Uptake and processing of liposomal phospholipids by Kupffer cells in vitro

Transcription

1 Eur. J. Biochem. 148, (1985) 0 FEBS 1985 Uptake and processing of liposomal phospholipids by Kupffer cells in vitro Jan DIJKSTRA, Mieke VAN GALEN, Djoeke REGTS and Gerrit SCHERPHOF Laboratory of Physiological Chemistry, University of Groningen (Received November 5, 1984/January 8,1985) - EJB We investigated the intracellular metabolic fate of [Me-'4C]choline-labeled phosphatidylcholines and sphingomyelin taken up by rat Kupffer cells in maintenance culture during interaction with large unilamellar liposomes composed of cholesterol, labeled choline-phospholipid and phosphatidylserine (molar ration 5 : 4 : 1). With both labeled compounds only small proportions of water-soluble radioactivity were found to accumulate in the cells and in the culture medium, suggesting limited phospholipid degradation. However, after a lag period of 30 min progressively increasing proportions of cell-associated liposomal phospholipid were found to be converted to cellular phospholipid, nearly all of which was phosphatidylcholine. This conversion as well as the limited release of watersoluble label from the cells was inhibited by the lysosomotropic agents ammonium chloride and chloroquine. With [Me-'4C]choline-labeled lysophosphatidylcholine, label was found to become cell-associated far in excess of an encapsulated liposomal label, [3H]inulin. Without a lag period virtually all of this was rapidly converted to phosphatidylcholine, a process which was not inhibited by the lysosomotropic agents.. It is concluded that Kupffer cells, after endocytosis of liposomes, degrade the liposomal phospholipids effectively but reutilize the choline moiety for de novo synthesis of cellular phosphatidylcholine. The dominant role of the resident liver macrophages (or Kupffer cells), in the clearance of intravenously administered phospholipid vesicles (liposomes) [I - 31 led us to initiate a study on the interaction of liposomes with Kupffer cells in a maintenance culture system [4, 51. Such a system in vitro provides the most direct approach to investigate the mechanism of liposome-kupffer cell interaction and to answer questions arising from studies in vivo [2, 61. From a previous study we concluded that the major mechanism by which liposomes, at 37"C, interact with Kupffer cells is a process of adsorptive endocytosis [4]. After internalization of the vesicles, radioactively labeled lipid and encapsulated protein were both effectively degraded by the cells. Fluorescence and electron microscopic observations indicated that the lysosomes are the intracellular degradation sites of the liposomal components and a study of the effects of lysosomotropic amines on uptake and degradation of liposomal constituents confirmed the involvement of the lysosomal system in the breakdown of internalized vesicles [55]. In comparison with the almost complete degradation of entrapped albumin, liposomal phospholipid was degraded to a much lower extent [5]. This observation prompted us to investigate in more detail the intracellular fate of the liposomal phospholipid label after uptake. In this study we report substantial reutilization of the choline moiety of the liposomal phospholipid for de novo synthesis of cellular phosphatidylcholine, after the intralysosomal degradation of sphingomyelin-containing liposomes. It is concluded that liposomal phospholipid degradation much more closely parallels degradation of encapsulated protein than becomes apparent from the accumulation of degradation products. Correspondence to G. L. Scherphof, Laboratorium voor Fysiologische Chemie, Rijksuniversiteit te Groningen, Bloemsingel 10, NL-9712-KZ Groningen, The Netherlands MATERIALS AND METHODS Lipids The phospholipids egg 3-sn-phosphatidylcholine (Sigma, type V-E, from egg-yolk), dimyristoyl-sn-glycero-3-phosphocholine (Calbiochem) and sphingomyelin (Sigma, from bovine brain) were labeled in the choline moiety with [ 14C]methyliodide (Amersham) according to Stoffel [7]. The labeled products were purified by thin-layer chromatography on Silica Gel HF (Merck); contaminating silica particles were removed by partition chromatography on a Sephadex G-25 column [8]. The radioactive lipids obtained had a specific activity of about 3 pci/pmol lipid. Egg 3-sn-2-lysophosphatidyl['4C]choline was prepared by phospholipase A2 (Sigma, from bee venom) treatment of sonicated egg 3-snphosphatidyl-[ 14C]choline. The products were extracted by the method of Bligh and Dyer [9] and purified by thin-layer chromatography on Silica Gel HF. 3-sn-Phosphatidylserine, isolated from bovine brain and converted to the sodium salt as described previously [lo, 111, was a gift from Dr. J. C. Wilschut of this laboratory. Cholesterol was obtained from Sigma (type CH-S). The lipid solutions in chloroform/methanol (1 : 1) were stored under nitrogen at - 20 "C. Routinely, the lipids were checked for purity and, if necessary, repurified by thin-layer chromatography Liposomes Large unilamellar vesicles were prepared in 135 mm NaCl, 10 mm Hepes, ph 7.4, according to Szoka and Papahadjopoulos [I21 as described in more detail before [4]. The liposomes were composed of a radioactively marked choline phospholipid, cholesterol and phosphatidylserine (molar ratio 4: 5 : 1) unless otherwise indicated. In some experiments [3H]inulin (Amersham) was encapsulated in the vesicles as a metabolically inert marker of the aqueous contents [4, 51. All

2 392 liposome preparations were sized by extrusion through 0.4 pm polycarbonate filters (Uni-Pore, Bio-Rad) [13] and freed from non-encapsulated marker on a Sephadex G-100 column [4]. Isolation and culture of Kupffer cells The Kupffer cells were isolated by pronase digestion of livers of female Wistar rats and purified by centrifugal elutriation [14] with some minor modifications as described in detail previously [4]. The cells were kept in maintenance culture as described before [15] in Dulbecco's modified Eagle's medium with L-glutamine, containing 20 mm Hepes and 10 mm NaHC03, ph 7.4. The medium was supplemented with penicillin (100 U/ml), streptomycin (100 pg/ml) and heat-inactivated fetal calf serum (20% v/v); the osmolarity was 285 mosm. For experiments in which analysis of labeled lipids was required 12.5 x lo6 cells in 10 ml culture medium were seeded on plastic Petri dishes of 9 cm diameter. In experiments where only the extent and rate of label association were studied 1.8 x lo6 cells in 1.5 ml medium were seeded on 3.5-cm Petri dishes. lo6 initially seeded cells yield a protein content of about 65 pg per dish. Further culture conditions were as described before [5]. Incubation of liposomes with the cells The cells were used for incubation experiments after 2 days in maintenance culture [4]. Incubations on 9-cm dishes were done in 7 ml incubation medium and those on 3.5-cm dishes in 1 ml incubation medium. The latter were done in duplicate. The incubation medium consisted of culture medium without antibiotics or serum. After 20 min of preincubation the liposomes were added as pl aliquots of concentrated suspensions. Ammonia and chloroquine, when added, were already present during the preincubation period. After incubation, the 9-cm dishes were washed six times with cold phosphate-buffered saline and then the cells were scraped from the dishes in ice-cold distilled water. Aliquots were taken for determination of protein content and radioactivity measurements. The remainder was lyophilized and the lipids were extracted according to Folch et al. [16]. Both the chloroform and the aqueous methanol phase were sampled for radioactivity determination. The lipids were separated by chromatography on a 0.25-mm layer of Silica Gel (Merck, 60 F 254) with chloroform/methanol/conc. ammonia/water (90: 54: 5.5 : 5.5, by vol.) as a solvent system. After identification, the spots were scraped from the plates and transferred into liquid scintillation vials containing 0.5 ml water. 5 ml Hydroluma (LUMAC BV, The Netherlands) was added as a scintillation fluid. When the 3.5-cm dishes were used, the cells were washed as described above and digested for 2 h at room temperature in 0.5 ml 0.5 M NaOH. Aliquots were taken for protein and radioactivity measurements. Miscellaneous methods Leakage of [3H]inulin from the liposomes into the incubation medium was determined by gel filtration [4]. The release of water-soluble lipid degradation products from the cells was detected by extraction of the lyophilized incubation medium according to the method of Folch, as described in the previous section. Phospholipid phosphorus was assayed according to Ames and Dubin [17] and protein was determined by the method of Lowry et al. [18] in 1% sodium dodecyl sulfate, with bovine serum albumin as a standard incubation time (min) Fig. 1. Fate of choline-labeled eggphosphatidylcholine during uptake of liposomes by cultured Kupffer cells. Liposomes (500 nmol of total lipid) consisting of egg phosphatidyl[' 4C]choline, cholesterol and phosphatidylserine (molar ratio 4: 5: 1) were incubated with 13 x lo6 Kupffer cells at 37 C. After 60 min (arrow) the medium of a number of dishes was replaced by a liposome-free medium and the incubation was continued. Incubation of the remaining dishes was continued without medium change. At the times indicated the medium was collected and the cells were rinsed with cold buffered saline and further processed as described in the methods section. (A) Total cell-associated label, ( x ) chloroform-soluble label, ( W) water-soluble label, (0) label associated with phosphatidylcholine. The dashed curves represent label association after medium replacement. The measured amounts of label were all converted to liposome equivalents, i.e. nmol of total liposomal lipid/mg of cell protein RESULTS Conversion of liposomal phospholipid into cellular phosphatidylcholine The kinetics of uptake of egg phosphatidyl [14CJcholinecontaining liposomes by cultured Kupffer cells is shown in Fig. 1. After a rapid initial phase the rate of uptake gradually levelled off during continued incubation. When after 65 min the liposome-containing medium was removed and replaced by a liposome-free medium, about one-third of the label was released from the cells during the following 2.5 h of incubation. Simultaneously, proportional amounts of label appeared in the medium, predominantly in degraded form (96% of the released label was water-soluble). Up to 15 min after the start of the incubation nearly all cell-associated label was still chloroform-soluble, but during prolonged incubation a gradually increasing discrepancy between chloroform-soluble and total cell-associated label became apparent. This was nearly completely accounted for by the amount of watersoluble radioactivity accumulating in the cells, probably as a result of enzymatic hydrolysis. Apparently, after a lag phase of about 15 min, the cells start to degrade the liposomal phosphatidylcholine and water-soluble products are, at least partially, released by the cells. Of the label in the chloroform phase at least 95% was found to migrate with the phosphatidylcholine during thin-layer chromatography, the remaining 5% being distributed between the lysophosphatidylcholine and, to a lesser extent, the sphingomyelin spot. Because hydrolysis of liposomal phospholipid as observed in these experiments was not nearly as extensive as we found for degradation of encapsulated protein [5], we considered the possibility that we were underestimating phospholipid hydrolysis, because degradation products were reutilized for

3 393 incubation time (mid Fig. 3. Intracellular conversion of liposomal choline-labeled sphingomyelin into phosphatidylcholine at 37 C after binding of liposornes at 4 C. 11 x lo6 Kupffer cells were pre-incubated at 4 C with liposomes (490 nmol total lipid) consisting of 14C-labeled sphingomyelin, cholesterol and phosphatidylserine (molar ratio 4: 5: 1). After 90 min (indicated as zero time) the medium was replaced by a liposome-free medium and the incubation was continued at 37 C. At the times indicated the relative amounts of label in the chloroform-soluble cell extracts associated with sphingomyelin (0) and phosphatidylcholine (0) were determined incubation time (mid Fig. 2. Conversion of choline-labeled liposomal sphingomyelin into cellularphosphatidylcholine during uptake of liposomes by Kupffer cells. Liposomes (455 nmol lipid) composed of I4C-labeled sphingomyelin, cholesterol and phosphatidylserine (molar ratio 4: 5 : 1) were incubated at 37 C with 11.5 x lo6 Kupffer cells. After 60 min (arrow) the medium of a number of dishes was replaced by a liposome-free medium and the incubation was continued (dashed lines). The other dishes were incubated without interruption (solid lines). At the times indicated the medium was collected and the cells were further processed for determination of the amounts of total cell-associated label (A), chloroform-soluble label ( x ), water-soluble label (H), label in sphingomyelin (0) and label in phosphatidylcholine (0).(A) Label association expressed as nmol of total lipid/mg of cell protein. (B) Relative amounts of label in the chloroform-soluble cell extracts associated with sphingomyelin and phosphatidylcholine cellular phosphatidylcholine synthesis. To investigate this possibility, we repeated the foregoing experiment with liposomes in which the egg phosphatidylcholine was replaced by [' 4C]choline-labeled sphingomyelin. The kinetics of uptake and release of the sphingomyelin label (Fig.2A) were essentially similar to those found for the phosphatidylcholine label. Thus, most of the cell-associated label remained chloroform-soluble, while the label released into the medium after medium replacement was mostly (91 YO) water-soluble. The chloroform-soluble label in the cells, however, turned out to consist of phosphatidylcholine and sphingomyelin; the relative amount of phosphatidylcholine increased linearly at the expense of the proportion of sphingomyelin. This conversion of labeled sphingomyelin into labeled phosphatidylcholine became manifest after a lag period of about 30 min and resulted, after 3.5 h of incubation, in a label distribution of 30% phosphatidylcholine and 70% sphingomyelin (Fig. 2B). The conversion became even more prominent when, after an initial 1-h incubation at 4 C during which the liposomes adsorbed to the cells [5], incubation was continued at 37 C in liposome-free medium (Fig. 3). Nearly one half of the label which became cell-associated during the 4 C incubation was released in the medium in degraded, water-soluble, form during the subsequent incubation at 37 "C (results not shown). This release of label from the cells, and the appearance of water-soluble degradation products in the cells, as well as the incorporation of label into endogenous phosphatidylcholine, all started after a lag phase of min. It is likely that this is the period needed for the liposomes to reach the intracellular degradation sites, presumably, as we concluded previously [5], the lysosomal system. Apparently, the transfer of label from sphingomyelin to phosphatidylcholine is preceded by intralysosomal degradation of the liposomal lipids. Endogenous phosphatidylcholine can also become labeled during incubation with liposomes containing dimyristoylglycerophospho['4c]choline; the relatively short acyl chains of this phosphatidylcholine allow appropriate chromatographic separation from the cellular phosphatidylcholine. The kinetics of uptake, release and conversion of label were very similar to those found for sphingomyelin (Fig. 4A). The release of degraded label (about 90% water-soluble) from the cells after medium replacement, was somewhat higher than for sphingomyelin or egg phosphatidylcholine (about 50% released in 2.5 h), indicating a slightly more efficient degradation of the synthetic phospholipid. This conclusion is supported by the relatively high rate of conversion of dimyristoylglycerophosphocholine into cellular phosphatidylcholine as presented in Fig. 4B (compare Fig. 2B). The 2-4% of chloroform-soluble radioactivity not associated with dimyristoylglycerophosphocholine or cellular phosphatidylcholine, co-chromatographed with lysophosphatidylcholine and sphingomyelin. Effects of ammonia and chloroquine on the lipid conversion To ascertain that intralysosomal degradation is essential for the process of incorporation of the choline moiety from the liposomal sphingomyelin or dimyristoylglycerophosphocholine label into endogenous phosphatidylcholine, we examined the effects of ammonia and chloroquine on this process. Both agents were found to inhibit intracellular degradation of liposomal lipid and protein [5]. Fig.5A shows the accumulation of liposomal sphingomyelin label in the absence and presence of the inhibitors. In the absence of inhibitor the conversion of sphingomyelin to phosphatidylcholine during uptake amounted again to about 40% in 2.5 h (Fig. 5B) and the accumulation of water-soluble label to 15%. In the pres-

4 394 *ra I ence of ammonia nearly all cell-associated radioactivity remained chloroform-soluble during the entire incubation period (Fig. 5C) and almost all of this label remained associated with sphingomyelin (Fig. 5C). With chloroquine the results were essentially the same as with ammonia (Fig. 5D). As far as the dimyristoylglycerophosphocholine-containing liposomes are concerned the effects of ammonia and chloroquine were essentially similar to those found with sphingomyelin, i.e. the kinetics of uptake and the conversion into cellular phosphatidylcholine were all similarly influenced although the inhibitory effects were slightly less pronounced than with the sphingomyelin liposomes (results not shown). The somewhat lower extent of inhibition observed with the dimyristoylglycerophosphocholine may, in part, be due to a small contamination with its lyso-derivative as will be demonstrated in the following section. The effects of the lysosomotropic amines on the conversion of liposomal sphingomyelin or dimyristoylglycerophosphocholine into cellular phosphatidylcholine strongly support the notion that this process involved the intralysosomal degradation of the liposomal phospholipid. incubation time (mid Fig. 4. Conversion of liposomal choline-labeled dimyristoyglycerophosphocholine inlo cellular phosphatidylcholine during uptake of liposomes by Kupffer cells. Liposomes (475 nmol lipid) composed of dimyristoylglycerophospho[ ''C-]choline, cholesterol and phosphatidylserine (molar ratio 4: 5: 1) were incubated at 37 C with 11.8 x lo6 Kupffer cells. After 60 min incubation the medium of a number of dishes was replaced by a liposome-free medium and the incubation continued (dashed lines). Solid lines represent uninterrupted incubations. At the times indicated the medium was collected and the cells were processed for determination of the amounts of total cellassociated label (A), chloroform-soluble label ( x ), water-soluble label (m), label in dimyristoylglycerophosphocholine (0) and label in phosphatidylcholine (0).(A) Label association was expressed as nmol of total liposomal lipid/rng of cell protein. (B) Relative amounts of label in the chloroform-soluble cell extracts associated with dirnyristoylglycerophosphocholine and phosphatidylcholine Con version of lysophosphatidylcholine into phosphatidylcholine. Effects of lysosomotropic amines Contamination of phosphatidylcholine preparations with small amounts of lysophosphatidylcholine is hard to avoid. Lysophospholipids are relatively water-soluble and may conceivably be transferred independently from liposomes to cells through the aqueous phase. Therefore we investigated the fate of labeled lysophosphatidylcholine, purposely incorporated in liposomes consisting of otherwise unlabeled lipids (0.2 mol% lysophosphatidylcholine in a 4: 1 : 5 molar ratio of egg phosphatidylcholine, phosphatidylserine and cholesterol, but containing [3H]inulin in the encapsulated volume. Fig. 6 shows that apparent liposome uptake, when calculated from lysophosphatidylcholine label, exceeded uptake calculated from inulin label by a factor of 5 to 7. This demonstrates that the lysophospholipid is preferentially transferred from the liposomes to the cells and, therefore, does not represent a realistic measure of liposome uptake; incubation time (minl Fig. 5. Effects of ammonia and chloroquine on the intracellular conversion of liposomal choline-labeled sphingomyelin into cellular ~hosphatidylchffl~ne during uptake of liposomes. Liposomes (435 nmol lipid) consisting of 14C-labeled sphingomyelin, cholesterol and phosphatidylserine (molar ratio 4:5: 1) were incubated with 12.6 x lo6 Kupffer cells at 37 C in the absence of lysosomotropic agents (A), in the presence of 10 mm NH4CI (A) or in the presence of 40 pm chloroquine (0).(A) At the times indicated the amounts of total cell-associated label were measured. The cell extracts were further processed for determination of the amounts of chloroform-soluble label ( x ), water-soluble label (W), label in sphingomyelin (0) and label in phosphatidylcholine (0) during incubation in the absence of inhibitors (B), in the presence of ammonia (C) and in the presence of chloroquine (D), respectively

5 incubotion ti(min) Fig. 6. Preferential transfer of choline-labeled lysophosphatidylcholine from liposomes to Kupffer cells. Effects of ammonia and chloroquine. Liposomes (900 nmol lipid) composed of egg phosphatidylcholine, cholesterol, phosphatidylserine and lys~phosphatidyl['~c]choline (molar ratio 4: 5: 1 : 0.15 ) and containing encapsulated [3H]inulin were incubated at 37 C with 1.8 x lo6 Kupffer cells without further additions (A), with 10 mm NH4Cl (x) or with 10 pm chloroquine (0). At the times indicated the total amounts of the cell-associated labels were determined and expressed as nmol of total liposomal lipid per mg protein. Uptake was calculated from I4C label (-) and from 'H label (- ---) particularly since the uptake values based on inulin are of the same magnitude as those in the foregoing figures. The transfer of lysophosphatidylcholine to the cells was not significantly affected by the presence of ammonia or chloroquine, while inulin uptake, after prolonged incubation, was somewhat reduced in the presence of ammonia or chloroquine, compatible with our previous observations that vesicle uptake is slightly inhibited by such compounds [5]. The lack of effect of ammonia and chloroquine indicates that the conversion of liposomal lysophosphatidylcholine into cellular phosphatidylcholine does not involve a degradation step in the lysosomal system. In experiments not shown we observed that lysophosphatidylcholine thus associating with the cells is rapidly and without delay converted into phosphatidylcholine with only a minor fraction of the label accumulating as watersoluble product. This conversion of lysophosphatidylcholine to phosphatidylcholine was essentially insensitive to ammonia or chloroquine. DISCUSSION In this study we showed that, in addition to being released from the cells, the labeled choline moiety of sphingomyelin or dimyristoylglycerophosphocholine-containing liposomes is incorporated into endogeneous Kupffer cell phosphatidylcholine. This incorporation became manifest after a lag period of min and was prevented by the presence of the lysosomotropic agents ammonia and chloroquine, indicating 395 that intralysosomal degradation of the liposomes is a prerequisite for the lipid conversion [5]. The capacity of lysosomes to degrade phospholipids to water-soluble products has been studied extensively during recent years. The degradation of sphingomyelin is initiated by the removal of phosphocholine through the action of sphingomyelinase [19, 201. Sphingomyelinase activity has been found in plasma membranes and microsomal membranes, as well as in lysosomes from rat liver [21, 221. Rat liver lysosomes also contain a phosphomonoesterase activity capable of hydrolyzing substrates like phosphocholine [23]. Thus, during degradation of sphingomyelin within the lysosomal compartment, it seems likely that both choline and phosphocholine are produced. The catabolism of phosphoglycerides such as phosphatidylcholine is more complex, as the degradation can start at different positions in the lipid molecule. In lysosomes two pathways can be distinguished, one initiated by phospholipase C [24, 251, the other by phospholipase A activity, either Al or A2 [ The remaining fatty acid can be removed by lysophospholipase A activities [23, 25, 31, 321. The water-soluble product glycerophosphocholine is not further degraded within the lysosomes [23] but probably passes the lysosomal membrane and is attacked by glycerophosphocholine diesterase activity liberating choline [33]. This enzyme is abundant in the cytosol and also present in the plasma membrane [23, 341. Alternatively, lysophospholipase C [24, 251 may produce phosphocholine from lysophosphatidylcholine. Irrespective of the pathway involved, degradation ultimately results in the formation of phosphocholine or choline from phosphatidylcholine. In liver, the predominant metabolic route by which choline is incorporated into phosphatidylcholine is the CDP-choline pathway [35, 361. Another possibility for incorporation of choline into phosphatidylcholine is the exchange of choline with the base of another phosphoglyceride molecule [35]. This reaction was shown to occur in liver microsomal fractions, but is believed not to be important in the in vivo incorporation of choline into phosphatidylcholine [35,37]. It is also conceivable that after degradation of liposomal phosphatidylcholine within the lysosomes the resulting lysoderivative is reacylated by a lysophosphoglyceride-acyltransferase. These acyltransferases, however, are not associated with lysosomal membranes [38]. Moreover, such a reaction would fail to explain the transfer of choline from sphingomyelin to endogenous phosphatidylcholine. Finally, the recently reported transfer of a phosphocholine moiety from phosphatidylcholine to ceramide as a major pathway for sphingomyelin synthesis [39], should be considered here. If this reaction, which has been localized in the plasma membrane of rat liver cells [39] and fibroblasts [40] and which is possibly also present in microsomes derived from mouse liver [41], is readily reversible, cellular phosphatidylcholine might, conceivably, become labeled by liposomal choline-labeled sphingomyelin through this mechanism. However, this would be difficult to reconcile with all the evidence we produced in this and preceding papers in favor of an endocytic route of liposome uptake and of an involvement of the lysosomal system. We tend to interpret our results, therefore, as resulting from partial reutilization of choline or phosphocholine, originating from the intralysosomal degradation of liposomal phospholipids, for de novo phosphatidylcholine synthesis by the CDP-choline pathway. Chloroquine and ammonia, at concentrations as present in our experiments, have been reported to raise the intralysosomal ph in peritoneal macrophages to comparable values [42, 431. This would be compatible with

6 396 our observation that both agents were, at these concentrations, also very similar in their effects on the reutilization of the choline moiety. On the other hand, chloroquine can also inhibit lysosomal hydrolases in a direct manner, as was reported for the phospholipases A and C and the lysophospholipases A and C [44, 451. The reutilization of choline label from sphingomyelin was almost completely blocked by both lysosomotropic agents, while the incorporation of choline from dimyristoylglycerophosphocholine was less efficiently inhibited. Probably, the activity of the lysosomal sphingomyelinase is more strongly dependent on ph than the rate-limiting enzyme(s) in the breakdown of phosphoglycerides. The experiments with liposomes containing small amounts of labeled lysophosphatidylcholine revealed that this marker was transferred to the cells far in excess of the other liposomal markers (Fig. 6). In contrast to the reutilization of the choline moiety of liposomal sphingomyelin or phosphatidylcholine, the conversion of liposomal lysophosphatidylcholine to phosphatidylcholine proceeded without a corresponding lag phase and was not inhibited by lysosomotropic amines, indicating that the lysosomal system is not involved in this process. These observations suggest that the disproportionally transferred lysophosphatidylcholine was quickly acylated after entering the cells. When liposome uptake by cells is assessed only with vesicles marked with labeled phosphoglycerides, the presence of small amounts of contaminating labeled lysoderivatives may lead to a considerable overestimation of real vesicle association. Particularly when only a very small fraction of the added liposomes is taken up, this may cause highly erroneous results. In our experiments as much as 2% of the total labeled phosphatidylcholine present during the incubation was taken up by the cells after 3.5 h. With 1 Yn of the label in liposomes present as lysophospholipid and about 10% of all lysophosphatidylcholine label present in the incubation mixture being transferred to the cells, it can be calculated that no more than 5% of the total label uptake was due to lysophosphatidylcholine transfer. This, however, would increase to as much as 50% when total label association amounts to only 0.2% of the amount added to the cells. Up to now little attention has been paid to the metabolic fate of liposomal lipids following liposome-cell interaction. In vivo, Ellens and coworkers [46] observed, after intraperitoneal administration of sphingomyelin-containing large unilamellar vesicles to rats, a similar conversion of liposomal sphingomyelin into liver phosphatidylcholine as was demonstrated to occur in vitro in the present study. A subsequent investigation indicated that probably Kupffer cells as well as parenchymal cells are involved in the conversion of sphingomyelin into phosphatidylcholine after hepatic uptake of liposomes [47]. We are indebted to Dr Jan Damen for the suggestion to use dimyristoylglycerophosphocholine in this study. We also wish to thank Jan Wijbenga and Bert Dontje for technical assistance and Rinske Kuperus for preparing the manuscript. These investigations were carried out under the auspices of the Netherlands Foundation for Medical Research (FUNGO) with financial support from the Netherlands Organization for the Advancement of Pure Research (ZWO). REFERENCES 1. Rahman, Y. E., Lau, E. H. &Wright, B. J. (1980) in Liposomes and immunobiology (Tom, B. H. & Six, H. R., eds.) pp , Elsevier/North-Holland, New York. 2. Roerdink, F., Dijkstra, J., Hartman, G., Bolscher, B. & Scherphof, G. (1981) Biochim. Biophys. Acta 677, Poste, G., Bucana, C., Raz, A,, Bugelski, P., Kirsh, R. & Fidler, I. J. (1982) Cancer Res. 42, Dijkstra, J., Van Galen, W. J. M., Hulstaert, C. E., Kalicharan, D., Roerdink, F. H. & Scherphof, G. (1984) Exp. Cell Res. 150, Dijkstra, J., Van Galen, M. & Scherphof, G. (1984) Biochim. Biophys. Acta 804, Freise, J., Muller, W. H., Brolsch, Ch. & Schmidt, F. W. (1980) Biomedicine (Paris) 32, Stoffel, W. (1975) Methods Enzymol. 35, Rouser, G., Kritchevsky, G. & Yamamoto, A. (1967) in Lipid chromatographic analysis (Marinetti, G. V., ed.) vol. 1, pp , Marcel Dekker, New York. 9. Bligh, E. G. & Dyer, W. J. (1959) Can. J. Biochem. Physiol. 37, Papahadjopoulos, D. & Miller, N (1967) Biochim. Biophys. Acta 135, Papahadjopoulos, D., Jacobson, K., Nir, S. & Isac, T. (1973) Biochim. Biophys. Acta 311, Szoka, F. C. & Papahadjopoulos, D. (1978) Proc. Nut1 Acad. Sci. USA 75, Szoka, F. C., Olson, F., Heath, T., Vail, W., Mayhew, E. & Papahadjopoulos, D. (1980) Biochim. Biophys. Acta 601, Knook, D. L. & Sleyster, E. Ch. (1976) Exp. Cell Res. 99, Brouwer, A. & Knook, D. L. (1977) in Kupffer cells and other liver sinusoidal cells (Wisse, E. & Knook, D. L., eds.) pp , Elsevier/North-Holland, Amsterdam. 16. Folch, J., Lees, M. & Sloane Stanley, G. H. (1957) J. Biol. Chem. 226, Ames, B. N. & Dubin, D. T. (1969) J. Biol. Chem. 235, Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall, R. J. (1951) J. Biol. Chem. 193, Heller, M. & Shapiro, B. (1966) Biochem J. 98, Kanfer, J. N., Young, 0. M., Shapiro, D. & Brady, R. 0. (1966) J. Biol. Chem. 241, Weinreb, N. J., Brady, R. 0. & Tappel, A. L. (1968) Biochim. Biophys. Acta 159, Hostetler, K. Y. & Yazaki, P. J. (1979) J. Lipid Res. 20, Fowler, S. & De Duve, C. (1969) J. Biol. Chem. 244, Matsuzawa, Y. & Hostetler, K. Y. (1980) J. Biol. Chem. 255, Kunze, H., Hesse, B. & Bohn, E. (1982) Biochim. Biophys. Acta 711, Stoffel, W. & Trabert, U. (1969) Hoppe-Seyler sz. Physiol. Chem. 350, Franson, R., Waite, M. & La Via, M. (1971) Biochemistry 10, Waite, M., Scherphof, G. L., Boshouwers, F. M. G. & Van Deenen, L. L. M. (1969) J. Lipid Res. 10, Nachbaur, J., Colbeau, A. & Vignais, P. M. (1972) Biochim. Biophys. Acta 274, Franson, R. C. & Waite, M. (1973) J. Cell Biol. 56, Mellors, A. & Tappel, A. L. (1967) J. Lipid Res. 8, Hostetler, K. Y., Yazaki, P. J. & Van den Bosch, H. (1982) J. Biol. Chern. 257, Dawson, R. M. C. (1956) Biochem. J. 62, Lloyd-Davies, K. A., Michel, R. H. & Coleman, R. (1972) Biochem. J. 127, Van Golde, L. M. G. & Van den Berg, S. G. (1977) in Lipid metabolism in mammals (Snyder, F., ed.) vol. 1, pp , Plenum Press, New York. 36. Kennedy, E. P. & Weiss, S. B. (1956) J. Biol. Chem. 222, Sundler, R., Arvidson, G. & Akesson, B. (1972) Biochim. Biophys. Acta 280, Kaulen, H. D., Henning, R. & Stoffel, W. (1979) Hoppe Seyler s Z. Physiol. Chem. 351,

7 Voelker, D. R. & Kennedy, E. P. (1982) Biochemistry 21, Marggraf, W. D., Anderer, F. A. & Kanfer, J. N. (1981) Biochim. Biophys. Acta 664, Bernert, J. T. & Ullman, M. D. (1981) Biochim. Biophys. Acta 666, Ohkuma, S. & Poole, B. (1978) Proc. Natl Acad. Sci. USA 75, Poole, B. & Ohkuma, S (1981) J. Biol. Chem. 90, Matsuzawa, Y. & Hostetler, K. Y. (1980) J. Biol. Chem. 255, Kunze, H., Hesse, B. & Bohn, E. (1982) Biochim. Biophys. Acta 713, Ellens, H., Morselt, H. & Scherphof, G. (1981) Biochirn. Biophys. Acta674, Scherphof, G., Roerdink, F., Dijkstra, J., Ellens, H., De Zanger, R. & Wisse, E. (1983) Biol. Cell 47,