Immunomicelles: Targeted pharmaceutical carriers for poorly soluble drugs

Abstract

To prepare immunomicelles, new targeted carriers for poorly soluble pharmaceuticals, a procedure has been developed to chemically attach mAbs to reactive groups incorporated into the corona of polymeric micelles made of polyethylene glycol–phosphatidylethanolamine conjugates. Micelle-attached antibodies retained their ability to specifically interact with their antigens. Immunomicelles with attached antitumor mAb 2C5 effectively recognized and bound various cancer cells in vitro and showed an increased accumulation in experimental tumors in mice when compared with nontargeted micelles. Intravenous administration of tumor-specific 2C5 immunomicelles loaded with a sparingly soluble anticancer agent, taxol, into experimental mice bearing Lewis lung carcinoma resulted in an increased accumulation of taxol in the tumor compared with free taxol or taxol in nontargeted micelles and in enhanced tumor growth inhibition. This family of pharmaceutical carriers can be used for the solubilization and enhanced delivery of poorly soluble drugs to various pathological sites in the body.

The requirements of pharmaceutical drug carriers for i.v. administration include small size, biodegradability, good loading capacity, high content of the drug in a final preparation, prolonged circulation, and ability to accumulate in required areas. These requirements are reasonably well met by some drug carriers (microcapsules, liposomes) used predominantly for water-soluble drugs (1–4). Although liposomes can entrap poorly soluble drugs in the hydrophobic bilayer, their loading capacity is limited because of possible membrane destabilization. Thus, the development of drug carriers displaying all of the named properties specifically for the delivery of poorly soluble pharmaceuticals continues to represent a challenge.

Low solubility in water tends to be an intrinsic property of many drugs, including anticancer agents, which often represent polycyclic compounds (5). The membrane permeability and efficacy of such drugs increases with increasing hydrophobicity (6, 7). On the other hand, parenteral administration of those intrinsically hydrophobic agents is associated with some problems. Thus, i.v. administration of aggregates formed by undissolved drug in aqueous media can cause embolization of blood capillaries (≤5 μm) before drug penetrates a tumor (8). Additionally, low solubility of hydrophobic drugs in combination with excretion and metabolic degradation hinders the maintenance of therapeutically significant systemic concentrations.

Micelles are spherical colloidal nanoparticles into which many amphiphilic molecules self-assemble. In water, hydrophobic fragments of amphiphilic molecules form the core of a micelle, which may then be used as a cargo space for poorly soluble pharmaceuticals (9, 10). Hydrophilic parts of the molecules form the micelle corona. Micelle encapsulation increases bioavailability of poorly soluble drugs, protects them from destruction in biological surroundings, and beneficially modifies their pharmacokinetics and biodistribution (11). Because of their small size (usually 5–50 nm), micelles demonstrate spontaneous accumulation in pathological areas with leaky vasculature, such as infarct zones (12) and tumors (13, 14). This phenomenon is known as the enhanced permeability and retention effect (15, 16).

Polymeric micelles are formed by amphiphilic copolymers and demonstrate a number of attractive properties (17–19). Because microparticulate drug carriers are removed from the blood via opsonization-mediated phagocytosis by cells and organs of the reticuloendothelial system (20), the micelle corona formed by hydrophilic polymer blocks provides longevity to micelles in vivo by preventing their opsonization and capture (21). Amphiphilic polymers have a low critical micelle concentration, which makes polymeric micelles stable and prevents their rapid dissociation in vivo. The use of lipid moieties as hydrophobic blocks provides an additional stability, because the existence of two hydrocarbon chains contributes considerably to the increased hydrophobic interactions in the micelle's core. Micelles prepared from conjugates of polyethylene glycol (PEG) and diacyllipids, such as phosphatidylethanolamine (PE) are of particular interest (22).

We have reported some data on polymeric micelles formed by PEG–PE conjugates (23). Such micelles are stable and long-circulating and can be loaded with a variety of poorly soluble drugs including anticancer drugs, such as taxol and tamoxifen (24, 25). Because of their small size, these preparations are capable of delivering their load even into poorly permeable tumors in mice with a higher efficiency than other long-circulating micropraticulate carriers (26, 27).

As with other carriers, the drug delivery potential of polymeric micelles may be enhanced by attaching targeting ligands, including antibodies, to the micelle surface (19). Such a development would represent the next generation of drug carriers for sparingly soluble pharmaceuticals. Certain attempts to prepare micelles with attached ligands (including antibodies) are known (28–30). However, no systematic study on antibody-targeted polymeric micelles (immunomicelles) has been undertaken thus far. This may be explained by the absence of suitable and broadly applicable protocols allowing for the covalent attachment of various antibodies to polymeric micelles.

By adapting the coupling technique we recently developed (31, 32), we have prepared PEG–PE-based immunomicelles modified with mAb 2C5 with nucleosome-restricted specificity reactive toward a variety of different cancer cells (33) and antimyosin mAb 2G4 specific for infarcted areas of the myocardium (34). Here, we present our results on the preparation and some properties of these carriers for sparingly soluble pharmaceuticals including the ability of 2C5 immunomicelles loaded with a poorly soluble anticancer drug, taxol, to deliver increased quantities of the drug into experimental tumors and provide improved tumor growth inhibition.

Materials and Methods

Materials.

PE, PEG-2000–PE, and PE-(lissamine-rhodamine B) were from Avanti Polar Lipids. p-Nitrophenylcarbonyl (pNP)–PEG–PE was synthesized as described (31). Diethylenetriaminepentaacetic acid–PE conjugate for radiolabeling micelles with ¹¹¹In was synthesized as in ref. 35. RPMI medium 1640, Eagle's MEM, DMEM, serum-free medium, and heat-inactivated FBS were from Cellgro (Herndon, VA). ¹¹¹In with specific radioactivity of 395 Ci/mg was from Perkin–Elmer. Cancer-specific antinucleosome 2C5 is routinely produced in our laboratory. Taxol was purchased from Sigma and dissolved in Cremophor EL (BASF Bioresearch, Mount Olive, NJ) mixed with ethanol (1:1 by volume) and then further in saline as described (36). Antimyosin 2G4 was provided by B.-A. Khaw (Northeastern University).

Cell Cultures.

Murine Lewis lung carcinoma (LLC) and EL4 T cell lymphoma and human BT20 breast adenocarcinoma cell lines were purchased from the American Type Culture Collection. LLC and BT20 cells were maintained in DMEM with 10% FBS, penicillin/streptomycin, pyruvate, l-glutamine, and nonessential amino acids. EL4 cells were grown in RPMI medium 1640 with the same additives as above. Cells were grown at 37°C in 5% CO₂.

Labeling of Antibodies with Carboxyfluorescein (CF).

A 50-μl aliquot of 2 mM 5-CF succinimidyl ester (Molecular Probes) in DMSO was added to 2 ml of a 12 μM solution of 2C5 in 50 mM Tris-buffered saline (TBS), pH 9.0. The mixture was incubated overnight at 4°C. Free CF was removed by dialysis against TBS.

Preparation of Immunomicelles.

A lipid film was prepared by removing chloroform from the mixed solution of PEG₂₀₀₀–PE and 2–8 mol percent of pNP–PEG–PE under vacuum. To load micelles, taxol dissolved in methanol was added to a chloroform solution of PEG–PE and pNP–PEG–PE (1.5 mg taxol per 80 mg PEG-PE). When required, trace amounts of diethylenetriaminepentaacetic acid–PE and/or 0.5 mol percent of PE–(lissamine-rhodamine B) were added to these preparations. To form micelles, the film was rehydrated at 50°C in a 5 mM Na citrate-buffered saline, pH 5.0, and vortexed for 5 min.

When required, 0.5 ml of a 12 μM solution of 2C5 or 2G4 in borate, pH 9.0 was added to 0.5 ml of pNP–PEG–PE–containing micelles with a PEG–PE concentration of 1.5 mM. The mixture was incubated for 3 h at room temperature and dialyzed against 10 mM Hepes-buffered saline, pH 7.4 (HBS), using cellulose ester membranes with a cut-off size of 300,000 Da (Spectrum Medical Industries). To assess the quantity of the micelle-bound protein, carboxyfluorescein-labeled antibody was used. Protein concentration was measured by associated fluorescence at an excitation wavelength of 490 nm and an emission of 520 nm on a F-2000 spectrofluorimeter (Hitachi, Tokyo). The micelle size was measured by dynamic light scattering with a N4 Plus Submicron Particle System (Coulter) at a PEG–PE concentration of 2–10 mM.

Freeze–Fracture Electron Microscopy.

The sample was quenched by using the sandwich technique and liquid nitrogen-cooled propane. A cooling rate of 10,000 K/s avoids ice crystal formation and cryofixation-caused artifacts. The fracturing process was carried out in JEOL JED-9000 freeze-etching equipment, and the exposed fracture planes were shadowed with Pt for 30 s at an angle of 25–35° followed with carbon for 35 s (2 kV, 60–70 mA, 1 × 10⁻⁵ Torr). The replicas were cleaned with fuming HNO₃ for 24–36 h followed by repeated agitation with chloroform/methanol (1:1 vol/vol) at least five times and examined with a JEOL 100 CX electron microscope.

Radiolabeling of Micelles.

Diethylenetriaminepentaacetic acid–PE-containing micelles in HBS were mixed with 50–100 μCi of ¹¹¹In in 0.1 M Na-citrate, pH 3.7. The mixture was incubated for 1 h at room temperature and dialyzed against at least 3,000-fold excess of HBS overnight at 4°C to remove unbound ¹¹¹In.

Binding of Immunomicelles to Substrate Monolayers.

ELISA plates were coated with 50 μl of 10 μg/ml nucleosomes (Worthington) for testing 2C5 immunomicelles or with 50 μl of 10 μg/ml cardiac myosin for testing 2G4 immunomicelles and incubated overnight at 4°C. The rinsed plates were coated with 1% FBS in HBS. To substrate-coated plates, 50 μl of 2C5 or 2G4 immunomicelles at 20 μg/ml PEG–PE was added and incubated for 4 h at room temperature. The plates were washed with HBS and coated with horseradish peroxidase-antimouse IgG conjugate (ICN) following the manufacturer's recommendations. The conjugate was removed after 3 h at room temperature, and the plates were washed with HBS. Bound peroxidase was quantified by degradation of diaminobenzidine (Neogen, Lexington, KY) supplied as a ready-for-use solution. Color intensity was analyzed by a Multiscan 340 ELISA reader (Labsystems, Helsinki).

Interaction of 2C5 Immunomicelles with Cancer Cells in Vitro.

After initial passage in tissue culture flasks, LLC and BT20 cells were grown on cover slips placed in 6-well tissue culture plates. After the cells reached a confluence of 60–70%, the plates were washed with Hank's buffer and treated with 1% BSA in Eagle's MEM (2 ml per well) and incubated for 1 h at 37°C, 5% CO₂. To these cells, PE-(lissamine-rhodamine B)-labeled 2C5 immunomicelles were added to a final concentration of PEG–PE of 0.15 mg/ml and incubated for 1 h at 37°C, 5% CO₂. After incubation, the cells were washed with Hanks' buffer, and the cover slips were mounted cell-side down on glass slides by using fluorescence-free glycerol-based mounting medium (Fluoromount-G; Southern Biotechnology Associates).

EL4 cells were grown in suspension to the density of ≈2 × 10⁴ cells per ml, centrifuged at 700 × g for 10 min, and transferred to Hanks' buffer. The cells were washed and resuspended in Hanks' buffer at ≈1 × 10⁵ cells per ml density. PE-(lissamine-rhodamine B)-labeled immunomicelles were added to the EL4 cell suspension to a final PEG–PE concentration of 0.15 mg/ml. The cells were incubated with immunomicelles for 1 h at 37°C, 5% CO₂, washed with Hanks' buffer, concentrated to a cell density of 1 × 10⁶ cells per ml, transferred to glass slides, and mounted as described. Mounted cells were studied with a Nikon Eclipse E400 microscope under bright light or under epifluorescence with a rhodamine filter.

2C5 Immunomicelles in Vivo.

All experiments were performed in 6- to 8-week-old female C57BL/6J mice (Charles River Breeding Laboratories) following protocol 011022 approved by the Institutional Animal Care and Use Committee in accordance with the National Institutes of Health's Principles of Laboratory Animal Care (publication 85–23, revised in 1985). The animals were allowed free access to food and water.

For blood clearance experiments, the mice were injected with 100 μl of 0.5 mM ¹¹¹In-labeled micelle formulations via the tail vein. At the required time points between 0.5 and 17 h postinjection, mice were anesthesized with ether and killed by cervical dislocation. Blood was collected and analyzed for the presence of the micelle-associated ¹¹¹In radioactivity with a γ-counter (GAMMA 5500, Beckman Instruments, Fullerton, CA).

For tumor accumulation experiments, LLC tumors were initiated in mice by s.c. injection of 20,000 LLC cells in 50 μl of 10 mM HBS into the left rear flank. When tumor diameters reached 3–7 mm (8–12 days postinoculation), the mice were injected with 100 μl of 0.5 mM ¹¹¹In-labeled micellar formulations via the tail vein. At 30 min and 2 h postinjection, mice were killed, and tumors and muscle samples were collected and analyzed for the presence of ¹¹¹In radioactivity. There were five animals per group for each time point.

To estimate the accumulation of free and micellar taxol in tumors, LLC-bearing mice were injected with the same quantity of taxol in Cremophor EL/ethanol/saline mixture (see above) or in plain PEG–PE micelles or 2C5 immunomicelles (≈100 μg of taxol per animal). At 30 min and 2 h postinjection, mice were killed, and tumors were removed, washed with saline, and homogenized in the presence of 2.5 times the tumor weight of saline by using a PowerGen 125 tissue homogenizer (Fisher). Taxol was extracted and quantified by HPLC as in ref. 37. tret- Butyl methyl ether used to extract taxol from homogenates contained 30 μg/ml of N-octylbenzamidine used as an internal HPLC standard and synthesized as in ref. 38. The HPLC was run on a reverse-phase LiChrospher RP18-5 column (Merck) at flow rate of 0.9 ml/min with the detection of taxol and the internal standard by optical density at 227 nm.

Inhibition of Tumor Growth with Different Taxol Preparations.

Mice with LLC tumors were injected with different taxol formulations on day 10 postinoculation (≈100 μg of taxol per animal per injection). On day 5 after the first administration, the injections were repeated. Twenty days after the first injection, the mice were killed, and tumors were extracted, rinsed twice with saline, wiped, and weighed. There were five mice in each experimental group.

All experimental results are shown as mean values ± standard deviations.

Results and Discussion

Preparation of Activated Micelles and Antibody Attachment.

Immunomicelles were prepared by using a procedure developed by us for the attachment of ligands to liposomes (31, 32). The procedure uses PEG–PE with the free PEG terminus activated with pNP. Micelles were prepared from PEG–PE with the addition of a small fraction of pNP–PEG–PE. The PE residues form the micelle core, whereas pNP groups allow for fast and efficient attachment of aminogroup-containing ligands via the formation of the urethane (carbamate) bond. Micelle structure and ligand attachment are illustrated in Fig. 1.

Figure 1.

(A) Schematic structure of PEG–PE micelles containing a small addition of the pNP–PEG–PE component. (B) Coupling of amino group-containing ligands (antibodies) with pNP groups.

A typical result of micelle-bound antibody quantification is shown in Fig. 2. About 30% of added 2C5 attached to micelles containing 2 mol percent of pNP–PEG–PE. This corresponds to a ≈60% reaction yield because a 2-fold molar excess of protein over pNP–PEG–PE was used to avoid antibody inactivation caused by modification of multiple NH₂ groups. This yield is close to the results obtained for liposomes in a similar procedure (31). From the yield value, it can be calculated that up to 10 antibody molecules bind to a single micelle. Protein binding to micelles without pNP–PEG–PE was negligible.

Figure 2.

(Upper) Attachment of 2C5 antibodies to PEG–PE micelles via micelle-incorporated pNP–PEG–PE. (Lower) 2C5 attachment yield as a function of pNP–PEG–PE content in the micelle.

Antibody attachment yield may be increased by increasing the molar fraction of pNP–PEG–PE in micelles (Fig. 2). The yield reached as high as 50% when the micelles contained 8 mol percent of pNP–PEG–PE. Excessive amount of pNP–PEG–PE, however, may cause overmodification of a protein molecule and its inactivation. Therefore, micelles with 2 mol percent of pNP–PEG–PE were used to prepare immunomicelles for all further experiments.

Properties of Immunomicelles.

One of the advantages of micelles as drug carriers is their small size. The transport and accumulation efficiency of microparticulates into the tumor interstitium is limited by their ability to penetrate tumor vascular endothelium (39, 40). Diffusion and accumulation of microparticles are determined by the cutoff size of the tumor vasculature, and this cutoff size varies for different tumors (41). If this size is <200 nm, drug carriers with comparable or larger sizes (liposomes) will not extravasate (42). Thus, it is important that the modification of micelles with an antibody does not increase their size.

The size distribution of micelles before and after attachment of 2C5 or 2G4 is shown in Fig. 3. Protein attachment does not affect the micelle size significantly. Dynamic light scattering data were confirmed by the results of freeze–fracture electron microscopy (Fig. 3). Both the original and 2C5-modified micelles have a spherical shape and a uniform size of ≈20 nm. Thus, protein-modified micelles should retain the ability to cross the vasculature even of tumors with a relatively small cutoff sizes that are not accessible with other particulate delivery systems (25, 26, 41).

Figure 3.

(Upper) Size distribution of “plain” PEG–PE/pNP–PEG–PE micelles (Left) and the same micelles after the attachment of 2C5 (Center) or 2G4 (Right). (Lower) Freeze–fracture electron images of plain PEG–PE/pNP–PEG–PE micelles (A) and 2C5 immunomicelles of the same composition (B).

The binding of 2C5 immunomicelles to the 2C5-specific substrate, nucleosomes, and the binding of 2G4 immunomicelles to the 2G4-specific substrate, cardiac myosin, is shown in Fig. 4. Both 2C5 and 2G4 immunomicelles bound effectively to monolayers of corresponding antigens. The binding of control micelles (containing no pNP–PEG–PE but prepared following the same protocol and incubated with antibody) was much lower (only background binding can be seen because of some noncovalently adsorbed antibodies). The control with pNP–PEG–PE-containing antibody-free micelles is not required here, because pNP groups spontaneously hydrolyze under experimental conditions, yielding plain PEG–PE micelles. These results show that 2C5 and 2G4 antibodies preserved their specific activity upon attachment to the micelle surface.

Figure 4.

(Upper) Binding of 2C5 immunomicelles to a monolayer of nucleosomes. (Lower) Binding of 2G4 immunomicelles to a monolayer of myosin.

Interaction of Immunomicelles with Cancer Cells in Vitro.

To enhance tumor accumulation, drugs and drug carriers were modified with tumor-specific mAbs (43–45). These antibodies are usually tumor type-specific and unable to react with different tumors. Recently, we have shown that certain nonpathogenic monoclonal antinuclear autoantibodies, with 2C5 among them, recognize the surface of numerous tumor, but not normal, cells via tumor cell surface-bound nucleosomes (33, 46, 47). Because these antibodies bind a broad variety of cancer cells, they may serve as specific ligands for the delivery of drugs and drug carriers into tumors.

The results in Fig. 5 clearly show that rhodamine-labeled 2C5 immunomicelles effectively bind to the surface of several unrelated tumor cells lines: human BT20 and murine LLC and EL4 cells. The incubation of antibody-free micelles with the same cells results in virtually no micelle-to-cell association. No binding of 2C5 or 2C5-modified drug carriers with any normal cells was reported (33). Thus, immunomicelles bearing an antinucleosomal antibody, which recognizes a variety of cancer cells, can specifically attach to these cells in vitro.

Figure 5.

Microscopy data on the binding of Rh-labeled 2C5 immunomicelles to EL4 T lymphoma cells (Top), LLC cells (Middle), and BT20 mammary adenocarcinoma cells (Bottom). (Left) Bright-field light microscopy. (Right) Fluorescent microscopy.

Longevity of Immunomicelles in the Blood.

Prolonged circulation provides a drug carrier with a better chance to extravasate into the tumor interstitium and/or interact with ligands on the tumor cell surface. Direct correlation between the ability of a drug carrier to stay in the circulation and its accumulation in tumors was observed for water-soluble polymers (16) and liposomes (13).

Micelles prepared from PEG–PE are long-circulating (27). However, antibody attachment to drug carriers might provoke their faster clearance from the circulation because of uptake by Fc receptor-bearing Kupffer cells. To test whether the antibody attachment affects the blood clearance of PEG–PE micelles, we compared clearance characteristics of plain and 2C5-modified micelles in mice. The data in Fig. 6 clearly show that micelle modification with 2C5 had a very small effect on their blood clearance. Elimination profiles of plain and 2C5-modified micelles are almost identical. The data are consistent with earlier observations made with PEG liposomes modified with antimyosin antibodies (48).

Figure 6.

Blood clearance of plain micelles and 2C5 immunomicelles in mice.

Specific Accumulation of Immunomicelles in the Tumor in Vivo.

The enhanced permeability and retention effect-mediated accumulation of drug carriers in tumors depends, among other factors, on tumor vasculature cutoff size (39). Low vascular permeability prevents many drug carriers, including long-circulating liposomes, from entering certain tumors, such as LLC (41, 42). Because PEG-based micelles are much smaller than liposomes, they can provide an alternative and more efficient way of drug delivery (26, 27). The accumulation of plain and immuno-PEG–PE micelles was investigated in LLC tumor in mice. The results shown in Fig. 7 demonstrate that although PEG–PE micelles accumulate in the tumor much better than in normal tissue, their accumulation can be further improved if micelles are additionally modified with tumor-specific antibodies, i.e., converted into immunomicelles. 2C5 immunomicelles accumulate in LLC significantly better (by ≈30%) than plain micelles at both investigated randomly chosen time points (Fig. 7). An enhanced accumulation of 2C5-targeted micelles over plain micelles in the tumor was observed at 30 min (4.1 ± 0.2% dose per g of tumor vs. 3.4 ± 0.3% dose per g of tumor, P < 0.05) and 2 h (6.4 ± 0.1 dose per g of tumor vs. 4.1 ± 4 dose per g of tumor, P < 0.005) postinjection. Thus, 2C5-targeted micelles are capable of specific recognition and binding tumor cells in vivo. One can expect that, opposite to plain micelles, 2C5 immunomicelles should be capable of delivering their load not only to tumors with a mature vasculature, but also to tumors at the earlier stages of their growth and to metastases. Immunomicelles might be also better internalized by tumor cells similar to antibody-targeted liposomes (49).

Figure 7.

(Upper) Accumulation of PEG–PE micelles in s.c. LLC tumor in mice at different time points. (Lower) Accumulation of free and micellar taxol in LLC tumor at the same time points.

Experiments on taxol delivery into tumors have revealed an additional advantage of immunomicells. The use of taxol-loaded 2C5 micelles resulted in the highest quantity of tumor-accumulated taxol at both 30-min and 2-h time points compared with free taxol or taxol in plain micelles. Interestingly, at 2 h one can see a bigger difference than at 30 min in accumulation of taxol in tumors with immunomicelles compared with other taxol preparations. The difference might be explained by the accumulation of free taxol or taxol delivered by plain PEG–PE micelles in the interstitial space of the tumor and its eventual clearance from there (in case of micellar taxol, after gradual micelle degradation). At the same time, taxol-loaded 2C5 immunomicelles are internalized by cancer cell and thus keep the drug inside the tumor in a way similar to what was observed with drug-loaded anti-her2 immunoliposomes (49). The internalization by tumor cells might be highly useful therapeutically for many antitumor agents. For example, a much higher tumor regression rate was observed with a carrier capable of intracellular drug delivery for an equal doxorubicin dose delivered to the tumor (49).

Tumor Growth Inhibition by Taxol-Loaded 2C5 Immunomicelles.

The results of our preliminary proof-of-principle experiments on LLC growth inhibition by different taxol preparations have also confirmed the highest efficiency of taxol-loaded 2C5 immunomicelles among all tested preparations. Even within the model with a randomly chosen and nonoptimized treatment schedule, the average weight of excised tumors in the group treated with taxol incorporated in immunomicelles was 0.67 ± 0.35 g compared with 1.58 ± 0.48 g and 1.37 ± 0.36 g in groups treated with free taxol or taxol in plain PEG–PE micelles, respectively (P < 0.05 in both cases). The weight of untreated tumors was 2.00 ± 0.60 g. Taxol-free micelles of any composition do not affect tumor growth (Fig. 8).

Figure 8.

Inhibition of LLC tumor growth in mice with different taxol preparations.

Thus, stable PEG–PE-based micelles with an enhanced ability to carry a variety of poorly soluble pharmaceuticals can be transformed into immunomicelles by attaching various specific antibodies to their surface with our method. These micelle-coupled antibodies preserve their specific activity. Immunomicelles prepared with cancer-specific 2C5 antibody specifically bind to different cancer cells in vitro and demonstrate increased accumulation in experimental tumors in vivo. This family of pharmaceutical carriers should be very useful for the enhanced delivery of poorly soluble pharmaceuticals to various pathological sites in the body.

Abbreviations

PEG

polyethylene glycol

PE

phosphatidylethanolamine

LLC

Lewis lung carcinoma

pNP

p-nitrophenylcarbonyl

HBS

Hepes-buffered saline