HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) All Versions of this Article: M500144-JLR200v1 46/11/2325    most recent Alert me when this article is cited Alert me if a correction is posted Citation Map Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Pütz, G. Articles by Blum, H. E. Search for Related Content PubMed PubMed Citation Articles by Pütz, G. Articles by Blum, H. E. Social Bookmarking                      What's this?

Journal of Lipid Research, Vol. 46, 2325-2338, November 2005
Copyright © 2005 by American Society for Biochemistry and Molecular Biology

Synthesis of phospholipid-conjugated bile salts and interaction of bile salt-coated liposomes with cultured hepatocytes¹

G. Pütz^2,*, W. Schmider, R. Nitschke, G. Kurz^*,1 and H. E. Blum^*

^* University Medical Clinic Freiburg, D-79106 Freiburg, Germany
Aventis Pharma Germany GmbH, Sanofi-Aventis Group, D-65929 Frankfurt/M, Germany
Life Imaging Center, Institute of Biology I, University of Freiburg, D-79104 Freiburg, Germany

¹ This paper is dedicated to the memory of Prof. Dr. G. Kurz, deceased May 29, 2002.

Published, JLR Papers in Press, September 8, 2005. DOI 10.1194/jlr.M500144-JLR200

² To whom correspondence should be addressed. e-mail: puetz{at}medizin.ukl.uni-freiburg.de

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
To examine the possibility of targeting liposomes to hepatocytes via bile salts, the bile salt lithocholyltaurine was covalently linked to a phospholipid. The isomeric compounds disodium 3-(2-(1,2-O-distearoyl-sn-glycero-3-phospho-2'-ethanolamidosuccinyloxy)ethoxy)-5ß-cholan-24-oyl-2'-aminoethansulfonate and disodium 3ß-(2-(1,2-O-distearoyl-sn-glycero-3-phospho-2'-ethanolamidosuccinyloxy)ethoxy-5ß-cholan-24-oyl-2'-aminoethansulfonate (DSPE-3ß-LCT) were synthesized and incorporated into liposomal membranes. Confocal laser scanning microscopy studies showed that bile salt-bearing liposomes (BSLs) attach to the surface of rat hepatocytes in culture. Studies with radioactively labeled liposomes revealed that the bile salt linked via the 3ß-conformation resulted in a higher attachment efficiency than that with the 3-derivative. In the presence of BSLs corresponding to 2 mM liposomal phosphatidylcholine, uptake of 50 µM cholyltaurine (CT) into hepatocytes was reduced by 40% by the 3ß-derivative and by 17% by the 3-derivative. When added simultaneously with the liposomes, CT up to 75 µM inhibited the binding of DSPE-3ß-LCT-bearing liposomes. By contrast, increasing concentrations reversed this inhibition and resulted in an increased bile salt-mediated binding. The same was true when CT was added 10 min before the liposomes were added.

The attachment of BSLs to the surface of hepatocytes opens up promising possibilities for hepatocyte-specific drug delivery. More generally, not only substrates for cellular endocytosing receptors but also substrates for cellular carrier proteins should be suitable ligands for the cell-specific targeting of nanoscale particles such as liposomes.

Abbreviations: BSL, bile salt-bearing liposome; CH, cholesterol; COL, control liposome; CT, cholyltaurine; DCCI, N,N'-dicyclohexylcarbodiimide; DPPC, 1,2-O-dipalmitoyl-sn-glycero-3-phosphocholine; DPPS, 1,2-O-dipalmitoyl-sn-glycero-3-phosphoserine; DSPE, 1,2-O-distearoyl-sn-glycero-3-phosphoethanolamine; DSPE-3-LCT, disodium 3-(2-(1,2-O-distearoyl-sn-glycero-3-phospho-2'-ethanolamidosuccinyloxy)ethoxy)-5ß-cholan-24-oyl-2'-aminoethansulfonate; DSPE-3-LCT, disodium 3-(2-(1,2-O-distearoyl-sn-glycero-3-phospho-2'-ethanolamidosuccinyloxy)ethoxy)-5ß-cholan-24-oyl-2'-aminoethansulfonate; DSPE-3ß-LCT, disodium 3ß-(2-(1,2-O-distearoyl-sn-glycero-3-phospho-2'-ethanolamidosuccinyloxy) ethoxy-5ß-cholan-24-oyl-2'-aminoethansulfonate; DSPG, 1,2-O-distearoyl-sn-glycero-3-phospho-rac-(1-glycerol); LSM, laser scanning microscopy; NBD, 4-nitrobenzo-2-oxa-1, 3-diazol; MP, melting point; MS, mass spectrum; Ntcp, Na⁺/cholyltaurine-cotransporting polypeptide; PC, phosphatidylcholine; Ph₂-DiOC18, 3,3'-dioctadecyl-5,5'-diphenyloxacarbocyanine chloride; PSL, phosphatidylserine-bearing liposome; RT, room temperature; SS, solvent system

Supplementary key words hepatic drug delivery • bile salt carrier • modified liposomes • hepatocyte-specific drug targeting

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Hepatocyte pathology is central to many severe liver diseases. For optimal treatment and minimal adverse effects, drug concentrations in the target cell should be high while the systemic exposure to the drug should be as low as possible. For hepatocyte-specific drug delivery, hepatotropic liposomes represent an attractive strategy. Such hepatotropic liposomes need a signal on their surface targeting them selectively and specifically to hepatocytes. Hepatocyte-specific targeting of liposomes has been achieved by covalently linking the ligand of a hepatocyte-specific receptor to a lipid moiety. The most prominent targeted receptor is the hepatic asialoglycoprotein receptor, which has been targeted by various galactosyllipids (1, 2) or by asialofetuin, a glycoprotein (3). Glycyrrhizin (4) and a soybean-derived sterylglucoside mixture (5) have been used to target hepatocytes via receptors that have not yet been identified at the molecular level. To date, all of these receptors lead to the endocytosis of bound liposomes.

After targeting the liposomes to endocytosing receptors, the physiological path of internalization usually ends up in the lysosomes, where most of the liposomal content is degraded. To avoid this problem, the ligand used for homing of the liposomes should target a cellular receptor that mediates only the binding of the liposomes to the hepatocytes but not endocytotic internalization. Binding of liposomes to the cells without endocytosis may allow the delivery of the liposomal content into the cytoplasm by diffusion over adjacent membranes (6) or by direct fusion of the liposomal membrane with the sinusoidal membrane of the hepatocytes. Cellular carrier proteins are one group of proteins that may act as receptors for modified liposomes without mediating the endocytosis of the bound liposomes.

A main problem of liposomes with a homing moiety is the immunogenicity of nonphysiological ligands. Any immune reaction to surface-modified liposomes does not allow their repeated administration. Linking a physiological substrate of a cellular carrier protein to the liposomal surface should avoid an immune response while introducing an appropriate homing device. To evaluate the concept of using a physiological ligand of hepatocyte-specific carrier proteins to target liposomes to hepatocytes, lithocholic acid was used targeting the hepatic bile salt transport system.

Bile salts are removed from the bloodstream specifically by the hepatocytes, which use certain well-defined transport systems (7–9). The bile salt uptake into hepatocytes is very effective and characterized by half-saturation transport constants in the micromolar range (10, 11). By covalently coupling a drug to a bile salt, the bile salt-transporting systems have been already used for hepatic drug delivery. Bile salt-drug conjugates are readily taken up by the hepatocytes, but uptake is followed by a fast excretion of the drug conjugate into the bile (12, 13).

Bile salt carrier systems have not been addressed as targets for liposome-based drug delivery. The half-saturation transport constant of a carrier protein does not equal its affinity for the substrate, and even though bile salt transport is very effective, it had to be proven whether bile salts linked to the liposomal membrane may allow an effective binding and anchoring of a huge particle like a bile salt-bearing liposome (BSL) to the hepatocytes. In this study, we addressed the interaction of liposomes bearing a covalently linked bile salt-lipid conjugate with cultured hepatocytes and the question of whether a substrate of a cellular carrier may be useful as a ligand for particle-based drug delivery.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Animals
Male Wistar S 300 rats (Interfauna, Tuttlingen, Germany) weighing 200–300 g were used in all experiments. They had free access to tap water and the standard rat diet Altromin 300^® (Altromin GmbH, Lage, Germany) and were housed under a constant temperature and an alternating 12/12 h light/dark cycle.

Materials
Phospholipids were obtained from Sygena, Ltd. (Liestal, Switzerland). Cholyltaurine (CT), N,N'-dicyclohexylcarbodiimide (DCCI), lithocholic acid, cholesterol (CH), Williams E medium, and supplements were from Sigma Chemie GmbH (Deisenhofen, Germany). N-(7-nitrobenzo-2-oxa-1,3-diazol-4-yl)-1,2-dihexadecanoyl-sn-glycero-3-phosphoethanolamine (triethylammonium salt), 3,3'-dioctadecyl-5,5'-diphenyloxacarbocyanine chloride (Ph₂-DiOC18), and calcein-blue-AM were from Molecular Probes (Leiden, The Netherlands). Fetal calf serum was from PAA Laboratories GmbH (Linz, Austria). Collagenase Worthington CLS II (specific activity of 220–230 U/mg protein) was obtained from Biochrom KG (Berlin, Germany). [1,2-³H]cholesterylhexadecylether (1.9 TBq/mmol) was from NEN Life Science Products GmbH (Frankfurt, Germany), and [³H-(G)]CT was from Biotrend Chemikalien GmbH (Cologne, Germany). All other chemicals were of the highest quality available commercially.

Determination of protein concentration and radioactivity
Protein content of cultured hepatocytes was determined by a modified Lowry procedure (14) using the DC protein determination assay (Bio-Rad Life Science Group, München, Germany) and BSA, dissolved in the solution used for dissolving the cells, as a standard. Radioactivity was determined with a Wallac 1411 (Berthold, Wildbad, Germany) liquid scintillation counter using Ultima Gold^® (Canberra Packard GmbH, Frankfurt, Germany) as the liquid scintillation cocktail.

Preparation and characterization of liposomes
The liposomes were prepared by the extrusion method (15, 16) in the medium used for isolation of hepatocytes, omitting CaCl₂, D-glucose, and saturation with carbogen. The extrusion device (LiposoFast^TM; Milsch-Equipment, Laudenbach, Germany) used was equipped with polycarbonate filters (Milsch-Equipment); the samples were first extruded 21 times through one filter with a pore diameter of 200 nm and subsequently 21 times through two stacked filters having a pore diameter of 100 nm. BSLs were composed of 26% 1,2-O-dipalmitoyl-sn-glycero-3-phosphocholine (DPPC), 25% POPC, 40% CH, 8% 1,2-O-distearoyl-sn-glycero-3-phospho-rac-(1-glycerol) (DSPG), and 1% disodium 3-(2-(1,2-O-distearoyl-sn-glycero-3-phospho-2'-ethanolamidosuccinyloxy)ethoxy)-5ß-cholan-24-oyl-2'-aminoethansulfonate (DSPE-3-LCT) or disodium 3ß-(2-(1,2-O-distearoyl-sn-glycero-3-phospho-2'-ethanolamidosuccinyloxy) ethoxy-5ß-cholan-24-oyl-2'-aminoethansulfonate (DSPE-3ß-LCT). Control liposomes (COLs) were composed of 25% DPPC, 25% POPC, 40% CH, and 10% DSPG, and phosphatidylserine-bearing liposomes (PSLs) were composed of 17.5% DPPC, 17.5% POPC, 40% CH, and 25% 1,2-O-dipalmitoyl-sn-glycero-3-phosphoserine (DPPS) (given as molar ratios). As radioactive label, [1,2-³H]cholesterylhexadecylether was used. Fluorescence-labeled liposomes were prepared by adding 0.3% Ph₂-DiOC18. Freshly prepared liposomes were stored at 4°C and used within 7 days. The size distributions and the zeta potentials of the liposomes were determined with a Zetasizer III (Malvern, UK). The phosphatidylcholine (PC) concentration in the liposomal suspension was measured by colorimetric determination using ammonium ferrothiocyanate (17). The CH concentration was controlled enzymatically (18, 19) using the CH reagent kit INFINITY^TM (Sigma Diagnostics).

To evaluate the lamellarity of the liposomes, the amount of encapsulated lipid was estimated with liposomes containing 0.1 mol% 4-nitrobenzo-2-oxa-1, 3-diazol (NBD)-1,2-O-dipalmitoyl-sn-glycero-3-phosphoethanolamine by determining the decrease of NBD fluorescence attributable to reduction with dithionite (20). Fluorescence was measured with a Perkin-Elmer LS 50B fluorescence spectrophotometer (Perkin-Elmer, Überlingen, Germany) using 3 cm fluorescence cuvettes equipped with a stirring bar. The fluorescence intensities were measured at 540 nm using excitation at 470 nm and a slit width of 10 nm and were corrected for autobleaching. In all experiments, 20 µl of a 1 M solution of dithionite in 1 M Tris-HCl buffer, pH 10.0, was added to 2 ml of the appropriate liposomal suspension kept at 25°C and stirred at a constant rate of 750 rpm. After the decay in intensity had stopped, 10 µl of 400 µM cholate was added. The amount of encapsulated lipid is given by the difference between the lipid in the liposomal membrane (F_membrane) and the total amount of liposomal lipid (F_total). At a liposomal diameter of 122 or 118 nm, the ratio of lipid in the outer and inner membrane leaflet of the liposomal membrane is 1:0.85. The parameters F_total and F_outside are measured (Fig. 3), and the amount of encapsulated lipid is calculated by the following equation:

View larger version (16K): [in this window] [in a new window]   Fig. 3. Measurement of encapsulated lipid. The graph shows a typical measurement of DSPE-3-LCT-bearing liposomes. DSPE-3-LCT-bearing liposomes, DSPE-3ß-LCT-bearing liposomes, and control liposomes (COLs) showed the same general behavior. After the addition of sodium-dithionite (1), the fluorescent lipid in the outer membrane layer was reduced and the respective fluorescence disappeared (Foutside). After the addition of cholate to the liposomes (2), the liposomal membrane became leaky and the dithionite was able to reduce the fluorescent lipid inside the liposomes (Ftotal). NBD, 4-nitrobenzo-2-oxa-1,3-diazol.

Determination of binding of CT to liposomes
The binding of CT to the liposomal membrane was determined as described (21). In Ultra-Clear^TM centrifugation tubes (Beckman Instruments, Inc., Palo Alto, CA), 2 ml of a liposomal suspension containing 2 mM liposomal PC and 50 µM [³H]CT was centrifuged at 125,000 g for 2 h. Free CT was determined in the clear supernatant, and bound CT was calculated by subtracting the amount of free CT from the total amount of CT found with the concentrated liposomes.

Isolation and primary culture of rat hepatocytes
Hepatocytes from rat liver were prepared by a two-step procedure of the collagenase perfusion method (22, 23) using a standard medium containing 118 mM NaCl, 4.74 mM KCl, 1.2 mM MgCl₂, 0.59 mM KH₂PO₄, 0.59 mM Na₂HPO₄, 10 mM HEPES, 1.25 mM CaCl₂, and 5.5 mM D-glucose, which was saturated with carbogen (95% O₂/5% CO₂) and adjusted to pH 7.4. Cell viability before cultivation was 90%, as determined by trypan blue exclusion.

Hepatocyte cultures used for binding and uptake studies were prepared on tissue culture six-well plates; the wells (35 mm) were precoated with collagen type I (Biocoat^®; Becton Dickinson Labware, Bedford, UK). Cell cultures used for confocal laser microscopic studies were cultivated on collagen-coated glass cover slips with a diameter of 30 mm. To replace the medium used for cell preparation by culture medium I, a modified Williams E medium without phenolsulfonphthalein and containing 10% fetal calf serum, 1 µM dexamethasone, 25 U/l bovine insulin, 200 U/l penicillin, and 0.2 g/l streptomycin was used. The cells were collected by centrifugation at 35 g for 2 min and resuspended in medium I. The centrifugation was repeated, and the number of hepatocytes in medium I was adjusted to 1.2–1.5 x 10⁵ cells per milliliter. Two milliliters of the hepatocyte suspension was transferred into each well, and the cells were allowed to attach to the collagen layer in a humidified incubator at 37°C in an atmosphere of 95% air and 5% CO₂. After 4 h, medium I was replaced by medium II, a phenolsulfonephthalein-free Williams E medium supplemented with 1 µM dexamethasone, removing unattached and damaged cells. Dexamethasone was used to stimulate bile salt transport activity in primary hepatocyte culture (24). After 24 h of cultivation, the confluence of the hepatocyte monolayers was controlled by light microscopy and the viability of the cells was determined. Only cultures with confluence 90% and viability 95% were used for the experiments.

Binding of liposomes to primary rat hepatocytes
Binding of liposomes to primary cultured hepatocytes was studied in the same standard medium that was used for the preparation of freshly isolated hepatocytes. To compensate for the missing amount of D-glucose and CaCl₂ in the liposomal stock solutions, the appropriate amount of a solution containing 125 mM CaCl₂ and 550 mM D-glucose was added. The medium used to wash the culture wells and the liposomal suspensions prepared for incubation were kept at 37°C. After 24 h in culture, the cells were washed three times with 2 ml of medium. After removing the medium as completely as possible, 800 µl of the respective liposomal suspension was added and the cells were incubated for 5 min at 37°C with gentle shaking at 100 rotations per minute in a water bath. Subsequently, the liposomal suspension was removed completely and the cells were cautiously washed three times with 2 ml of medium. The cells and the attached liposomes were dissolved in 5% (w/v) SDS and 1% (v/v) Triton X-100 in 0.2 N NaOH (1 ml/well) by gentle mixing and avoiding foaming. Radioactivity and protein content in the wells were determined as described.

The effect of CT on the attachment of liposomes to the cultured hepatocytes was analyzed under exactly the same experimental conditions, except that CT at different concentrations was present in both the liposomal suspensions used for the incubation of the cells and the medium used for the washing procedure. In some experiments, cells were kept for 10 min in standard medium containing 100 µM CT before adding liposomal suspensions containing 2 mM liposomal PC and 100 µM CT.

Inhibition of CT uptake into cultured primary rat hepatocytes by liposomes
Uptake of CT into cultured primary hepatocytes was compared in the absence and presence of liposomal suspensions corresponding to 2 mM PC. Uptake was measured from a 50 µM CT solution containing 10 kBq [³H]CT/ml. The culture plates were kept at 37°C in a water bath at 100 rpm. Uptake was started by the addition of 800 µl of the corresponding [³H]CT-containing solutions and stopped after 10, 20, 30, 40, 50, or 60 s by quickly removing the supernatants and quickly washing the cells three times with 1.25 ml of ice-cold medium. The cells were then processed as described above.

All uptake experiments were performed in duplicate. Initial uptake rates were calculated by linear regression. For each individual experiment, the uptake rate of the liposome-free solution was determined and the inhibition of the uptake of CT by the different kinds of liposomes was calculated by dividing the measured uptake rates in the presence of liposomes by the uptake rate of CT in the absence of liposomes.

Confocal laser scanning microscopy
Fluorescence images were recorded with a confocal laser scanning microscope (LSM 510 UV; Zeiss, Jena, Germany) equipped with a galvanometer-driven scanning stage for fast focusing in the z direction (HRZ-200; Zeiss). The scanning speed was set to a pixel time of 2.24 µs. The usual pixel size, depending on the optical magnification, was 0.2–0.4 µm in the x and y directions. The z-focus step size was usually set to values between 0.05 and 0.15 µm. The confocal pinhole was set to achieve a full-width, half-maximum z resolution between 0.5 and 1 µm. Ph₂-DiOC was excited at 488 nm with an argon laser. Fluorescence emission signals >505 nm were collected. At the given time points after incubation with fluorescence-labeled liposomes, images were acquired sequentially, whereas the differential interference contrast images were taken simultaneously with 488 nm excitation and a water-immersion 40x objective (C-APO 40/1.2 water; Zeiss).

Model calculations
A liposome with a diameter of 122 nm has a surface of 4.9 x 10^–14 m². As a membrane is 4.5 nm thick, the surface of the inner membrane leaflet is 4.2 x 10^–14 m². Assuming the average space of a lipid molecule to be 44 x 10^–20 m², as determined for a 6:4 mixture of DPPC/CH (25), each liposome is made up of 2.1 x 10⁵ lipid molecules and has a mass of 1.4 x 10⁸ Da. Thus, each liposome consists of 1.7 x 10^–19 mol of PC, or 1 nmol of liposomal PC is equivalent to 5.9 x 10⁹ liposomes. As 1 g of hepatocyte protein equals 5 x 10⁵ cells, 1 nmol liposomal PC/mg protein equals 12,000 liposomes/cell. As 1 mol% of BSLs are disodium 3-(2-(1,2-O-distearoyl-sn-glycero-3-phospho-2'-ethanolamidosuccinyloxy)ethoxy)-5ß-cholan-24-oyl-2'-aminoethansulfonate (DSPE-3-LCT), the surface area of each DSPE-3-LCT is 4.3 x 10^–17 m² or 43 nm². As 78.5% of a surface may be covered by circles, each DSPE-LCT is surrounded by a circle of 33.8 nm², having a radius of 3.3 nm. Thus, the average distance between two DSPE-3-LCT molecules on the liposomal surface is 6.6 nm. With this distance between two molecules, the corresponding concentration of bile salts on the liposomal surface can be calculated. As spheres in close packing cover a maximum of 74% of total volume, each molecule in solution is surrounded by a total volume of 2.0 x 10^–25 m³ or 2.0 x 10^–22 liters. Thus, 5 x 10²¹ molecules, having a distance of 6.6 nm between each other, are found in a volume of 1 liter, corresponding to a concentration of bile salts on the liposomal surface of 8 mM.

Analysis of organic compounds
Melting points (MPs) were determined with a hot-stage apparatus (Büchi, Flawil, Switzerland) and are uncorrected. ¹H-NMR spectra and ¹³C-NMR spectra were recorded on a Bruker AM 400 spectrometer (Bruker GmbH, Karlsruhe, Germany) with tetramethylsilane as an internal standard. Chemical shifts are reported as ppm (), and signals are expressed as s (singlet), d (doublet), t (triplet), q (quartet), m (multiplet), or br (broad). Mass spectra were determined with a TSQ 7000 ESI (Finnigan, Sunnyvale, CA). The compounds were ionized by electron spray ionization at a temperature of 200–250°C and a voltage of 4–5 kV. Nitrogen at a pressure of 30 pounds per square inch was used as inert gas. Bile salt derivatives were visualized after TLC on silica-coated 60 F₂₅₄ plates (E. Merck, Darmstadt, Germany) by spraying with concentrated H₂SO₄ containing 1% (w/v) vanillin followed by heating at 100°C for 5 min (26), and phospholipids were visualized by spraying with a solution of 25 N H₂SO₄ containing molybden salts (27).

Syntheses
Preparative column chromatography was carried out under hydrostatic pressure on 20 x 5 cm columns of silica gel 60, 230–400 mesh (ICN Biomedicals, Meckenheim, Germany), if not stated otherwise. Solvent systems (SSs) for chromatographic separations were as follows: SS 1, cyclohexane-ethyl acetate (30:1, v/v); SS 2, cyclohexane-ethyl acetate (5:1, v/v); SS 3, cyclohexane-ethyl acetate-acetic acid (80:20:1, v/v/v); SS 4, cyclohexane-ethyl acetate-acetic acid (70:30:1, v/v/v); SS 5, cyclohexane-ethyl acetate-acetic acid (5:15:1, v/v/v); SS 6, chloroform-methanol (3:1, v/v); SS 7, chloroform-methanol (2:1, v/v); SS 8, chloroform-methanol-water (10:10:0.1, v/v/v); SS 9, chloroform-methanol-water (9:3:0.1, v/v/v); SS 10, chloroform-methanol-water (8:2:0.1, v/v/v). If chloroform-methanol mixtures were used as solvents for chromatography, the isolated products were separated from eluted silica gel by adsorption chromatography on a 30 x 3 cm column of Serdolit^® PAD-1 (Serva Feinbiochemica, Heidelberg, Germany) using 50 g of Serdolit^® PAD-1 per gram of product. The silica gel-containing product was suspended in warm water (250 ml/g), and the suspension was passed through the column at 1 ml/min. The column was washed twice with the same volume of water, then the product was eluted with methanol.

3-Hydroxy-5ß-cholan-24-oic acid methyl ester (1) For esterification, 14 g of 3-lithocholic acid was dissolved in 150 ml of dried methanol and 70 µl of concentrated HCl was added. After heating the reaction mixture for 4 h under reflux, the solvent and the HCl were removed under reduced pressure. The resulting crude product was purified by recrystallization in methanol-water yielding 13.6 g (34.8 mmol, 94% yield) of compound (1). MP, 115°C; TLC: relative mobility (R_f) = 0.18 (SS 3), R_f = 0.66 (SS 5); ¹H-NMR (CDCl₃): 0.68 (s, 3H, CH₃-18); 0.91 (d, J = 7 Hz, 3H, CH₃-21); 0.92 (s, 3H, CH₃-19); 3.64 (m b, CH-3); 3.69 (s, 3H, OCH₃); ¹³C-NMR (CDCl₃): 12.1 (C-18); 18.3 (C-21); 51.5 (OCH₃); 71.9 (C-3); 174.9 (C-24); mass spectrum (MS): m/z 373 (M-H₂O+H)⁺, 413 (M+Na)⁺.

3-(2-Propenoxy)-5ß-cholan-24-oic acid methyl ester (2) Under an atmosphere of argon, 5 g (12.8 mmol) of compound (1) was dissolved in 15 ml of dry N,N-dimethylformamide and 6.5 ml of N,N-diisopropylethylamine was added. Under refluxing, 15 ml (173 mmol) of allyl bromide was added dropwise over a period of 8 h. After removing excess allyl bromide and amine by distillation at normal pressure, the solvent was removed under reduced pressure. The crude product was purified by chromatography using SS 1. The yield of pure product was 4.2 g (9.8 mmol, 77% yield). MP, 74°C; TLC: R_f = 0.19 (SS 1); R_f = 0.48 (SS 2); ¹H-NMR (CDCl₃): 0.63 (s, 3H, CH₃-18); 0.91 (d, J = 7 Hz, 3H, CH₃-21); 0.92 (s, 3H, CH₃-19); 3.29 (m, 1H, CH-3); 3.68 (s, 3H, OCH₃); 4.02 (d, J = 6 Hz, 2H, -CHCH₂O); 5.18 (d, ³J_cis = 10 Hz, 1H, -CHCH₂); 5.28 (d, ³J_trans = 16 Hz, 1H, -CHCH₂); 5.95 (m, 1H, H₂CCHCH₂-); ¹³C-NMR (CDCl₃): 12.1 (C-18); 18.3 (C-21); 51.5 (-OCH₃); 69.0 (=CHCH₂O-); 78.7 (C-3); 116.4(-CHCH₂); 135 (H₂CCHCH₂-); 174.7 (C-24); MS: m/z 453 (M+Na)⁺, 373 (M-CH₂CHCH₂O)⁺.

3-(2-Oxoethoxy)-5ß-cholan-24-oic acid methyl ester (3) A sample of 3.5 g of compound (2) (8.1 mmol) was dissolved in 50 ml of dioxane, and 56 mg of OsO₄ (0.21 mmol) in 10 ml of water was added. Then, 4.5 g (21 mmol) of NaIO₄ was added in small portions at room temperature (RT) over a period of 45 min. After stirring the reaction mixture at RT for 10 h, the precipitated salt was removed by filtration. The solvent was removed under reduced pressure, and the residue was purified by chromatography using SS 2. The yield of pure compound (3) was 2.75 g (6.3 mmol, 78% yield). MP, 111°C; TLC: R_f = 0.22 (SS 2), R_f' = 0.74 (SS 5); ¹H-NMR (CDCl₃): 0.67 (s, 3H, CH₃-18); 0.93 (d, J = 7 Hz, 3H, CH₃-21); 0.95 (s, 3H, CH₃-19); 3.36 (m, 1H, CH-3); 3.70 (s, 3H, -OCH₃); 4.17 (s, 2H, OHCCH₂O-); 9.75 (s, 1H, OCH-); ¹³C-NMR (CDCl₃): 12.1 (C-18); 18.3 (C-21); 51.6 (-OCH₃); 73.9 (OHCCH₂O-); 80.6 (C-3); 174.7 (C-24); 201.6 (OCH-); MS: m/z 455 (M+Na)⁺, 373 (M-OCHCH₂O)⁺.

3-(2-Hydroxyethoxy)-5ß-cholan-24-oic acid methyl ester (4) In 40 ml of a mixture of tetrahydrofuran and water (4:1, v/v), 3.5 g (8.0 mmol) of compound (3) was dissolved and 250 mg of NaBH₄ was added. The reaction mixture was stirred at RT for 14 h. After destroying the excess borohydride with 0.1 N HCl, the organic compounds were extracted with ether. The crude product was purified by chromatography in SS 4. The yield of pure product was 2.1 g (4.8 mmol, 62% yield). MP, 116°C; TLC: R_f = 0.25 (SS 4), R_f = 0.60 (SS 5); ¹H-NMR (CDCl₃): 0.65 (s, 3H, CH₃-18); 0.91 (d, J = 7 Hz, 3H, CH₃-21); 0.92 (s, 3H, CH₃-19); 3.31 (m, 1H, CH-3); 3.60 (t, J = 6 Hz, 2H, -CH₂CH₂O-); 3.69 (s, 3H, -OCH₃); 3.71 (t, J = 6 Hz, 2H, HOCH₂-); ¹³C-NMR (CDCl₃): 12.1 (C-18); 18.3 (C-21); 51.5 (-OCH₃); 62.2 (HOCH₂-); 69.0 (-CH₂CH₂O-); 79.7 (C-3); 174.8 (C-24); MS: m/z 457 (M+Na)⁺, 373 (M-HOCH₂CH₂O)⁺.

3-(2-Hydroxyethoxy)-5ß-cholan-24-oic acid (5) A mixture of 1 g (2.3 mmol) of compound (4) and 260 mg (4.6 mmol) of KOH in 20 ml of methanol and 5 ml of water was stirred under reflux for 4 h. After acidification of the reaction mixture with 250 ml of 0.1 N HCl, the product was extracted with ether and purified by chromatography using SS 4. The yield of pure compound (5) was 0.89 g (2.1 mmol, 92% yield). MP, 131–133°C; TLC: R_f = 0.13 (SS 4), R_f = 0.61 (SS 5); ¹H-NMR (d₆-DMSO): 0.67 (s, 3H, CH₃-18); 0.95 (s, 3H, CH₃-19); 0.95 (d, J = 7 Hz, 3H, CH₃-21); 3.31 (m, 1H, CH-3); 3.62 (t, J = 6 Hz, 2H, -CH₂CH₂O-); 3.72 (t, J = 6 Hz, 2 H, HOCH₂-); ¹³C-NMR (d₆-DMSO): 11.8 (C-18); 18.1 (C-21); 62.8 (HOCH₂-); 70.3 (-CH₂CH₂O-); 78.5 (C-3); 174.7 (C-24); MS: m/z 443 (M+Na)⁺, 359 (M-HOCH₂CH₂O)⁺.

3ß-(2-Hydroxyethoxy)-5ß-cholan-24-oic acid (6) Under an atmosphere of argon, 5 g (13.3 mmol) of lithocholic acid was dissolved in 10 ml of dried and freshly distilled pyridine. After cooling the solution to 0°C, 1.3 ml (16 mmol) of methansulfonylchloride was added dropwise over 30 min. The reaction mixture was allowed to warm to RT and stirred for 2 h. Then, 20 ml of dried glycol was added and the mixture was stirred at 100°C for another 2 h. After cooling to RT, the mixture was poured into 500 ml of 0.1 N HCl. The organic compounds were extracted with ether. The crude product was purified by chromatography using SS 4, to give 1.2 g (2.9 mmol, 22%) of compound (6). MP, 135–137°C; TLC: R_f = 0.14 (SS 4), R_f = 0.63 (SS 5); ¹H-NMR (d₆-DMSO): 0.62 (s, 3H, CH₃-18); 0.89 (s, 3H, CH₃-19); 0.89 (d, J = 7 Hz, 3H, CH₃-21); 3.35 (t, J = 5 Hz, 2H, -CH₂O-); 3.48 (t, J = 5 Hz, 2H, HOCH₂-); 3.56 (m, 1H, CH-3); 4.46 (m, 1H, HOCH₂-); 11.93 (m, 1H, -COOH); ¹³C-NMR (d₆-DMSO): = 11.8 (C-18); 18.1 (C-21); 60.6 (HOCH₂-); 69.0 (-CH₂CH₂O-); 73.6 (C-3); 174.7 (C-24); MS: m/z 443 (M+Na)⁺, 359 (M-HOCH₂CH₂O)⁺.

3- and 3ß-(2-acetyloxyethoxy)-5ß-cholan-24-oic acids [(7) and (8)] To a solution of 1 g (2.4 mmol) of compound (5) or (6) in 20 ml of dried and freshly distilled pyridine, 4 ml of acetic acid anhydride was added, and the reaction mixture was stirred at RT for 1 h. Excess acetic acid anhydride was hydrolyzed by adding 10 ml of water. After removing the solvent as azeotrope with toluene under reduced pressure, the residue was dissolved in SS 3 and purified by chromatography. Compound (7) yielded 780 mg (1.7 mmol, 71% yield). MP, 107°C; TLC: R_f = 0.21 (SS 3), R_f = 0.70 (SS 5); ¹H-NMR (d₆-DMSO): 0.62 (s, 3H, CH3-18); 0.89 (d, J = 7 Hz, 3H, CH3-21); 0.90 (s, 3H, CH3-19); 2.02 (s, 3H, H₃CCOO-); 3.26 (m, 1H, CH-3); 3.59 (t, J = 6 Hz, 2H, -CH₂O-); 4.08 (t, J = 6 Hz, 2H, AcOCH₂-); ¹³C-NMR (d₆-DMSO): 11.8 (C-18); 18.1 (C-21); 20.3 (H₃CCO-); 63.6 (-COOCH₂-); 65.0 (-CH₂CH₂O-); 78.4 (C-3); 170.2 (H₃CCO-); 174.7 (C-24); MS: m/z 485 (M+Na)⁺, 463 (M+H)⁺, 359 (M-H3CCOOCH2CH2O)⁺. Compound (8) yielded 980 mg (2.1 mmol, 88% yield). MP, 109°C; TLC: R_f = 0.18 (SS 3), R_f = 0.71 (SS 5); ¹H-NMR (d₆-DMSO): 0.62 (s, 3H, CH₃-18); 0.90 (s, 3H, CH₃-19); 0.90 (d, J = 7 Hz, 3H, CH₃-21); 2.01 (s, 3H, H₃CCOO-); 3.51 (t, J = 5 Hz, 2H, -CH₂O-); 3.58 (m, 1H, CH-3); 4.10 (t, J = 5 Hz, 2H, HOCH₂-); 11.96 (m, 1H, -COOH); ¹³C-NMR (d₆-DMSO): 11.8 (C-18); 18.1 (C-21); 20.3 (H₃CCO-); 63.7 (-COOCH₂-); 65.1 (-CH₂CH₂O-); 73.8 (C-3); 170.2 (H₃CCO-); 174.7 (C-24); MS: m/z 485 (M+Na)⁺, 463 (M+H)⁺, 359 (M-H₃CCOOCH2CH2O)⁺.

Sodium (3- and 3ß-(2-hydroxyethoxy)-5ß-cholan-24-oyl)-2'-aminoethanesulfonate [(9) and (10)] Under an atmosphere of argon, 660 mg (1.4 mmol) of compound (7) or (8), 430 mg (2.1 mmol) of DCCI, and 240 mg (2.1 mmol) of N-hydroxysuccinimide were dissolved in 10 ml of dry dioxane, and the mixtures were stirred at RT for 12 h. The precipitated N,N-dicyclohexylurea was removed by filtration, and a solution of 263 mg (2.1 mmol) of taurine and 177 mg (2.1 mmol) of sodium hydrogen carbonate in 5 ml of water was added to the filtrates. After stirring at RT for 16 h, the reaction mixture was poured into 500 ml of 0.1 N NaOH and the organic compounds were extracted with ether. The corresponding bile acid derivative was extracted out of the aqueous phase by adsorption with Serdolit^® PAD-1. After washing, the product was eluted with methanol. The methanol was evaporated under reduced pressure, and the residue was heated for 2 h in 500 ml of 1 N NaOH under reflux to remove the protective acetyl group. After the reaction mixture had been cooled to RT, adsorption chromatography was repeated. Finally, the crude product was purified by chromatography using SS 6. Compound (9) yielded 550 mg (1.0 mmol, 72% yield). MP, 223°C; TLC: R_f = 0.33 (SS 6), R_f = 0.71 (SS 8); ¹H-NMR (CD₃OD): 0.70 (s, 3H, CH₃-18); 0.94 (d, J = 7 Hz, 3H, CH₃-21); 0.96 (s, 3H, CH₃-19); 2.97 (t, J = 7 Hz, 2H, -CH₂SO₃-), 3.32 (m, 1H, CH-3); 3.54–3.68 (m b, 6H); ¹³C-NMR (CD₃OD): 12.6 (C-18); 18.9 (C-21); 33.1 (-NHCH₂-); 51.6 (-CH₂SO₃-); 62.6 (HOCH₂-); 70.4 (-CH₂CH₂O-); 80.9 (C-3); 176.7 (C-24); MS: m/z 572 (M+Na)⁺, 550 (M+H)⁺. Compound (10) yielded 610 mg (1.1 mmol, 79% yield). MP: 214°C; R_f = 0.35 (SS 6), R_f = 0.76 (SS 8); ¹H-NMR (CD₃OD): 0.69 (s, 3H, CH₃-18); 0.98 (s, 3H, CH₃-19); 0.98 (d, J = 7 Hz, 3H, CH₃-21); 2.96 (t, J = 7 Hz, 2H, -CH₂SO₃-); 3.47 (t, J = 5 Hz, 2H, -CH₂O-); 3.60 (m, 1H, CH-3); 3.61–3.71 (m, 4H); ¹³C-NMR (CD₃OD): 12.6 (C-18); 18.9 (C-21); 33.1 (-NHCH₂-); 51.5 (-CH₂SO₃-); 62.7 (HOCH₂-); 70.3 (-CH₂CH₂O-); 76.3 (C-3); 176.5 (C-24); MS: m/z 572 (M+Na)⁺, 550 (M+H)⁺.

Sodium (3- and 3ß-(2-succinyloxyethoxy)-5ß-cholan-24-oyl)-2'-aminoethanesulfonate [(11) and (12)] To a solution of 200 mg (0.38 mmol) of compound (9) or (10) in 5 ml of dry pyridine, 120 mg (1.2 mmol) of succinic anhydride was added under an argon atmosphere. The reaction mixture was stirred at RT for 12 h, then 2 ml of water was added. Water and pyridine were removed under reduced pressure, and the residue was chromatographed on silica gel with SS 7. Compound (11) yielded 170 mg (0.27 mmol, 71% yield). MP, 178°C; TLC: R_f = 0.22 (SS 7), R_f = 0.51 (SS 8); ¹H-NMR (CD₃OD): 0.69 (s, 3H, CH₃-18); 0.95 (s, 3H, CH₃-19); 0.96 (d, J = 7 Hz, 3H, CH₃-21); 2.61 (m b, 4H, -OCCH₂CH₂CO-), 2.96 (t, J = 7 Hz, 2H, -CH₂SO₃-); 3.35 (m, 1H, CH-3); 3.61 (t, J = 7 Hz, 2H, -HNCH₂-); 3.69 (t, 2H, J = 6 Hz, -CH₂OCH-); 4.19 (t, 2H, J = 6 Hz, -COOCH₂-); ¹³C-NMR (CD₃OD): 12.5 (C-18); 18.9 (C-21); 30.3 (-OOCCH₂-); 33.1 (-NHCH₂-); 51.5 (-CH₂SO₃-); 65.3 (-COOCH₂-); 66.9 (-CH₂OCH-); 81.0 (C-3); 174.3 (HOOC- and -COOCH₂-); 176.5 (C-24); MS: m/z 672 (M+Na)⁺, 650 (M+H)⁺. Compound (12) yielded 190 mg (0.30 mmol, 79% yield). MP, 176°C; TLC: R_f = 0.33 (SS 7), R_f = 0.49 (SS 8); ¹H-NMR (CD₃OD): 0.69 (s, 3H, CH₃-18); 0.96 (s, 3H, CH₃-19); 0.96 (d, J = 7 Hz, 3H, CH₃-21); 2.62 (m b, 4H, -OCCH₂CH₂CO-), 2.96 (t, J = 7 Hz, 2H, -CH₂SO₃-); 3.44 (t, 2H, J = 6 Hz, -CH₂OCH-); 3.60 (m, 3H); 4.21 (t, 2H, J = 6 Hz, -COOCH₂-); ¹³C-NMR (CD₃OD): 12.5 (C-18); 18.9 (C-21); 30.3 (-OOCCH₂-); 33.1 (-NHCH₂-); 51.6 (-CH₂SO₃-); 65.2 (-COOCH₂-); 66.9 (-CH₂OCH-); 76.3 (C-3); 174.4 (HOOC- and -COOCH₂-); 176.5 (C-24); MS: m/z 672 (M+Na)⁺, 650 (M+H)⁺.

DSPE-3-LCT and DSPE-3ß-LCT A solution of 250 mg (0.39 mmol) of compound (11) or (12), 116 mg (0.58 mmol) of DCCI, and 65 mg (0.58 mmol) of N-hydroxysuccinimide in 3 ml of anhydrous N,N-dimethylformamide was stirred at RT for 16 h under an atmosphere of argon. The reaction mixtures were cooled to 4°C and filtered. The precipitate was washed with 2 ml of ice-cold N,N-dimethylformamide. The clear filtrate was then poured into 50 ml of cold ether and centrifuged (5 min at 2,000 g). The ether was decanted and the precipitate was dissolved in a mixture of 10 ml of CHCl₃, 5 ml of methanol, and 200 µl of water. To this solution, 300 mg (0.4 mmol) of 1,2-O-distearoyl-sn-glycero-3-phosphoethanolamine (DSPE) and 400 µl of 1 N NaOH were added. The mixture was stirred at RT for 2 h, then the solvent was removed under reduced pressure and the crude product was chromatographed with SS 9. DSPE-3-LCT yielded 170 mg (0.12 mmol, 31% yield). MP, 211°C; TLC: R_f = 0.08 (SS 9), R_f = 0.05 (SS 10); ¹H-NMR: (CDCl₃-CD₃OD-D₂O 2:1:0.1 (v/v/v)) 0.69 (s, 3H, CH₃-18); 0.90 (t, J = 8 Hz, 6H, 2 x H₃CCH₂-); 0.95 (s, 3H, CH₃-19); 0.95 (d, J = 7 Hz, 3H, CH₃-21); 2.32 (m, 4H, 2 x -CH₂CH₂CH₂OCO-); 2.56 (t, J = 7 Hz, 2H, -COCH₂CH₂COO-); 2.69 (t, J = 7 Hz, 2H, -NHCOCH₂CH₂CO-); 3.01 (t, J = 7 Hz, 2H, -CH₂SO₃-); 3.35 (m, 1H, CH-3); 3.44 (m, 2H, -OCH₂CH₂NH-); 3.61 (t, J = 7 Hz, 2H, -HNCH₂CH₂SO₃Na); 3.70 (t, 2H, J = 6 Hz, -CH₂OCH-); 3.81–4.18 (m b, 6H); 4.21 (t, 2H, J = 6 Hz, -COOCH₂-); 5.23 (m, -OCH₂HCOHCH₂O-); ¹³C-NMR (CDCl₃-CD₃OD-D₂O 2:1:0.1 (v/v/v)): 12.3 (C-18); 14.5 (stearoyl-CH₃); 18.5 (C-21); 25.2 (stearoyl-C-3') 30.0 (stearoyl-CH₂); 30.8 (succinyl-CH₂); 33.4 (taurine-NHCH₂-); 50.6 (-CH₂SO₃-); 63.1 (glyceryl-C-3'); 64.0 (glyceryl-C-1'); 64.7 (-COOCH₂CH₂-O); 66.0 (-CH₂OCH-); 71.2 (glyceryl-C-2'); 80.2 (C-3); 173.9 (stearoyl-CO); 174.0 (succinyl-CO); 174.4 (C-24); MS: m/z 1,424 (M+Na)⁺, 1,402 (M+H)⁺, 1,380 (M-Na⁺2H)⁺, 1,378 (M-Na)^–, 677 (M-2Na) (2-). DSPE-3ß-LCT yielded 220 mg (0.16 mmol, 41% yield). MP, 206°C; TLC: R_f = 0.08 (SS 9), R_f = 0.03 (SS 10); ¹H-NMR (CDCl₃-CD₃OD-D₂O 2:1:0.1 (v/v/v)): 0.68 (s, 3H, CH₃-18); 0.92 (t, J = 8, 6H, 2 x H₃CCH₂-); 0.96 (d, J = 7 Hz, 3H, CH₃-21); 0.97 (s, 3H, CH₃-19); 2.35 (m, 4H, 2 x -CH₂CH₂CH₂OCO-); 2.56 (t, J = 7, 2H, -COCH₂CH₂COO-); 2.68 (t, J = 7 Hz, 2H, -NHCOCH₂CH₂CO-); 3.01 (t, J = 7 Hz, 2H, -CH₂-SO₃-); 3.44 (m, 2H, -OCH₂CH₂NH-); 3.58 (t, J = 7 Hz, 2H, -HNCH₂CH₂SO₃Na); 3.62 (t, 2H, J = 6 Hz, -CH₂OCH-); 3.67 (m, 1H, CH-3); 3.92–4.21 (m b, 6H); 4.21 (t, 2H, J = 6 Hz, -COOCH₂-); 5.27 (m, -OCH₂HCOHCH₂O-); ¹³C-NMR (CDCl₃-CD₃OD-D₂O 2:1:0.1 (v/v/v)): 12.5 (C-18); 14.5 (stearoyl-CH₃); 18.7 (C-21); 25.5 (stearoyl-C-3') 30.8 (succinyl-CH₂); 30.0; 30.2 30.3 (stearoyl-CH₂); 33.9 (taurine-NHCH₂-); 50.8 (-CH₂SO₃-); 63.4 (glyceryl-C-3'); 64.2 (glyceryl-C-1'); 64.9 (-COOCH₂CH₂-O); 66.0 (-CH₂OCH-); 71.2 (glyceryl-C-2'); 75.8 (C-3); 174.1 (stearoyl-CO); 174.4 (succinyl-CO); 176.0 (C-24); MS: m/z 1,424 (M+Na)⁺, 1,402 (M+H)⁺, 1,380 (M-Na⁺2H)⁺, 1,378 (M-Na)^–, 677 (M-2Na) (2-).

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Syntheses
To determine whether bile salts may serve as ligands to direct liposomes to hepatocytes, a bile salt had to be linked to the liposomal membrane. Therefore, a bile salt-phospholipid conjugate was synthesized and characterized. In a first step, the secondary 3-hydroxyl group of lithocholic acid was converted in a stereoselective manner to the 3- or 3ß-glycol ether derivative (Fig. 1). With slight modifications, this synthesis followed the synthesis for the corresponding cholic acid derivatives (28). The modified bile salts were then conjugated with taurine (Fig. 2). Activating the carboxyl group with N-hydroxysuccinimide gave higher yields under milder conditions than the methods used previously (29). A spacer between the lipid anchor and the homing device is necessary for efficient binding (30), and succinate was chosen as a physiological linker. The succinyl-bile salt derivatives were linked to DSPE via the N-hydroxysuccinimide method. The low solubility of the two amphipathic molecules required N,N-dimethylformamide as a solvent. Subsequent to the formation of the glycol ether, all reactions were performed under mild conditions. Thereby, the fixed stereochemistry at the 3 position of the bile salt was preserved, as confirmed by ¹H- and ¹³C-NMR.

View larger version (18K): [in this window] [in a new window]   Fig. 1. Synthesis of 3- and 3ß-(2-hydroxyethoxy)-5ß-cholan-24-oic acid. DMF, N,N-dimethylformamide; DPEA, N,N-diisopropylethylamine.

View larger version (18K): [in this window] [in a new window]   Fig. 2. Synthesis of disodium 3-(2-(1,2-O-distearoyl-sn-glycero-3-phospho-2'-ethanolamidosuccinyloxy)ethoxy)-5ß-cholan-24-oyl-2'-aminoethansulfonate (DSPE-3-LCT) and disodium 3ß-(2-(1,2-O-distearoyl-sn-glycero-3-phospho-2'-ethanolamidosuccinyloxy)ethoxy-5ß-cholan-24-oyl-2'-aminoethansulfonate (DSPE-3ß-LCT).

The prepared BSLs had a mean diameter of 122 ± 14 nm, and COLs had a mean diameter of 118 ± 6 nm. The polydispersity of the liposomes was 0.050 ± 0.035 for BSLs and 0.073 ± 0.026 for COLs. Diameter and size distribution were as expected for the extrusion method used. The prepared BSLs and COLs were stable over a period of several days (as measured by changes in diameter). The zeta potential of BSLs was 31.2 ± 3.6 mV, and that of COLs was 35.3 ± 2.7 mV. The difference was slightly significant (P = 0.06). The amount of encapsulated lipid (lamellarity) was measured according to a modified method described by McIntyre and Sleight (20). Instead of using detergent to destroy the liposomes, cholate was used to increase the permeability of the liposomal bilayer while the liposomal structure was maintained (21). Adding cholate to a final concentration of 2 mM allows the dithionite anion to reach the interior of the liposomes and to reduce the NBD lipids inside the liposomes. At this concentration, no quenching of fluorescence by the addition of cholate was observed, and the amount of light scattering by the liposomes (baseline in Fig. 3) was not altered. This allows the continuous measurement of lamellarity. A typical measurement is shown in Fig. 3. Approximately 6 ± 2% of total lipid was encapsulated within the COLs, and 2 ± 2% was encapsulated within the BSLs (n = 4). Thus, the BSLs as well as the COLs were almost unilamellar.

Binding of CT to liposomes
For bile salt uptake studies, CT at a concentration of 50 µM was used as a reference. Therefore, the binding of 50 µM CT to the liposomes was measured. The amount of bound CT was determined by ultracentrifugation under equilibrium conditions as described (31). Binding of CT to the liposomes was very low. Approximately 0.4 ± 0.2 or 0.2 ± 0.2 mmol CT/mol liposomal PC was bound to BSLs or COLs (n = 3). The difference between BSLs and COLs was almost significant (P = 0.1). Between DSPE-3-LCT-bearing and DSPE-3ß-LCT-bearing liposomes, no difference in binding of CT was observed. Bile salts tend to interact with each other. Depending on the concentration, they form dimers, multimers, or micelles (32), and probably CT interacts weakly with bile salts protruding from the liposomal membrane.

Inhibition of bile salt uptake
To examine whether bile salts covalently linked to the surface of a liposome are able to interact with the bile salt carriers of the hepatocytes, the inhibition of the uptake of CT in the presence of liposomes was examined (Fig. 4). Liposomes and CT were added simultaneously to the cells. The uptake rates were measured under initial rate conditions within the first 60 s. The concentration of free CT was reduced from 50 to 49.6 µM for COLs and from 50 to 49.2 µM for BSLs as a result of binding of CT to the liposomes (see above). To consider this effect on uptake rates, the kinetic constants published by Schramm et al. (10) were used. Binding of CT to the liposomes reduces uptake rates of CT by 0.2% (COLs) or 0.4% (BSLs) in the experimental setting applied. The measured uptake rates in the absence of liposomes were reduced by 0.2% or 0.4%, respectively, before the calculation of relative uptake rates in the presence of liposomes.

View larger version (45K): [in this window] [in a new window]   Fig. 4. Inhibition of cholyltaurine (CT) uptake by liposomes. The uptake of 3H-labeled CT was measured in the absence and the presence of liposomes in a concentration of 2 mM liposomal phosphatidylcholine (PC). The concentration of CT was 50 µM. Liposomes and CT were added simultaneously to the cells. The graph shows the uptake in the presence of liposomes (Jlip) related to the uptake in the absence of liposomes (JCT). JCT was 0.5 ± 0.2 nmol/min/mg protein. The means and SD of five individual experiments are shown. BSL, bile salt-bearing liposome.

Binding of liposomes to the surface of hepatocytes
To determine whether BSLs bind to the surface of hepatocytes, confocal laser scanning microscopy (LSM) was performed. Liposomes were labeled with the fluorescent dye Ph₂-DiOC18. Ph₂-DiOC18 is a very hydrophobic membrane marker that is not exchanged between liposomes and the cytoplasmic membrane during the experiments (10–15 min). Fluorescent liposomes were added to hepatocytes, and after 5 min the liposomes were removed. The cells were washed with buffer and examined by confocal LSM. Typical images are shown in Fig. 5. Fig. 5A, B show COLs, and Fig. 5C, D show DSPE-3ß-LCT-bearing liposomes. Fig. 5A, C show differential interference contrast images combined with fluorescence images along the plane of the cover slips the cells were grown on (x-y plane). The confocal plane in the images focuses on the surface of the cover slip. Fig. 5B, D show cross-sections (x-z plane) of the cells along the white lines shown in Fig. 5A, C.

View larger version (104K): [in this window] [in a new window]   Fig. 5. Binding of liposomes to the surface of hepatocytes. Liposomes in a concentration of 2 mM liposomal PC were added to hepatocytes cultured for 24 h. After 5 min, the cells were carefully washed and subjected to laser scanning microscopy. A, C: Interference contrast images along the x-y plane, focusing on the surface of the cover slips on which the cells were grown. B, D: x-z cross-sections along the white lines indicated in A, C, respectively. Liposomes without DSPE-LCT (COLs) are shown in A, B, whereas DSPE-3ß-LCT-bearing liposomes (BSLs) are shown in C, D.

To quantify the amount of bound liposomes, liposomes were labeled with [³H]cholesterylhexadecylether, a nonexchangeable membrane marker (33). Liposomes were added to the cells for 5 min. After washing (three times), the amount of radioactivity remaining associated with the cells was quantified. The amount of liposomes associated with the cells was calculated as nanomoles of liposomal PC per milligram of cellular protein. As described in Materials and Methods, a rough estimate was made to calculate the number of liposomes bound per cell. Hepatocytes were cultured on six-well plates covered with collagen. Because there is some significant binding of liposomes to collagen not covered with cells (Fig. 5), only cultures with a confluence >90% were used. The binding of liposomes to the collagen was quantified under identical conditions, and the amount of liposomes associated with the cells was corrected for nonspecific binding to 10% of the surface area of a culture well.

In Fig. 6, means ± SD are shown for 10–12 individual experiments. In all experiments, more BSLs than COLs remained associated with the cells. The difference in binding was highly significant (P < 0.01). Again, DSPE-3ß-LCT-bearing liposomes showed a stronger interaction with the hepatocytes than DSPE-3-LCT-bearing liposomes (P < 0.01).

View larger version (31K): [in this window] [in a new window]   Fig. 6. Binding of liposomes to hepatocytes. 3H-labeled liposomes in a concentration of 2 mM liposomal PC were added for 5 min to primary rat hepatocytes kept in culture for 24 h. Cells were washed, and the amount of associated radioactivity was quantified as described. In a rough estimation, 1 nmol liposomal PC/mg protein = 12,000 liposomes/cell (see model calculations for details). The means and SD of 10–12 individual cell culture experiments are shown.

View larger version (30K): [in this window] [in a new window]   Fig. 7. Binding of liposomes during different washing steps. 3H-labeled liposomes in a concentration of 2 mM liposomal PC were added for 5 min to primary rat hepatocytes cultured for 24 h on six-well plates (35 mm/well). The cells were washed different times with 2 ml of buffer per well. The amount of radioactivity remaining in the well was quantified as described. COLs and BSLs prepared with DSPE-3ß-LCT were used. To calculate bile salt-mediated binding (BMB), the amount of COLs associated with the cells was subtracted from the amount of BSLs associated with the cells. The means and SD of 3 individual cell culture experiments are shown. SD for BMB was calculated by adding the SD for COL and BSL.

View larger version (33K): [in this window] [in a new window]   Fig. 8. Binding of anionic liposomes to hepatocytes. Liposomes in a concentration of 2 mM liposomal PC were added for 5 min to primary rat hepatocytes kept in culture for 24 h. Cells were washed, and the amount of associated radioactivity was quantified as described. The means and SD of 3 individual cell culture experiments are shown. BSL, liposomes prepared with 8 mol% 1,2-O-distearoyl-sn-glycero-3-phospho-rac-(1-glycerol) and 1 mol% DSPE-3ß-LCT; BSL/PSL, 3H-labeled BSLs and unlabeled PSLs, each in a concentration of 2 mM liposomal PC, were added simultaneously to the cells; PSL, liposomes prepared with 25 mol% 1,2-O-dipalmitoyl-sn-glycero-3-phosphoserine.

CT up to a concentration of 500 µM had no significant effect on the binding of COLs. The effect of increasing CT concentrations on the binding of BSLs to hepatocytes is shown in Fig. 9. The bile salt-mediated binding was calculated as described before. Within each test series, the amount of bile salt-mediated binding in the presence of CT was divided by the amount of bile salt-mediated binding in the absence of CT.

View larger version (42K): [in this window] [in a new window]   Fig. 9. Influence of CT on the association of BSLs with hepatocytes. Liposomal suspensions containing 3H-labeled liposomes in a concentration of 2 mM liposomal PC and different concentrations of CT were added for 5 min to primary rat hepatocytes in culture for 24 h. Cells were washed, and the amount of associated radioactivity was quantified as described. The means and SD of 6 individual cell culture experiments are shown. BSLs prepared with DSPE-3ß-LCT were used. To calculate bile salt-mediated binding, the amount of COLs associated with the cells was subtracted from the amount of BSLs associated with the cells.

The transport of bile salts is regulated by a feed-forward mechanism (36, 37), and the presence of bile salts leads to an increasing amount of bile salt carrier in the hepatocyte membrane (38, 39). The activation of bile salt transport is transient and time-dependent. For primary hepatocytes cultivated for 24 h, the maximum activation of intracellular signal cascades by CT was observed 10 min after adding the bile salts to the culture dishes (37). Binding of BSLs in the presence of CT was observed within 5 min, so the activation of bile salt transport by the presence of CT had presumably not reached its peak. To determine whether there is CT-mediated activation of BSL binding, hepatocytes were treated for 10 min with medium containing 100 µM CT, then BSLs were added in the presence of 100 µM CT and binding was observed as before. The amounts of bound liposomes are shown in Fig. 10.

View larger version (28K): [in this window] [in a new window]   Fig. 10. Influence of CT on the association of liposomes with hepatocytes. 3H-labeled liposomes in a concentration of 2 mM liposomal PC were added for 5 min to 24 h cultured primary hepatocytes. The liposomal suspensions contained either no CT or CT in a concentration of 100 µM. Some hepatocytes were pretreated for 10 min with medium containing 100 µM CT. This medium was removed followed by immediate addition of CT-containing liposomal suspensions. The amount of associated liposomes was measured as described. COLs without DSPE-3-LCT and BSLs prepared with DSPE-3ß-LCT were used. The means and SD of 6 individual experiments are shown.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Synthesis
To determine whether bile salts may serve as ligands to direct liposomes to hepatocytes, a bile salt-lipid conjugate was synthesized. The synthesis path allows the stereoselective preparation of DSPE-3-LCT and DSPE-3ß-LCT in a convenient way with an acceptable yield. DSPE was chosen as the lipid component of the conjugate because a phospholipid as a natural building block of lipid bilayers should not alter the membrane properties of the liposomes. Lithocholic acid was chosen as the bile salt because it bears only one hydroxyl group at the C-24 position of the bile acid skeleton, which allows chemical modification without the necessity to protect further functional groups. Other bile salts having two or three hydroxyl groups, such as cholic acid or the therapeutically used ursodeoxycholic acid, may be useful as ligands as well. Higher amounts of these bile salts are less toxic than lithocholic acid, but synthesis with these bile salts is more complex. The aim of this study was a proof of principle, and the demonstration of other bile salts as potential ligands is the subject of further work. Taurine-conjugated bile salts were favored because they bear a permanent negative charge at physiological pH. This charge should prevent the insertion of the hydrophobic bile salt backbone into the liposomal membrane, leading to a buried anchor, and this charge should prevent the insertion of the hydrophobic bile salt backbone into cellular membranes, leading to nonspecific binding of the liposomes to cells.

Different from several other potential ligands for hepatocyte-specific targeting, DSPE-3-LCT may be synthesized at low cost. The lipid-coupled ligand allows the addition of desired amounts of ligand to the lipid mixture of which the liposomes are prepared, without any further elaborate coupling procedure. BSLs may be produced at much lower cost than other targeting systems, using antibodies, for example, as a targeting moiety. Because bile salts are physiological molecules of low molecular weight, it is very unlikely that BSLs trigger any bile salt-specific immune response.

Influence of stereochemistry
The interaction of BSLs with the carrier systems for CT is dependent on the stereochemical orientation of the bridging 3-O of the bile salt. The different degree of inhibition of CT uptake by COLs and BSLs bearing either the stereoisomer DSPE-3-LCT or DSPE-3ß-LCT indicates that inhibition of CT transport is bile salt-specific with interaction between liposomal DSPE-3-LCT and bile salt carrier proteins.

Conjugation of the fluorescent dye NBD to CT via the 3ß position of the bile salt leads to a compound that readily interferes with CT, whereas the isomeric 3 compound does not (32). As shown by Kramer et al. (13), a 3ß-conjugated drug-bile salt is secreted faster into bile than its respective 3 derivative. In accordance with these data for dissolved bile salt conjugates, liposomes bearing DSPE-3ß-LCT interfere to a greater extent with CT transport than liposomes bearing DSPE-3-LCT. Furthermore, higher amounts of DSPE-3ß-LCT-bearing liposomes bind to hepatocytes than DSPE-3-LCT-bearing liposomes. Thus, DSPE-3ß-LCT is more suitable for hepatocyte-specific targeting than its 3 derivative. In general, conjugation via the 3ß position seems to be favorable for using hepatocyte-specific bile salt-transporting systems for drug delivery.

Interaction of BSLs with bile salt carrier systems
DSPE-3ß-LCT-bearing liposomes inhibited the uptake of 50 µM CT by 40%. At a concentration of 2 mM liposomal PC, the overall concentration of liposomal bile salt in the outer membrane layer is 20 µM. Thus, the observed inhibition is in the same range as expected for a simple competitive mechanism. The mechanism of inhibition is definitely more complex than a simple competitive inhibition, but the affinity of liposomal DSPE-3-LCT for bile salt carrier must be of the same order of magnitude as the affinity for CT.

CT is transported by different carrier systems, such as Na⁺/cholyltaurine-cotransporting polypeptide (Ntcp) and different organic anion-transporting polypeptides (7). Because CT uptake is not inhibited completely in the presence of BSLs, at least under the experimental conditions used in the described experiments, it cannot be definitely determined whether these interactions occur with all or certain different carrier systems. Expression, activity, and kinetic parameters of the different carrier systems are not well defined for primary hepatocyte cultures. Other model systems than cultured hepatocytes might be useful to obtain further kinetic data on the interference of BSLs with bile salt transport.

Binding of BSLs to hepatocytes
Bile salts covalently attached to the surface of hepatocytes are able to anchor liposomes on the surface of hepatocytes. As revealed by confocal LSM, the BSLs are more or less evenly distributed on the surface of a single cell, and all vital hepatocytes bind BSLs. Thus, the potential binding sites for BSLs are distributed all over the cell surface. As expected, no endocytosis of bound liposomes was seen during the experiments (10–15 min for binding, washing, and processing). Whether this is also true for longer time periods has to be established and will be discussed in detail in a subsequent publication.

Anionic liposomes are bound by scavenger receptors (34), such as scavenger receptor class B type I, which is found on the surface of hepatocytes. The binding is dependent on the lipid providing the anionic charges and the net charge of the liposomes. COLs had the same net charge and an even higher zeta potential as BSLs, and the binding of COLs and BSLs to scavenger receptors should be comparable, but anionic charges of DSPE-3-LCT protrude from the liposomal surface. So it is possible that binding of BSLs to scavenger receptors may be facilitated by DSPE-3-LCT. To rule out the possibility that the observed binding of BSLs to the hepatocytes is mediated by scavenger receptors only, highly anionic liposomes (PSLs) were used as inhibitors that readily bind to scavenger receptors. When given at the same concentration of liposomal PC, the inhibition of BSL binding by PSLs was significant, but much less than expected for a direct competition at the same receptor(s). Thus, BSLs and PSLs may compete for binding at some binding sites where highly anionic liposomes show a high affinity, but the binding of BSLs is mediated to a major extent by binding sites where BSLs show a much higher affinity than PSLs. Because the PSLs bound to a slightly higher extent to the hepatocytes than the BSLs did, these binding sites cannot be scavenger receptors. Furthermore, DSPE-3-LCT- and DSPE-3ß-LCT-carrying liposomes show no difference in charge distribution. Because the