Small and Large Unilamellar Vesicle Membranes as Model System for Bile Acid Diffusion in Hepatocytes

Archives of Biochemistry and Biophysics Vol. 368, No. 1, August 1, pp. 198 –206, 1999 Article ID abbi.1999.1295, available online at http://www.idealibrary.com on

Small and Large Unilamellar Vesicle Membranes as Model System for Bile Acid Diffusion in Hepatocytes Michael Hofmann, Dimitrios Zgouras, Panagiotis Samaras, Carsten Schumann, Karin Henzel, Guido Zimmer, and Ulrich Leuschner 1 Center of Internal Medicine, Medical Clinic II, Building 11, University Clinics, Theodor-Stern-Kai 7, 60590 Frankfurt/Main, Germany

Received April 21, 1999

Uptake of bile acids into the liver cell occurs via active transport or passive diffusion. In a model system, passive diffusion was studied in liposomes using pyranine fluorescence. Rate constants for the diffusion of diverse more polar or more apolar bile acids were examined. Hydrophobic lithocholic acid (LCA) revealed a maximal rate constant of 0.057 s 21 ; with the polar ursodeoxycholic acid (UDCA), the value was 0.019 s 21 . UDCA (3 mol%) effectively decreased the rate constant of 0.1 mM chenodeoxycholic acid (CDCA), whereas cholesterol reached a similar decrease only between 5 and 10 mol%. At higher concentrations of CDCA (above 1 mM) or LCA (0.3– 0.4 mM), breaking up of liposomal structure was confirmed by light-scattering decrease and increase of carboxyfluorescein fluorescence. Changes in lipid composition of phosphatidylcholine (PC)- small unilamellar vesicles (SUVs) or large unilamellar vesicles (LUVs) also caused decreasing rate constants. For a cardiolipin (CL):PC ratio of 1:20 the CDCA (0.1 mM) rate constant was 71% lower (0.015 s 21 ) and for a sphingomyelin (SM):PC ratio of 2:1 the rate constant was 50% lower (0.026 s 21 ). Changes in membrane fluidity were detected using membrane anisotropy measurements with the 1,6-diphenyl-1,3,5hexatriene (DPH) method. Membrane fluidity was reduced with cholesterol- but not with CL- or SMcontaining SUVs (ratio: cholesterol, CL, SM:PC of 1:5). This model system is currently used for the analysis of more complex lipid vesicles resembling the plasma/hepatocyte membrane, which is either stabilized or destabilized by appropriate conditions. The results should become clinically relevant. © 1999 Academic Press

1 To whom correspondence should be addressed. Fax: 0049-696301-6448.

198

Key Words: liposomes; SUV; LUV; bile acid diffusion; rate constant; sphingomyelin; cardiolipin; cholesterol; pyranine fluorescence; DPH anisotropy.

Cholesterol appears to have several functions in cells. It influences the passive permeability of ions, water, and other substances and exerts a stabilizing effect for biological membranes (1). The incorporation of cholesterol and ursodeoxycholic acid (UDCA) 2 into model membranes causes structural effects. The interaction of steroids with membrane lipids reduced the sensitivity of small unilamellar vesicles (SUVs) or large unilamellar vesicles (LUVs) against toxic damages (2, 3). The molar ratio cholesterol:phospholipid in rat liver cell plasma membranes is 0.83, and in erythrocytes it is 0.9. Physiologically cholesterol amounts to about 20 –30% of total lipid in plasma membranes of hepatocytes (4). Interactions between sterols and fatty acids appear to be affected by double bonds and side-chain groups (4, 5). Cholesterol is also known as membrane protective in phospholipid bilayers (6, 7). Increasing cholesterol in membranes reduces membrane fluidity above and increases it below the transition temperature. Using electron paramagnetic resonance spectroscopy and differential calorimetry our group has shown previously that in membranes UDCA acts similar to 2

Abbreviations used: UCDA, ursodeoxycholic acid; SUVs, small unilamellar vesicles; LUVs, large unilamellar vesicles; CDCA, chenodeoxycholic acid; LCA, lithocholic acid; EYL, egg yolk lecithin; DPH, 1,6-diphenyl-1,3,5-hexatriene; CF, 6-carboxyfluorescein; CL, cardiolipin; SM, sphingomyelin; TUDCA, tauroursodeoxycholic acid; GUDCA, glycoursodeoxycholic acid; GCA, glycocholic acid; glycochenodeoxycholic acid; HDCA, hyodeoxycholic acid; HCA, hyocholic acid; CA, cholic acid; DCA; deoxycholic acid. 0003-9861/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.

UNILAMELLAR VESICLES AS MODEL SYSTEM FOR BILE ACID DIFFUSION

cholesterol. UDCA stabilizes membranes against chemical or mechanical stress. Furthermore, there is protection against a solubilizing influence of apolar bile acids (7, 8). The aim of the present study was to investigate the specificity of bile acid treatment as far as the rate of diffusion of chenodeoxycholic acid (CDCA) into model membranes is concerned. In some cases the efficiency of the more apolar lithocholic acid (LCA) diffusion rates was compared to CDCA diffusion rates. The influence of varying the lipid components of model membranes toward CDCA diffusion rates was studied. Gu¨ldu¨tuna et al. showed that 14C-labeled UDCA was bound to the interface of the membrane and tauroursodeoxycholic acid (TUDCA), because of its taurine head, exhibited more polar interactions (8). We tried to corroborate these experiments in liposomal vesicles and to reveal how these effects correspond with the membrane composition. MATERIALS AND METHODS Materials. Phosphatidylcholine from EYL (egg yolk lecithin), DPH (1,6-diphenyl-1,3,5-hexatriene), CF (6-carboxyfluorescein), CL (cardiolipin, bovine heart, 98%), and SM (sphingomyelin, bovine brain, 99%; cholesterol, 99%) were purchased from Sigma (St. Louis, MO). Bile acids CDCA, LCA, UDCA, TUDCA (tauroursodeoxycholic acid), GUDCA (glycoursodeoxycholic acid), GCA (glycocholic acid), GCDCA (glycochenodeoxycholic acid), HDCA (hyodeoxycholic acid), HCA (hyocholic acid), CA (cholic acid), and DCA (deoxycholic acid) and Sephadex G-25 Fine were obtained from Calbiochem–Novabiochem (San Diego, CA); purity was over 95%. Pyranine (6-hydroxy1,3,8-pyrenetrisulfonic acid) was obtained from Lancester Synthesis Ltd. (Windham). Pyranine method. Kamp and Hamilton (9) have described the observation of fluorescence of the pH-sensitive fluorophore pyranine inside liposomes. Bile acids diffusing through the lipid layer decrease intraliposomal pH. The 8-hydroxy group of pyranine becomes protonated decreasing fluorescence intensity of pyranine. Alteration of fluorescence intensity vs time during bile acid diffusion results in rate constants of diffusion. According to this method rate constants for passive diffusion of bile acids were determined. Preparation of pyranine-containing liposomes. EYL liposomes [5 ml, 12 mg/ml lipid, 100 mM KOH/Hepes (N-[2-hydroxyethyl]-piperazine-N9[2-ethanesulfonic acid]), pH 7.4] with entrapped pyranine (0.5 mM) were prepared by sonication for 10 min at 40 W (Branson Sonifier B15P) according to the method described by Kamp and Hamilton (9). The lipid solution in chloroform was dried under vacuum for 16 h and stored under a nitrogen gas atmosphere. Untrapped pyranine, bile acids, sterols, and cholesterol were removed by gel filtration (Sephadex G-25 column). Liposomes were treated with bile acids at different concentrations and the change of fluorescence intensities (excitation, 460 nm; emission, 509 nm) was recorded over 25 s (5 s after start of measurements bile acids were added) at 25°C using a Shimadzu RF-5001 PC fluorescence spectrophotometer. The rate constant was calculated for a first-order reaction (rate constant [s 21] 5 ln(F 1) 2 ln(F 2)/t 2 2 t 1). The lipid (10) and cholesterol (11) concentrations were detected photometrically at 25°C. Ratios are always given by weight (w/w). The free bile acids and the bile acid sodium salts were dissolved in ethanol or tridistilled water at a concentration of 100 mM (stock solution).

199

Vesicle preparation. Two methods for preparation of liposomes are available: (i) sonication (SUVs) and (ii) using the extruder method (LUVETs). (i) SUVs were prepared by sonication (5 3 3 min, 40 W, Branson Sonifier B15P, 8°C). Sizes of SUVs are nearly 40 – 60 nm. (ii) LUVETs (large unilamellar vesicles by extrusion techniques) were prepared according to the extrusion method (12, 13). Depending on filter pores LUVETs resulted in a vesicle size of 100 to 400 nm (50 ml lipid solution, 10 –50 mg/ml lipid). Interestingly, lipids in liposomes, prepared by the extruder method, are arranged asymmetrically, resembling biological membranes where phospholipids as well as proteins are similarly arranged in an asymmetric way (14). For the determination of vesicle size, light-scattering experiments were carried out using an Hitachi UV/VIS spectrophotometer at 500 nm (25°C). Carboxyfluorescein method. Measurements with CF were done at 492 and 515 nm for excitation and emission to evaluate membrane leakage. Complete membrane disintegration was achieved with 1% Triton X-100 and under these conditions CF release was set to 100%. The CF-containing liposomes were prepared according to Liu et al. (15). CF (100 mM) was dissolved in Hepes buffer titrated with KOH to a pH of 7.4. Ten milligrams of lipid was mixed with 1 ml CF and sonicated (3 min, 40 W). Untrapped CF was removed by gel filtration. Fluorescence intensity was detected over 270 s (the baseline was detected for 30 s and subsequently the bile acids were added). CF release was calculated as CF release [%] 5 100 * (F 2 F i)/(F t 2 F i), where F is intensity after 270 s, F i is intensity after 30 s, and F t is 100% CF release [incubation with 10% (v/v) Triton X-100]. DPH method. Membrane fluidity was measured with DPH at wavelengths 360 nm for excitation and 430 nm for emission (16, 17). Using a Shimadzu polarization filter two fluorescence intensities, I 0 and I 90, were detected. Anisotropy r was calculated according to the formula r 5 I 0 2 I 90/(I 0 1 2 * I 90). Measurements were done at 25°C with 0.1 mg/ml lipid. The liposomes were incubated with 1 mM DPH for 3 min and the fluorescence intensities were recorded. Statistics. Statistical evaluations were carried out using Student’s t test. If not described otherwise, significance (P) was calculated for n 5 6.

RESULTS

Rate constants of the investigated bile acids in EYL liposomes (SUVs) are shown in Fig. 1. Rate constants refer to the time necessary for passive diffusion of the investigated bile acids from the buffer solution through the lipid bilayer (18). For 0.1 mM CDCA and for 0.1 mM LCA the rate constants are increased to values above 0.05 s 21, whereas UDCA, GCDCA, HDCA, and DCA reach values just above 0.02 s 21. Constants of GUDCA, TUDCA, GCA, and CA stay below 0.005 s 21. Experiments were carried out using all bile acids shown in Fig. 1. To be able to compare with previous work (7, 8), rate constants obtained with CDCA were mainly used in this paper. The effect of increasing amounts of cholesterol or UDCA in SUVs on the rate constant of CDCA diffusion is presented in Fig. 2. Similar to earlier studies (7) we were interested to see whether a cholesterol-like effect of UDCA could be found. Between concentrations of 3 and 20 mol% both steroid molecules significantly decreased the rate constants for 0.1 mM CDCA; however,

200

HOFMANN ET AL.

FIG. 1. Rate constants of all investigated bile acids in pure EYL (egg yolk lecithin) liposomes detected with pyranine fluorescence. This method estimates the rate of permeation of the different bile acids through the bilayer. Control means spontaneous fluorescence decrease without bile acids in pure liposomes containing pyranine (SUVs, 0.1 mM bile acid, 0.6 mg/ml lipid, Hepes buffer, 100 mM, pH 7.4, 25°C). SD values are shown. Significance values for all figures: *P , 0.05, **P , 0.01, ***P , 0.002.

UDCA acts more intensely than cholesterol, decreasing the value to about 0.02 s 21 at 20 mol%. Introducing cardiolipin at increasing ratios into EYL liposomes (LUVETs) revealed an even stronger influence on the rate constant of 0.1 mM CDCA: At a ratio of 1:10 6 CL/PC the rate constant was decreased to about 0.03 s 21 and at ratios of 1:5 or 1:3 (w/w) the rate constant was decreased to about 0.01 s 21 (Fig. 3a). Sphingomyelin, by contrast, became only efficient at a ratio of about 1:1 (Fig. 3b). Membrane fluidity measurements with DPH showed no change with CL-containing membranes [(CL/PC ratio: 1:5), r 5 0.074, P , 0.05 vs pure PC, n 5 6] and

sphingomyelin-containing membranes [(SM/PC ratio: 1:5), r 5 0.069, P , 0.05 vs pure PC, n 5 6]. This is in contrast to the known decrease in fluidity with PC/ cholesterol membranes [(cholesterol/PC ratio 1:5), r 5 0.12, P , 0.0001, n 5 6] (Fig. 4). Comparing the rate constants of CDCA and LCA exemplifies the different diffusional behaviors of these more apolar bile acids. Increasing concentrations of CDCA or LCA above 0.1 mM influenced the CDCA or LCA rate constant for their respective diffusions in SUVs in different ways: Clearly, the maximum rate constant of about 0.07– 0.08 s 21 is reached at 0.7 mM CDCA, whereas with LCA the maximum

FIG. 2. The rate constants of CDCA in EYL liposomes containing 0 –20% UDCA (black bars) or cholesterol (gray bars). UDCA and cholesterol decreased the rate constant (SUVs, 0.1 mM CDCA, Hepes buffer, 100 mM, pH 7.4, 25°C). SD values are shown. Significance values were calculated according to mol% cholesterol or UDCA.

UNILAMELLAR VESICLES AS MODEL SYSTEM FOR BILE ACID DIFFUSION

201

FIG. 3. (a) Cardiolipin-containing EYL liposomes (CL/PC ratio: 1:2 to 1:10 6) decreased the CDCA rate constant (LUVETs 100 nm, 0.1 mM CDCA, Hepes buffer, 100 mM, pH 7.4, 25°C). SD values are shown. (b) Sphingomyelin-containing EYL liposomes (SM/PC ratio: 1:2 to 3:1, LUVETs 100 nm, 0.1 mM CDCA, Hepes buffer, 100 mM, pH 7.4, 25°C) decreased the CDCA rate constant. SD values are shown.

of 0.06 s 21 is already attained at 0.1 mM (Fig. 5 and Fig. 6). Analogous experiments were carried out with UDCA, HDCA, and HCA (not shown). All bile acids revealed an increase up to a maximum and thereafter a decrease of values for the rate constants. Position of the maximums appeared characteristic for each bile acid [maximal values: UDCA (2 mM), 0.12s 21; HDCA (1.5 mM), 0.082 s 21; HCA (3 mM), 0.08 s 21]. The decrease of rate constants after the maximums in Figs. 5 and 6 can be explained by a disintegration of vesicles as revealed by light-scattering experiments. At a concentration of 1.0 mM CDCA, light-scattering intensity is significantly decreased; between 3 and 3.5 mM CDCA light-scattering intensity in the EYL liposomes approached zero (Fig. 7). Similary, release of the fluorochrome CF from vesicles (which is 100% in the presence of Triton X-100)

approaches 90% at 0.9 mM CDCA, reaching 100% at 1 mM (Fig. 8a). For LCA, 0.3– 0.4 mM concentration reach 90 –100% of the CF release value in the presence of Triton X-100 (Fig. 8b). Measurements of carboxyfluoresceine release cannot be directly compared with those experiments using pyranine fluorescence for determination of rate constants. Diffusion of bile acids is complete within less than 25 s, whereas the leakiness of liposomes for carboxyfluoresceine is measured over a period of 270 s. Therefore, bile acid diffusion is always determined in stable liposomes. Also in the experiments of Kamp and Hamilton (9), a pH gradient generated by addition of CA remained constant for minutes, showing that CA did not disrupt the vesicles and did not enhance proton leak. Finally, we investigated influences of pH and temperature on the rate constants for 0.1 mM CDCA com-

202

HOFMANN ET AL.

FIG. 4. Measurement of membrane fluidity (anisotropy) with DPH at pure (hatched bar) and CL- (light gray bar), SM- (dark gray bar), and cholesterol (black bar)-containing LUVETs. The CL/PC-, SM/PC-, and cholesterol/PC ratio was 1:5 (Hepes buffer, 100 mM, pH 7.4, 25°C). SD values are shown.

pared to the rate constants for 0.1 mM UDCA (Figs. 9a and 9b). Rate constants decreased with pH (5.4 –9.4) and increased with temperature (10 –37°C). Differences in CDCA/UDCA were maximal at physiological pH 7.4 and between 10 and 25°C. DISCUSSION

In this paper we have used the rate constant of CDCA diffusion into liposomes as a diagnostic tool for the investigation of the influence of lipid composition as well as of different bile acids on this parameter. In particular, we were interested to learn about the influence of more polar compared to more apolar bile

acids. The kinetics of bile acid movement is dependent on both number and position of hydroxyl groups; also, the overall hydrophobicity of the bile acid is important in determining its rate of transbilayer movement (18). In biliary liver disease an increase in the amount of apolar bile acids has been observed; their rate of bilayer penetration appears important. More polar bile acids have been experimentally and clinically used to diminish the deteriorating action of the more apolar bile acids in membranes. The rate constant provides us with an insight into resistance of the model membrane toward the penetrating power of the apolar bile acid under different experimental conditions.

FIG. 5. Concentration dependency of CDCA diffusion rates (0.1–1.5 mM) shown in SUVs. The maximum was detected at 0.7– 0.9 mM CDCA. Control means spontaneous fluorescence decrease without CDCA in pure liposomes containing pyranine (Hepes buffer, 100 mM, pH 7.4, 25°C). SD values are shown.

UNILAMELLAR VESICLES AS MODEL SYSTEM FOR BILE ACID DIFFUSION

203

FIG. 6. Concentration dependency of LCA diffusion rates (0.01–1 mM) shown in SUVs. The maximum was found at 0.1 mM LCA (Hepes buffer, 100 mM, pH 7.4, 25°C). Control means spontaneous fluorescence decrease without LCA in pure liposomes containing pyranine. SD values are shown.

The rate constant varied with the type of bile acids in the expected way: More polar bile acids (GCA, TUDCA, CA) revealed the lowest rate constants, while apolar bile acids (CDCA, LCA) revealed the highest. UDCA, GCDCA, HDCA, and DCA appeared in between. A characteristic concentration dependency of rate constants could be established for all investigated bile acids. Such behavior can be explained by different hydrophobicities and CMC (critical micellar concentration) values of the different bile acids (19, 20). Decrease of the rate constant after having reached the individual maximum can be explained by loss of pyranine due to membrane leakiness and subsequent solubilization. In the present work we have used the easily manageable method for preparation of SUVs. Moreover,

analogous results (Figs. 3a and 3b) were obtained with LUV (same as LUVETs) and SUV preparations. SUVs and LUVETs differ in dimension of vesicles: Preparation by sonication results in SUVs 40 – 60 nm in diameter. LUVETs prepared according to the extrusion method resulted in a vesicle size of 100 – 400 nm (12, 13). In our hands using SUVs, the results obtained with additions of CL- or SM-containing vesicles were irreproducible. Cardiolipin or sphingomyelin could, however, become accommodated by the extrusion technique (LUVETs). Attenuation of the rate constant increased by CDCA was found in the presence of UDCA (Fig. 2). This is of particular theoretical and clinical interest (8). Nowadays due to its considerable side effects, the therapeu-

FIG. 7. Light-scattering experiments with CDCA (0 – 4 mM) on EYL liposomes. Liposomal damage is complete at 3.5 mM CDCA (l 5 500 nm, SUVs, Hepes buffer, 100 mM, pH 7.4, 25°C). Control means light-scattering intensity in buffer without liposomes. SD values are shown.

204

HOFMANN ET AL.

FIG. 8. (a) 6-Carboxyfluorescein (CF) release with 0.1–1.0 mM CDCA. Complete liposomal damage was detected with 1% Triton X-100. This value was set to 100% CF release. Control means CF release without CDCA (SUVs, Hepes buffer, 100 mM, pH 7.4, 25°C). SD values are shown. (b) CF release with 0.01– 0.4 mM LCA. Control means CF release without LCA (SUVs, Hepes buffer, 100 mM, pH 7.4, 25°C). Otherwise, see a. SD values are shown.

tic use of more apolar bile acids, such as CDCA, is nearly abandoned (21, 22). UDCA is very much preferred for dissolution of gallstones as well as for treatment of primary biliary cirrhosis, due to its polarity and much fewer side effects (23, 24). On the other hand, it is interesting that the difference between CDCA and UDCA is exclusively due to the a- or b-position of the 7-OH groups. Recent infrared and Raman spectroscopic investigations allow a better understanding of the structural consequences in solution (25). In apolar bile acids such as CDCA, 3a-OH and 7b-OH form stable intermolecular interactions. Polar bile acids such as UDCA instead in general have their carboxyl group involved in stable dimeric association (25).

Therefore, two CDCA molecules with their inflexible sterol cores sandwiched to each other by means of their 3- and 7-a-OHs reaching at least double thickness will— unlike cholesterol—perturb and interrupt hydrophobic interactions between lipid alkyl chains. By contrast, the firm association between two carboxyl– carbonyl regions in the case of two UDCA molecules will allow better accommodation in the membrane. This dimer is flexible due to the sterically unrestricted rotatory competence between C 18 and C 20. Similar to cholesterol, the 3a- and 7b-OHs will interact with the phospholipid phosphate region or the membrane interface. In addition to its diverse structural consequences during interaction with phospholipids, higher concentrations of cholesterol contribute to rearrangement of

UNILAMELLAR VESICLES AS MODEL SYSTEM FOR BILE ACID DIFFUSION

205

FIG. 9. (a) pH dependency of bile acid diffusion of 0.1 mM CDCA (light gray bars) and 0.1 mM UDCA (dark gray bars) at pH 5.4, 7.4, and 9.4. Control means spontaneous fluorescence decrease in pyranine-containing liposomes without addition of CDCA or UDCA for all pH values (SUVs, Hepes buffer, 100 mM, pH 5.4, 7.4, and 9.4, 25°C). SD values are shown. (b) Temperature dependency of bile acid diffusion of 0.1 mM CDCA (light gray bars) and 0.1 mM UDCA (dark gray bars) at 10, 25, and 37°C. Control means spontaneous fluorescence decrease in pyranine-containing liposomes without addition of CDCA or UDCA for all temperatures (SUVs, Hepes buffer, 100 mM, pH 7.4, 10, 25, and 30°C). SD values are shown.

phosphatidylethanolamine, which changes its position in the membrane from inside to outside (26). Due to similar efficiency in the plasma membrane, UDCA might cause analogous changes in phospholipid arrangement. This, however, cannot occur in our present system, which is based on phosphatidylcholine-containing liposomes. Consequently, the rate constant of CDCA in EYL liposomes decreases in a linear way with increasing cholesterol or UDCA concentrations (Fig. 2). Otherwise, cholesterol–lipid interactions are multifold (27). Sensitivity of membranes in general against “toxicity” of hydrophobic bile acids depends on their cholesterol content (8, 28). A previous investigation (7) has revealed that UDCA in addition to stabilization of phospholipid-rich membranes mimics the effect of cholesterol in LUVETs.

Membrane stabilization of SUVs by UDCA against CDCA toxicity was found to be even more intense than that by cholesterol. On the other side, membrane stabilization by means of addition of cardiolipin is extremely efficient against CDCA’s toxicity, reducing the rate constant by about 70% at a cardiolipin/PC ratio of 1:20 to 1:30 (w/w). A similar effect is obtained only at 20 mol% UDCA. Negative charge of the cardiolipin polar head may prohibit the occurrence in plasma membranes (4) but in tiny amounts (1%). In particular for hepatocytes, rate diffusion of bile acids depends on their concentration at the membrane surface. A surplus of negative charges will hinder accumulation of bile acids at the surface, which slows the diffusion process (29).

206

HOFMANN ET AL.

The method described here may allow one to judge membrane stability on its lipid constituents as well as the membrane-labilizing/stabilizing power of bile acids. REFERENCES 1. Demel, R. A., and De Kruijff, B. (1976) Biochim. Biophys. Acta 457, 109 –132. 2. Schubert, R., and Schmidt, K.-H. (1988) Biochemistry 27, 8787– 8794. 3. Needham, D., and Nunn, R. S. (1990) Biophys. J. 58, 997–1009. 4. Dorling, P. R., and Le Page, R. N. (1973) Biochim. Biophys. Acta 318, 33– 40. 5. Demel, R. A., Bruckdorfer, K. R., and van Deenen, L. L. (1972) Biochim. Biophys. Acta 255, 311–320. 6. Parasassi, T., Giusti, A. M., Raimondi, M., Ravagnan, G., Sapora, O., and Gratton, E. (1995) Free Radical Biol. Med. 4, 511–516. 7. Gu¨ldu¨tuna S., Deisinger, B., Weiss, A., Freisleben, H.-J., Zimmer, G., Sipos, P., and Leuschner, U. (1997) Biochim. Biophys. Acta 1326, 265–274 8. Gu¨ldu¨tuna, S., Zimmer, G., Imhof, M., Bhatti, S., Tiangeng, Y., and Leuschner, U. (1993) Gastroenterology 104, 1736 –1744. 9. Kamp, F., and Hamilton, A. (1993) Biochemistry 32, 11074 – 11086. 10. Takayama, M., Itch, S., Nakasaki, T., and Tanimuru, I. (1977) Clin. Chim. Acta 79, 93–98. 11. Siedel, J., Ha¨gele, E. O., Ziegenhorn, J., and Wahlefeld, A. W. (1983) Clin. Chem. 29, 1075–1080. 12. Hope, M. J., Bally, M. B., Webb, G., and Cullis, P. R. (1985) Biochim. Biophys. Acta 812, 55– 65. 13. Mayer, L. D., Hope, M. J., and Cullis, P. R. (1986) Biochim. Biophys. Acta 858, 161–168.

14. Roy, M. T., Gallardo, M., and Esterich, J. (1997) Bioconj. Chem. 8, 941–945. 15. Liu, Z., Solo, R., and Hu, V. W. (1988) Biochim. Biophys. Acta 945, 253–262. 16. Shinitzky, M., and Inbar, M. (1976) Biochim. Biophys. Acta 433, 133–149. 17. Shinitzky, M., and Barenholz, Y. (1978) Biochim. Biophys. Acta 515, 367–394. 18. Cabral, D. J., Small, D. M., Lilly, H. S., and Hamilton, J. A. (1987) Biochemistry 26, 1801–1804. 19. Lamcharfi, L., Meyer, C., and Lutton, C. (1997) Biospectroscopy 3, 393– 401. 20. Lasch, J. (1995) Biochim. Biophys. Acta 1241, 269 –293. 21. Iser, J. H., and Sali, A. (1981) Drugs 21, 90 –119. 22. Leuschner, U., Leuschner, M., Sieratzki, J., Kurtz, W., and Hubner, K. (1985) Dig. Dis. Sci. 30, 642– 649. 23. Lim, A. G., Jazrawi, R. P., and Northfield, T. C. (1995) Gut 37, 301–304. 24. Brodanova, M., and Perlik, F. (1997) Casopis Lekaru Ceskych 136, 215–220. [English abstract] 25. Lamcharfi, E., Cohen-Solal, Parquet, M., Lutton, Dupre´, J., and Meyer, C. (1997) Eur. Biophys. J. 25, 285–291. 26. Boesze-Battaglia, K., and Schimmel, R. (1997) J. Exp. Biol. 200, 2927–2936. 27. McMullen, T. P. W., Lewis, R. N. A. H., and McElhaney, N. (1993) Biochemistry 32, 516 –522. 28. van de Heijning, B. J. M., van den Broek, A. M. W. C., and Berge-Henegouwen, G. P. (1997) Eur. J. Gastroenterol. Hepatol. 9, 473– 479. 29. Shibata, A., Ikawa, K., Shimooka, T., and Terada, H. (1994) Biochim. Biophys. Acta 1192, 71–78.