advanced
drugdelivery reviews ELSEVIER
Advanced
Drug Delivery
Reviews
24 (1997) 201-213
Biological factors involved in blood clearance of liposomes by liver Dexi Department
qf Pharmaceutical
Liu ’
Sciences, University
of Pittsburgh,
Pittsburgh,
PA 15261, USA
Abstract The mechanisms underlying the removal of liposomes from the blood by liver were investigated by employing an in situ single-pass liver perfusion technique. Using liposomes with different composition and diameter, we demonstrated that two different mechanisms are involved in liposome clearance, one requires complement activation, called serum-dependent mechanism, and the other, named serum-independent mechanism, does not require serum components. For serum dependent mechanism, activated C3 component bound to the surface of liposomes is likely to serve as the opsonin to lead the binding of liposomes to liver macrophages. For serum-independent liposome uptake, it appears that liver uptake of liposomes involves multiple receptors, some of which may be shared by different types of liposomes. Fluorescence microscopic studies using a dual fluorescence label system in combination with a Kupffer cell elimination technique revealed that, in addition to a dominant role of Kupffer cells for taking up liposomes, non-Kupffer cells are also involved in taking up liposomes with specific lipid composition. Keywords:
Serum opsonin;
Kupffer cells; Liposome
clearance
Contents
I. Introduction __......__...................................................................................................................................................,,,........,.., 2. Summary of results . . . . . . . . . .._................................................................................................................................,,.......,..,,,...,.. and serum-independent liposome uptake by the liver . . . . . . . . . ..__................................................................... 2.1. Serum-dependent 2.2. Uptake of liposomes by liver involves different receptors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . liver uptake of liposomes involves activation of the complement system . . . . . . . . . .._......................................... 2.3. Serum-enhanced 2.4. Liver uptake of liposomes involves different cell types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.5. Liposome clearance by the liver: different animal species have different mechanisms . . . . . . . . . . . . . . . .._......................................... 2.6. Mechanism of the size effect on liposome clearance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .._...._......................... 3. Conclusion and discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ............................... Acknowledgments ..__.............................................................................,.......,,........,.,.......................,.............,....,..,,,..........,,,.... .. .. .. . .__. ... _.__. __.__. .. __._. ._ References ‘Tel.: (412) 648
[email protected]
8553;
fax:
(412)
648
2116;
e-mail:
Abbreviations: aGM,, gangliotetraosyl ceramide; Chol, cholesterol; CL, cardiohpin; Cl,MBP, dichloromethylene bisphosphonate; DCP, dicetylphosphate; DiI, 1.1 ‘-dioctadecyl-3,3,3,‘,3’-tetramethyhndocarbocyanine perchlorate; DPGS, 1,2-dipalmitoyl-snglycerol-3-succinate; DTPA-SA, diethylenetriaminepentaacitic acid stearylamide; GD, h, disialoganglioside; GM,, monosialoganglioside; GT,,, trisialoganglioside; PA, phosphatidic acid; PC, phosphatidylcholine; PE, phosphatidylethanolamine; PG, phosphatidylglycerol; PI, phosphatidylinositol; PS, phosphatidylserine
0169-409X/97/$32.00 PII SO 169-409X(
Copyright 96)00459-O
0
201 202 202 203 204 205 206 209 210 212 212
1. Introduction . ” . r 1 r *. n 3uccessmt use of nposomes as a carrier tar drug delivery often requires liposomes with sufficient circulation time in blood [ 1,2]. Although significant progress has been made in the past to extend the resident time of liposomes in the circulation by manipulating liposome composition and liposome size [3-71, the mechanisms by which liposomes are
1997 Elsevier Science BY All rights reserved
202
D. Liu I Advanced Drug Delivery Reviews 24 (1997) 201-213
removed from the blood stream by the reticuloendothelial system (RES), mainly the macrophages of the liver and spleen, are still unknown. It remains to be elucidated how liposomes with some subtle difference in their surface structure and diameter could have dramatically different half lives in the blood. In an attempt to elucidate the mechanisms of liposome clearance, we have focused our attention on the identifying biological factors that are involved in regulating the liposome uptake by the liver. Different from many similar studies (for review, see Ref. [S]), the technique we used in the study was an in situ single-pass liver perfusion. Compared to the conventional methods for this type of studies, our liver perfusion system has many advantages. Firstly, unlike the approaches that purified macrophages or established cell lines are used, the liver perfusion technique does not involve cell purification and therefore is more convenient. Secondly, instead of studying liposome uptake in a static environment in cell culture, which is different from the in vivo situation where liposomes are taken up by liver cells when flowing through the liver, such a liver perfusion system examines the liposome uptake in a dynamic environment that is similar or identical to that of live animals. Thirdly, under the perfusion conditions, the anatomic structure of the liver in taking up liposomes is maintained. Lastly, since liposomes under our perfusion conditions are only allowed to pass the liver once in a relatively short period of time, it is possible, using this system to study the relationship between the surface structure of liposomes and the cell surface receptors that are involved in taking up liposomes with specific surface characteristics.
2. Summary
of results
2.1. Serum-dependent and serum-independent liposome uptake by the liver To examine whether liver uptake of liposomes involves serum components or so called opsonins [8], liposomes with different lipid compositions were incubated with freshly collected mouse serum and the mixture was then perfused through pre-washed
mouse liver via portal vein. Liposomes preincubated with buffer were used as a control. All of these liposomes contained PC/Chol (105, molar ratio) as the matrix lipid and were labeled with ’ “In-DTPASA [6]. In the absence of serum as shown in Table 1, about 10% of perfused liposomes composed of PC/ Chol were taken up by the perfused liver. Inclusion into liposome bilayer with 6.25 molar% of PE, PG, PI, GM,, aGM, , GD,,, GT,, or PEG2000-PE with final lipid ratio of 10:5:1 (PC/Chol/additional lipid, molar ratio) decreased liposome uptake by the liver. Conversely, inclusion in the same liposomes of the same molar percentage of PS, DCP, PA, CL or DPGS increased the liposome uptake by the perfused liver. Serum appeared to have a different effect on the level of liposome uptake by the perfused liver. Three patterns were observed, including a positive (increasing the level of liposome uptake), negative (decreasing the level of liposome uptake) and no effect. Compared to the level of liver uptake in the absence of serum, serum enhanced the liver uptake of liposomes containing PE (from 1.4 to 18.7%) PG (from 1.4 to 4.8%) aGM, (from 2.5 to 22.1%) or GD,, (from 2.5 to 5.1%). A negative effect of serum on the level of liposome uptake by the liver was observed with liposomes containing PA, or CL. For example, about 20% of total perfused PA liposomes were taken up by the liver in the absence of serum, while the total liver uptake dropped to about 10% when serum was included. A similar level (50%) of reduction by serum treatment was also seen in liposomes containing CL (from 60% to about 30%). The last pattern observed was that in which serum did not affect the level of liposome uptake by the perfused liver. Liposomes, belonging to this category, include those containing PI, GM,, GT,, or PEG-PE, PS, DCP or DPGS. Liposomes composed of PC/Chol also belongs to this group. Considering the mechanisms of liposome uptake by the liver, these results would suggest that two independent mechanisms are involved in liposome uptake by the liver, one does not and the other does require serum components. We called the former a serum-independent clearance mechanism [9], and the later a serumdependent clearance mechanism [lo]. Considering the in vivo clearance of liposomes by the liver, liposomes could be cleared from the blood through one of the two mechanisms.
203
D. Liu I Advanced Drug Delivery Reviews 24 (1997) 201-213 Table 1 Effect of serum on liposome Liposome
uptake by the perfused
composition
PC/Chol PC/Chol/PE PC/Chol/PG PC/Chol/PI PC/Chol/GM, PC/Chol/aGM, PC/ChoI/GD,, PC/Chol/GT,, PC/Chol/PEG-PE PC/Chol/PS PC/Chol/DCP PC/Chol/PA PC/Chol/CL PC/Chol/DPGS
mouse liver
% Liver uptake -Serum
+ Serum
9.42 1.2 1.42 1.6 1.4tl.6 1.7?1.6 3.0+0.2 2.5+2.0 2521.5 6.720.4 3.020.2 22.722.2 28.921.6 22.428.5 60.9rt4.9 17.1”1.3
8521.3 18.723.8 4.822.3 2.0t2.1 3.5t0.1 22.lZ3.2 5.1t1.5 6.021.5 2.5tO.l 20.4+ 1.4 29.222.3 1 I .4t3.8 29.220.7 16.921.6
Ratio ( + serum/ - serum) 0.9 13.4 3.4 1.2 I .2 8.8 2.0 0.9 0.8 0.9 1.0 0.5 0.5 1.0
” ‘In-labeled liposomes (0.12 pmol total lipids in 12 pl) were mixed with 200 Pl of either mouse serum (freshly collected) or buffer (Krebs-Hensleit buffer, pH 7.4) and incubated for 10 min at 37°C. The mixture was then diluted to 2.4 ml with buffer and 2 ml of the diluted mixture was perfused by the portal vein through the mouse liver which was prewashed with 3 ml of buffer (prewarmed to 37°C). Mice were anesthetized by 2,2,2-Tribromoethanol (0.6 ml, 20 mg/ml, i.p.) during the experiment. The inferior vena cava was cut at the beginning of the experiments to drain the blood, immediately followed by washing and subsequent perfusion with liposomes. Liposomes were only allowed to pass through the liver once. Unbound liposomes in the liver were removed by washing the perfused liver with 5 ml of buffer via the same route. The perfusion rate was kept constant at 2 ml/min. The amount of liposomes taken up by the liver was analyzed by measuring the “‘In-radioactivity in the liver and presented as percentage of total amount of liposomes perfused. The average diameter of liposomes was about 300-500 nm as determined by light scattering using a submicron particle analyzer (Coulter N4SD). The lipid ratio used was PC/Chol/X = 10:5:1 (molar ratio).
2.2. Uptake of liposomes receptors
by liver involves different
Theoretically, liposome uptake by the liver must involve the binding of liposomes to the surface of the involved liver cells. The molecules involved in such binding may be considered as receptors. Since liposomes, under the perfusion conditions, only passed through the liver once over a very short time, the majority of liposomes taken up by the liver are those that are still on the cell surface and not internalized. Therefore, a higher level of liposome uptake by the perfused liver would suggest that there are more receptors in the liver for this type of liposomes. Alternatively, it would also suggest that the affinity of this type of liposome is higher than that of liposomes that are taken up by the liver to a lower extent. Such an alternative interpretation is under the assumption that a given number of receptors are responsible for taking up all types of liposomes. To explore the nature of the receptors involved in taking up different types of liposomes,
we used a competition assay to check whether receptors involved in taking up one type of liposome are also responsible for the liver uptake of other types. In these experiments, “‘In-labeled liposomes (0.1 Fmol total lipids) were mixed with 2 l.i,rnol non-labeled liposomes of the same or different lipid composition. The resulting liposome mixture was then perfused through the mouse liver and the total liver uptake for the test liposomes was analyzed. As seen in Table 2, the effect of the excess amount of competing liposomes on liver uptake of test liposomes varies among the liposomes. Taking PS-containing liposomes as an example, the total level of liposome uptake by the liver was decreased in the presence of the excess amount of competing liposomes. For instance, about 80% (from 33% liver uptake to 6%) of the liver uptake for ’ “In-labeled PS-containing liposomes was blocked when 2 kmol of same liposomes were included. Liposomes containing CL exhibited lower but significant activity in inhibiting the liver uptake of PS-containing liposomes (50% inhibition). Liposomes containing PA,
D. Liu I Advanced
204 Table 2 Inhibition
of liver uptake of liposomes
Test liposomes
PC/Chol/PS PC/Chol/CL PCICholIPA PC/Chol/DPGS PC/Chol/DCP
by liposomes
% Inhibition
Drug Delivery Reviews 24 (1997) 201-213
of various composition
of liver uptake for test liposomes
in the presence
of liposomes
containing
PS
CL
PA
DPGS
DCP
80?1.2 45.5t5.6 83.3-c 1.5 91.420.9 25.4? 15.1
51.3-+9.3 62.323 92.2?1.1 89.92 1.7 88.92 1.2
34.5 -CO.8 28.923.8 80.9-cg.3 90.6?2.1 63.6-tg.6
17.7?11.7 19.3tl0.8 61.828.4 86.720.8 51.1?12.3
28.8?3.5 30.827.8 61.7t6.7 77.623.9 72.72 12.8
The experiments were done in identical procedures as described in Table 1 except that test hposomes ( ’ ’ ‘In-labeled) were mixed with competing liposomes (2 pmol total lipids, without labeling) before perfused through the liver. The percent inhibition was calculated by dividing the difference between the total liver uptake of the liposomes in the absence and presence of competing liposomes by the total liposome uptake without competing liposomes.
DCP and DPGS showed similar inhibition activity with the inhibition level ranging from 17 to 34%. A similar inhibition pattern was also seen when CLcontaining liposomes were used as the test liposomes. Sixty percent of liver uptake was inhibited when 2 kmol of same liposomes were used. The presence of PS-containing liposomes inhibited liver uptake by about 45% for CL-containing liposomes. Again, liposomes containing either DCP, PA or DPGS showed much lower inhibition activity (1930% inhibition). In contrast to the inhibition pattern observed in PS- and CL-containing liposomes, the liver uptake of liposomes containing PA and DPGS was more easily inhibited by the competing liposomes. All liposomes tested for competition (PS-, CL- and DCP-containing liposomes) were effective in blocking the liver uptake of liposomes containing either PA or DPGS. When DCP-containing liposomes were used as the test liposomes, PS-containing liposomes showed a much lower inhibition activity compared to those containing CL, PA or DPGS. Similar competition experiments were also performed with liposomes containing PE or aGM, that showed high level uptake only in the presence of serum. As seen in Fig. 1, a minimal effect on the liver uptake of liposomes containing either PE or aGM, was observed in the presence of excess amount of liposomes (2 pmol total lipids) containing either PS, CL or DCP, suggesting that the receptors responsible for serum-dependent liposome uptake are different from those involved in liposome uptake via the serum-independent mechanism. All of the above results would indicate that liposome uptake by the liver involves multiple
receptors. Some of these receptors are shared by different types of liposomes, but others may be unique to a particular liposome composition. 2.3. Serum-enhanced liver uptake of liposomes involves activation of the complement system The involvement of the complement system serum enhanced liposome uptake was examined.
for As
I””
80
20
0 PClChollPE
PC/Chol/aGM
I
TEST LIPOSOMES Fig. 1. Inhibition of serum dependent liposome uptake. Liposomes containing either PE or aGM, were incubated with mouse serum under the conditions described in Table 1. The mixture was then mixed with 2 p,mol liposomes of different composition and liver perfusion was then performed. Competitive liposomes containing PS (II), CL (I), and DCP (W). Percentage inhibition was calculated by dividing the difference between the total liver uptake in the absence and presence of competing liposomes by the total liposome uptake without competing liposomes. Data represent the mean?SD.
D. Liu I Advanced Drug Delivery Reviews 24 (1997) ZOI-Z/3
20 1
15
2 g
lo
s 5
0i PC/Chol/aGMl
PC/Chol/PE LIPOSOMES
Fig. 2. Effect of blocking the complement system on the serummediated liposome uptake by the perfused liver. (0) Serum without treatment, (m) serum treated with EDTA, ( treated with EGTA/Mg*’ and (m) serum treated at 56°C for 30 min. Data represent mean?SD.
shown in Fig. 2, serum activity in enhancing liver uptake of liposomes containing either PE or aGM, was abolished under the conditions that are known to block the complement activation pathway [ 111. For example, over 90% of serum-mediated liposome uptake by the perfused liver was blocked by pretreatment of serum with either EDTA, EGTA/Mg2+ or high temperature (56°C for 30 min). Identical results were obtained for both types of liposomes, suggesting that complement activation is responsible for serum-enhanced liposome uptake by the liver. 2.4. Liver uptake of liposomes cell types
involves different
To test whether liposome uptake by the liver involves different types of liver cells, two independent approaches were taken. The first approach was based on the known fact that the uptake of polystyrene beads by the liver is exclusively mediated by the Kupffer cells [12]. In these experiments, fluorescence labeled polystyrene beads were injected into animals. One hour after the injection, liver perfusion was performed using different types of liposomes labeled with DiI, a commonly used fluorescence marker for membrane studies [ 131. The co-localization of both fluorescence markers was then examined
205
using a standard fluorescence microscopic technique. It is evident in Fig. 3 that two distinct distribution patterns of the fluorescence probes were obtained. For liposomes containing PS, PE and aGM, , DiI (the red color) was well co-localized with that of Fluoresbrite YG Microspheres (Fig. 3A,C,D), suggesting that the Kupffer cells are responsible for the uptake of these liposomes. For CL and DCP-containing liposomes, however, only partial co-localization between DiI-labeled liposomes and the fluorescencelabeled beads was observed (Fig. 3B,E). Such partial co-localization, as shown in Fig. 3B,E, would suggest the involvement of both Kupffer and non-Kupffer cells in taking up CL and DCP-containing liposomes. If both Kupffer and non-Kupffer cells are involved in the uptake of CL- and DCP-containing liposomes, one would predict that elimination of Kupffer cells from the liver should only partially affect the total liposome uptake by the liver. By the same notion, for liposomes that are taken up solely by the Kupffer cells, liposome uptake by the liver should be inhibited if these cells are eliminated from the liver. To test this hypothesis and confirm the conclusion drawn from data in Fig. 3, we have used a macrophage suicide reagent, liposomal Cl,MBP developed by Van Rooijen [14], to eliminate the Kupffer cells from the liver. In these experiments, animals were pre-injected with 90 Fg of Cl,MBP encapsulated in liposomes composed of PC/Chol/PS. Twenty-four hours later, standard perfusion experiments were performed to check the total liposome uptake by the liver. As shown in Fig. 4, the total liver uptake for CL-containing liposomes decreased only about 50% in comparison to the control animals (treated with empty liposomes). The total liver uptake for DCPcontaining liposomes was also partially inhibited, being about 20% in treated animals compared to about 28% in the control group. However, the liver uptake for other types of liposomes, including those containing PS, PE and aGM,, was completely abolished in animals pretreated with liposomal Cl,MBl? These results strongly suggest the involvement of non-Kupffer cells in taking up liposomes containing CL and DCP. The Kupffer cells, however, are solely responsible for liver uptake of PS, PE, and aGM,-containing liposomes. These results support the conclusion obtained from our microscopic studies (Fig. 3).
Fig. 3. Fluorescence microscopic studies of liposome localization in the liver. Fluorescence labeled polystyrene beads with a diameter of 1 pm (1.2 X IO” beads/mouse) were injected to mouse via tail vein to label the Kupffer cells. One hour later, a standard liver perfusion was performed on these animals with DiI-labeled liposomes (2 pmol total lipid) of various lipid compositions. Cryosections of the liver were made using Cryostat (Jung Frigocut 2800N) and examined under a fluorescence microscope. Pictures were taken using double exposure technique. Yellow color is from polystyrene beads and red color indicates the location of liposomes. Small arrow points the colocalization of polystyrene beads with liposomes and large arrow shows the liposome location different from polystyrene beads. The bar in C represents 20 pm. (A) PS containing liposomes, (B) CL-containing liposomes, (C) aGMl-containing liposomes, (D) PE-containing liposomes and (E) DCP-containing liposomes. Liposomes containing aGM1 and PE were preincubated with mouse serum for 10 min at 37°C before the perfusion (Taken from Ref. [21]).
2.5. Liposome clearance by the liver: difSerent animal species have different mechanisms The differences in liposome pharmacokinetics and tissue distribution in different animal species were also examined. Fig. 5 summarizes the results of an experiment where the pharmacokinetics of four
different liposome compositions was compared in two animal species. It is evident from the results that liposomes with a given lipid composition can have very different blood half life in different animal species. For example, in mice (Fig. 5A), liposomes containing 10% of either GM, or PEG5000-PE exhibited relatively long circulation time. Inclusion
D. Liu I Advanced
Drug Delivery
207
Reviews 24 (1997) 201-213
70 T 60 50 “x q40 E
-
* 3 30s 20 10 0’
LIPOSOMECOMPOSI’IION
-
Fig. 4. Effect of Kupffer cell elimination using liposomal CllMBP on the liver uptake of liposomes. Mouse was injected with 3 ymol of plain liposomes (PC/Chol/PS) or liposomes containing about 90 pg of C12MBP. Twenty-four hours post injection, single pass liposomes was then performed on liver perfusion of “‘In-labeled these animals according to the standard procedure. (0) Treated with empty liposomes and (m) treated with ClzMBP containing liposomes. Data represent mean%SD, (n = 3-S) (Taken from Ref.
L7+ll).
'B .l
of the same molar percentage of PS into liposomes resulted in a rapid blood clearance with a half life of less than 10 min. Liposomes of PC/Chol without additional lipid components exhibited a circulation time shorter than those containing GM, or PEG5000-PE but longer than PS containing liposomes. The order of longevity of blood circulation time in mice for the various liposome types was PC/Chol/GM, > PC/Chol/PEG5000-PE > PC/ Chol > PC/Chol/PS. In contrast to what we observed in mice, GM,-containing liposomes showed a very short circulation time in rats (Fig. 5B) with an estimated half life in blood of less than 10 min; over lOO-fold less than that observed in mice. Liposomes with or without PEG-PE displayed almost identical clearance kinetics in rats. Four hours after injection, about 57% of the injected dose of liposomes composed of PC/Chol remained in the blood compared to about 52% for PEG-containing liposomes. The effect of PEG5000-PE in prolonging liposome circulation time as observed in mice was not obvious in rats within the 4-h time period.
0
.I’l’l’l.l’l.l-~J 30 60
90
120
150
180
210
240
TIME bin) Fig. 5. Blood clearance of liposomes composed of PC/Chol (O), PC/Chol/GMl (O), PC/Chol/PS (A) and PCICholIPEG-PE (A) in mice (A) and rats (B). ’ ’ ‘In-labeledliposomes with an average diameter of 100 nm were intravenously injected to animals at a dose of 4 pmol/kg. Blood concentration of liposomes at different time intervals was analyzed by measurement of the radioactivity of “‘In in blood samples (Taken from Ref. [lo]).
The mechanism of liposome clearance in rats was explored using the same liver perfusion technique. Two liposome compositions that exhibit short half lives in blood were selected to examine the serum effect on liposome uptake by the rat liver. Compared to similar experiments where the mouse system was used (Table l), rat liver showed a very low liver uptake for PS liposomes in the absence of serum. In contrast to the mouse system where mouse serum has no effect on the total liver uptake of PS-containing liposomes, total liposome uptake by the rat liver in the presence of serum was S-fold higher than that in
D. Liu I Advanced Drug Deliver) Reviews 24 (1997) 201-213
208 Table 3 Effect of serum on liposome Liposomes
PC/Chol/PS PC/Chol/GM,
uptake by the rat liver
% Uptake -Serum
+Serum
4.2-c 1.2 7.2k2.6
33.6i3.6 30.5’2.8
Ratio ( + serum/
- serum)
8.0 10.9
“‘In-labeled liposomes with a mean diameter of about 500 nm (1.2 pmol total lipids in 120 p.1) were incubated with 1 ml of either Krebs-Henseleit buffer (pH 7.4) or serum for 10 min at 37°C and then diluted to 24 ml with Krebs-Henseleit buffer. Twenty ml of this mixture were perfused by the portal vein through the rat liver that had previously been washed with 20 ml of buffer to remove the contaminating blood. Unbound liposomes were removed by washing the perfused liver with 50 ml of buffer. The perfusion rate used was 7 ml/min. The blood and the perfusate were drained by an incision at the inferior vena cava. The method of data analysis is the same as that described in Table 1.
the absence of serum. A similar observation was also obtained for liposomes containing GM,. These results would suggest that the mechanisms involved in liver uptake for PS-containing liposomes are different. In mouse, PS-containing liposomes are taken up by the liver through a serum-independent mechanism while in rats, it is the serum-dependent mechanism that regulates the removal of both GM, - and PScontaining liposomes from the blood. From the results described above, it was obvious that serum components are involved in mediating the liver uptake of PS and GM, containing liposomes in rats. Such dependence on the serum for a high level of liposome uptake is different from that in mouse. We wondered if the sera of other animal species
Table 4 Effect of human and bovine serum on liposome Liposome composition
PC/Chol PC/Chol/PS PC/Chol/GM, PC/Chol/PEG-PE
uptake by liver
Liver source
Mouse Rat Mouse Rat Mouse Rat Mouse Rat
would have similar activities. To test this possibility, human and bovine serum were selected as examples and their effects on liposome uptake by the liver, using the perfusion system, were tested. As summarized in Table 4, different degrees of serum effect on liposome uptake by the perfused liver were observed. For instance, when mouse liver was used as a model, the total liver uptake of liposomes composed of either PC/Chol or PC/Chol/GM, was greatly increased in the presence of either human or bovine serum. The total uptake for PC/Chol liposomes by the perfused mouse liver in the absence of serum is less than 2% (Table 4), which was increased to about 20% or 5% in the presence of human or bovine serum, respectively. A similar increase in liposome uptake was also obtained for GM ,-containing liposomes. Interestingly, both human and bovine sera decreased the total mouse liver uptake of liposomes containing PS (from 20% in buffer to about 14 and 13% in human and bovine serum, respectively). When rat liver was used, both human and bovine sera increased the total uptake of liposomes composed of either PC and Chol or liposomes containing additional PS or GM,. In comparison to liposomes not exposed to serum, a greater than 3-fold increase in the total liposome uptake was observed in perfused rat liver in the presence of either human or bovine serum. It is important to note that neither human nor bovine serum has a noticeable effect on the uptake of PEG-PE containing liposomes regardless of the tested animal species (Tables 1 and 4). The involvement of the complement system in the effect of human serum on the level of liposome uptake by the mouse liver was examined using a
% Uptake Buffer
H-serum
B-serum
1.2t0.1 1.350.4 19.522.6 2.1?0.1 1.920.5 4.320.4 2.650.1 1.51-0.1
20.02 1.3 8.1?3.1 13.9IT2.4 7.3-cl.l 11.6?0.3 14.22 1.3 2.020.1 1.3*0.1
5.420.6 7.9% 1.3 12.650.7 6.022.0 4.850.3 8.420.6 2.9+ 1.4 1.o*o. 1
Conditions used were identical to those described in Table 1 for mice and Table 3 for rats. The average diameter of hposomes experiments was around 100 nm. H-serum, human serum; B-serum, bovine serum.
used in the
209
D. Liu I Advanced Drug Delivery Reviews24 (1997) 201-213
PS containing liposomes ment activation. 2.6. Mechanism clearance
PC/Chol
PC/Chol/rS
l’C/Chol/GMl
LIPOSOMES Fig. 6. Effect of inhibition of the complement system in human serum on the liposome uptake by the perfused mouse liver. For serum treatments, 200 ~1 freshly collected human serum were preincubated with either 40 )*I of EDTA (60 mM), EGTA/Mg” (60 mMil5 mM) or cobra venom factor (160 units/ml) for 10 min prior to incubation with liposomes. (0) Without serum, (m) with serum (0) with serum treated with EDTA, (m) with serum treated with EGTA/Mg” and ( ) with serum treated with cobra venom factor (Taken from Ref. [lo]).
standard complement inactivation system [ 111. As shown in Fig. 6, treatment of human serum by EDTA, EGTA/Mg’+ or cobra venom factor abolishes the serum-enhanced liposome uptake, proving that the complement system in human serum plays an important role in mediating the liposome uptake by the liver. However, blocking complement activation did not have much effect on the activity of human serum in reducing the uptake of PS-containing liposomes. This would suggest that the activity of serum in reducing the total liver uptake of
Table 5 Serum effect on liver uptake of liposomes
with different
Liposome
% Uptake
diameter (nm)
Conditions
comple-
of the size effect on liposome
It has been well known that liposome size plays an important role in determining the rate of liposome clearance from blood [6,7]. Liposomes with large diameter are usually removed from the blood circulation faster than those of the same liposomes with a smaller diameter. To elucidate the size effect regarding the mechanism of liposome uptake by the liver, liposomes with different diameters were prepared and the level of liposome uptake by the liver after perfusion was compared between the presence and absence of serum. As shown in Table 5, the size effect of liposomes on liposome uptake by the liver is controlled by serum. In the mouse system, serum showed a negative effect on the liver uptake of PS-containing liposomes. Serum showed a decreased level of liver uptake for PS-containing liposomes. Such an inhibition activity is inversely proportional to the diameter of liposomes. For example, serum reduced the liver uptake of liposomes with an average diameter of 60-100 nm by about 50-60%. However, there was only about a 10% reduction observed when the diameter of liposomes increased to around 400 nm or greater. These results would suggest that one of the reasons that small liposomes in mouse usually circulate longer than the large ones in the blood is that serum provides a better protection on small liposomes against the liver uptake. This conclusion may be especially true for liposomes that are cleared by the liver through the serum-independent mechanism.
diameters
Mouse system
60 100 200 400 600
does not involve
Rat system
-Serum
+Serum
-Serum
+Serum
23.8% I .7 19.522.6 22.320.9 22.022.5 28.1 k4.5
9.620.3 10.3% 1.5 17.02 I .3 20.2k2.0 24.224.5
ND 2.950.1 3.220.2 4.2% 1.2 ND
ND 10.6+0.4 17.8t0.8 33.623.1 ND
used were the same as those described
in Table 1 for mouse and Table 3 for rat liver perfusion.
ND, not determined
210
D. Liu I Advanced
Drug Delivery Reviews 24 (1997) 201-213
In contrast to this mechanism, liposomes that exhibit higher liver uptake in the presence of serum (through the serum-dependent mechanism) appear to have a different mechanism. As shown in Table 5 using a rat system, it appears that the level of liposome uptake by the perfused liver in the presence of serum is directly proportional to the diameter of liposomes. While the level of liposome uptake by the rat liver in the absence of serum was minimal for all sized liposomes tested, a linear increase in the presence of serum was observed with increase of liposome size. These results would indicate that large sized liposomes are more capable of activating the complement system such that they are removed from the blood faster than the small ones and therefore exhibit a shorter circulation half-life in blood.
3. Conclusion
and discussion
As summarized in the above sections and in previous studies [ 15- 181, the mechanisms involved in liposome clearance by the liver are fairly complicated. However, data summarized in this paper and in those of our previous publications [9,10,19-211 show that liposomes, upon entering the blood stream, are removed from the circulation by the liver through two mechanisms. Depending on liposome composition and animal species, liposomes can be taken up by the liver through a serum-independent clearance pathway or a pathway that requires complement activation. For liposomes taken up by the liver through the serum-independent mechanism, serum decreases the liver uptake of liposomes with small diameter. For serum-dependent liposome uptake by the liver, the serum effect of increasing liposome uptake is directly proportional to the size of liposomes. While Kupffer cells play a dominant role in taking up most types of liposomes, other types of cells in the liver may be also involved. Concerning the receptors involved in liver uptake of liposomes, it is likely that liposome uptake involves multiple receptors, some are shared by liposomes with different composition and others may be more specific to one type of liposome composition than the other. The high level liposome uptake by the liver in the absence of serum for liposomes containing PS, DCP, PA, CL and DPGS (Table 1) in mice appears to support the hypothesis that liver uptake of negatively
charged liposomes involves the direct recognition of the negatively charged head groups (phosphate group in DCP, PA and CL, and carboxyl group of PS and DPGS) of the phospholipids by membrane receptors of the liver cells. One common feature for these liposomes is that the negatively charged head groups of these phospholipids are directly exposed to the liposome surface. The reason for a low level of liver uptake for PG-containing liposomes in the absence of serum, even though they contain a negatively charged phosphate group, is probably because the phosphate groups of PG on the liposome surface are not recognized by these membrane receptors. This may be due to the steric hindrance provided by glycerol head group on PG. The low level of liver uptake for glycolipid and PEG-PE containing liposomes may be due to the same mechanism. By the same notion, a 3-fold increase in liver uptake for GT, ,-containing liposomes, compared to the level for PI-containing liposomes, would suggest that there are certain amounts of negatively charged carboxyl groups on GT,,-containing liposomes that are available for recognition. Such a low level of exposure, although sufficient to result in a decreased liposome circulation time in vivo [4], is not enough to result in a high level of liver uptake in our perfusion system. Such an explanation is consistent with the hypothesis that the hindrance effect of the sugar moieties on the negative charge plays a key role in the prolongation of liposome circulation time in blood [4]. It is evident that the clearance of liposomes containing PE and aGM, in mouse (Table l), PS and GM, liposomes in rat (Table 3) involves serum opsonins. Such serum-mediated liposome uptake has been reported by many laboratories (for review, see Ref. [8]) including ours [19-211. The loss of serum activity in increasing the liver uptake by EDTA and cobra venom factor and high temperaEGTA/Mg’+, ture treatment would suggest that activation of complement system by these liposomes is through the classic pathway. Activation of complement system by PE [22] and other glycolipid such as cetyl mannoside [23] in the form of liposomes has been reported previously. It has also been reported that the complement system is involved in the clearance of PG-containing liposomes in vivo [24]. Such a conclusion also seems to be true for other types of liposomes in different animal systems [lo]. It is not clear at this moment, however, whether the comple-
D. Liu I Advanced
Drug Delivery Reviews 24 (1997) 201-213
ment activation by these lipids involves antibodies specific to these lipids. Limited evidence suggests that C3b of the complement system is responsible for the serum-mediated liposome uptake by the liver (8,15,20]. In addition to the fact that serum is capable of increasing the liver uptake of liposomes of a certain composition, data in Table 1 and Table 5 show that serum can also decrease the level of liver uptake for liposomes. Since such a negative effect by the serum was only observed in liposomes that exhibit high liver binding in the absence of serum and it is not sensitive to treatment designed to inactivate the complement system (Fig. 6) it is highly possible that such an effect results from a non-specific binding of serum proteins to the surface of these liposomes. Under such conditions, upon the exposure of liposomes to the serum, some serum proteins will nonspecifically adsorb to the liposome surface and provide some level of steric hindrance that can interfere with the binding of liposomes to the cells through the specific membrane receptors. The types and amount of serum proteins that can bind to the liposome surface will depend on the surface properties of the liposomes, thus some negatively charged liposomes (CL, PA liposomes, Fig. 1) show signihcant reduction in liposome uptake by the liver and some liposomes such as those containing DPGS do not when serum components are present. Different levels of protein binding to liposomes of various lipid compositions and liposomes with different diameters have been well documented in the literature [25,26]. Data in Fig. 3 and Fig. 4 suggest that, while it is true that Kupffer cells are the major cell type in the liver that are responsible for liposome uptake, nonKupffer cells may also participate in the uptake of liposomes containing CL or DCP. To our knowledge, this is the first time, using large sized liposomes, that the involvement of non-Kupffer cells in the liposome uptake has been demonstrated. Taking the fact that the diameter of the liver fenestration is about 100 nm [27] and the average diameter of our liposomes is above 300 nm, it is unlikely that the hepatocytes, which are located outside the blood vessels, are involved. While it is possible that these non Kupffer cells are subclasses of macrophages that do not normally take up liposomes, it is also possible that other liver
21 I
cells such as endothelia cells forming the capillary wall of the blood vessel are involved. While it is evident that different types of liver cells are involved in liposome uptake, the molecules on the cell surface responsible for liposome uptake are still not clear. The fact that a large proportion of the liver uptake for liposomes containing PA or DPGS (Fig. 2) can be inhibited by excess amounts of liposomes containing either CL, PS or DCP would suggest that these molecules involved in binding of PA and DPGS are shared by CL, PS and DCP. The reduced effectiveness of these two types of liposomes in blocking the uptake of liposomes containing CL, PS and DCP indicates that the affinity of the receptors involved in the binding of PA and DPGS liposomes is lower. An alternative explanation for this observation, even though less likely, is that the number of binding sites for these types of liposomes is less than those for other liposomes. In this case, different types of liposomes are taken up through different molecules. The reason that one type of liposomes can block the binding of another is that the binding of competitive liposomes when they are in excess will sterically hinder the binding of the other. Under such circumstances, the cell surface is basically covered by the competing liposomes and there may not be enough room left on the cell surface for the other type of liposome to bind, thus they are ‘competed out’ from binding to the liver, resulting in a low liver uptake. The higher efficiency of CLcontaining liposomes in inhibiting the liver uptake of other types of liposomes is presumably related to the double negatively charged phosphate group in each molecule of CL. Theoretically, among the negatively charged lipids tested, CL may be the only one that contains two directly exposed negatively charged phosphate groups. Nevertheless, our data suggest the existence of multiple receptors that are involved in liposome clearance. The identity and physiological functions of these receptors are still not known. Concerning the fact that liver cells, especially Kupffer cells, are part of immune defense system that keep abnormal substances and cells out of circulation, it is, therefore, possible that these receptors are involved in the binding of these abnormal substances. For example, it has been reported that PS is exclusively located in the inner leaflet of the cell membrane. The appearance of PS on the outer leaflet of the cell membrane will result in a quick clearance
D. Liu ! Advanced Drug Delivery Reviews 24 (1997) 201-213
212
of the cells from the blood circulation by the RES. The mechanisms of such clearance may be similar, if not identical, to that of the clearance for PS-containing liposomes as discussed in this report. Studies to explore the physiological functions of the receptors involved in the uptake of other types of liposomes such as those containing CL, PA, DCP and possibly others are yet to be performed. Concerning the potential use of liposomes as a carrier for drug delivery, one of the important observations from our study is related to the demonstration that liposome circulation time in blood depends on animal species. Given the fact that animals are commonly used as a model to predict the clinical behavior of liposomes in humans, these results are not only important for our understanding of the mechanisms for liposome clearance but also for choosing the appropriate animal model for the development of liposome-based drug delivery systems.
Acknowledgments I wish to thank Mr Qingru Hu for critical reading of the manuscript.
References [ 1J Papahadjopoulos,
[2]
[3]
[4]
[5]
D., Allen, T.M., Gabizon, A., Mayhew, E., Mathay, K., Huang, SK., Lee, K.D., Woddle, M.C., Lasic, D.D., Redemane, C. and Martin, F.J. (1991) Sterically stabilized liposomes: improvements in pharmacokinetics and anti-tumor therapeutic efficacy. Proc. Natl. Acad. Sci. USA, 88, 11460-I 1464. Allen, T.M. (1989) Stealth liposomes: avoiding reticuloendothelial uptake, In Lopez-Berestein and I. Fidler (Eds), Liposomes in the Therapy of Infectious Diseases and Cancer, New Series, Vol. 89, Alan R. Liss, New York, pp. 405-415. Gregoriadis, G. (1995) Fate of liposomes in vivo and its control: a historical perspectives. In D. Lasic and F. Martin (Eds.) Stealth Liposomes. CRC Press, Boca Raton, FL, pp. 7-12. Gabizon, A. and Papahadjopoulos, D. (1988) Liposome formulations with prolonged circulation time in blood and enhanced uptake by tumors. Proc. Natl. Acad. Sci. USA, 85, 6949-6953. Allen, T.M. (1994) The use of glycolipids and hydrophilic polymers in avoiding rapid uptake of liposomes by the mononuclear phagocyte system. Adv. Drug. Dehv. Rev., 13, 285-309.
161Juhano, R.L. and Stamp, D. (1975) Effect of particle size and charge on the clearance encapsulated drugs. Biochem. 651-658.
of liposomes and liposomeBiophys. Res. Commun., 63,
[71 Liu, D., Mori, A. and Huang, L. (1992) Role of liposome size and RES blockade in controlling biodisttibution and tumor uptake of GMl-containing liposomes. Biochim. Biophys. Acta, 1104, 95-101. RI Pate1 H.M. (1992) Serum opsonin and liposomes: their interaction and opsonophagocytosis. Crit. Rev. Ther. Drug. Carrier Syst., 9, 39-90. [91 Liu, F. and Liu, D. (1996) Serum independent liposome uptake by mouse liver. Biochim. Biophys. Acta, 1278, 5-l 1. UOI Liu, D., Hu, Q. and Song, Y.K. (1995) Liposome clearance from blood: different animal species have different mechanisms. Biochim. Biophys. Acta, 1240, 277-284. 1111 Johnstone, A. and Thorpe, R. (1982) Immunochemistry Practice, Blackwell Scientific, Oxford, pp. 56-59.
in
1121 McCuskey, R.S. (1993) Endothelial cells and Kupffer cells. In A.V LeBouton (Ed.) Molecular and Cell Biology of the Liver. CRC Press, Boca Raton, FL, pp. 407-427. fluorescent labeling of 1131 Claassen, E. (1992) Post-formation liposomal membranes, In vivo detection, localization and kinetics J. Immunol. Methods, 147, 231-240. [I41 Van Rooijen, N. and Sanders, A. (I 994) Liposome mediated depletion of macrophages: mechanism of action, preparation of liposomes and application. J. Immunol. Methods, 174, 83-93. H., Sakata, K., Funato, K. and Kiwada, H. [I51 Harashima, (1993) Enhanced hepatic uptake of liposomes through complement activation depending on the size of liposomes. Pharm. Res., 11, 402-406. [I61 Allen, T.M., Austin, G., Chonn, A., Lin, L. and Lee, K.D. (1991) Uptake of liposomes by cultured mouse bone marrow macrophages: influence of liposome composition and size. Biochim. Biophys. Acta, 1061, 56-64. [I71 Nishikawa, K., Arai, H. and Inoue, K. (1990) Scanvenger receptor-mediated uptake and metabolism of lipid vesicles containing acidic phospholipids by mouse peritoneal macrophages. J. Biol. Chem., 265, 5226-5231. D. (1992) Recog[I81 Lee, K.D., Hong, K. and Papahadjopoulos, nition of hposomes by cells: in vitro binding and endocytosis mediated by specific lipid headgroups and surface charge density. Biochim. Biophys. Acta, 1103, 185-197. and L191 Liu, D., Liu, F. and Song, Y.K. (1995) Recognition clearance of liposomes containing phosphatidylserine are mediated by serum opsonin. Biochim. Biophys. Acta. 1235, 140- 146. WI
Liu, D., Liu, F. and Song, Y.K. (1995)Antibody complement mediated liver uptake of liposomes GMI. Pharm. Res., 12, 508-512.
dependent, containing
[211 Hu, Q. and Liu, D. (1996) Co-existence of serum-dependent and serum-independent mechanisms for liposome clearance and involvement of non-Kupffer cells in liposome uptake by mouse liver. Biochim. Biophys. Acta, 1284, 153-161. on GQI Mold, C. (1989) Effect of membrane phospholipids activation of the alternative complement pathway. J. Immunol., 143, 1663-1668.
D. Liu I Advanced
Drug Delivery
[23] Funato, K., Yamashita, C., Kamada, .I., Tominaga, S. and Kiwada, H. (1994) Plasma factor triggering alternative complement pathway activation by liposomes. Pharm. Res., 11, 372-376. [24] Loughrey, H.C., Bally, M.B., Renish, L.W. and Cullis, P.R. (1990) The binding of phosphatidylglycerol liposomes to rat platelets is mediated by complement. Thromb. Haemost., 64, 172-176. [25] Juliano, R.L. and Lin, G. (1980) The binding of human
Reviews 24 (1997) 201-21.3
213
plasma proteins. In: H. Six and B. Tom (Eds.), Liposomes and Immunology, Elsevier, Amaterdam, pp. 49-66. [26] Chonn, A., Semple, S.C. and Cullis, P.R. (1992) Association of blood proteins with large unilamellar liposomes in viva. J. Biol. Chem., 267, 18759- 18765. [27] Wisse, E. (1970) An electron microscopic study of the fenestrated endothelial lining of rat liver sinusoids, J. Ultrastrue. Res., 31, 125-150.
