Characterization of lipid DNA interactions. I. Destabilization of bound lipids and DNA dissociation

We have recently described a method for preparing lipid-based DNA particles (LDPs) that form spontaneously when detergent-solubilized cationic lipids are mixed with DNA. LDPs have the potential to be developed as carriers for use in gene therapy. More importantly, the lipid-DNA interactions that give rise to particle formation can be studied to gain a better understanding of factors that govern lipid binding and lipid dissociation. In this study the stability of lipid-DNA interactions was evaluated by measurement of DNA protection (binding of the DNA intercalating dye TO-PRO-1 and sensitivity to DNase I) and membrane destabilization (lipid mixing reactions measured by fluorescence resonance energy transfer techniques) after the addition of anionic liposomes. Lipid-based DNA transfer systems were prepared with pInexCAT v.2.0, a 4.49-kb plasmid expression vector that contains the marker gene for chloramphenicol acetyltransferase (CAT). LDPs were prepared using N-N-dioleoyl-N,N-dimethylammonium chloride (DODAC) and either 1, 2-dioleoyl-sn-glycero-3-phosphocholine (DOPC) or 1, 2-dioleoyl-sn-glycero-3-phosphoethanolamine (DOPE). For comparison, liposome/DNA aggregates (LDAs) were also prepared by using preformed DODAC/DOPE (1:1 mole ratio) and DODAC/DOPC (1:1 mole ratio) liposomes. The addition of anionic liposomes to the lipid-based DNA formulations initiated rapid membrane destabilization as measured by the resonance energy transfer lipid-mixing assay. It is suggested that lipid mixing is a reflection of processes (contact, dehydration, packing defects) that lead to formulation disassembly and DNA release. This destabilization reaction was associated with an increase in DNA sensitivity to DNase I, and anionic membrane-mediated destabilization was not dependent on the incorporation of DOPE. These results are interpreted in terms of factors that regulate the disassembly of lipid-based DNA formulations.

Comparison of different hydrophobic anchors conjugated to poly(ethylene glycol): effects on the pharmacokinetics of liposomal vincristine

Poly(ethylene glycol) (PEG) conjugated lipids have been used to increase the circulation longevity of liposomal carriers encapsulating therapeutic compounds. PEG is typically conjugated to distearoylphosphatidylethanolamine (DSPE) via a carbamate linkage that results in a net negative charge on the phosphate moiety at physiological pH. It was anticipated that the presence of this negative charge could have deleterious effects on liposome pharmacokinetic characteristics. We describe here the synthesis of a new class of neutrally charged PEG-lipid conjugates in which the PEG moiety was linked to ceramide (CER). These PEG-CER conjugates were compared with PEG-DSPE conjugates for their effects on the pharmacokinetics of liposomal vincristine. PEG-CER (78% palmitic acid, C16) and PEG-DSPE achieved comparable increases in the circulation lifetimes of sphingomyelin/cholesterol (SM/chol) liposomes. However, PEG-DSPE significantly increased the in vitro and in vivo leakage rates of vincristine from SM/chol-based liposomes compared to vincristine leakage observed when PEG-CER was used. The increase in drug leakage observed in vitro that was due to the presence of PEG-DSPE was likely due to the presence of a negative surface charge. Analysis of the electrophoretic mobilities of these formulations suggested that the negative surface charges were shielded by approx. 80% by the PEG layer extending from the membrane surface. In contrast, formulations containing PEG-CER had no surface charge and no electrophoretic mobility. A comparison of the effects of the ceramide acyl chain length (C8 through C24) on the pharmacokinetics of SM/chol/PEG-CER formulations of vincristine demonstrated that longer acyl chains on the PEG-CER were associated with longer circulation lifetimes of the liposomal carriers and, consequently, higher plasma vincristine concentrations. These data suggest that the short chain PEG-ceramides underwent rapid partitioning from the vesicles after i.v. administration, whereas the longer chain PEG-ceramides had stronger anchoring properties in the liposome bilayers and partitioned slowly from the administered vesicles. These data demonstrate the utility of ceramide-based steric stabilizing lipids as well as the potential for developing controlled release formulations by manipulating the retention of the PEG-ceramide conjugate in liposome bilayers.

Pharmacokinetic behavior of vincristine sulfate following administration of vincristine sulfate liposome injection

The pharmacokinetic behavior of vincristine sulfate (VINC) following administration of vincristine sulfate liposome injection (VSLI), 0.16 mg/ml, as an intravenous infusion over 60 min in 24 of 25 patients enrolled in a phase I clinical study of this drug is described. Plasma samples for determination of the pharmacokinetic behavior of VINC were collected during the infusion at 15, 30 and 60 min as well as at 2, 4, 8, 12, 48 and 72 h postinfusion. Total VINC concentration was determined using a validated high-performance liquid chromatographic (HPLC) assay. Patients receiving doses of 0.5 to 1.5 mg/m2 VSLI did not provide useful pharmacokinetic data at late time-points owing to the limit of quantitation of the HPLC assay (28.6 ng/ml). Sufficient concentration-time data were available for seven of the patients receiving doses of VSLI from 2.0 to 2.8 mg/m2 for compartmental modelling. A two-compartment open model (PCNONLIN Model 10) was the best fit for the observed VINC plasma data for these patients. The mean maximum observed concentration values were significantly greater for patients receiving VSLI at 2.8 mg/m2 (2260 +/- 212 ng/ml, n = 2) than for those receiving 2.0 mg/m2 and 2.4 mg/m2 (891 +/- 671 ng/ml, n = 6; 679 +/- 634 ng/ml, n = 6, respectively). No significant differences were observed in maximum concentration values between patients at 2.0 mg/m2 and those at 2.4 mg/m2. A trend towards higher parametric AUC (0 to infinity) values with increasing dose (on a milligram per meter squared basis) was observed but statistical significance was not reached. Comparison of the pharmacokinetic behavior of VSLI observed in this study with nonencapsulated VINC demonstrated that (1) the variability observed for VSLI pharmacokinetic parameters was similar to nonencapsulated VINC, (2) although variability in absolute concentration was observed between patients, the behavior of VSLI in individual patients followed a two- rather than a three-compartment open model, and (3) VINC plasma concentrations were significantly greater following administration of VSLI than described for nonencapsulated VINC. Overall, the results for patients treated with VSLI from 2.0 to 2.8 mg/m2 suggest that this formulation protects VINC from the early phase of rapid elimination seen with nonencapsulated drug, resulting in significantly elevated VINC plasma concentrations over extended periods of time.

Cationic liposome--plasmid DNA complexes used for gene transfer retain a significant trapped volume

The goal of this study is to determine whether cationic liposomes retain any trapped volume after their complexation to plasmid DNA. This serves two purposes: to further the understanding of the physical nature of liposome/plasmid DNA complexes used in gene therapy and to investigate the potential for codelivery of other encapsulated molecules with the liposome-DNA complexes. Cationic liposomes composed of N,N-dioleoyl-N,N-dimethylammonium chloride and dioleoylphosphatidylethanolamine (DODAC/DOPE, 50/50 mol %) encapsulating an aqueous trap marker were used to prepare liposome-DNA complexes at various charge ratios. The trapped volume before and after DNA binding was measured by two methods: dialysis and filtration. The effect of tissue culture medium on trapped volume was also investigated. A lipid-mixing assay was employed to further characterize the aggregation events that influence trap volume. The trapped volume (Vt) of neutral control liposomes was 1.1 +/- 0.04 microL/mumol, which was not affected by the addition of DNA. For cationic liposomes in the absence of DNA the Vt was 1.45 +/- 0.46 and 1.54 +/- 0.08 microL/mumol, as measured by the filtration and dialysis methods, respectively. After addition of DNA, the residual trapped volume (RVt) decreased to 0.43 +/- 0.1 microL/mumol and 0.47 +/- 0.05 microL/mumol, as determined by each method, respectively. RVt increased as the ratio of cationic lipid to DNA (nmol of lipid/mg of DNA) was increased above 10, a ratio that corresponds to a charge ratio (positively charged lipids to negatively charged phosphate groups) of 1.62. Aggregation and lipid-mixing were greatest at charge ratios coinciding with the lowest trapped volume. In the presence of tissue culture medium, the Vt of cationic liposomes but not neutral liposomes was reduced, suggesting that the salts have a direct effect on cationic liposomes in the absence of DNA. The RVt of both neutral and cationic liposomes in the presence of DNA, however, was not different from that of the liposomes in the absence of DNA. These results suggest that a significant trapped volume is retained by cationic liposomes after binding to plasmid DNA. This is an important finding with regard to the potential use of DNA/liposome complexes in the codelivery of other bioactive molecules at the time of cell transfection.

An immune response to ovalbumin covalently coupled to liposomes is prevented when the liposomes used contain doxorubicin

It is now well established that liposomes with surface associated proteins are immunogenic. Repeated administration of protein coated liposomes elicits the generation of antibodies and the elimination of proteoliposome increases markedly in animals 'immunized' with such liposomes. This immune response compromises the therapeutic potential of liposomal formulations that rely on the use of protein- or peptide-based targeting ligands to enhance cell specificity. Strategies to suppress or inhibit such immune responses must be developed if this technology is going to prove therapeutically viable. This study evaluates whether an immune response to a protein, covalently attached to liposomes by a thioether bond between N-succinimidyl-3-(2-pyridyldithio)propionate (SPDP)-modified-protein and N-(4-(P-maleimidophenyl)butyryl) (MPB)-activated lipids, can be suppressed when the liposomes used contain the anti-cancer drug doxorubicin. To assess this, the highly immunogenic protein ovalbumin was conjugated onto liposomes composed of distearoylphosphatidylcholine/cholesterol (DSPC/Chol) with sufficient poly(ethylene glycol)-modified distearoyl phosphatidylethanolamine (PEG-DSPE) (2 mol%) to prevent liposome aggregation during protein coupling and to engender increased circulation lifetimes. The immune response to these liposomes with and without encapsulated doxorubicin was measured by: (1) monitoring liposome elimination after 3 weekly i.v. injections in C3H/HeJ mice and (2) measuring the anti-ovalbumin antibody levels by an ELISA assay. One week after a single dose of ovalbumin-coated PEG liposomes (50 microg protein/mouse) the immune response resulted in rapid elimination of a second dose of ovalbumin-coated PEG liposomes. Rapid liposome elimination was correlated to generation of high levels (> 9 microg/ml plasma) of circulating anti-ovalbumin IgG. In contrast, anti-ovalbumin antibodies were not detected when the liposomes used contained doxorubicin. Plasma elimination of these drug loaded protein coated liposomes decreased following repeated weekly i.v. doses, an effect that is consistent with liposomal doxorubicin mediated suppression of phagocytic cells in the liver.

Validation of a high-performance liquid chromatographic assay method for quantification of total vincristine sulfate in human plasma following administration of vincristine sulfate liposome injection

The validation of a high performance liquid chromatographic (HPLC) assay method for quantitation of total vincristine sulfate (VINC) in human plasma is described. VINC was extracted from plasma using BondElut CBA solid phase cartridges with vinblastine as the internal standard. Chromatography was accomplished using a Waters Symmetry C8 (250 mm x 4.6 mm i.d.) analytical column, a Waters Delta-Pak ODS guard column with a mobile phase of 34.9% water-0.1% diethylamine (pH 7.0)-40% acetonitrile-25% methanol pumped isocratically at 1.0 ml min(-1) with ultraviolet detection at 297 nm. Above the limit of quantitation of 28.6 ng ml(-1), the area ratio precision (R.S.D. range 3.33-11.6%) and accuracy of predicted values (R.S.D. range 8.56-23.8% with the limit of quantitation being the only value above 20%) were acceptable. The assay was linear from 28.6-2860 ng ml(-1) VINC in plasma. Recovery of VINC from plasma and VINC from plasma spiked with vincristine sulfate liposome injection ranged from 74.9-87.1%. Stability of VINC in plasma stored at -20 degrees C for at least 49 days and of extracted plasma samples was demonstrated. Potential interference in quantitation of VINC from commonly co-administered drugs was evaluated along with day-to-day variability. The assay procedure was found suitable for evaluation of VINC clinical pharmacokinetics in plasma following administration of vincristine sulfate liposome injection prepared using distearoylphosphatidylcholine (DSPC)/cholesterol liposomes for injection.

Analysis of cationic liposome-mediated interactions of plasmid DNA with murine and human melanoma cells in vitro

Lipid-based DNA transfer formulations are typically selected on the basis of in vitro transfection studies where the activity of specific formulations is defined by transgene expression. It is unclear, however, whether expression is directly related to the efficiency of DNA transfer. In an attempt to correlate DNA transfer with transgene expression, we used a simple assay consisting of measuring DNA (3H-plasmid encoding for beta-galactosidase) binding to murine (B16/BL6) and human (KZ) melanoma cells in vitro at 4 and 37 degrees C. The difference in cell association at these temperatures was assumed to be a consequence of DNA uptake, an assumption that was confirmed by protease removal of cell surface-associated DNA. DNA associated with B16/BL6 melanoma cells (up to 30 ng or 12% of the added DNA) following incubation with dioleoyldimethylammonium chloride/dioleoylphosphatidylethanolamine (DOPE) liposome-DNA aggregates was comparable to that achieved with 1,2-dioleoyloxypropyl-3-trimethylammonium bromide/DOPE or dimethyldioctadecylammonium bromide/DOPE liposomes; however, transgene expression was 2- and 5-fold less for the latter two formulations, respectively. Similarly, equivalent amounts of DNA delivery were achieved with B16/BL6 and KZ melanoma cells, yet the level of transgene expression in the KZ cells was undetectable. It was demonstrated that the lack of transgene expression was not a consequence of cell-specific differences in DNA degradation.

Influence of drug release characteristics on the therapeutic activity of liposomal mitoxantrone

The influence of liposome drug release on the therapeutic activity of encapsulated mitoxantrone was investigated. Liposomes prepared from 1,2-distearoyl-sn-glycero-3-phosphocholine (DSPC)/cholesterol (Chol) (55:45, molar ratio) or 1,2 dimyristoyl-sn-glycero-3-phosphocholine (DMPC)/Chol (55:45, molar ratio) were loaded with mitoxantrone using the transmembrane pH gradient loading procedure. In vivo studies demonstrated that DMPC/Chol liposomes released drug faster (1.7 microg drug/microg lipid/hr) than did DSPC/Chol liposomes (<0.025 microg drug/microg lipid/hr). In BDF1 mice, the acute toxicities of DMPC/Chol and DSPC/Chol liposomal mitoxantrone were similar, with a maximum tolerated dose of approximately 30 mg drug/kg, in comparison with the maximum tolerated dose of free drug, which was approximately 10 mg/kg. Efficacy studies were conducted in BDF1 mice inoculated i.v. with murine P388 cells or L1210 tumor cells. These cells seed in the liver and spleen of animals after i.v. inoculation, and a single dose of DMPC/Chol liposomal mitoxantrone of 10 mg drug/kg resulted in 100% of the treated animals surviving for >60 days. In contrast, no long-term survivors were obtained in any other treatment group, even when drug doses were escalated to the maximum tolerated dose. Pharmacodynamic studies with DMPC/Chol liposomal mitoxantrone and DSPC/Chol liposomal mitoxantrone illustrate the importance of achieving a balance between drug release characteristics and drug delivery to the site of tumor progression.

Accumulation of liposomal lipid and encapsulated doxorubicin in murine Lewis lung carcinoma: the lack of beneficial effects by coating liposomes with poly(ethylene glycol)

The efficiency of drug accumulation in tumors was measured after intravenous administration of doxorubicin encapsulated in distearoyl phosphatidylcholine/cholesterol liposomes prepared in the presence or absence of 5 mol % polyethylene glycol-modified phosphatidylethanolamine (PEG-PE). These liposomal formulations of doxorubicin were administered at the maximum tolerated dose in female BDF-1 mice bearing subcutaneously established Lewis Lung carcinoma. The parameters used to determine tumor targeting efficiency (T(e)) included area under the doxorubicin plasma (AUC(P)) and tumor (AUC(T)) concentration-time curves. Extended time-course studies evaluating lipid and drug levels in plasma and tumors during 7 days after administration indicated that the T(e) (AUC(T)/AUC(P)) was greater for liposomes that did not contain PEG-PE. The AUC(P) after administration of free doxorubicin, doxorubicin encapsulated in distearoyl phosphatidylcholine/cholesterol liposomes and doxorubicin encapsulated in distearoyl phosphatidylcholine/cholesterol/PEG-PE-stabilized liposomes were 0.087 micromol x ml(-1) x h, 50 micromol x ml(-1) x h and 78 micromol x ml(-1) x h, respectively. Maximum drug levels achieved in the tumors were similar for both liposomal doxorubicin formulations, 140 microg (250 nmol)/g tumor; however, this level was achieved faster when the liposomes did not contain PEG-PE. Maximum levels measured after administration of free drug were less than 5 microg/g tumor, and these were achieved within 15 min. The results suggest that some of the benefits associated with the use of PEG-modified liposomes, such as increased blood levels and enhanced circulation lifetime, may be of little advantage in terms of maximizing liposomal drug accumulation in sites of tumor growth.

The role of tumor-associated macrophages in the delivery of liposomal doxorubicin to solid murine fibrosarcoma tumors

Murine fibrosarcoma tumors arising from subcutaneous inoculation of FSa-N cells exhibit 4-fold higher tumor-associated macrophage (TAM) levels than those from the FSa-R line. These solid tumors were used to assess the role of TAMs in the accumulation of liposomal anticancer drugs. Two liposomal formulations of doxorubicin were investigated: a conventional formulation composed of distearoylphosphatidylcholine (DSPC) and cholesterol and a sterically stabilized liposomal formulation composed of DSPC/cholesterol/poly (ethylene glycol)-modified distearoylphosphatidyethanolamine (PEG-PE). Circulating concentrations of PEG-PE containing liposomes 24 h after i.v. administration were 3-fold greater than those observed after administration of conventional liposomes. No differences were observed in drug retention or tumor (FSa-R or FSa-N) drug and liposomal lipid delivery when comparisons were made between different liposomal formulations. However, tumor doxorubicin concentrations were increased as much as 4-fold for liposomal formulations relative to free drug. Further, there was a 1.5- to 2-fold increase in doxorubicin delivery to TAM-enriched FSa-N tumors compared with FSa-R tumors. Fluorescence microscopy studies revealed a poor correlation between CD11b (Mac-1) positive cells (TAMs) and the appearance of doxorubicin fluorescence. These results suggest that uptake of liposomal drugs by TAMs does not account for the enhanced accumulation of liposomal drugs in solid tumors. Rather, the increased tumor drug delivery may be related to alternative TAM-mediated processes that increase tumor vascular permeability. Therapeutic studies demonstrated that increased tumor drug uptake observed for the liposomal doxorubicin formulations led to marginal improvements in antitumor activity, and it is suggested that much of the drug delivered in liposomal form is not biologically available.

Self-assembling DNA-lipid particles for gene transfer

PURPOSE

We have demonstrated that a heteromolecular complex consisting of cationic lipids and DNA can be prepared and isolated (1). Cationic lipids bind DNA through electrostatic interactions. However, when sufficient lipids are bound to DNA the physical and chemical properties of the complex are governed by hydrophobic effects. Here we describe an approach where this hydrophobic complex is used as an intermediate in the preparation of lipid-DNA particles (LDPs).

METHODS

The approach relies on the generation of mixed micelles containing the detergent, n-octyl beta-D-glucopyranoside (OGP), the cationic lipid, N-N-dioleoyl-N, N-dimethylammonium chloride (DODAC), and selected zwitterionic lipids, 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine (DOPE) or egg sphingomyelin (SM).

RESULTS

When these micelles were prepared at low detergent concentrations (20 mM OGP) and combined with pCMV beta DNA, LDPs spontaneously formed. The mean diameter of these particles as measured by quasielastic light scattering was 55-70 nm, a result that was confirmed by negative stain electron microscopy. Further characterization of these LDPs showed that DNA within the particles was inaccessible to the small fluorochrome TO-PRO-1 and protected against DNase I degradation. LDPs could also be prepared in high concentrations of OGP (100 mM), however particles formed only after removal of OGP by dialysis. Particles formed in this manner were large (> 2000 nm) and mediated efficient transfection of Chinese hamster ovary cells. Transfection activity was greater when the lipid composition used consisted of SM/ DODAC. Small particles (< 100 nm) prepared of SM/DODAC were, however, inefficient transfecting agents.

CONCLUSIONS

We believe that LDP formation is a consequence of the molecular forces that promote optimal hydrocarbon-hydrocarbon interactions and elimination of the hydrocarbon-water interface.

Plasma clearance, biodistribution and therapeutic properties of mitoxantrone encapsulated in conventional and sterically stabilized liposomes after intravenous administration in BDF1 mice

Mitoxantrone can be efficiently loaded into large unilamellar vesicles using a transmembrane pH gradient. Release studies indicate that these drug-loaded carriers are highly stable and even after dissipation of the residual pH gradient retain more than 85% of encapsulated mitoxantrone following dialysis at 37 degrees C for 5 days. In murine studies we have compared the plasma clearance and biodistribution of both mitoxantrone and liposomal lipid following intravenous administration of free drug or mitoxantrone encapsulated in either conventional or sterically stabilized liposomes. In contrast to the rapid blood clearance observed for free mitoxantrone, both liposomal systems provided extended circulation lifetimes, with over 90% of the drug present 1 h after administration and 15-30% remaining at 24 h. In agreement with previous reports, longer plasma half-lives were observed for sterically stabilized liposomes than for conventional systems. In addition, a strong correlation between drug and carrier biodistribution was seen, with uptake occurring mainly in the liver and spleen and paralleling plasma clearance. This would suggest that tissue disposition reflects that of drug-loaded liposomes rather than the individual components. Liposomal encapsulation also significantly reduced mitoxantrone toxicity, allowing administration of higher, more efficacious drug doses. In a murine L1210 tumour model, for example, no long-term survivors were seen in animal groups treated with free drug, whereas at the maximum therapeutic dose of liposomal mitoxantrone survival rates of 40% were observed.

Intratumor distribution of doxorubicin following i.v. administration of drug encapsulated in egg phosphatidylcholine/cholesterol liposomes

PURPOSE

A pharmacological evaluation of an egg phosphatidylcholine/cholesterol (55:45 mole ratio, EPC/Chol) liposome doxorubicin formulation was carried out. The objective was to define liposomal lipid and drug distribution within sites of tumor growth following intravenous (i.v.) administration to female BDF1 mice bearing either Lewis lung carcinoma, B16/BL6 melanoma, or L1210 ascitic tumors.

METHODS

Mice were injected i.v. with EPC/Chol liposomal doxorubicin, and plasma and tumor levels of lipid and drug were determined 1, 4 and 24 h late with radiolabeled lipid and fluorimetry or fluorescence microscopy, respectively. In addition, single-cell suspensions of the Lewis lung and B16/BL6 tumors were prepared and the presence of macrophages was determined with an FITC-labeled rat antimouse CD11b (MAC-1) antibody.

RESULTS

For mice bearing the Lewis lung solid tumors, there was a time-dependent accumulation of liposomal lipid, with a plateau of approximately 500 micrograms lipid/g tumor at 48 h. In contrast, the apparent plateau (microgram doxorubicin/g tumor) for doxorubicin was achieved at 1 h and remained constant over a 72-h time course. In comparison with free drug administered at the maximum tolerated dose (MTD, 20 mg/kg) doxorubicin levels in tumors were two- to threefold greater when the drug was administered in liposomal form. The increase in drug delivery was comparable for both solid tumors. With animals bearing the L1210 ascitic tumor, drug exposure was as much as ten times greater (in comparison with free drug) when doxorubicin was administered in liposomes. An evaluation of single-cell suspensions prepared from the two solid tumors suggested that more than 98% of the tumor-associated drug and liposomal lipid was not tumor cell-associated. Histological studies with the Lewis lung carcinoma, however, revealed that a proportion of the drug did colocalize with tumor-associated macrophages. Analysis of cells obtained from mice bearing ascitic tumors showed that more than 80% of the cell-associated drug could be removed by procedures designed to remove adherent cells.

CONCLUSION

The results summarized here suggest drug concentrations within a solid tumor, such as the Lewis lung carcinoma, are constant over time when the drug is given in a "leaky" EPC/Chol formulation. The results also suggest that liposomal lipid within sites of tumor growth is primarily localized within the interstitial spaces or tumor-associated macrophages.

Cationic lipid binding to DNA: characterization of complex formation

We recently demonstrated that cationic lipids, added in monomer or micellar form, bind to DNA, resulting in the formation of a hydrophobic complex. This complex can serve as a well-defined intermediate in the preparation of DNA-lipid particles (DLPs) with many potential applications for delivery of polynucleotides in vitro and in vivo. To develop a better understanding of the factors governing complex formation, we have characterized the cationic lipid/DNA binding reaction. This was evaluated by measuring DNA and cationic lipid (DODAC) complex formation using the Bligh and Dyer extraction procedure. Efficient recovery of DNA (> 95%) in the organic phase was achieved when sufficient monocationic lipids interact with DNA phosphate groups. The rate of binding depends on the amount of DNA or cationic lipid present in the system. The time required to generate the hydrophobic complex was increased when < 10 micrograms of DNA or < 40 nmol of DODAC was present. Surprisingly, the rate of complex formation was contingent on the incubation period after partitioning the DNA/lipid mixture into organic and aqueous phases. These results suggest that the cationic lipid/DNA complex forms at the aqueous/organic interface and that DNA/lipid binding is dependent on multivalent interactions at this interface. A Scatchard analysis of DNA/DODAC binding demonstrated that the binding reaction exhibits a high degree of positive cooperativity. The apparent dissociation constant (Kn), using data obtained under conditions where DODAC binding to DNA approached saturation, indicated a high-affinity reaction (Kn > 10(-11) mol L-1). At this point, approximately 8400 mol of DODAC was bound per mole of DNA, which is equivalent to a charge ratio (+/-) of 0.585 for the 7.2 kb plasmid used and suggests that formation of the hydrophobic complex occurs at a stage prior to charge neutralization. The influence of other lipids on DNA/cationic lipid binding at the aqueous/organic interface was also studied. Cholesterol and DOPC had little effect on DNA/DODAC binding while the anionic lipids LPI, DOPS, and DMPG inhibited complex formation. The zwitterionic lipid DOPE, however, had a concentration-dependent effect on cationic lipid binding that was also dependent on the mixing order. We believe that this approach for evaluating lipid/DNA binding provides an effective procedure for assessing factors which control the dissociation of lipids from DNA and may be beneficial in the selection of lipids for effective use in gene transfection studies.

Plasmid DNA is protected against ultrasonic cavitation-induced damage when complexed to cationic liposomes

Cationic liposomes bound to plasmid DNA are currently used for in vitro and in vivo gene therapy applications, but such complexes readily form large, heterogeneous aggregates that are not appropriate for pharmaceutical development. More importantly, size heterogeneity makes studies focused on optimizing gene transfer to cells difficult to conduct or understand. For this reason we have evaluated the effect of microprobe sonication on these complexes in an effort to achieve process-controlled size homogeneity. Complexes were prepared using a 7.2 kb reporter plasmid and the following liposomal lipid combinations: DDAB/DOPE (50:50 mol %), DDAB/DOPE/PEG-PE (50:45:5 mol %), DDAB/EPC (50:50 mol %), DDAB/EPC/PEG-PE (50:45:5, 50:40:10, 50:35:15 mol %), DODAC/DOPE (50:50 mol %), and DODAC/EPC (50:50 mol %) (DDAB, dimethyldioctadecylammonium bromide; DOPE, dioleoylphosphatidylethanolamine; PEG-PE, monomethoxypolyethylene glycol2000 succinate- distearoylphosphatidylethanolamine; EPC, egg phosphatidylcholine; DODAC, dioleoyldimethylammonium chloride). The influence of complex composition and lipid:DNA ratio was evaluated. Particle size was determined before and after complexation and again after sonication using the quasi-elastic light scattering technique. DNA integrity was assessed via agarose gel electrophoresis. Finally, gene transfection was evaluated using CHO cells that were transfected in vitro with sonicated and unsonicated complexes. It is established in this study that size reduction can occur, but this is dependent on cationic and neutral lipid composition and, in some cases, lipid:DNA ratio. Surprisingly, the process of sonication leaves a significant percentage of the plasmid DNA intact and capable of in vitro transfection. This study shows that plasmid DNA can be protected from damage due to sonication by liposome complex formation. This may indicate that more common pharmaceutical methods for size reduction which subject particles to mechanical stress may be applicable in preparation of liposome/DNA formulations for in vivo application.

Vincristine-induced dermal toxicity is significantly reduced when the drug is given in liposomes

A problem associated with the intravenous delivery of vincristine concerns drug extravasation at the site of injection or infusion. This can result in extensive local soft-tissue damage. A new formulation of vincristine has recently been developed based on encapsulation of the drug in liposomes. The liposomal drug is somewhat less toxic and substantially more efficacious than free drug. The studies described here assessed, using a murine model of drug extravasation, whether vincristine encapsulation in liposomes influences drug-induced dermal toxicity. It was shown that subcutaneous injection of vincristine in liposomes does not result in the gross skin necrosis and ulceration observed following injection of free drug. Histological analysis of the dermal tissue surrounding the injection site suggests that free drug induces a pronounced inflammatory reaction as judged by the presence of infiltrating leukocytes. In contrast, the liposomal formulation of vincristine engenders a mild prolonged inflammatory condition. These toxicological studies were correlated with an evaluation of drug retention at the site of administration. It was shown using radiolabelled vincristine as a drug marker, that free vincristine is rapidly eliminated from the injection site. In contrast, the level of drug at the site of injection was far greater when the drug was given in liposomal form.

Accumulation of protein-coated liposomes in an extravascular site: influence of increasing carrier circulation lifetimes

The primary objective of this work was to test whether increased blood levels and circulation lifetimes result in increased passive targeting of protein-coated liposomal drug carriers. The system used to evaluate this was based on i.v. injection of 100 nm of distearoyl phosphatidylcholine/cholesterol liposomes with covalently bound streptavidin. The circulation lifetime of these liposomes was increased by procedures that involved blockade of liposome uptake by phagocytic cells in the liver and/or the incorporation of a poly(ethylene glycol)-modified phospholipid [poly(ethylene glycol)2000-modified distearoyl phosphatidylethanolamine]. Blockade of liver phagocytic cells with a low predose (2 mg/kg of drug) of liposomal doxorubicin increased the circulation half-life of the streptavidin liposomes from less than 1 hr to greater than 3 hr. A further 2-fold increase in circulating half-life (to approximately 7.5 hr) was achieved by using liposomes with 2 mole % of poly(ethylene glycol)2000-modified phosphatidylethanolamine. In combination with RES blockade, the circulation lifetimes of poly(ethylene glycol)phosphatidylethanolamine containing streptavidin liposomes could be increased to greater than 12 hr. The ability of these liposomes to move from the plasma compartment to an extravascular compartment was measured by using the peritoneal cavity as a convenient, accessible, extravascular site. The tendency for liposomes to accumulate in this site was not, however, clearly dependent on circulating blood levels. Comparable levels of liposomes in the peritoneal cavity were achieved when using systems that exhibited significantly different circulation lifetimes.

Liposomal cyclosporine. Comparison of drug and lipid carrier pharmacokinetics and biodistribution

In a preceding paper (Ouyang et al., 1995, this issue), we have characterized cyclosporine incorporation into well-defined liposomal systems, large unilamellar vesicles. This study demonstrated that only modest drug levels could be accommodated within the membrane, particularly for cholesterol-containing liposomes, and that rapid drug exchange could occur between vesicles. This raised the possibility that following intravenous administration, drug migration to other blood components might negate the potential benefits arising from liposomal delivery. We have, therefore, examined the pharmacokinetics and biodistribution of both cyclosporine and its liposomal carrier. We show that whereas liposomes, as expected, are only slowly cleared from the blood, redistribution of cyclosporine occurs much more rapidly. Further we have shown that liposomal loss of cyclosporine in blood results from drug migration to the lipoproteins and, to a lesser extent, the erythrocytes. As a result, while liposomes accumulate preferentially in organs of the reticuloendothelial system after intravenous administration, tissue cyclosporine levels, in general, do not reflect the distribution profile obtained for the liposomal carrier.

Formation of novel hydrophobic complexes between cationic lipids and plasmid DNA

An ability to generate a well defined lipid-based carrier system for the delivery of plasmid DNA in vivo requires the characterization of factors governing DNA/lipid interactions and carrier formation. We report that a hydrophobic DNA/lipid complex can be formed following addition of cationic lipids to DNA in a Bligh and Dyer monophase consisting of chloroform/methanol/water (1:2.1:1). Subsequent partitioning of the monophase into a two-phase system allows for the extraction of DNA into the organic phase. When using monovalent cationic lipids, such as dimethyldioctadecylammonium bromide, dioleyldimethylammonium chloride, and 1,2-dioleyl-3-N,N,N-trimethylaminopropane chloride, greater than 95% of the DNA present can be recovered in the organic phase when the lipid is added at concentrations sufficient to neutralize DNA phosphate charge. When the polyvalent cationic lipids 2,3-dioleyloxy-N-[2(sperminecarboxamido)ethyl]-N,N-dimethyl- 1- propanaminium trifluoroacetate and diheptadecylamidoglycyl spermidine are used, efficient extraction of the DNA into the organic phase is also achieved when the charge ratio between lipid and DNA is approximately equal. Formation of the hydrophobic DNA complex can only be achieved with cationic lipids. In the absence of added cations or in the presence of excess Ca2+, L-lysine, or poly(L-lysine), 100% of the DNA is recovered in the aqueous fraction. The monovalent cationic lipid/DNA complexes can also be prepared in the presence of detergent; however, low concentrations of NaCl (< 1 mM) lead to dissociation of the complex. Importantly, these results clearly demonstrate that cationic lipid binding does not lead to DNA condensation. The methods described, therefore, enable DNA/lipid complexes to be characterized in the absence of DNA condensation.(ABSTRACT TRUNCATED AT 250 WORDS)

Sphingomyelin-cholesterol liposomes significantly enhance the pharmacokinetic and therapeutic properties of vincristine in murine and human tumour models

This study reports on the development of a liposomal formulation of vincristine with significantly enhanced stability and biological properties. The in vitro and in vivo pharmacokinetic, tumour delivery and efficacy properties of liposomal vincristine formulations based on sphingomyelin (SM) and cholesterol were compared with liposomes composed of distearoylphosphatidylcholine (DSPC) and cholesterol. SM/cholesterol liposomes had significantly greater in vitro stability than did similar DSPC/cholesterol liposomes. SM/cholesterol liposomes also had significantly improved biological properties compared with DSPC/cholesterol. Specifically, SM/cholesterol liposomes administered intravenously retained 25% of the entrapped vincristine after 72 h in the circulation, compared with 5% retention in DSPC/cholesterol liposomes. The improved retention properties of SM/cholesterol liposomes resulted in plasma vincristine levels 7-fold higher than in DSPC/cholesterol liposomes. The improved circulation lifetime of vincristine in SM/cholesterol liposomes correlated with increased vincristine accumulation in peritoneal ascitic murine P388 tumours and in subcutaneous solid A431 human xenograft tumours. Increased vincristine delivery to tumours was also accompanied by increased anti-tumour efficacy. Treatment with SM/cholesterol liposomal formulations of vincristine resulted in greater than 50% cures in mice bearing ascitic P388 tumours, an activity that could not be achieved with the DSPC/cholesterol formulation. Similarly, treatment of mice with severe combined immunodeficiency (SCID) bearing solid human A431 xenograft tumours with SM/cholesterol vincristine formulations delayed the time required for 100% increase in tumour mass to > 40 days, compared with 5 days, 7 days and 14 days for mice receiving no treatment or treatment with free vincristine or DSPC/cholesterol formulations of vincristine respectively.

The cationic lipid stearylamine reduces the permeability of the cationic drugs verapamil and prochlorperazine to lipid bilayers: implications for drug delivery

The therapeutic activity of a wide variety of drugs is significantly improved when their longevity in the circulation is extended by encapsulation in liposomes. To improve the retention of cationic drugs in liposomes, we have investigated the effect of the cationic lipid stearylamine on the permeability of the calcium channel blocker verapamil and the antipsychotic drug prochlorperazine, both of which are also multidrug resistance modulators. Both drugs were efficiently incorporated into liposomes composed of DSPC/cholesterol that possessed a transmembrane pH gradient (inside acidic). However, the efflux of the loaded drugs was relatively rapid (i.e., 50% of the encapsulated verapamil was released after 4 h at 37 degrees C), despite the presence of a 3 unit pH gradient (pHi = 4.0, pHo = 7.5). Drug retention within the liposomes was improved by increasing the magnitude of the transmembrane pH gradient to approx. 5 units (pHi = 2.0, pHo = 7.5). Further improvements in drug retention were achieved by the addition of 10 mol% of the cationic lipid stearylamine in the DSPC/cholesterol liposomes. The combination of the 5 unit pH gradient and stearylamine resulted in increases of the retention of verapamil and prochlorperazine by approx. 20- and 5-fold, respectively. Calculation of the permeability coefficients for the charged (cationic) and neutral forms of the drugs indicated that the neutral forms of both drugs were approx. 10(4)-fold more permeable than were the cationic forms of the drugs. Further, the presence of stearylamine reduced the permeability coefficient for the cationic species of the drugs by approximately an order of magnitude, but had no effect on the neutral species of the drugs. The efflux curves observed for both verapamil and prochlorperazine could be mathematically modeled by assuming that the primary influence of stearylamine was on the development of a positive surface charge density on the inner monolayer of the liposome. Taken in sum, these results indicate that stearylamine is effective at decreasing the leakage of cationic drugs from liposomes, and may prove to be a valuable component of liposomal drug formulations.

Poly(ethylene glycol)-modified phospholipids prevent aggregation during covalent conjugation of proteins to liposomes

Liposome aggregation is a major problem associated with the covalent attachment of proteins to liposomes. This report describes a procedure for coupling proteins to liposomes that results in little or no change in liposome size. This is achieved by incorporating appropriate levels of poly(ethylene glycol)-modified lipids into the liposomes. The studies employed thiolated avidin-D coupled to liposomes containing the thio-reactive lipid N-(4-(p-maleimidophenyl)butyryl)dipalmitoyl phosphatidylethanolamine (1 mol % of total lipid) and various amounts of MePEG-S-POPE (monomethoxypoly(ethylene glycol) linked to phosphatidylethanolamine via a succinate linkage). The influence of PEG chain length and density was also assessed. The presence of PEG on the surface of liposomes is shown to provide an effective method of inhibiting aggregation and the corresponding increase in liposome size during the covalent coupling of avidin-D. A balance between the size of the PEG used and the amount of PEG-lipid incorporated into the liposome had to be achieved in order to maintain efficient coupling. Optimal coupling efficiencies in combination with minimal aggregation effects were achieved using 2 mol % MePEG2000-S-POPE (PEG of 2000 MW) or 0.8 mol % MePEG5000-S-POPE (PEG of 5000 MW). At these levels, the presence of PEG did not affect the biotin binding activity of the covalently attached avidin. The ability of the resulting liposomes to specifically target to biotinylated cells is demonstrated.

A two-step targeting approach for delivery of doxorubicin-loaded liposomes to tumour cells in vivo

A two-step targeting approach was used to deliver doxorubicin-loaded liposomes to a murine tumour cell (P388 leukaemia) grown in culture and, more importantly, in vivo. Targeting was mediated through the use of an antibody specific for the Thy 1.2 antigen that is highly expressed on P388 cells. Briefly, the approach consists of prelabeling target cells with biotinylated anti-Thy 1.2 antibody prior to administration of drug-loaded liposomes that have streptavidin covalently attached to their surface. Results from in vitro studies demonstrate that a 30-fold increase in cell-associated lipid and a 20-fold increase in cell-associated doxorubicin can be achieved over control liposomes using this two-step procedure. Flow-cytometry and fluorescent-microscopy data were used to confirm that P388 cells can be stably labeled with the biotinylated anti-Thy 1.2 antibody in vivo. Subsequently, liposome-targeting studies were initiated in vivo, where target cell binding was assessed following i.p. or i.v. injection of doxorubicin-loaded liposomes into animals bearing P388 tumours prelabeled with biotinylated antibody. A streptavidin-mediated 3.7-fold increase in cell-associated lipid and drug was achieved when the liposomes were given i.p. When doxorubicin-loaded streptavidin liposomes were injected i.v., P388 cells located in the peritoneal cavity were specifically labeled, although the efficiency of this targeting reaction was low. Less than a 2-fold increase in cell-associated lipid was achieved through the use of target-specific (streptavidin-coated) liposomes. These studies demonstrate that the presence of a well-labeled target cell population within the peritoneal cavity will not promote accumulation of an i.v. injected, targeted liposomal drug. Furthermore, the importance of separating target-cell-specific binding from non-specific uptake by tumour-associated macrophages is discussed.

Pharmacology of liposomal vincristine in mice bearing L1210 ascitic and B16/BL6 solid tumours

Vincristine pharmacokinetic, tumour uptake and therapeutic characteristics were investigated here in order to elucidate the processes underlying the enhanced efficacy observed for vincristine entrapped in small (120 nm) distearoylphosphatidylcholine/cholesterol liposomes. Plasma vincristine levels after intravenous (i.v.) injection are elevated more than 100-fold in the liposomal formulation compared with free drug in tumour-bearing as well as non-tumour-bearing mice over 24 h. Biodistribution studies demonstrate that the extent and duration of tumour exposure to vincristine is dramatically improved when the drug is administered i.v. in liposomal form. Specifically, 72 h trapezoidal area under the curve values for liposomal vincristine in the murine L1210 ascitic and B16/BL6 solid tumours are 12.9- to 4.1-fold larger, respectively, than observed for free drug. Similar to previous results with the L1210 model, increased drug delivery to the B16 tumour results in significant inhibition of tumour growth, whereas no anti-tumour activity is observed with free vincristine. Comparisons of drug and liposomal lipid accumulation in tumour and muscle tissue indicate that the enhanced efficacy of liposomal vincristine is related predominantly to drug delivered by liposomes to the tumour site rather than drug released from liposomes in the circulation. Consequently, improvements in liposomal vincristine formulations must focus on factors that increase uptake of liposomes into tumour sites as well as enhance liposomal drug retention in the circulation.

Polyethylene glycol modified phospholipids stabilize emulsions prepared from triacylglycerol

A stable lipid-based carrier system containing a triacylglycerol core has been developed. This has been achieved by homogenization of corn oil (primarily triacylglycerol) in the presence of phosphatidylcholine (PC), preformed 100 nm liposomes prepared from PC/cholesterol (55: 45; mol:mol) and polyethylene glycol modified phosphatidylethanolamine (PEG2000-PE). The lipid/liposome mixtures were emulsified using a microfluidizer, and the resulting particles could, depending on the phospholipids used and the addition of cholesterol, be designed to exhibit a uniform mean particle size of less than 100 nm (as measured by quasielastic light scattering). The presence of an oil core within the emulsified lipid preparation was confirmed by freeze-fracture and cryoelectron microscopy. Stability of the resulting PEG2000-PE-coated triacylglycerol emulsion was determined by several techniques including (1) time dependent changes in light scatter determined by measuring changes in absorbance at 600 nm (decreases in absorbance are indicative of unstable oil emulsions), (2) column chromatography procedures evaluating the migration of incorporated lipids, specifically PEG2000-PE and triolein, after emulsions were incubated at 37 degrees C in the presence and absence of serum, and (3) in vivo plasma clearance data demonstrating that the lipid mixtures were maintained at ratios specified prior to i.v. administration. This emulsion technology has been used to prepare formulations of several water insoluble compounds, such as the drugs taxol and dibucaine and the lipophilic dye sudan IV.

Liposome encapsulated vincristine: preclinical toxicologic and pharmacologic comparison with free vincristine and empty liposomes in mice, rats and dogs

A preclinical toxicology study of liposome encapsulated vincristine, free vincristine and empty liposomes was carried out in mice and dogs by single and multiple (daily for 5 days) intravenous injection. Single and multiple dose intravenous injection studies in mice showed the encapsulated form of vincristine to be less toxic than free vincristine. Empty liposomes injected intravenously into dogs were without significant toxicity. In dogs, the toxicities seen with liposomal vincristine were qualitatively similar to those of free vincristine with only minor quantitative differences. The principal toxicities of free and liposomal vincristine in dogs were anorexia, weight loss, pyrexia, myelosuppression and gastrointestinal toxicity. After single high doses of either formulation gastrointestinal toxicity was the dose-limiting toxicity, while either hematologic or gastrointestinal toxicity was dose limiting after multiple dose administration of either drug. Histopathologic lesions of importance were bone marrow atrophy, necrosis and atrophy of the lymphoproliferative tissues, necrosis of gastrointestinal tract mucosa, liver and pancreas, and hemorrhage. Distribution studies in rats showed significantly higher vincristine levels in serum, spleen, liver, trachea, jejunum, cerebrum, lung, ischiatic nerve and heart, and significantly lower levels in colon, stomach, salivary gland, thymus esophagus and pancreas after injection of the liposome-associated agent. No toxicities were seen that should preclude safe clinical trial of liposomal vincristine in man.

Liposomal vincristine which exhibits increased drug retention and increased circulation longevity cures mice bearing P388 tumors

Prolonged exposure to vincristine correlates with improved therapeutic activity. In this work, two methods are used to increase the circulation longevity of liposomal formulations of vincristine. The first involves incorporation of the ganglioside GM1, which acts to increase the circulation longevity of liposomal carriers, while the second approach relies on a modification of the vincristine encapsulation procedure which enhances drug retention. It is shown that these approaches are synergistic and increase the circulation half-life of vincristine from approximately 1 h to greater than 12 h. This results in a dramatic improvement in the therapeutic activity of liposomal vincristine as measured using a murine P388 lymphocytic leukemia model. At doses above 2 mg/kg, the optimized liposomal vincristine formulation cures greater than 50% of mice bearing the P388 tumor, whereas free vincristine results in no cures.

Transfer of liposomal drug carriers from the blood to the peritoneal cavity of normal and ascitic tumor-bearing mice

Previously we have demonstrated that the L1210 antitumor activity of liposomal doxorubicin increased significantly as the size of the liposomal carrier was reduced from 1.0 to 0.1 micron. It is demonstrated herein that empty and drug-loaded small (0.1-micron diameter) liposomes accumulate efficiently into the peritoneal cavity of normal and ascitic L1210 tumor-bearing animals following i.v. administration. In normal mice injected with 100 nm DSPC/chol liposomal doxorubicin (drug-to-lipid ratio of 0.2; wt/wt) approximately 2.8 micrograms drug could be recovered from the peritoneal cavity following peritoneal lavage at 24 h. Although this represents only 0.7% of the injected doxorubicin dose, this level of drug is 2 orders of magnitude greater than that achieved following administration of an equivalent dose of free drug (20 mg/kg). The drug levels achieved within the peritoneal cavity are dependent on the physical characteristics (size, drug-to-lipid ratio and lipid composition) of the liposomes employed. Optimal delivery is obtained employing 100 nm DSPC/chol liposomal doxorubicin, a vesicle system that is known to retain entrapped drug following i.v. administration and exhibits extended circulation lifetimes. Analysis of drug and liposome distribution within the peritoneal cavity of normal mice indicates that as much as 50% of the measured doxorubicin and liposomal lipid is cell-associated. Flow cytometric analysis of the peritoneal cells demonstrated that cell-associated doxorubicin resides almost exclusively within resident peritoneal macrophages. The increased delivery of doxorubicin to the peritoneal cavity of normal mice following i.v. administration of small (0.1-micron) liposomal doxorubicin is correlated with a pronounced (> 90%) and prolonged (> 14-day) suppression of resident peritoneal cells. Liposomal drug accumulation increased dramatically in animals with an established L1210 ascitic tumor. More than 5% of the injected dose was found in the peritoneal cavity of these animals 24 h after treatment with DSPC/chol liposomal doxorubicin as compared with a value of 0.03% of the injected dose achieved with free drug. It is proposed that accumulation of liposomes into the peritoneal cavity of normal and tumor-bearing mice may serve as a useful model for characterizing factors mediating the transfer of liposomes from the vascular compartment to extravascular sites.

The presence of GM1 in liposomes with entrapped doxorubicin does not prevent RES blockade

The incorporation of ganglioside GM1 or phosphatidylethanolamine-polyethyleneglycol conjugates into liposomes can result in extended circulation lifetimes in vivo. This has been attributed to an ability to avoid uptake by the reticuloendothelial system (RES), specifically the phagocytic cells of the liver and spleen. Here we examine whether a representative large unilamellar vesicle (LUV) formulation which contains GM1 (distearoylphosphatidylcholine/cholesterol/GM1, 45:45:10 mol/mol), actually does avoid the RES. It is shown that a pre-dose of LUVs which contain GM1 and entrapped doxorubicin blocks the accumulation of subsequently injected empty distearoylphosphatidylcholine/cholesterol liposomes in liver. It is therefore concluded that liposomes exhibiting extended circulation lifetimes can induce RES blockade and do not avoid uptake by liver phagocytes.

Identification of vesicle properties that enhance the antitumour activity of liposomal vincristine against murine L1210 leukemia

The influence of vesicle lipid composition, size and drug-to-lipid ratio on the antitumour activity of liposomal vincristine was assessed in the murine L1210 ascitic leukemia model. A pH gradient-dependent entrapment procedure was used to encapsulate vincristine and allowed such vesicle properties to be independently varied. Free vincristine delivered i.v. at the maximum tolerated dose (2.0 mg/kg) resulted in a 27.8% increase in the life span (ILS) of mice inoculated i.p. with L1210 cells. Encapsulation of the drug in egg phosphatidylcholine/cholesterol vesicles did not significantly increase the antitumour efficacy of vincristine (ILS, 38.9%). In contrast, administration of vincristine entrapped in vesicles composed of distearoylphosphatidylcholine (DSPC)/cholesterol resulted in ILS values as high as 133%. This enhanced antitumour activity of the DSPC/cholesterol formulations was sensitive to the size of the liposomes; increasing the vesicle size from 100 nm to 1 micron decreased the ILS from 133.3% to 55.6% at a drug dose of 2.0 mg/kg. Decreasing the drug-to-lipid ratio from 0.1:1 to 0.05:1 (w/w) had negligible effects on the activity of liposomal vincristine; however, a further decrease in the drug-to-lipid ratio to 0.01:1 (w/w) decreased the antitumour potency at all drug doses studied. Pharmacology studies indicated that the antitumour activities of free and various liposomal forms of vincristine correlated well with the residence time of the drug in the circulation. These studies indicate that efforts to enhance the therapeutic activity of vincristine through liposome encapsulation must address not only the circulation lifetime of the vesicle systems but also the capacity of the liposomes to retain entrapped drug in vivo.

Protein-liposome conjugates with defined size distributions

Conjugation of protein to liposomes by two coupling protocols is shown to result in vesicle aggregation. The degree of aggregation is directly related to the levels of protein conjugated to the liposomes. In an attempt to develop a method of generating stable, homogeneously sized protein-conjugated vesicles, highly aggregated liposome-protein conjugates were extruded through filters of defined pore size distributions, with no loss of protein binding. The extruded samples are relatively stable with respect to size and are easily prepared for various protein to lipid ratios. Liposome size has been shown to be a major factor in determining the in vivo blood circulation times of liposomes. A corresponding, significant enhancement in the blood circulation lifetimes for extruded versus aggregated streptavidin-liposome conjugates is observed. Furthermore, the stability of streptavidin-liposome conjugates in vivo was shown by the binding of biotin to liposomes isolated from plasma 1 and 4 h post-injection. In conclusion, extrusion of the aggregated systems obtained on coupling proteins to liposomes provides a convenient and general method for generating homogeneously sized protein-liposome conjugates.

Comparison of free and liposome encapsulated doxorubicin tumor drug uptake and antitumor efficacy in the SC115 murine mammary tumor

Tumor drug uptake and antitumor efficacy of free and liposomal doxorubicin (DOX) were determined in the SC115 Shionogi mouse mammary tumor. Liposomal DOX systems were prepared by pH gradient-driven drug encapsulation in 170 nm egg phosphatidylcholine/cholesterol (55:45, mol ratio) vesicles. Intravenous injection of free DOX at 6.5 mg/kg, the maximum tolerated dose for free drug in the multiple dose therapy regimen, resulted in tumor-associated drug levels of 2.0 micrograms/g tissue at 1 h which remained constant over 24 h. Liposomal DOX injected at 6.5 mg/kg led to an accumulation of drug in the tumor from 2.6 micrograms/g tissue to 5.5 micrograms/g tissue between 1 h and 24 h, respectively. Increasing the dose of liposomal DOX to 13.0 mg/kg increased tumor drug uptake levels to 5.7 micrograms/g and 10.2 micrograms/g tissue at 1 h and 24 h, respectively. Administration of free or liposome encapsulated DOX every 7 days for 3 weeks resulted in a dose-dependent decrease in tumor growth rate. However, liposomal DOX injected at 6.5 mg/kg exhibited enhanced tumor growth inhibition compared to an equivalent dose of free drug. Further, the ability to administer increased doses of the less toxic liposomal DOX not only resulted in a greater inhibition of tumor growth but also significantly reduced tumor weight. Tumors weighing as much as 5 g were diminished to less than 0.5 g upon treatment with liposomal DOX at a dose of 13 mg/kg. In addition, groups receiving the highest liposomal DOX dose exhibited 25% complete tumor regression which persisted over the 50-day study period. These results demonstrate the ability of appropriately designed liposomal DOX systems to significantly enhance the delivery and retention of drug at solid tumor sites, resulting in increased therapeutic activity.

Optimized procedures for the coupling of proteins to liposomes

A general, optimized method for coupling proteins to liposomes is presented. This procedure utilizes streptavidin covalently coupled to liposomes to allow the subsequent attachment of a variety of biotinated proteins of interest. In the first part of this study, covalent methods for coupling proteins to liposomes which contain the lipid derivatives MPB-PE and PDP-PE were examined. The maleimide lipid derivative MPB-PE was found to allow more efficient coupling. Thin layer chromatography however revealed that during the standard synthesis of MPB-PE, an impurity was generated which can constitute 40% or more of the derivatized PE. An improved method for the synthesis and isolation of pure MPB-PE is presented here. Subsequently, optimized conditions for the covalent coupling of streptavidin to liposomes containing pure MPB-PE were determined. The flexibility of the streptavidin-liposome system for the preparation of various types of ligand bearing liposomes is demonstrated by the rapid association of a variety of biotinated proteins to streptavidin-liposome systems. The ability of these conjugates to target to specific cell populations in vitro as directed by defined biotinated monoclonal antibodies is demonstrated.

The binding of phosphatidylglycerol liposomes to rat platelets is mediated by complement

Previous work has shown that intravenous administration of phosphatidylglycercol (PG) containing liposomes to rats results in a rapid transient decline in platelet count (1). Here the interactions of PG liposomes with rat platelets in vitro have been examined with the aim of characterizing factors associated with the decline. It is shown that PG liposomes induce formation of rat (but not human) platelet-liposome microaggregates in vitro. The PG liposome dependent thrombocytopenia observed in vivo can therefore be attributed to sequestration of PG liposome-platelet aggregates. Further, the aggregation of platelets with PG liposomes, which can be monitored as a reduction in platelet count using a coulter counter, is shown to be mediated by a serum complement factor, likely C3b. This is indicated by a requirement of plasma for the in vitro reduction in platelet count induced by PG liposomes, and the inhibition of this effect by heat treatment of plasma, by incubation of plasma with purified cobra venom factor, or by removal of C3 from plasma.

Method for rapid separation of liposome-associated doxorubicin from free doxorubicin in plasma

To understand and predict the efficacy and/or toxicity of liposomal drugs in vivo, it is essential to have rapid, reliable methods of separating and quantitating both the free and the liposomal forms of the drug. A method using solid-phase extraction chromatography columns was developed to separate and quantitate unencapsulated doxorubicin and liposome-associated doxorubicin in plasma following the intravenous injection of liposomal doxorubicin. The method facilitated the recovery and quantitation of free and liposomal drug. The separation and recovery of doxorubicin were linear across the entire range of possible mixtures (0 to 100%) of the two forms of the drug in plasma. Free drug and liposomal drug were readily separated for liposomal doxorubicin systems varying in size (0.1-1.0 microns) and lipid composition (egg yolk phosphatidylcholine/cholesterol and distearylphosphatidylcholine/cholesterol). The method is rapid and allows for multiple samples to be processed simultaneously.

Characterization of liposomal systems containing doxorubicin entrapped in response to pH gradients

Studies from this laboratory (Mayer et al. (1986) Biochim. Biophys. Acta 857, 123-126) have shown that doxorubicin can be accumulated into liposomal systems in response to transmembrane pH gradients (inside acidic). Here, detailed characterizations of the drug uptake and retention properties of these systems are performed. It is shown that for egg phosphatidylcholine (EPC) vesicles (mean diameter of 170 nm) exhibiting transmembrane pH gradients (inside acidic) doxorubicin can be sequestered into the interior aqueous compartment to achieve drug trapping efficiencies in excess of 98% and drug-to-lipid ratios of 0.36:1 (mol/mol). Drug-to-lipid ratios as high as 1.7:1 (mol/mol) can be obtained under appropriate conditions. Lower drug-to-lipid ratios are required to achieve trapping efficiencies in excess of 98% for smaller (less than or equal to 100 nm) systems. Doxorubicin trapping efficiencies and uptake capacities are related ito maintenance of the transmembrane pH gradient during encapsulation as well as the interaction between doxorubicin and entrapped citrate. This citrate-doxorubicin interaction increases drug uptake levels above those predicted by the Henderson-Hasselbach relationship. Increased drug-to-lipid ratios and trapping efficiencies are observed for higher interior buffering capacities. Retention of a large transmembrane pH gradient (greater than 2 units) after entrapment reduces the rate of drug leakage from the liposomes. For example, EPC/cholesterol (55:45, mol/mol) liposomal doxorubicin systems can be achieved which released less than 5% of encapsulated doxorubicin (drug-to-lipid molar ratio = 0.33:1) over 24 h at 37 degrees C. This pH gradient-dependent encapsulation technique is extremely versatile, and well characterized liposomal doxorubicin preparations can be generated to exhibit a wide range of properties such as vesicle size, lipid composition, drug-to-lipid ratio and drug release kinetics. This entrapment procedure therefore appears well suited for use in therapeutic applications. Finally, a rapid colorimetric test for determining the amount of unencapsulated doxorubicin in liposomal systems is described.

Liposomes with entrapped doxorubicin exhibit extended blood residence times

The blood residence time of liposomes with entrapped doxorubicin is shown to be significantly longer than for identically prepared empty liposomes. Liposomal doxorubicin systems with a drug-to-lipid ratio of 0.2 (w/w) were administered at a dose of 100 mg lipid/kg. Both doxorubicin and liposomal lipid were quantified in order to assess in vivo stability and blood residence times. For empty vesicles composed of phosphatidylcholine (PC)/cholesterol (55:45, mole ratio) and sized through filters of 100 nm pore size, 15-25% of the administered lipid dose was recovered in the blood 24 h after i.v. injection. The percentage of the dose retained in the circulation at 24 h increased 2-3-fold when the liposomes contain entrapped doxorubicin. For 100 nm distearoyl PC/chol liposomal doxorubicin systems, as much as 80% of the injected dose of lipid and drug remain within the blood compartment 24 h after i.v. administration.

The accumulation of drugs within large unilamellar vesicles exhibiting a proton gradient: a survey

We have shown previously that transmembrane proton gradients can be used to efficiently accumulate biogenic amines [M.B. Bally et al. (1988) Chem. Phys. Lipids 47, 97-107] and doxorubicin [L.D. Mayer, M.B. Bally and P.R. Cullis (1986) Biochim. Biophys. Acta 857, 123-126] to high concentrations within liposomes. To determine the generality of this loading procedure, representative drugs from a variety of different classes (antineoplastics, local anaesthetics, antihistamines, etc.) were examined as to their ability to redistribute in response to a proton gradient. While the majority of drugs examined, all of which are weak bases, were accumulated by large unilamellar vesicles exhibiting a pH gradient (interior acid) the extent of uptake varied considerably between different pharmaceuticals. These differences are discussed in the context of various factors which will likely influence drug accumulation including its membrane/water partition coefficient and its solubility in the intravesicular medium.

Liposomal vincristine preparations which exhibit decreased drug toxicity and increased activity against murine L1210 and P388 tumors

The toxicity and antitumor activity of liposomal vincristine preparations have been examined. Vincristine was encapsulated inside egg phosphatidylcholine (EPC)/cholesterol (55/45, mol/mol) and distearoylphosphatidylcholine (DSPC)/cholesterol (55/45, mol/mol) vesicles utilizing transmembrane pH gradient (inside acidic) drug uptake processes. Trapping efficiencies approaching 100% were achieved for this procedure using drug:lipid ratios as high as 0.2:1 (w/w). Although both EPC/cholesterol and DSPC/cholesterol liposomal systems yielded high trapping efficiencies, DSPC/cholesterol vesicles exhibited superior drug retention properties. This ability to retain entrapped vincristine was related to maintenance of the transmembrane pH gradient as well as the membrane permeability properties. Thirty-day dose-response survival studies in mice indicated that vincristine encapsulated in DSPC/cholesterol liposomes was less toxic than free drug. The 50% lethal dose of 1.9 mg/kg in CD-1 mice observed for free vincristine increased to 4.8 mg/kg upon administration of the drug in liposomal form. Liposome encapsulation of vincristine also enhanced the antitumor activity against murine P388 and L1210 lymphocytic leukemia models. This resulted from increased efficacy for liposomal vincristine at doses equal to free drug (liposomal/free drug median survival times greater than 1.0) as well as the ability to administer increased doses of liposomal vincristine. The combined effects of decreased toxicity and increased antitumor efficacy of liposomal vincristine over free drug suggest significant clinical utility of appropriate liposomal vincristine systems.

Studies on the myelosuppressive activity of doxorubicin entrapped in liposomes

The myelosuppressive activity of doxorubicin encapsulated in liposomes of differing lipid composition and size was quantified in mice by measurement of changes in spleen weight, peripheral white blood cells (WBC), and bone marrow nucleated cells. Following i.v. administration of free doxorubicin at a dose of 20 mg/kg, a 90% reduction in marrow cellularity was observed on day 3. The marrow nucleated cell count was similar to control values by day 7. Administration of an equivalent dose of doxorubicin that was encapsulated in large (diameter, approximately 1.0 microns) egg phosphatidylcholine/cholesterol (EPC/Chol)(molar ratio, 55:45) liposomes induced an 80% reduction in bone marrow cellularity that lasted for periods of greater than 7 days. Similar results were obtained following administration of large (1.0 microns) liposomal doxorubicin systems formulated with distearoylphosphatidylcholine/cholesterol (DSPC/Chol) (molar ratio 55:45). In contrast, liposomal doxorubicin prepared using small (diameter, approximately 0.1 micron) DSPC/Chol liposomes induced only a 40% reduction (day 3) in bone marrow cellularity, which returned to control values by day 7. Other indicators of doxorubicin-mediated myelosuppressive activity (spleen weight loss and peripheral leukopenia) correlated well with changes observed in marrow cellularity. An exception to this, however, was observed in animals treated with small (0.1 -micron) DSPC/Chol Liposomal doxorubicin, which displayed peripheral leukopenia for periods of greater than 14 days. This extended leukopenia was not observed following administration of small (0.1 -micron) EPC/Chol liposomal doxorubicin. Marrow-associated liposomal lipid and doxorubicin were quantified to determine if the extent of doxorubicin-mediated myeloid toxicity could be correlated to changes in biodistribution of the entrapped drug. It was demonstrated that 10-20 times more doxorubicin is delivered to the bone marrow when the drug is given encapsulated in large liposomes than when it is associated with small liposomes. These data are useful in defining characteristics of liposomal preparations that modulate the myelosuppressive behaviour of entrapped antineoplastic agents.

Influence of vesicle size, lipid composition, and drug-to-lipid ratio on the biological activity of liposomal doxorubicin in mice

The effects of vesicle size, lipid composition, and drug-to-lipid ratio on the biological activity of liposomal doxorubicin in mice have been investigated using a versatile procedure for encapsulating doxorubicin inside liposomes. In this procedure, vesicles exhibiting transmembrane pH gradients (acidic inside) were employed to achieve drug trapping efficiencies in excess of 98%. Drug-to-lipid ratios as high as 0.3:1 (wt:wt) could be obtained in a manner that is relatively independent of lipid composition and vesicle size. Egg phosphatidylcholine (EPC)/cholesterol (55:45; mol/mol) vesicles sized through filters with a 200-nm pore size and loaded employing transmembrane pH gradients to achieve a doxorubicin-to-lipid ratio of 0.3:1 (wt/wt) increased the LD50 of free drug by approximately twofold. Removing cholesterol or decreasing the drug-to-lipid ratio in EPC/cholesterol preparations led to significant decreases in the LD50 of liposomal doxorubicin whereas, the LD50 increased 4- to 6-fold when distearoylphosphatidylcholine was substituted for EPC. The results suggest that the stability of liposomally entrapped doxorubicin in the circulation is an important factor in the toxicity of this drug in liposomal form. In contrast, the antitumor activity of liposomal doxorubicin is not influenced dramatically by alterations in lipid composition. Liposomal doxorubicin preparations of EPC, EPC/cholesterol (55:45; mol:mol), EPC/egg phosphatidylglycerol (EPG)/cholesterol (27.5:27.5:45; mol:mol), and distearoylphosphatidylcholine/cholesterol (55:45; mol:mol) all demonstrated similar efficacy to that of free drug when given at doses of 20 mg/kg and below. Higher dose levels of the less toxic formulations could be administered, leading to enhanced increases in life span (ILS) values. Variations in vesicle size, however, strongly influenced the antitumor activity of liposomal doxorubicin. At a dose of 20 mg/kg, large EPC/cholesterol systems are significantly less effective than free drug (with ILS values of 65% and 145%, respectively). In contrast, small systems sized through filters with a 100-nm pore size are more effective than free drug, resulting in an ILS of 375% and a 30% long term (greater than 60 days) survival rate when administered at a dose of 20 mg/kg. Similar size-dependent effects are observed for distearoylphosphatidylcholine/cholesterol systems.

Proton flux in large unilamellar vesicles in response to membrane potentials and pH gradients

The transport of protons across liposomes composed of phosphatidylcholine in response to electrical potentials or pH gradients has been investigated. The results support three major conclusions. The first of these concerns the need for reliable measurements of electrical potentials and pH gradients. It is shown that the potential probe tetraphenylphosphonium and the pH probe methylamine provide accurate and self consistent measures of electrical potentials and pH gradients respectively in these systems. Second, it is shown by two independent techniques that the pH gradients induced in response to valinomycin and potassium dependent electrical potentials are significantly smaller than would be expected for electrochemical equilibrium. The pH gradients observed are stable over an 8 h time course and are sensitive to the ionic composition of the buffers employed, where the presence of external sodium results in the smallest induced pH gradients. These results are discussed in terms of current models of proton conductance across membranes. In a final area of investigation, it is shown that valinomycin and carbonyl cyanide m-chlorophenyl hydrazone (CCCP) can transport sodium ions in a synergistic manner.

Platelet distribution in rabbits following infusion of liposomes

This investigation determined the organ distribution of liposomes containing egg phosphatidylcholine and cholesterol with egg phosphatidylglycerol (PG liposomes) or without (PC liposomes) and the effect of each liposome on platelet distribution in rabbits. Eight minutes after 51chromium-labelled platelets were given intravenously, either saline (n = 7), iodinated PG liposomes (n = 5) or iodinated PC liposomes (n = 5) were infused. Two minutes later the organ distribution of 51Cr-platelets and 125I-liposomes were compared. The PG liposomes produced a 41 +/- 5% reduction in circulating platelet counts while PC liposomes did not. The PG liposomes decreased circulating 51Cr-platelets by a factor of 2 and increased platelet recoveries in the liver and lungs. The increased platelet recovery in the liver was associated with a greater PG liposome recovery. When animals receiving PG liposomes were studied over 60 minutes, both the labelled and unlabelled platelet counts returned to control values by 30 minutes and the 51Cr-platelet distribution between organs was similar to control values. These data indicate that platelets and PG liposomes initially sequester together and that this platelet-liposome interaction is specific for PG liposomes. However, the platelet sequestration is transient and by 60 minutes the platelets were released and circulating.

Analysis of the effect of liposome encapsulation on the vesicant properties, acute and cardiac toxicities, and antitumor efficacy of doxorubicin

Numerous studies have demonstrated that liposomal encapsulation decreases the life-threatening chronic and acute toxicities of doxorubicin in the face of unaltered or improved antitumor activity. Minimal attention has been paid to the encapsulation effect on the lesser toxicities of the drug, specifically the vesicant properties. In this report we assess the effect of the encapsulation of doxorubicin in an egg-yolk phosphatidylcholine (EPC) cholesterol liposome on the drug's topical toxicity. In addition, to ensure acceptable activity and reduction in toxicity comparable with those of previously assessed formulations, the cardiac and acute toxicities and antitumor activity of the liposomal doxorubicin complex were also investigated. Antitumor efficacy was assessed using the metastatic murine P815 mastocytoma model. Equivalent doses of free and encapsulated doxorubicin possessed the same antitumor activity in the prolongation of animal survival in 14-day survival studies conducted to assess the effect of liposomal encapsulation on the acute toxicity of this drug. The LD50 of liposomal doxorubicin was found to be 40 mg/kg, 53% higher than that of free doxorubicin (26 mg/kg). Histologic examination of cardiac sections taken from DBA/2J mice 7 days after a single i.v. injection of free or liposomal doxorubicin (25 mg/kg) revealed that the liposomal preparation was much less cardiotoxic. In animals receiving the free drug, edema, monocytic infiltration, and cell necrosis were evident. In contrast, those receiving the liposomal preparation demonstrated slight cellular edema but showed no evidence of cellular necrosis. To assess vesicant properties, DBA/2J mice were given a single s.c. injection (0.2 ml) of free or liposomal doxorubicin (2 mg/ml). Those receiving the free drug immediately developed erythema and edema at the injection site, which progressed to ulceration. Those receiving the liposomal complex developed slight erythema and edema but did not ulcerate at any time. All signs of irritation in this group had subsided 3 weeks postinjection. In summary, the liposomal complex used eliminated the vesicant properties of doxorubicin as well as significantly decreasing its cardiac and acute toxicities in the face of unaltered antitumor activity.

Interactions of liposomes and platelets

Rats were injected intravenously with liposomes of various compositions and sizes and blood platelet count measured. It was found that negatively-charged liposomal systems produced a transient reduction in platelet count in the first 5 minutes after injection which recovered by 60 minutes post-injection. This effect was most striking for multilamellar vesicles (MLV's) containing phosphatidylglycerol (PG). Dose levels of 25 mg/kg of MLV's containing 10 mole% PG caused the platelet count to drop from a control value of 1,086 +/- 21 X 10(9)/l to 193 +/- 14 X 10(9)/l by 2 minutes post-injection, an 82% decline. This thrombocytopenic effect was observed to diminish as vesicle size or vesicle dose was decreased. Positively-charged liposomes produced a less pronounced transient reduction in platelet count while neutral liposomes caused only a mild, transient platelet decline. This transient thrombocytopenic effect was not blocked by common anticoagulants and fibrinolytic agents but was prevented by liposomal pretreatment. Radiolabeled platelet studies revealed that transient sequestration of platelets occurs in the liver and spleen 2 minutes after PG:EPC:CHOL MLV injection with a normalization of platelet distribution by 60 minutes post-injection. In vitro studies, using an automated blood counter, suggest a transient association of liposomes and platelets occurring following injection. Liposomally-induced transient thrombocytopenia suggests a role for platelets in the biodistribution of liposomes.

Dopamine accumulation in large unilamellar vesicle systems induced by transmembrane ion gradients

Transmembrane movement of dopamine in response to K+ or H+ ion gradients has been investigated. It is shown that dopamine can accumulate rapidly into large unilamellar vesicles (LUVs) composed of egg phosphatidylcholine exhibiting either a K+ diffusion potential (delta psi; negative inside) or a pH gradient (inside acidic). This can result in entrapped dopamine concentrations of 30-40 mM and inside-outside concentration gradients of nearly 300-fold. The transmembrane dopamine gradients formed in LUV systems exhibiting delta pH (inside acidic) indicate that the transport process can be dictated by movement of the neutral form of dopamine which redistributes according to a simple Henderson-Hasselbach equilibrium. The mechanism of dopamine transport in response to a valinomycin-induced K+ potential is more complex. Although generation of a K+ diffusion potential results in acidification of the vesicle interior, the magnitude of the induced delta pH (approx. 1 pH unit) is insufficient to account for the dopamine concentration gradient achieved (greater than 200-fold). Further, data presented here suggest that higher uptake levels of dopamine can be achieved when certain anions (ATP and citrate) are entrapped within the LUV system. These anions may complex with the protonated form of dopamine creating a non-equilibrium trapping phenomena resulting in interior concentrations of dopamine in excess of that predicted by a simple Henderson-Hasselbach equilibrium.