INTRODUCTION

Liposomes have repeatedly shown to be able to improve the therapeutic index of a variety of drugs. Intravenously administered liposomes generally undergo extensive opsonization and are therefore rapidly cleared by macrophages of the mononuclear phagocyte system (MPS), particularly Kupffer cells in the liver and spleen macrophages. As a result, liposome targeting to pathological tissues is impeded. This drawback could be amended by coating liposomes with hydrophilic, neutral polymers such as polyethylene glycol (PEG) (1). PEG increases the hydrophilicity of the liposome surface and provides a steric barrier against opsonization (1, 2). The resulting long-circulating liposomes, also referred to as ‘stealth’ or ‘sterically stabilized’ liposomes, are still removed from the blood circulation, however, removal occurs at a much lower rate and with less involvement of the hepatosplenic macrophages. Consequently, they circulate in the bloodstream for a prolonged period of time, enabling their extravasation into solid tumors and sites of inflammation by virtue of the presence of capillary discontinuities. This so-called enhanced permeability and retention (EPR) effect allows for increased local drug concentrations in the target region (3).

Besides PEG, several other hydrophilic polymers have been successfully applied as coatings for long-circulating liposomes, among them conjugates based on poly(oxazoline) (4), polyglycerol (5), poly(N-(2-hydroxypropyl)methacrylamide)) (6), poly-N-vinylpyrrolidone (7, 8) and polyvinyl alcohol (9). Our group proposed liposome coatings based on poly(amino acid)s (PAA): poly(hydroxyethyl l-glutamine)-N-succinyldioctadecylamine (PHEG-DODASuc) and poly(hydroxyl-ethyl l-asparagine)-DODASuc (PHEA-DODASuc) (Fig. 1) (10, 11). PAAs have the advantage of being degradable by lysosomal proteases, allowing for a complete elimination from the body, and reducing the risk of side effects associated with polymer accumulation, as described to occur in case of some non-degradable polymers (12–15). Their degradability can also be exploited for enzyme-induced drug release and/or target cell interaction, as the coating may be shed by proteolytic cleavage in the extracellular space of tumors and sites of inflammation. It has been demonstrated that PAA coatings can prolong circulation times of liposomes to an equal extent as a PEG coating (10).

Fig. 1
figure 1

Structures of PHEA-DODASuc (average degree of polymerization (n = 15) and PHEG-DODASuc (n = 18).

Full size image

An understanding of the effect of physicochemical characteristics of long-circulating liposomes on their pharmacokinetics and biodistribution is crucial for the rational development of therapeutic applications. Liposome clearance is a very complex mechanism and is mediated by a diversity of protein molecules such as immunoglobulins, complement factors and nonimmune opsonins, and the opsonization process can be very different for different types of liposomes (16–18). For long-circulating liposomes with PEG as a ‘stealth' coating, it has been reported that circulation times are relatively independent of factors such as lipid dose, lipid composition, surface charge and liposome size, which makes them versatile drug carriers (19). In the present study, we investigated if this also holds true for PAA-liposomes. Therefore, the influence of PAA grafting density, cholesterol inclusion, surface charge, particle size and lipid dose on the circulation kinetics and biodistribution of PAA-liposomes was evaluated.

MATERIALS AND METHODS

Materials

PHEA-DODASuc (average MW = 3,000 Da, determined by NMR and MALDI-ToF MS, corresponding with an average degree of polymerization of 15) and PHEG-DODASuc (average MW = 4,000 Da, determined by NMR and MALDI-ToF MS, corresponding with an average degree of polymerization of 18) were synthesized as described previously (10). Dipalmitoyl phosphatidylcholine (DPPC), egg phosphatidylcholine (EPC) and egg phosphatidylglycerol (EPG) were kindly provided by Lipoid GmbH, Ludwigshafen, Germany. 1,2-Dioleoyl-3-trimethylammonium-propane (DOTAP) and l-α-phosphatidyl serine (PS, porcine brain, sodium salt) were purchased from Avanti Polar Lipids, Inc., Alabaster, AL, USA. Cholesterol was purchased from Sigma-Aldrich Chemie BV, Zwijndrecht, The Netherlands. [3H]-Cholesteryl oleylether was a product of Amersham, Roosendaal, The Netherlands. Ultima Gold liquid scintillation cocktail and Solvable tissue solubilizer were purchased from Perkin Elmer BioScience B.V., Groningen, The Netherlands. All other reagents were of analytical grade.

Animals

Male outbred Wistar rats (body weight approximately 250 g) were obtained from Harlan Nederland, Horst, The Netherlands. Rats were housed in groups of four under standard laboratory conditions and had free access to rat chow and water. All animal experiments were performed according to national regulations and approved by the local animal experiments ethical committee.

Liposome Preparation

Liposomes were prepared as described previously by a lipid film hydration method (20). Appropriate amounts of DPPC, EPC, EPG, cholesterol, PS, or DOTAP were dissolved in ethanol in 50 ml round-bottom flasks (for lipid compositions see Table I). To each formulation 20 μl of the lipid phase marker [3H]-cholesteryl oleylether (approximately 740 kBq) were added. This label has been shown to be non-metabolizable and non-exchangeable (21, 22). The solvent was evaporated under a stream of nitrogen. The obtained lipid film was hydrated by gentle shaking in 2.5 ml HBS (5 mM HEPES, 0.9% NaCl, pH 7.4) resulting in a final concentration of 20 μmol total lipid/ml, except for two formulations to study the effect of the lipid dose: besides the formulation of 20 μmol total lipid/ml, a second formulation was prepared at a concentration of 0.2 μmol/ml and divided into two equal parts. One part was diluted 1:10 in HBS, yielding a lipid concentration of 0.02 μmol total lipid/ml. Liposomes of the different sizes were obtained by multiple extrusion through two stacked polycarbonate filters (Poretics, Livermore, CA, USA, 400, 200, and 100 nm pore size) with a high-pressure extruder. Formulations were stored at 4°C and used within 1 week after preparation.

Table I Liposome Characteristics
Full size table

Liposome Characterization

The mean particle size and polydispersity index (PD) of the liposome dispersions (diluted 1:100 with HBS) were measured by dynamic light scattering on a Malvern ALV/CGS-3 Goniometer, detected at an angle of 90° to the laser beam. The polydispersity index gives information on the particle size distribution, it ranges from 0 for an entirely monodisperse up to 1 for a polydisperse system. The phospholipid content of the liposomal dispersions was determined colorimetrically according to Rouser et al. (23). Radioactivity of the liposomal dispersions was analyzed in an Ultima Gold liquid scintillation cocktail (dilution of 1:1,000) and counted in a Packard Tricarb 2200 CA liquid scintillation counter.

Pharmacokinetics

Rats were put under light isoflurane anesthesia and liposomes were injected via the tail vein. To study the effect of lipid dose, injections of 25, 0.25, and 0.025 μmol total lipid/kg were given. For all other studies the injected dose was 20 μmol total lipid/kg. Blood samples (approximately 150 μl) were drawn under isoflurane anesthesia from the tail vein of each rat immediately after injection (from the opposite vein) and at 1, 4, 8, 24 and 48 h after injection. Hundred microliters of each blood sample were mixed with 100 μl Solvable tissue solubilizer and 100–200 μl 35% hydrogen peroxide to decolorize the samples and incubated overnight. Samples were diluted in 10 ml Ultima Gold scintillation cocktail and radioactivity was subsequently assessed in a Packard Tricarb 2200CA liquid scintillation counter. Forty-eight hours after injection, the rats were sacrificed by cervical dislocation and for the studies involving variations in surface charge, liposome size and lipid dose, liver and spleen were dissected and homogenized in 25 and 5 ml water, respectively. To 0.5 ml of the liver homogenate and 1 ml of the spleen homogenate, 200 μl of Solvable tissue solubilizer and 200 μl 35% hydrogen peroxide were added and mixtures were incubated until samples were dissolved and colorless. Radioactivity of the samples was assessed in 10 ml Ultima Gold scintillation cocktail as described above. Besides tissue and blood samples, also liposome samples were analyzed to determine the radioactivity of the injected amount of liposomes.

Data Analysis

Blood concentrations of liposomes at the different time points were calculated from the radioactivity of the blood samples as the percentage of the radioactivity measured immediately after injection (% injected dose). For the blood samples, a Q-test for identification of outliers was applied. From the blood concentration-time curves, area under the curve values were calculated for the analyzed time interval from zero to 48 hours using the trapezoidal rule [AUC0–48 h (% injected dose × h)]. The percentage of the injected dose (% ID) in the organs was calculated by division of the radioactivity of the organs by the radioactivity of the injected liposome amount. Differences in AUCs and liver and spleen values were compared by a one-way analysis of variance. The Bonferroni method was used to correct for multiple comparison. Differences were considered significant when the p-value was <0.05. All analyses were performed using Graph Pad Prism 4 software.

RESULTS AND DISCUSSION

To study the effect of liposome characteristics on the pharmacokinetics and biodistribution, different liposome types were prepared. Liposome compositions, sizes and polydispersity indices are presented in Table I. Different grafting densities of PAAs on the liposome surface were studied. The bilayer fluidity of liposomes was varied by inclusion of cholesterol. The effect of charge was examined by inclusion of the negatively charged lipids egg phosphatidylglycerol (EPG) and phosphatidylserine (PS) and the positively charged lipid dioleoyl phosphatidyltrimethylammonium propane (DOTAP). The effect of particle size was assessed with liposomes of sizes between 120 and 360 nm. The lipid doses needed in therapeutic settings can differ substantially. PEG-liposomes show dose-independent pharmacokinetics over a broad dose range, however at low doses (<1 μmol/kg) rapid clearance has been reported. To investigate if this also holds true for PAA-liposomes, the effect of lipid dose on the pharmacokinetics of PAA-liposomes was examined. Except for liposome formulations prepared to study the effect of liposome size, the hydrodynamic diameter of liposomes varied between 105 and 160 nm and the polydispersity index was ≤0.14, indicating a relatively narrow size distribution.

Effect of Grafting Density

In Table II the effect of three different grafting levels of PHEA-DODASuc (i.e. 2.5, 7.5 and 15 mol%) on the circulation kinetics of liposomes is shown. For this study the composition of DPPC and cholesterol (2:1) was chosen to be able to compare the data to earlier published results on the effect of grafting density of PHEG-DODASuc (1, 2.5, 7.5 and 15 mol%) on liposome circulation times (10). The blood concentrations at 4 and 24 h post-injection show no significant differences between the formulations. This is in line with the findings on the effect of grafting density of PHEG-DODASuc on liposome circulation times (10). The density of stealth polymers on the liposome surface has been shown to be of importance in the design of long-circulating liposomes. For example, for PEG-liposomes with a PEG2000-conjugate a grafting density of 5 to 7.5 mol% is often used in preclinical studies (24, 25). At PEG levels >4 mol%, a transition from the mushroom state, where PEG chains do not interact laterally, to the denser and thicker brush state of PEG occurs, the latter being the condition for optimal steric stabilization of the liposomes (26). It has been shown that a further increase of the density of PEG up to 20 mol% did not change circulation kinetics and biodistribution. However, considerable foaming made liposome preparation difficult; the foaming was likely related to the presence of PEG-DSPE micelles (24). Our studies show that at grafting densities between 2.5 and 15%, PAAs are able to prolong liposome circulation times similar to PEG. In previous studies a PAA grafting density of 7.5% was used, however, the present study indicates that a lower grafting density of 2.5% is also sufficient to prolong liposome circulation times.

Table II Percentage of Injected Dose of PHEA-Coated DPPC/Cholesterol Liposomes in the Circulation at 4 and 24 h Post-Injection
Full size table

Effect of Cholesterol Inclusion

The circulation kinetics of EPC-based liposomes without or with 33 mol% cholesterol and coated with PHEA or PHEG were determined (Fig. 2). For both PHEA- and PHEG-liposomes, only small differences in circulation profiles and AUC values of liposomes with or without cholesterol were observed. Cholesterol is known to influence the bilayer fluidity, resulting in increased rigidity and lateral packing and an increased phase transition temperature (27). For conventional PC-based liposomes interaction with proteins was reduced and circulation times were extended upon inclusion of cholesterol (28, 29). For PEGylated liposomes controversial observations have been made. Addition of 20–50 mol% cholesterol resulted in improved circulation kinetics (25). In other studies, however, no differences in circulation times and biodistribution have been observed when 33 mol% cholesterol were added, indicating that differences in bilayer rigidity have only little influence on the circulation kinetics of liposomes that are shielded by a ‘stealth’ coating (30, 31). Our findings are in agreement with this observation made with PEG-liposomes: shielding of liposomes by PAAs diminished the influence of bilayer fluidity on circulation times.

Fig. 2
figure 2

Pharmacokinetic behavior of EPC-liposomes coated with PHEA (top left) and PHEG (top right) without (empty squares) or with (filled squares) 33% cholesterol (% injected dose). Area under the curve (AUC0–48 h) values (bottom) calculated from blood concentration time curves. All results are expressed as mean ± SD (n = 3–4). *p < 0.05; n.s. Not significant.

Full size image

Effect of Charged Lipids

Charged lipids are often added to liposome formulations to improve drug loading and/or stability of the formulation against aggregation during storage (30, 31). Changing the surface charge of conventional liposomes has been shown to greatly influence their pharmacokinetic behavior (32, 33). For PEG-liposomes it has been reported that the polymer is able to shield liposome charge. For example, PEG-liposomes containing up to 30 mol% of the negatively charged phospholipid EPG showed no altered pharmacokinetic behavior (30, 31). This is in line with our findings that inclusion of 20 mol% of EPG in PAA-coated liposomes did not result in significant changes in circulation kinetics and biodistribution (Fig. 3 and Table III). When 47.5 mol% of this lipid were included, only a slight decrease in circulation times was observed.

Fig. 3
figure 3

Pharmacokinetic behavior of EPC liposomes with a negative charge. EPC-liposomes (filled diamonds), EPC/EPG-liposomes in a molar ratio of 4:1 (filled circles) or 1:1 (filled squares) and EPC/PS-liposomes (9:1) (filled triangles) coated with PHEA (top left) and PHEG (top right) (% injected dose). Area under the curve (AUC0-48h) values (bottom) calculated from blood concentration time curves. All results are expressed as mean ± SD (n = 3–4). **p < 0.01; *p < 0.05; n.s. Not significant.

Full size image
Table III Percentage of Injected Dose of PHEA- and PHEG-Coated EPC-Liposomes with or without a Negative Charge in Liver and Spleen 48 h Post-Injection
Full size table

When the negatively charged phospholipid PS was included in PHEA- and PHEG-liposomes at a level of 10 mol%, a very rapid removal from the circulation was observed. Apparently, at the grafting density level of 5 mol%, PAAs are not able to sufficiently shield PS-containing liposomes from opsonization. This is in line with the findings that 5 mol% of PEG was not effective in protecting PS-containing liposomes from rapid clearance (24). PS containing PAA-liposomes showed an enhanced liver distribution when compared to PAA-liposomes containing EPG. The results indicate that PAA-coatings are able to shield the surface of EPG-containing liposomes more effectively than that of PS-containing liposomes. It has been suggested that the position of the negative charge within the molecule might play a role in the differences in shielding capacity (33). The negatively charged carboxyl group of PS is located at the terminal position of the hydrophilic part of the molecule, directed to the outer fluid, whereas the negatively charged phosphate group of EPG is located closer to the hydrophobic anchor and therefore shielded more effectively by the coating. In addition, the observed differences in pharmacokinetics of both types of liposomes may be related to the involvement of a receptor recognition mechanism in case of the PS-containing PAA-liposomes (33). PS has important physiological functions, it serves as signal for recognition and removal of aged erythrocytes and apoptotic cells by macrophages, which display a PS-specific receptor (34). In addition, it binds to clotting factors and initiates coagulation (35, 36). PS-containing liposomes are known to bind high amounts of plasma proteins. They are rapidly cleared from the circulation by the MPS, show in particular high uptake by Kupffer cells of the liver and activate the complement system (37, 38).

The differences in pharmacokinetics between the differently charged formulations were not reflected in the distribution to liver and spleen. However, it should be realized that the presented data likely give an underestimated impression of the actual hepatosplenic accumulation, as at 48 h post-injection liposomes may have already been subject to substantial degradation and consequently redistribution of the label may have occurred.

Besides liposomes with a negative surface charge, PAA-coated liposomes prepared with 47.5 mol% of the positively charged lipid DOTAP were studied. These cationic liposomes were very rapidly cleared from the bloodstream, independent of the type of PAA-coating. Less than 1% of the administered dose was detected in the blood circulation one hour after injection of PHEA- or PHEG-liposomes. Cationic liposomes have been reported to interact strongly with blood components and may even induce aggregate formation in the circulation (29, 39, 40). PHEA and PHEG at a grafting density of 5 mol% are apparently not capable to shield and prolong the circulation times of cationic liposomes containing high levels of DOTAP (47.5 mol%).

Effect of Particle Size

The circulation kinetics and AUC values of PHEA- and PHEG-liposomes of approximately 360, 230 and 120 nm are shown in Fig. 4. Table IV summarizes their distribution to liver and spleen. Liposomes with a mean size of 120 nm were removed from the bloodstream at a lower rate than liposomes of 230 and 360 nm, respectively, for both PHEA- and PHEG-coated liposomes. This observation has also been reported in independent studies for conventional and PEGylated liposomes (25, 30, 32, 41–43). PEG-liposomes with a diameter of more than 200 nm are more rapidly cleared than liposomes with a diameter between 100 and 200 nm. A decrease in size has been shown to also reduce recognition by the complement system (44–46). At 48 hours after administration, the liposome formulation tested showed only small differences regarding the degree of uptake by liver and spleen. As mentioned above, differences may have been obscured as at this late time point redistribution of the label may already have occurred. Our results indicate that, as observed for PEG-liposomes, also for PHEA- and PHEG-liposomes a particle size below 150 nm is the most optimal to achieve long-circulation properties.

Fig. 4
figure 4

Pharmacokinetic behavior of EPC/EPG (1:1) liposomes coated with PHEA (top left) and PHEG (top right) of 360 nm (filled squares) 230 nm (filled circles) and 120 nm (filled diamonds) (% injected dose). Area under the curve (AUC0–48 h) values (bottom) calculated from blood concentration time curves. All results are expressed as mean±SD (n = 4). **p < 0.01; n.s. Not significant.

Full size image
Table IV Percentage of Injected Dose of PHEA- and PHEG-Coated EPC/EPG Liposomes in Liver and Spleen 48 h Post-Injection
Full size table

Effect of Lipid Dose

The effect of the administered lipid dose of PHEG-liposomes on the circulation kinetics and AUC values is shown in Fig. 5. No differences in clearance were observed when liposomes were administered at doses of 25, 0.25, and 0.025 μmol/kg. At 48 hours after administration, liver uptake was also independent of the lipid dose but spleen uptake was lower at both lower lipid doses (Table V). For PEG-liposomes dose-independent kinetics have been observed over a dose range between 4 and 400 μmol/kg (47). However, at doses <1 μmol/kg, which are used in case of for instance diagnostic imaging of tumors and sites of inflammation, PEG-liposomes are much more rapidly cleared from the circulation (48, 49).

Fig. 5
figure 5

Pharmacokinetic behavior of EPC/EPG-liposomes coated with PHEG at 25 μmol/kg (filled squares), 0.25 μmol/kg and (filled diamonds) 0.025 μmol/kg (filled squares) (% injected dose) (left). Area under the curve (AUC0–48 h) values (right) calculated from blood concentration time curves. All results are expressed as mean±SD (n = 4). n.s. Not significant.

Full size image
Table V Percentage of Injected Dose of PHEG-Coated EPC/EPG Liposomes in Liver and Spleen 48 h Post-Injection
Full size table

The mechanism behind this ‘enhanced clearance effect’ at low lipid dose is not yet fully understood. It has been suggested that rapid clearance is mediated by a limited pool of circulating opsonic factors that interact with PEG-liposomes and induce phagocytosis. At higher lipid doses, the influence of this relatively low amount of opsonic factors becomes more restricted and the majority of the administered liposomes will be long-circulating (49, 50). For PHEA-liposomes, a recent study showed dose-independent pharmacokinetics for doses as low as 0.025 μmol/kg (51). The present study demonstrates that this is also the case for PHEG-liposomes. This benefit of PAA-liposomes over PEG-liposomes can be advantageous in situations in which only low lipid doses of liposomes are required for the desired application.

CONCLUSION

This study demonstrates that the prolonged circulation kinetics of PAA-liposomes can be maintained upon variation of liposome characteristics and the lipid dose given. This allows the possibility to tailor the liposome composition to the required properties for achieving efficient drug targeting, without the risk of negative effects on the long-circulation property of the PAA-liposomes. However, the use of relatively high amounts of strongly charge-inducing lipids and a too large mean size is to be avoided.

In addition to their advantage of being enzymatically degradable, PAAs are able to sterically stabilize a broad range of liposomes varying in physicochemical properties and therefore PAA-liposomes represent a versatile drug carrier system for a wide variety of applications.