. 252, G374--383. 7 ltalpern, Z., Dudley, M.A., Lynn, M.P., Nader, J.M., Breuer, A.C. and Holzbach. R.T. (1986) J. Lipid Res. 27, 2t95-306.

8 McLean, L,R. and Phillips, M.C. (19841 Biochim. Biophys. Acta. 776, 21-26. 9 Haipern, Z., Dudley, M.A., Kibe, A,, Lyre, M.P., Breuer, A.C., and -Holzbach, R.T. (19861 Gastroenterology 9(1, 875-885. II) Amigo, L., Covarrubias, C. and Nervi, F. (19901J. Lipid Res. 31, 341-347. 11 Sedaght, A. and Grundy, S.M. (19801 N. Engl. J. Med. 302, 1274-1277. 12 Holan, K.R., Holzbach, R.T Hermann, R.E., Cooperman, A.F. and Claffey, N.J. (19791 Gastroenterology 77, 611-617. 13 Burnstein, M.J,, llson, R.G., Petrunka, C,N., Taylor, R.D. and Strasberg, S.M. (I983) Gastroenterology 85, 801-807. 14 Groen, A.K., Stout, J.P.J., Drapers, J.A.G., Haek, F.J., Grijm, R. and Tytgai, G.N.J. (1988)Hepatology 8, 347-352. 15 Levy, P.F., Smith. B.F. and LaMont, J.T. (1984) Gastroenterology 87, 270-275. 16 Gallinger, S., Taylor, R.D., Harvey, P.R.C., Petrunka, C.N. and Strasberg, S.M. (1985) Gastroenterology 89, 648-658. I7 Rieder, H., Ramadori, G., Dienes, H.P. and Biisehenfelde, K.H.M. (1987) Hepatology 7(5), 856-864. 18 Tamkun. J.W. and Hynes. R.O. (1983) J. Biol. Chem. 258(7), 4641-4647. 19 Wilton, P.B., Dalmasso, A.P. and A!!en, M.O. (1987) J. Surg. Res. 42, 434-439. 20 Rossi, J.D. and Wallace, B.A. (1983) J, Biol. Chem. 258(5), 3327-3331. 21 Johansson, S., Forsberg, E. and Lundgren, B. (1987) J. Biol. Chem. 262, 7819-7824. 22 Chijiiwa, K., Kiyosawa, R. and Nakayama, F. (1988) Ciinica Chimiea Acta 178, 181-192. 23 Yamashita, N., Yanagisawa, J. and Nakayama, F. (1988) Dig. Dis. Set. 33, 449-453. 24 Mashige, F., Imai, K, and Osuga, T. (1976) Clin. Chim. Aeta. 70, 79. 25 Bartlett, G.R. (1958)J. Biol. Chem. 234, 466-468. 26 Chijiiwa, K (1987)Lipids 22, 121-125, 27 Chijiiwa, K. and Nakayama, F. (1988) J. Chromatogr. 431, 17-25. 28 Carey, M.C. (1978) 3, Lipid Res. 19, 945-955. 29 Shibata, Y., Izumi, T, and Yamamoto, T. (1987) J. Microse. 1448: 97-101. 30 Groen, A,K., O!tenhoff, R., Noordam, C., Zuyderhoudt, F.M,J., Hock, F.J. and Tytgat, G.N.J. 1987 Gastroenterology 92, 1737. 31 Groen, A.K., Ottenhoff, R., Jansen, P.L.M., Van Made, J., and Tytgat. (1989) J. Lipid Res. 30, 51-58. 32 Harvey, P,R.C., Rupar, C.A., Gailinger, S., Petrunka, C.N. and Strasberg, S.M. (1986) Gut 27, 374-381. 33 Harvey, P.R.C., Somjen, G., Gilat, T., Gallinger, S, and Strasberg, S.M. (1988) Bioehim, Biophys. Acta 958, 10-18. 34 Chijiiwa, K., Hirota, !., Noshiro, H., Yamasaki, T., Nakano, K. and Nakayama, F. (1991) Gastroenterology 100, A313. 35 Holzbaeh, R.T., Kibe, A., Thiel, E., Howell, J.H., Marsh, M. and Holan, K. (1984) J. Clin. invest. 73, 35-45.