Search

We studied the kinetics of hepatic uptake of liposomes during serum-free recirculating perfusion of rat livers. Liposomes consisted of phosphatidylcholine, cholesterol and phosphatidylserine in a 6:4:0 or a 3:4:3 molar ratio and were radiolabelled with [3H]cholesteryl oleyl ether. The negatively charged liposomes were taken up to a 10-fold higher extent than the neutral ones. Hepatic uptake of fluorescently labelled liposomes was examined by fluorescence microscopy. The neutral liposomes displayed a typical Kupffer cell distribution pattern, in addition to weak diffuse staining of the parenchyma, while the negatively charged liposomes showed a characteristic sinusoidal lining pattern, consistent with an endothelial localization. In addition, scattered Kupffer cell staining was distinguished as well as diffuse parenchymal fluorescence. The mainly endothelial localisation of the negatively charged liposomes was confirmed by determining radioactivity in endothelial and Kupffer cells isolated following a 1-h perfusion. Perfusion in the presence of polyinosinic acid, an inhibitor of scavenger receptor activity, reduced the rate of uptake of the negatively charged liposomes twofold, indicating the involvement of this receptor in the elimination mechanism. These results are compatible with earlier in vitro studies on liposome uptake by isolated endothelial cells and Kupffer cells, which showed that in the absence of serum also endothelial cells in situ are able to take up massive amounts of negatively charged liposomes. The present results emphasize that the high in vitro endothelial cell uptake in the absence of serum from earlier observations was not an artifact induced by the cell isolation procedure.

Adsorption of serum proteins to the liposomal surface plays a critical role in liposome clearance from the blood. The aim of this study was to investigate the role of liposome-adsorbed serum proteins in the interaction of liposomes with hepatocytes. We analyzed the serum proteins adsorbing to the surface of differently composed small unilamellar liposomes during incubation with human or rat serum, and found that one protein, with a molecular weight of around 55 kDa, adsorbed in a large amount to negatively charged liposomes containing phosphatidylserine (PS) or phosphatidylglycerol (PG). The binding was dependent on the liposomal charge density. The approximately 55-kDa protein was identified as beta2-glycoprotein I (beta2GPI) by Western blotting. Despite the high affinity of beta2GPI for strongly negatively charged liposomes, in vitro uptake and binding experiments with isolated rat hepatocytes, Kupffer cells or liver endothelial cells, and with HepG2 cells showed no enhancing effect of this protein on the association of negatively charged liposomes with any of these cells. On the contrary, an inhibitory effect was observed. We conclude that despite abundant adsorption to negatively charged liposomes, beta2GP1 inhibits, rather than enhances, liposome uptake by liver cells.

A monoclonal antibody against the rat colon carcinoma CC531 was covalently coupled to liposomes containing a dipalmitoylated derivative of the anticancer drug FUdR as a prodrug in their bilayers. We investigated the in vitro interaction of these liposomes with CC531 target cells and the mechanism by which they deliver the active drug FUdR intracellularly to the cells by monitoring the fate of the liposomal bilayer markers cholesterol-[(14)C]oleate and [(3)H]cholesteryloleylether as well as the (3)H-labeled prodrug and colloidal gold as an encapsulated liposome marker. After binding of the immunoliposomes to the cell surface, only limited amounts were internalized as demonstrated by a low level of hydrolysis of liposomal cholesterol ester and by morphological studies employing colloidal gold-labeled immunoliposomes. By contrast, already within 24 h immunoliposome-incorporated FUdR-dP was hydrolyzed virtually completely to the parent drug FUdR intracellularly. This process was inhibited by a variety of endocytosis inhibitors, indicating that the prodrug enters and is processed by the cells by a mechanism involving an endocytic process, resulting in intracellular FUdR concentrations up to 3000-fold higher than those in the medium. Immunoliposomes containing poly(ethyleneglycol) (PEG) chains on their surface, with the antibody coupled either directly to the bilayer or at the distal end of the PEG chains were able to deliver the prodrug into the tumor cells at the same rate as immunoliposomes without PEG. Based on these observations, we tentatively conclude that during the interaction of the immunoliposomes with the tumor cells the lipophilic prodrug FUdR-dP is selectively transferred to the cell surface and subsequently internalized by constitutive endocytic or pinocytic invaginations of the plasma membrane, thus ultimately delivering the prodrug to a lysosomal compartment where hydrolysis and release of parent drug takes place. This concept allows for an efficient delivery of a liposome-associated drug without the need for the liposome as such to be internalized by the cells.

We describe the synthesis of biodegradable poly(ethyleneglycol)-coupled galactolipids in which the galactose moiety is separated from a diacylglyceride lipid anchor by poly(ethylene glycol) chains of 10, 20 or 40 oxyethylene residues (PEG10/20/40). These Gal-PEG lipids (Gal-PEG-Lip) were incorporated in the bilayer of liposomes. The surface exposure of the galactose was investigated by aggregation experiments with ricinus communis agglutinin 120. Only the liposomes containing the PEG10 galactolipid aggregated with the lectin. Therefore liposomes were prepared containing Gal-PEG10-Lip and a trace amount of [3H]cholesteryl oleyl ether with an average diameter of approximately 100 nm and injected intravenously into rats. The Gal-PEG10-Lip liposomes were cleared from plasma with a T1/2 of 0.3 h. Identically sized and composed control liposomes without the Gal-PEG10-Lip had a T1/2 of approximately 12 h. The rapid plasma elimination of the Gal-PEG10-Lip liposomes could be attributed entirely to increased uptake by the liver amounting to more than 90% of injected dose. Uptake by the spleen was decreased to less than 1% of injected dose. A single injection of N-acetylgalactosamine 1 min prior to Gal-PEG-Lip liposome administration reduced the initial rate of plasma clearance to control levels. The increased liver uptake was almost entirely attributable to increased uptake by the Kupffer cells. Incorporation of PEG-DSPE in the Gal-PEG10-Lip liposomes only partially reversed the effect of the galactolipid with respect to liver and spleen uptake as well as intrahepatic distribution. These experiments demonstrate that liposome surface-exposed galactose residues, even if attached at the distal end of a poly(ethyleneglycol) chain anchored in the liposomal bilayer are effectively recognized by the galactose particle receptor on the Kupffer cells but fail to achieve significant targeting to the asialoglycoprotein receptor on the hepatocytes.

Liposomes can very efficiently deliver immunomodulators to macrophages so as to induce tumor cytotoxicity. Liposomes most widely used for that purpose contain negatively charged lipids, in particular phosphatidylserine (PS), to enhance liposome uptake by the macrophages. We investigated the effect of three negatively charged liposomal lipids on the in vitro activation of liver macrophages to tumor cytotoxicity by muramyl dipeptide (MDP) and lipopolysaccharide (LPS). Both MDP- and LPS-induced tumor cytotoxicity towards murine colon adenocarcinoma cells were strongly inhibited by PS-containing liposomes. Under comparable conditions phosphatidylglycerol (DPPG)-containing or dicetyl phosphate (DCP)-containing liposomes did not inhibit or only marginally inhibited the induction of tumor cytotoxicity. We did not observe PS-mediated inhibition of tumor cell toxicity when the exposure of the macrophages to PS-liposomes was limited to the 4-h activation period prior to addition of the tumor target cells, suggesting that the inhibitory effect is accomplished at the level of the later stages of the activation process. Previously, we showed that macrophages which are activated to tumor cytotoxicity during a 24-h incubation with MDP become refractory to a second activation with MDP. Now we observed that simultaneous incubation with PS-containing liposomes partially prevents this refractoriness, which is also compatible with an interfering action of PS at a relatively late stage in the activation process. We conclude that PS, despite its reported stimulatory effect on liposome uptake by macrophages, can seriously antagonize the effectiveness of immunomodulating agents acting on macrophages. This bears relevance to the use of PS-containing liposomes as a vehicle for such agents. The results are discussed in perspective of earlier reported pharmacological effects of PS and its metabolites.

Human serum albumin (HSA) derivatized with cis-aconitic anhydride (Aco-HSA) that was earlier shown to inhibit replication of human immunodeficiency virus type 1 (HIV-1), was covalently coupled to conventional liposomes, consisting of phosphatidylcholine, cholesterol and maleimido-4-(p-phenylbutyryl)phosphatidylethanolamine, using the heterobifunctional reagent N-succinimidyl-S-acetylthioacetate (SATA). The amount of HSA that could be coupled to the liposomes depended on derivatization of the HSA and ranged from 64.2 +/- microgram HSA/micromol total lipid for native HSA to 29.5 +/- 2.7 microgram HSA/micromol total lipid for HSA in which 53 of the epsilon amino groups of lysine were derivatized with cis-aconitic anhydride (Aco53-HSA). Incorporation of 3.8 mol% of total lipid of a poly(ethylene glycol) derivative of phosphatidylethanolamine (PEG-PE) in the liposomes resulted in a lower coupling efficiency of Aco-HSA. The elimination and distribution of the liposomal conjugates in rats in vivo was largely dependent on the modification of the HSA coupled to the liposomes. With native HSA-liposomes, more than 70% of the conjugate was still found in the blood plasma 30 min after i.v. injection in rats, while at this time Aco-HSA-liposomes were completely cleared from the circulation. The rapid clearance of conventional Aco-HSA-liposomes was due to a rapid uptake into the liver and could be considerably decreased by incorporating PEG-PE in the liposomal bilayer. After 3 h 60% of Aco-HSA-PEG-liposome conjugates were found in the blood. In an in vitro anti-HIV-1 assay, the 50% inhibitory concentrations (IC50) for Aco39-HSA-liposomes and Aco53-HSA-liposomes expressed as protein weight, were 2.87 microgram/ml and 0.154 microgram/ml, respectively. When PEG-PE was incorporated, the Aco53-HSA-liposomes retained anti HIV-1 activity (IC50:3.13 microgram/ml). The possibility to modulate the residence time in the bloodstream of Aco-HSA-liposomes and the potent anti-HIV-1 activity of these conjugates, may allow the development of an intrinsically active drug carrier system. By incorporating anti HIV-1 drugs such as AZT into such liposomes a drug delivery system can be designed that might act simultaneously on the virus/cell binding by virtue of the coupled Aco-HSA and on the RNA/DNA transcription of the HIV-1 replication cycle through the nucleoside analogue.

The water-soluble antineoplastic agent 5-fluoro-2'-deoxyuridine (FUdR) was encapsulated in the water phase of liposomes of different lipid compositions. The retention of this drug upon storage and during contact with plasma was assessed. It was found that, upon refrigeration, diffusion of FUdR across the liposome bilayer was considerably faster when the drug was encapsulated in fluid-type liposomes (egg PC/PS/CHOL) than in solid-type liposomes (DSPC/DPPG/CHOL). With either composition, leakage of the drug from the liposomes was accelerated upon contact with plasma. To achieve improved liposomal retention of the drug, FUdR was converted to a lipophilic prodrug by esterifying the free hydroxyl groups in the deoxyribose moiety with fatty acids of different chain lengths. Thus FUdR-dipalmitate (C-16) and FUdR-dioctanoate (C-8) were synthesized and incorporated in liposomes. The dipalmitoyl derivative could be incorporated upto 13 mol% in solid-type liposomes but to only 2 mol% in fluid-type liposomes. Freeze-fracture electron microscopy revealed no major differences between control liposomes and those containing the prodrug. FUdR-dipalmitate was found to be firmly associated with the liposomal bilayer in both liposome-types: no exchange of the pro-drug with blood constituents or hydrolysis by serum esterases could be registered when the liposomes were incubated with serum. On the other hand, liposome-incorporated FUdR-dioctanoate was found to be readily extracted from the liposomes by serum components (predominantly albumin) and was found to be degraded rapidly by serum esterase activity. The antitumor activity of FUdR-prodrugs was determined using C26 colon adenocarcinoma cells. This cell line was found to be highly sensitive to FUdR. Liposomal FUdR-dioctanoate inhibited cell growth in the same concentration range as unesterified FUdR. FUdR-dipalmitate, however, was more than two orders of magnitude less potent in inhibiting cell proliferation. Its antiproliferative activity was dependent on the liposome-type used: when incorporated in fluid-type liposomes, antiproliferative activity of FUdR-dipalmitate was several-fold higher than in solid-type liposomes. The difference in antitumor activity between FUdR-dipalmitate and FUdR-dioctanoate and between FUdR-dipalmitate in the fluid- and solid-type liposomes could be explained by differences in the rate of hydrolysis of the prodrugs to FUdR by esterase activity in the tumor cells or in the growth medium.

We studied the hepatic processing and biliary excretion of metabolites of the radiolabeled cytostatic agent 5-fluoro,-2'-deoxy[6-3H]uridine (FUdR) and its lipophilic derivative FUdR-dipalmitate incorporated in liposomes. After intracardial injection in rats, free FUdR was cleared from the circulation within minutes. When FUdR or FUdR-dipalmitate was encapsulated in multilamellar vesicles (MLVs) composed of distearoylphosphatidylcholine/dipalmitoylphosphatidylglycerol (DSPC/DPPG/CHOL, 10:1), as expected, the clearance of 3H label was substantially delayed; incorporation of 50 mol% cholesterol in the liposomal bilayer caused a 2-fold further reduction in elimination rate. Incorporation of FUdR-dipalmitate in small unilamellar vesicles (SUV) of similar composition produced a several-fold further decrease in elimination rate: more than 40% of the injected dose was still circulating after 6 h. The plasma concentration of free FUdR after administration of liposomal FUdR-dipalmitate was below the detection limit (5 x 10(-8) M) at any time. Although only about 9% of the administered radioactivity was excreted into the bile within 48 h after injection of [3H]FUdR, a rapid initial excretion rate was observed (4% of the injected dose in the first 2 h). The bile-associated radioactivity was identified mainly as the catabolite alpha-fluoro-beta-alanine (FBAL), conjugated with the three major bile acid species present in rat bile, i.e., muricholic acid, cholic acid and chenodeoxycholic acid in a ratio of 1:3:1. Liposome incorporation of FUdR or FUdR-dipalmitate did not affect the nature of the excretory products but caused a significant decrease in the initial rate at which label appeared in the bile (< 2% in 6 h).

We compared the metabolic fate of [3H]cholesteryl[14C]oleate, [3H]cholesteryl hexadecylether, 125I-labeled bovine serum albumin and [3H]inulin as constituents of large immunoglobulin-coupled unilamellar lipid vesicles following their internalization by rat liver macrophages (Kupffer cells) in monolayer culture. Under serum-free conditions, the cholesteryl oleate that is taken up is hydrolyzed, for the greater part, within 2 h. This occurs in the lysosomal compartment as judged by the inhibitory effect of the lysosomotropic agents monensin and chloroquin. After hydrolysis, the cholesterol moiety is accommodated in the cellular pool of free cholesterol and the oleate is reutilized for the synthesis mainly of phospholipids and, to a lesser extent of triacylglycerols. During incubation in plasma, however, substantial proportions of both the cholesterol and the oleate are shed from the cells, predominantly in the unesterified form. When the liposomes are labeled with the cholesteryl ester analog [3H]cholesteryl hexadecylether only a very small fraction of the label is released from the cells, even in the presence of plasma. Similar to the label remaining associated with the cells, the released label is identified in that case as unchanged cholesteryl ether. The liposomal aqueous phase marker 125I-labeled bovine serum albumin is also readily degraded intralysosomally and the radioactive label is rapidly released from the cells in a trichloroacetic acid-soluble form. Also, as much as 20% of the aqueous phase marker [3H]inulin that becomes cell-associated during a 2-h incubation with inulin-containing liposomes, is released from the cells during a subsequent 4-h incubation period in medium or rat plasma. The usefulness of the various liposomal labels as parameters of liposome uptake and intracellular processing is discussed.

Incorporation of 8 mol% lactosylceramide in small unilamellar vesicles consisting of cholesterol, dimyristoylphosphatidylcholine and phosphatidylserine in a molar ratio of 5:4:1 and containing [3H]inulin as an aqueous-space marker resulted in a 3-fold decreased half-life of the vesicles in blood and a corresponding increase in liver uptake after intracardial injection into rats. The increase in liver uptake was mostly accounted for by an enhanced uptake in the parenchymal cells, while the uptake by the non-parenchymal cells was only slightly increased. The uptake of both the control and the glycolipid-containing vesicles by the non-parenchymal cell fraction could be attributed completely to the Kupffer cells; no radioactivity was found in the endothelial cells. The effect of lactosylceramide on liver uptake and blood disappearance of the liposomes was effectively counteracted by desialylated fetuin, injected shortly before the liposome dose. This observation supports the notion that a galactose-specific receptor is involved in the liver uptake of lactosylceramide liposomes.

In this study we investigated the interaction of liposomes with rat Kupffer cells in maintenance culture by using the lysosomotropic amines ammonium chloride and chloroquine as inhibitors of intralysosomal degradation. The liposomes (large unilamellar vesicles) contained either the metabolically inert 3H-labeled inulin or the degradable 125I-labeled bovine serum albumin. In control incubations, the cells released nearly all accumulated protein label and about 30% of the lipid label when they were incubated in the absence of liposomes, after an initial uptake period of 1 h in the presence of liposomes. This release of label was, for the greater part, suppressed in the presence of ammonia or chloroquine. When the inhibitors were present during the initial uptake period, a several-fold increase in the amount of protein label accumulating in the cells and a smaller, but still marked, increase in lipid label accumulation were observed. The effect of ammonia when present during uptake was readily reversible in contrast to that of chloroquine. Experiments with encapsulated inulin revealed that both lysosomotropic agents also affected the uptake process per se to some extent, probably as a result of impaired membrane/receptor recycling. Labeled liposomes adsorbed to the cells at 4 degrees C were effectively internalized and processed intracellulary after shifting the temperature to 37 degrees C, even when a 500-fold excess of unlabeled liposomes was present in the medium during the 37 degrees C incubation. The observed effects of ammonia and chloroquine indicate that, after uptake, the liposomes are degraded within lysosomes, thus confirming our previous conclusion that endocytosis is the major uptake mechanism at 37 degrees C. From the temperature-change experiments we conclude that, at 4 degrees C, the liposomes are bound with high affinity to the cells, remaining firmly attached to the cell-surface structures which initiate their internalization when the temperature is raised to 37 degrees C.

Incorporation of 8 mol percent lactosylceramide into small unilamellar vesicles consisting of cholesterol and sphingomyelin in an equimolar ratio and containing [3H] inulin as a marker resulted in an increase in total liver uptake and a drastic change in intrahepatic distribution of the liposomes after intravenous injection into rats. The control vesicles without glycolipid accumulated predominantly in the hepatocytes, but incorporation of the glycolipid resulted in a larger stimulation of Kupffer-cell uptake (3.2-fold) than of hepatocyte uptake (1.2-fold). Liposome preparations both with and without lactosylceramide in which part of the sphingomyelin was replaced by phosphatidylserine, resulting in a net negative charge of the vesicles, were cleared much more rapidly from the blood and taken up by the liver to higher extents. The negative charge had, however, no influence on the intrahepatic distributions. The fast hepatic uptake of the negatively charged liposomes allowed competition experiments with substrates for the galactose receptors on liver cells. Inhibition of blood clearance and liver uptake of lactosylceramide-containing liposomes by N-acetyl-D-galactosamine indicated the involvement of specific recognition sites for the liposomal galactose residues. This inhibitory effect of N-acetyl-D-galactosamine was shown to be mainly the result of a decreased liposome uptake by the Kupffer cells, compatible with the reported presence of a galactose specific receptor on this cell type (Kolb-Bachofen et al. (1982) Cell 29, 859-866). The difference between the results on sphingomyelin-based liposomes as described in this paper and those on phosphatidylcholine-based liposomes as published previously (Spanjer and Scherphof (1983) Biochim. Biophys. Acta 734, 40-47) are discussed.

Crude mitochondrial preparations from Neurospora crassa contain high levels of lysophospholipase (EC 3.1.1.5) activity when assayed with lysophosphatidylcholine as a substrate. In mitochondria purified by centrifugation on a sucrose-density gradient this activity is virtually absent. The enzyme was shown to be linked to a contaminating cell fraction which mainly consists of cell-wall material as was demonstrated by electron microscopy and chemical analysis. The enzyme has no absolute Ca2+ requirement but it is slightly stimulated by 10 mM CaCl2. The pH optimum is 5.8 in presence of CaCl2 and is shifted to 4.2 when EDTA is present. In contrast to other lysophospholipases this enzyme is only slightly inhibited by deoxycholate. This detergent is able to release part of the lysophospholipase activity from the wall fragments without producing an increase in specific activity. The enzyme is possibly secreted by the cells as high lysophospholipase activities were also found in the culture medium.

Hepatic cholesterol metabolism was studied in rats with a permanent biliary drainage. Three cholesterol vehicles were used to discriminate between metabolic pathways of cholesterol in the liver. [3H]Cholesterol was administered intravenously associated with rat serum lipoproteins, multilamellar (MLV) or small unilamellar (SUV) liposomes. The liposomes were made from cholesterol, sphingomyelin and phosphatidylserine in a 5:4:1 molar ratio. Initial blood elimination differed markedly for the three vehicles: 15 min after injection the 3H radioactivity content of blood for MLV, SUV and lipoprotein was 3, 50 and 54% of the injected dose, respectively. After about 30 min, MLV-cholesterol label started to reappear in the blood, probably after processing of the vehicle by the Kupffer cells. For all vehicles about 80% of the cholesterol label had been excreted in bile after 120 h, predominantly as bile acids. Initial biliary excretion was highest for lipoproteins (5.7% at 1 h), followed by MLV and SUV (1.3 and 1.2%, respectively). No differences in the radioactivity of excreted bile acids were detectable between the three vehicles at 12 h after injection. However, at 1 h the radioactivity in the muricholic acid fraction was markedly increased, as compared to the other bile acids after injection of SUV-cholesterol, but not after injection of MLV- or lipoprotein-cholesterol. Also, the glycine/taurine conjugation ratio of bile acids was increased for SUV-cholesterol at 1 h as compared to that for the other two vehicles. Since SUV appear to donate their cholesterol to a pool which preferentially supplies cholesterol for muricholic acid synthesis, we conclude that more than one cholesterol pool exists in the hepatocytes from which cholesterol can be recruited for bile acid synthesis. Zonal heterogeneity might be responsible for the observed differences.

In vivo uptake and processing by liver macrophages (Kupffer cells) of liposomes, covalently coated with rabbit immunoglobulin (Ig liposomes) was studied following intravenous injection in rats. Rabbit Ig liposomes were labeled with trace amounts of cholesteryl[14C]oleate and [3H]cholesteryl hexadecyl ether. 1 h after injection of the liposomes, the non-parenchymal cells were isolated and subjected to centrifugal elutriation with stepwise-increasing flow rates; thus, five sub-fractions of Kupffer cells were obtained ranging in size from 9 to 14 micron in diameter. The cells were assayed for peroxidase activity and protein content. Rabbit Ig liposomes were taken up preferentially by Kupffer cells with diameters larger than 11 micron, which constitute less than 25% of the total Kupffer cell population. The intralysosomal degradation of the ingested liposomes was monitored by measuring the 3H/14C ratio of the cells. Due to the rapid release from the cells of the [14C]oleate formed from the cholesteryl[14C]oleate and the virtually complete retention of the non-metabolizable [3H]cholesteryl hexadecyl ether the 3H/14C ratio of the cells increases with proceeding hydrolysis of the liposomes. Thus, we were able to show that, in vivo, the Kupffer cells of the larger size classes, are not only more active in liposome uptake, but are also substantially more active in liposome degradation than smaller cells. The maintenance of the observed heterogeneity of rat liver Kupffer cells, with respect to liposome uptake under in vitro culture conditions, was examined. Subfractions were maintained in monolayer culture for 2 days and incubated with rabbit Ig liposomes. Binding and uptake of liposomes by the cells was monitored by measuring cell-associated radioactivity at 4 degrees C and 37 degrees C, respectively. In contrast to our in vivo results, we observed maximal in vitro liposome binding and uptake in those subfractions containing small cells (10-11 micron diameter), while the fractions containing cells larger than 12 micron, which were more active in vivo, were substantially less active than the smaller cells. The maximum we observed was even more pronounced when the liposome concentration was increased. We conclude that liver macrophage subfractions that barely participate in liposome uptake from the bloodstream in vivo, possess the potential to develop the capacity in vitro to phagocytose rabbit Ig-coated liposomes to extents equal to or even higher than the cells belonging to those subfractions containing the phagocytically most active cells under in vivo conditions.

We investigated the effect of cholesterol on the uptake and intracellular degradation of liposomes by rat liver and spleen macrophages. Multilamellar vesicles (MLV) consisting of distearoylphosphatidylcholine/phosphatidylserine (molar ratio 9:1) or distearoylphosphatidylcholine/cholesterol/phosphatidylserine (molar ratio 4:5:1) were labeled with [3H]cholesteryl hexadecyl ether and/or cholesteryl [14C]oleate. After i.v. injection the cholesterol-containing liposomes were eliminated less rapidly from the bloodstream and taken up to a lesser extent by the liver (macrophages) than the cholesterol-free liposomes. Assessment of the 3H/14C ratios in liver and spleen cells revealed that the cholesterol-containing liposomes are substantially more resistant towards intracellular degradation than the cholesterol-free liposomes. These results could be confirmed by measuring the release of 111In from liposomes after uptake by liver and spleen by means of gamma-ray perturbed angular correlation spectroscopy. Experiments with cultured Kupffer cells in monolayer also revealed that incorporation of cholesterol results in a decrease of the uptake and an increase of the intracellular stability of cholesteryl [14C]oleate-labeled liposomes. Finally, incubation of both types of liposomes with lysosomal fractions prepared from rat liver demonstrated a difference in susceptibility to lysosomal degradation: the cholesterol-free vesicles were much more sensitive to lysosomal esterase than the cholesterol-containing liposomes. These results may be relevant to the application of liposomes as a drug carrier system to liver and spleen (macrophages).

We studied the in vitro interaction between Zajdela ascites hepatoma cells and small unilamellar vesicles, consisting of 14C-labeled phosphatidylacholine, cholesterol, and phosphatidylserine (molar ratio 5 : 4 : 1), containing high intravesicular concentrations of carboxyfluorescein or fluorescein isothiocyanate tagged dextran. The entrapped markers were found to be associated with the cells to a lesser degree than the vesicle membrane marker. This discrepancy, which is slightly less pronounced for fluorescein isothiocyanate tagged dextran than for carboxyfluorescein, increases with incubation time and decreases with increasing vesicle lipid concentration in the incubation mixture. Vesicle-plasma membrane exchange of the vesicle lipid marker could not entirely explain the observed discrepancy. It is tentatively concluded that the gap mainly arises from a selective loss of entrapped dyes from vesicles actually interacting with the cell surface. Both spectrofluorimetry and fluorescence microscopic observations, as well as the relative insensitivity of vesicle uptake towards the presence of metabolic inhibitors, exclude a major contribution of endocytosis as a vesicle uptake route. We therefore conclude that vesicles are primarily internalized by a vesicle-cell fusion-like process. The observed discrepancy in uptake between entrapped materials and vesicle lipid is discussed in terms of a two-site vesicle-cell surface interaction model.

When the water-soluble cholesterol derivative, N-[tris [(beta-D-galactopyranosyloxy)methyl]methyl]-N alpha-[4-(5-cholesten-3 beta-yloxy)succinyl]glycinamide (tris-gal-chol) (Kempen et al. (1984) J. Medicin. Chem. 27, 1306-1312) is added as an aqueous micellar solution to a dispersion of small unilamellar phospholipid vesicles it rapidly associates with the vesicles, without causing significant leakage of liposome contents. Incorporation of 10 mol% tris-gal-chol in the liposomal membrane caused a substantial increase in the rate and extent of rat liver uptake and a shift in intrahepatic distribution of an intravenously administered dose of liposomes. For neutral liposomes composed of equimolar amounts of cholesterol and sphingomyelin incorporation of tris-gal-chol led to a 7-fold increase in total liver uptake, which was mainly accounted for by an increase in uptake by the Kupffer cells (12-fold) and by only a small increase in uptake by the hepatocytes (1.4-fold). The increased liver uptake is blocked by preinjection of N-acetyl-D-galactosamine and not affected by preinjection of N-acetyl-D-glucosamine. This indicates that the increased interaction of liposomes as a result of tris-gal-chol incorporation is mediated by galactose-specific recognition sites on both Kupffer cells and hepatocytes. Targeting of liposomes to the asialoglycoprotein receptor of the hepatocytes is thus frustrated by the highly active galactose-specific receptor on Kupffer cells. Comparable results on lactosylceramide incorporation into liposomes were recently reported by us (Spanjer et al. (1984) Biochim. Biophys. Acta 774, 49-55).

We investigated the intrahepatic distribution of small unilamellar liposomes injected intravenously into rats at a dose of 0.10 mmol of lipid per kg body weight. Sonicated liposomes consisting of cholesterol/sphingomyelin (1:1), (A); cholesterol/egg phosphatidylcholine (1:1), (B); cholesterol/sphingomyelin/phosphatidylserine (5:4:1), (C) or cholesterol/egg-phosphatidylcholine/phosphatidylserine (5:4:1), (D) were labeled by encapsulation of [3H]inulin. The observed differences in rate of blood elimination and hepatic accumulation (A much less than B approximately equal to C less than D) confirmed earlier observations and reflected the rates of uptake of the four liposome formulations by isolated liver macrophages in monolayer culture. Fractionation of the liver into a parenchymal and a non-parenchymal cell fraction revealed that 80-90% of the slowly clearing type-A liposomes were taken up by the parenchymal cells while of the more rapidly eliminated type-B liposomes even more than 95% was associated with the parenchymal cells. Incorporation of phosphatidylserine into the sphingomyelin-based liposomes caused a significant increase in hepatocyte uptake but a much more substantial increase in non-parenchymal cell uptake, resulting in a major shift of the intrahepatic distribution towards the non-parenchymal cell fraction. For the phosphatidylcholine-based liposomes incorporation of phosphatidylserine did not increase the already high uptake by the parenchymal cells while uptake by the non-parenchymal cells was only moderately elevated; this resulted in only a small shift in distribution towards the non-parenchymal cells. The phosphatidylserine-induced increase in liposome uptake by non-parenchymal liver cells was paralleled by an increase in uptake by the spleen. Fractionation of the non-parenchymal liver cells in a Kupffer cell fraction and an endothelial cell fraction showed that even for the slowly eliminated liposomes of type A endothelial cells do not participate to a measurable extent in the elimination process, thus excluding involvement of fluid-phase pinocytosis in the uptake process.

Activation of rat liver macrophages with free and liposome-encapsulated muramyl dipeptide (MDP) to a tumorcytotoxic state was characterized by employing various experimental conditions. Macrophage-mediated tumor cytotoxicity was determined using two standard assay systems: a [methyl-3H]thymidine release assay to measure the extent of tumor cell lysis and a [methyl-3H]thymidine incorporation assay to measure the combined effects of tumor cell lysis and stasis. The extent of cell lysis was not affected by the ratio of macrophages to tumor cells within the ratio range of 30:1 to 5:1, provided that the macrophages form a confluent monolayer. Tumor cell lysis, however, was significantly influenced by macrophage density; a low macrophage density for example resulted in a low percentage of tumor cell lysis. Tumor target cells used in this study, i.e., C26 adenocarcinoma, B16 melanoma and P815 mastocytoma, differed in their susceptibility towards macrophage-mediated cell lysis, whereas no differences were observed with respect to tumor cell stasis. Non-tumorigenic cell lines such as human fibroblastic cells and LLC monkey kidney cells were not lysed by activated macrophages, although proliferation of these cells was markedly inhibited. Additionally, the effects of liposomal lipid composition on macrophage activation were studied. With a basic composition of phospholipid/cholesterol/dicetylphosphate, we used either egg-yolk, dipalmitoyl-, distearoyl- or dihexadecylphosphatidylcholine as the bulk phospholipid constituent. Although these liposomes display a widely different susceptibility to lysosomal phospholipase activities, we could not detect any significant difference in either the extent or the duration of the tumoricidal activity induced by MDP encapsulated in these different types of liposomes.

Glycolithocholic acid and its sulfated derivative are major metabolites of the secondary bile acid lithocholic acid in man. Both compounds are known to induce cholestasis in experimental animals. We compared the effects of these endogenous hepatotoxins on bile production and biliary lipid composition in rats with chronic biliary drainage. The compounds were administered enterally at relatively low rates (5-50% of the rats' endogenous bile acid secretion in these experiments) to simulate enterohepatic circulation. Both compounds were substantially secreted into bile (more than 90% of dose); sulfated glycolithocholic acid unchanged and glycolithocholic acid after hepatic hydroxylation predominantly in the form of glyco-beta-muricholic acid (cf. Kuipers et al. (1986) Am. J. Physiol. 251, G189-G194). Neither glycolithocholic acid nor its sulfated derivative affected the biliary excretion of endogenous bile acids or bile flow in these experiments. In spite of this, phospholipid and cholesterol secretion were significantly reduced by sulfated glycolithocholic acid but were not altered by glycolithocholic acid. Phospholipid and cholesterol secretion rapidly decreased to 25 and 50% of their initial values, respectively, at biliary output rates of sulfated glycolithocholic acid up to 2 mumol/h, and did not further decrease when this output was increased to 6 mumol/h. Small unilamellar liposomes consisting of cholesterol, [Me-14C]choline-labeled phosphatidylcholine, phosphatidylserine and [3H]cholesteryl oleate in a 5:4:1:0.1 molar ratio were employed to label intrahepatic lipid pools. Administration of sulfated glycolithocholic acid slightly reduced bile acid synthesis from [3H]cholesteryl oleate, but significantly reduced the biliary secretion of [14C]phospholipid. Glycolithocholic acid did not affect the hepatic processing of liposomal lipids. It is concluded that sulfated glycolithocholic acid at low doses causes the uncoupling of biliary lipid secretion from that of bile acids, which might represent in initiating event in sulfated glycolithocholic acid hepatotoxicity.

Intravenously administered gadolinium chloride caused only a slight decrease in the rate of elimination of small unilamellar liposomes from the blood and had no influence on the total hepatic uptake of these vesicles, but did alter their intrahepatic distribution substantially. Uptake by the non-parenchymal cells was substantially decreased, whereas uptake by the parenchymal cells showed a concomitant increase. Our earlier observations (Roerdink et al. (1981) Biochim. Biophys. Acta 677, 79-89) on the effect of lanthanides on the in vivo distribution of multilamellar liposomes have been extended, in that we demonstrate, in addition to the drop in elimination rate from the blood and in the over-all hepatic uptake, a shift of liposome distribution within the Kupffer cell population. While the larger Kupffer cells, which normally take up a major fraction of an injected liposome dose, were strongly inhibited in liposome uptake, the more numerous small macrophages showed a 3-4-fold increase in uptake.