Biomacromolecules 2008, 9, 435–443
435
Surface Charge of Nanoparticles Determines Their Endocytic and Transcytotic Pathway in Polarized MDCK Cells Oshrat Harush-Frenkel,†,‡ Eva Rozentur,† Simon Benita,† and Yoram Altschuler*,‡ Departments of Pharmaceutics and Pharmacology, School of Pharmacy, The Hebrew University of Jerusalem, POB 12065, Jerusalem 91120, Israel Received May 16, 2007; Revised Manuscript Received September 10, 2007
A major challenge in drug delivery is the internalization through the apical plasma membrane of the polarized epithelial cells lining organs facing the external environment, e.g., lungs and the gastrointestinal tract. The reduced permeation of drugs entering through this pathway is in part due to the mucosal barrier and low rate of endocytosis at these membranes. We investigated the possible role of nanoparticle surface charge on its entry through the apical plasma membrane and its intracellular pathway. We found that both cationic and anionic nanoparticles are targeted mainly to the clathrin endocytic machinery. A fraction of both nanoparticle formulations is suspected to internalize through a macropinocytosis-dependent pathway. A significant amount of nanoparticles transcytose and accumulate at the basolateral membrane. Some anionic but not cationic nanoparticles transited through the degradative lysosomal pathway. Taken together, these observations indicate that cationic nanoparticles, in addition to their potential for drug delivery to epithelia, may be promising carriers for transcytosing drugs to the blood stream.
Introduction Nanoparticles (NPs) have demonstrated great potential as delivery vehicles of drugs, nucleic acids, and therapeutic proteins.1–5 Improved NP formulations may enable drug targeting and superior drug stability as well as drug availability and intracellular retention within the target tissue.6 Recently, progress has been made regarding the endocytic mechanism utilized by NPs for their entry into mammalian cells. NPs endocytosis was shown to be a concentration, time, and energy dependent process, suggesting an active saturable internalization mechanism. These studies indicate that NP entry takes place through one of several endocytosis mechanisms operating in mammalian cells, including adsorptive-type endocytosis, transcytosis, and endocytic processes that are neither clathrin- nor caveolin-1mediated dependent.7–12 Previous studies suggested that cationic NPs elicited a better cell penetration and action increase over anionic or neutral NPs owing to the attraction of cationic NPs to the negatively charged plasma membrane of cells.13–15 Drug internalization through the apical plasma membrane (PM) of polarized epithelial cells and desired intracellular fate are major therapeutic obstacles that need to be overcome for efficient drug delivery. The low permeation through apical membranes is partly due to the mucosal barriers and the low rate of endocytosis operating at the apical PM. This process is associated with the important role of epithelia and specifically of the apical PM to filter and repel exogenous substances in order to protect cells and organs from the external milieu. Endocytosis at the PM of nonpolarized cells and at the basolateral (BL) PM of polarized epithelial cells was found to be several-fold faster than that operating at the apical PM.16,17 Others as well as members of our group have performed in* Corresponding author. E-mail:
[email protected]. Telephone: +972-2-6757578. Fax: +972-2-6758927. † Department of Pharmaceutics, School of Pharmacy, The Hebrew University of Jerusalem. ‡ Department of Pharmacology, School of Pharmacy, The Hebrew University of Jerusalem.
depth studies on the regulation of the endocytic machinery operating at the apical PM of polarized epithelial cells. It was found that the low endocytic rate is mainly due to the activation of several signaling cascades that regulate the recruitment of the clathrin-mediated machinery that governs the main endocytic events.17–20 Six different endocytic machineries have been identified based on the involvement of structural proteins participating in endocytic events.21 The major endocytosis pathway in all cells has been shown to be clathrin and dynamin dependent. These proteins participate in the endocytic event generating a coated pit responsible for the membrane invagination followed by a scission event. A second important pathway utilizes caveolae in which dynamin plays a role in the scission event.22 Clathrin macromolecules, composed of three heavy and three light chains, assemble to generate the structural skeleton of the coated pit, which associates through adaptor proteins with receptors bound to their cargo. Following invagination, dynamin, a 100 kDa GTPase, polymerizes around the neck of the coated pit and facilitates the scission event, generating a vesicle within the cytoplasm. Mutated forms of either clathrin heavy chain or dynamin have been shown to arrest clathrin- and caveolinmediated endocytosis and have been extensively employed to characterize macromolecules and pathogens endocytosing through these pathways. Therefore, the aim of this study was to understand the fundamental type of endocytosis and membrane transport of differently charged poly(ethylene glycol)-D,L-polylactide (PEGPLA) NPs through the apical PM of epithelial cells and their fate within the cells. Gaining knowledge on the cellular internalization mechanisms of these specific NPs is crucial for improving NP efficacy, site-specific delivery, and intracellular targeting. As epithelial cells are coated with extensively charged mucus that could interact with the charged NPs and affect their association and absorption by these cells,8 we used epithelial Madin-Darby canine kidney (MDCK) cells that harbor a relatively thin mucus layer. MDCK cells enabled us to study the specific interaction between the NP and the endocytic
10.1021/bm700535p CCC: $40.75 2008 American Chemical Society Published on Web 01/12/2008
436 Biomacromolecules, Vol. 9, No. 2, 2008
machinery independent of the mucus layer. The effect of mucus was previously studied by Behrnes et al.,8 who showed that it presented a major barrier for the uptake of hydrophobic polystyrene NPs and even had a more profound effect upon the uptake of cationic chitosan NPs.8 On the basis of semiquantitative assays, we found that anionic and cationic formulations applied to the apical PM of polarized epithelial MDCK cells are targeted to the clathrin endocytic machinery. The exposed positive charge on the surface of the NP apparently increased the amount of NPs internalized compared to those with a negative charge. Additionally, a significant number of NPs transcytosed and accumulated at the lateral PM. Of great interest is our observation that a large fraction of the anionic NPs accumulated in the degradative lysosomal pathway while the cationic NPs avoided this route.
Materials and Methods Materials. D,L-Lactide (Purasorb) was purchased from Purac (Gorinchem, The Netherlands). Benzyl alcohol, dimethyl sulfoxide (DMSO), 2-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT), methoxy polyethylene glycol (mPEG) MW 5000, stannous 1-ethylhexanoate, and polysorbate 80 (Tween 80) were acquired from Sigma (St. Louis, MO). Acetone, ethyl acetate, methylene chloride, and water were purchased from J. T. Baker (Deventer, The Netherlands). PEG-15-hydroxystearate (Solutol HS 15) was obtained from BASF (Ludwigshafen, Germany). Coumarin-6 and LysoTracker DND-99 were purchased from Polysciences (Warrington, PA) and Molecular Probes (Eugene, OR), respectively. Cell culture media and reagents were purchased from Biological Industries (Beit Haemek, Israel), except MEM-Hepes media, which was from Gibco, Invitrogen (Carlsbad, California). All reagents were of pharmaceutical grade. Preparation of Nanoparticles. NPs were prepared by the solvent displacement method.23 The polymers PLA MW 100 000 and mPEGPLA MW 100 000 (2:1) were dissolved in 50 mL of acetone containing 0.2% (w/v) Tween 80, at a concentration of 0.6% (w/v) acetone, and coumarin-6 solution at a concentration of 0.0003% (w/v) was added to the organic phase. The organic phase was added to 100 mL of an aqueous solution of 0.25% (w/v) Solutol HS 15. The suspension was stirred at 900 rpm for 1 h and subsequently concentrated by evaporation to 10 mL. All formulations were diafiltrated with a 100 mL solution of 0.1% Tween 80 (Vivaspin 300 000 MWCO, Vivascience, Stonehouse, UK) and filtered through a 1.2 µm filter (FP 30/1.2 CA, Schleicher & Schuell, Dassel, Germany). A typical blank cationic NP formulation consisted (in % w/v) of PLA100 000 2, mPEG-PLA100 000 1, Solutol HS 15 2.5, stearylamine 0.2, Tween 80 1, and doubleddistilled water to 100 mL. The composition of the anionic NP formulation was identical to that of the cationic one without the inclusion of the cationic lipid stearylamine. Three batches of each NP formulation were prepared. Physicochemical Characterization. Particle size distribution and mean diameter measurements were carried out using an ALV noninvasive backscattering high-performance particle sizer (ALV-NIBS HPPS, Langen, Germany) at 25 °C. NPs formulations were diluted with double-distilled water as described before.24–26 Zeta potential measurements were carried out using a Malvern Zetasizer (Malvern Instruments, Ltd., Malvern, UK). The samples were diluted in double-distilled water as performed previously.24,1,27,26 Cell Culture, Adenoviruses, and Antibodies. MDCK cells (type II) were grown in MEM supplemented with 5% (v/v) FCS and 50 µg/ mL of each penicillin and streptomycin at 37 °C in an incubator with 90% humidity and 5% CO2. Cells were used for assay after 3 or 4 days at 100% confluence for epithelial polarization. Fresh medium was replaced daily. Recombinant adenoviruses encoding the dominant negative mutants dynamin-I K44A (HA tag at the N-terminus) and clathrin hub (T7 tag at the N-terminus) were used. We routinely used 70–90 pfu/cell to
Harush-Frenkel et al. obtain the minimal functional effect. These conditions minimized the possibility of toxic effects and produced an adequate signal.28 To prevent adenovirus-related toxic side effects, protein expression levels were regulated by adjusting the amount of virus and length of incubation time (16–24 h). Anti-T7 tag antibody was obtained from Novagen (San Diego, CA). (Anti-HA tag) 12CA5 antibody was purchased from Covance (Indianapolis, IN). Antimouse secondary antibody conjugated to horseradish peroxidase (HRP) was purchased from Jackson Laboratory (Bar Harbor, ME). Monoclonal P58 antibody was kindly provided by Keith Mostov (UCSF, San Francisco, CA), Rhodamine Red-X (RRX)-conjugated donkey antimouse secondary antibody was purchased from Jackson ImmunoResearch Laboratories Inc. (West Grove, PA). Cytotoxicity of PEG-PLA NPs. The cytotoxicity assay was carried out using the MTT method.29 During the MTT assay, Hank’s buffer salt solution (HBSS) was used as both a positive control and a sample dilution buffer, and a 0.1% (w/v) SDS solution was used as a negative control for cell viability. Five consecutive dilutions (0.015–0.6 mg polymer/mL) were prepared from each formulation, with five replicates for each dilution. The tightness of the MDCK monolayer was tested for NP related effects. MDCK cells were seeded on transwells 3 days prior to experiment. Growth media was replaced daily. Media leakage was tested from the apical to the BL compartment for 18 h prior and following the incubation of cells with NPs. No leak of cell media was observed, confirming the integrity of MDCK cell monolayer prior to and following the incubation of cells with NPs. Cells were then washed three times using MEM-BSA-Hepes prior to initial transepithelial electric resistance (TEER) measurement. TEER measurement was repeated immediately after the addition of NPs formulation diluted into MEM-BSA-Hepes to the apical compartment of the cells. A third measurement was performed following 60 min incubation of NPs with cells. Visualization of Cellular Uptake of NPs. Three days after MDCK seeding (0.6 million MDCK cells per well in a 12-well plate), cells were washed three times with MEM-0.6% BSA-Hepes, an acceptable cell media in endocytosis experiments in order to remove residual mucus and neutralize possible mucus effect. NP formulations in MEM-BSAHepes (dilution factor 1000 v/v) were applied to the cells at 37 °C for the specified incubation time, ranging from 0 to 60 min. Cells were then washed three times using ice-cold phosphate buffered saline (PBS) and fixed using fresh paraformaldehyde (PFA) 4% (w/v), washed extensively, mounted on glass slides, and then visualized with an Olympus 1 × 70 confocal laser scanning microscope (CLSM; Olympus Co. Ltd., Tokyo, Japan). All images were compiled using Adobe Photoshop, Adobe Illustrator, and/or Image J software. The images are representatives of the original data. Evaluation of Fluorescent NP Uptake into MDCK Cells Using Fluorescence-Activated Cell Sorter (FACS) Analysis and a Fluorescent Plate Reader. Three days after MDCK seeding, in a 12well plate, cells were washed and NP formulations in MEM-BSA-Hepes were applied at 37 °C for the specified incubation times and then washed using ice-cold PBS. Cells were detached using trypsin, centrifuged at 1000g for 3 min, and analyzed for cell-associated NPs by FACS scan with Cellquest software (Becton-Dickinson and Co., Franklin Lakes, NJ). Three days after MDCK seeding, in a 96-well plate, cells were washed and fluorescent NP formulations in MEM-BSA-Hepes were applied at 37 °C for the specified incubation time, then washed using ice-cold PBS and held at 4 °C, followed by microplate reading using FluoStar (BMG Labtechnologies, Offenburg, Germany) as was previously described,30 with the exception that our samples were not lysed prior to fluorescent reading. Fluorescence filters with an excitation wavelength of 485 nm and an emission wavelength of 520 nm were used. Evaluation of Fundamental Endocytic Polypeptides Involved in NP Internalization into Polarized MDCK Cells. Cells were infected with recombinant adenoviruses and then allowed to express
Charged Nanoparticle Uptake by Polarized Epithelial Cells
Biomacromolecules, Vol. 9, No. 2, 2008 437
Table 1. Physical Stability of Positively and Negatively Charged NPs Stored for 60 Days at 4 °C under Nitrogen Atmosphere (mean ( SD, N ) 3)a cationic nanoparticles mean diameter, nm ( SD
storage, day
89.8 ( 4 ND 91.86 ( 8 ND 73.24 ( 5
0 15 30 45 60 a
anionic nanoparticles
ζ potential, mV ( SD
droplet diameter, nm ( SD
ζ potential, mV ( SD
32.8 ( 8.19 39.47 ( 10.31 45.46 ( 2 34.9 ( 7.08 37 ( 1.3
96.36 ( 6 ND 93.52 ( 5 ND 91.56 ( 3
-26 ( 1 -21.2 ( 8.73 -22.17 ( 0.4 -29 ( 7.34 -24.57 ( 0.67
ND ) not determined.
recombinant endocytic polypeptide for 16-24 h. Subsequently, cells were washed and incubated at 37 °C with the NP formulations in MEMBSA-Hepes for 60 min, followed by washing with ice-cold PBS. NP uptake by MDCK cells was examined and evaluated using various methods: FACS analysis, Fluostar plate reader and CLSM, as described above. Levels of protein expression were assayed by Western blotting. Briefly, equal amounts of protein were denatured by boiling for 5 min and loaded onto a 12% SDS-polyacrylamide gel. Proteins were transferred to nitrocellulose membranes by electroblotting using Trisglycine buffer with 20% (v/v) methanol. Membranes were blocked for 2 h with 5% (w/v) nonfat dried milk in PBS containing 0.1% Tween 20 (PBS-T) and then incubated for 1 h with specific antibody diluted in PBS-T plus 5% nonfat dried milk. After washing with PBS-T three times, the membranes were incubated for 1 h with HRP-labeled secondary antibody in PBS-T plus 5% nonfat dried milk. The blots were then washed three times, and proteins were visualized by the enhanced chemiluminescence (ECL) method (Amersham, Uppsala, Sweden). Immunofluorescence. Cells were grown as described above. On the third day, cells were washed extensively and incubated with NPs for 60 min. Subsequently, cells were washed three times with ice-cold PBS, fixed with freshly prepared 4% PFA, and then blocked with 3% BSA to prevent unspecific antibodies from binding. Cells were permeabilized with 0.7% fish skin gelatin, 0.025% saponin in PBS (PFS solution) for 15 min at 37 °C, and then incubated with P-58 mouse monoclonal antibody for 2 h. Subsequently, cells were washed four times with PFS, incubated with RRX-conjugated secondary antibody for 2 h, washed again, fixed, and mounted for microscopic visualization. Intracellular Localization of NPs. Cells were grown as described above. On the third day, cells were washed three times with prewarmed MEM-BSA-Hepes media, followed by incubation with NPs and LysoTracker at a concentration of 50 nM over 60 min. Cells were then washed three times using MEM-BSA-Hepes media, fixed, and examined by CLSM. Statistical Analysis. Statistical analysis was performed by using the Student’s t test with *0.01 < p < 0.05 or **p < 0.01 as significant difference; p value is indicated when appropriate.
Results and Discussion PEG-PLA NPs have great potential for drug and gene delivery owing to their known “stealth” biodegradable and biocompatible properties.1–5 An appealing therapeutic application for these NPs is the delivery of drugs through the endocytic machinery within the apical epithelial monolayer lining organs utilized for drug delivery. In this study, the influence of the exposed electrostatic surface charge of PEG-PLA NPs on their endocytosis at the apical PM and fate within polarized epithelial MDCK cells was investigated. This specific cell line was chosen because it has been extensively used as the model system for studies focusing on endocytosis and membrane transport of epithelial cells in the last 20 years.18,31 Furthermore, the MDCK cells system enables the utilization of gene delivery and tight regulation of
protein expression with adenoviruses through both (1) amount of virus applied to cells and (2) promoter regulation by tetracycline. These features enable us to avoid toxic side effects that may result from ectopic gene expression. In addition, MDCK cells possess a thin mucus layer, making it possible to focus the study on the interaction between NPs and the endocytic system with limited interference of the mucus. Physicochemical Characteristics of PEG-PLA NPs. In the present study, differently charged PEG-PLA NPs were utilized. Adding stearylamine to the NP formulation resulted in positively charged NPs. NPs were manufactured by the solvent displacement method using coumarin-6, a well-known and suitable fluorescent marker with log P value of 5.2.1–3,30,32–34 It should be noted that equal amounts of fluorescent coumarin-6 were loaded into both formulations as analyzed using a high sensitivity fluorescence spectrometer (FP-6300, Jasco Corporation, Tokyo, Japan), thus enabling a comparative uptake study (data not shown). Previous research done on PLGA based NPs (50:50 MW of 143 000 Da) containing 4% w/w bovine serum albumin (BSA) labeled with coumarin-6 exhibited only minor (0.32% of the actual content) release of the free dye in bicarbonate Ringer’s solution during a 24 h period.33 Other authors have shown that in vitro free coumarin dye released from PLGA-PVA-BSA (50:50 MW of 143 000 Da) NPs accounted for only about 3% of the total uptake of the dye following incubation of the released medium with HUVEC cells.35 Furthermore, Caco-2 cells incubated with medium of the coumarin-6 released from the poly(lactic-coglycolic acid) (PLGA) NPs coated with polyvinyl alcohol or vitamin E succinated polyethylene glycol 1000 did not show any significant uptake (approximately 1%), demonstrating that coumarin-6 marker cannot be directly internalized by the cells.30 Therefore, in view of the extensive steps of NPs diafiltration in the present study, it was determined that the fluorescent signal following 60 min incubation with NPs is mainly associated with NPs internalization and not related to the free released dye in the medium. To establish the stability of cationic and anionic NPs over a period of 60 days in storage at 4 °C, we measured the ζ potential and particle size of both formulations. As shown in Table 1, both cationic and anionic formulations were stable and maintained their charge and size. The particle size of cationic and anionic NP formulations averaged 89.8 ( 4 and 96.4 ( 3 nm, respectively, and following 60 days of storage (4 °C), they were 73.24 ( 4 and 91.6 ( 3 nm, accordingly (Table 1). The difference of 18% between 0 and 60 days in Table 1 for average values of the particle size of cationic NPs cannot be considered as a significant finding when comparing the size of two populations of the same NPs formulation. There is no marked difference between two NP populations with average particle sizes of 89 and 73 nm. Both behave similarly. Cationic and anionic NP formulations exhibited initial ζ potential values of
438 Biomacromolecules, Vol. 9, No. 2, 2008
Figure 1. Charged NPs show no detectable cytotoxicity. MDCK cells were incubated with the indicated dilutions (0.015–0.6 mg PEG-PLA polymer/mL HBSS) of cationic (black) and anionic (gray) charged NP formulations for 180 min and then processed for MTT assay. Viability of cells shown compared to untreated cells (control) (mean ( SD, N ) 3). Positive control cells were incubated with HBSS. Negative control cells were incubated with 0.1% SDS/HBSS.
+32.8 ( 8.19 and -26 ( 1 mV, respectively. After 60 days in storage, these values were +37 ( 1.3 and -24 ( 0.6 mV, accordingly. In general, charge modification had no effect on particle size in either NP formulation, maintaining a diameter in the range of 100 nm, which is in the size range of the clathrincoated pit and endocytic vesicles.22,36 Cytotoxicity of PEG-PLA NPs in MDCK Cells. The potential cytotoxicity of the PEG-PLA-based NPs was evaluated by MTT assay. We incubated dilutions of both NP formulations and measured their effect on epithelial cell survival. At all dilutions tested, from 0.015 up to 0.6 mg PEG-PLA polymer in HBSS (mL), MDCK cell viability was relatively unchanged (Figure 1). Although cationic NPs seemed to slightly reduce cell viability, their effect was minor at all dilutions tested. Overall, both formulations exhibited high stability and were nontoxic. These results are in agreement with the NPs performance in nonpolarized HeLa cells.11 PEG-PLA NP Uptake through the Apical PM of Polarized MDCK Cells. Delivery of drugs requires internalization through the apical PM of the epithelial monolayer. To quantitatively examine NP internalization through the apical PM of polarized epithelial cells, we grew confluent monolayers of MDCK cells for 3 days. Subsequently, fluorescent NPs were incubated on the apical PM for up to 60 min. The tightness of MDCK monolayer in the presence of anionic and cationic NPs following 60 min incubation was verified and confirmed by the results of the TEER measurements. CLSM imaging of fluorescent NP uptake into MDCK cells revealed a significant initial accumulation of NPs after 30 min of incubation and a gradual accumulation thereafter (Figure 2A). Intracellular NP accumulation appeared to occur at the tight-junction level, in the lateral domain of MDCK cells. Beginning at 30 min, and significantly observed at 45 min, bright staining of cationic NPs appeared at the lateral PM, indicating that a fraction of the cationic NPs followed the transcytotic pathway and accumulated at the lateral PM (Figure 2A, 45 min, arrows; B). These results support the previous observation by Behrens and colleagues, who observed the apical to basal transcytosis of positively charged chitosan NPs in human intestinal Caco-2 cells.8 In contrast, a very small
Harush-Frenkel et al.
fraction of the anionic NPs transcytosed and appeared at the lateral PM (Figure 2E). The punctate staining of both formulations throughout the cell indicated that both NP formulations most likely remained attached to intracellular membranes or cluster into aggregates during this period (Figure 2A, 30–60 min, B and E). Additionally, a fraction of the MDCK cells incubated with both NPs formulations that exhibited elevated uptake compared to neighboring cells, also displayed honeycomblike circular organelles, the surfaces of which were decorated with NPs and looked like macropinosomes (Figure 2D). We think that the appearance of these structures results from concomitant saturation of the main internalization pathways operating at the PM and of numerous binding sites at the cell PM. Furthermore, Figure 2A reveals that cationic NPs accumulate within the cells to a higher degree than anionic NPs. The intracellular accumulation of both NPs formulations was also analyzed by confocal microscopy z-scans, revealing reduced amounts of anionic as compared to cationic NPs within the cells throughout the whole cellular scan (data not shown). To establish the difference between the two formulations, a semiquantitative analysis of the amount of NP uptake was performed using Image J software (Figure 2F) as described before by Kim et al.37 Following incubation times shorter than 30 min, the fluorescence intensities in the cells, albeit low, were stronger for cationic versus anionic NPs. Incubation periods longer than 30 min showed a marked increase in the uptake of both types of NPs. Most importantly, both formulations exhibited reduced uptake following the 30 min time point, indicating that cellular endocytic machineries and binding sites were saturated and newer binding sites were not yet available at the plasma membrane. These findings resemble the results of Panyam et al., where NP uptake increased with incubation time in the presence of NPs in the medium.9 Panyam and colleagues also observed an interesting phenomenon of rapid exocytosis of poly(D,L-lactide-coglycolide) NPs incubated with vascular smooth muscle cells.9 Measurements of NP uptake at the apical PM of MDCK cells were conducted by FACS analysis. To focus on NPs that have been endocytosed, we trypsinized the cells extensively to remove most of the PM-bound NPs. As shown in Figure 3A, cationic NPs exhibited approximately 20% more uptake than their anionic counterparts in a saturable uptake process. In agreement with our previous results, the reduction in uptake at later time points is indicative of a saturable endocytic process as well as binding factors at the apical PM that are present in limiting concentrations. The reduced fluorescence resulting from the use of trypsin may have been due to the elimination of NP fluorescence from the apical PM that would have been internalized at later time points. To overcome possible artifacts due to trypsin treatment, we quantitated NP uptake by live cells with a fluorescent plate reader (Figure 3B). This approach allows measurements to be performed on live cells that have not undergone any manipulation (e.g., trypsinization or fixation). On the other hand, this methodology does not allow discrimination between PM-bound and internalized NPs. As shown in Figure 3B, the association and internalization of cationic NPs was almost 2-fold higher than with anionic NPs. Moreover, in these experiments, both formulations revealed increasing uptake within the first 15 min, albeit to different degrees, followed by a plateau for about 30 min and then additional uptake. This result suggests saturation of cellular endocytic machineries as well as limited number of NP binding sites at the apical surface. These active endocytic machineries and binding sites are responsible for the initial (first 15 min) internalization of NPs. The second cycle of NP uptake, which
Charged Nanoparticle Uptake by Polarized Epithelial Cells
Biomacromolecules, Vol. 9, No. 2, 2008 439
Figure 2. Endocytosis of charged NPs by MDCK cells. MDCK cells were incubated for the indicated time periods with either cationic (left) or anionic (right) NPs. Subsequently, cells were fixed and processed for confocal microscopy. (A) Cationic and anionic NPs follow the transcytotic pathway and appear as endosomal punctate staining and on the lateral PM. The cationic NPs reveal higher apical endocytosis as reflected by the higher staining intensity throughout the endosomal and lateral PM. Bar, 10 µM. (B-E) Enlargement of MDCK cells incubated with cationic (B-D) or anionic (E) NPs for 30 min. Bar, 2 µM. In (D), honeycomb-like circular organelles (arrowheads), whose surfaces are decorated with NPs, are observed at the cell periphery, resembling macropinosomes. This phenomenon has been observed in cells incubated with either cationic or anionic NPs. These structures appear in cells showing high NP uptake. (F) Fluorescence intensity of cell-internalized NPs was quantified and used to calculate the amount of NPs endocytosed. Ten images of each time point, each containing approximately 25-30 cells, were analyzed by Image J software. Cationic NPs (solid line, diamonds), when applied to the apical PM, appear to internalize to a higher degree than anionic NPs (dashed line, squares).
takes place between 45 and 60 min, requires delivery of available endocytic polypeptides and binding factors to the apical PM through either the recycling or biosynthetic pathways of the cell. This delivery takes place during the 15-45 min period following initial entry and fits the kinetics of endocytosis and recycling through the clathrin-mediated pathway at the apical PM.38 The difference between the results obtained by FACS and those obtained by fluorescent plate reader was a higher number of cationic NPs observed in the latter, probably reflecting those that were eliminated by trypsin in the FACS experiment. Therefore, it is reasonable to assume that NP uptake is dependent on available endocytic polypeptides and binding
factors present at the apical PM, rate of endocytosis, and delivery of available endocytic polypeptides and binding sites to the apical PM. Recently, we have shown that in HeLa cells, NP charge has a major effect on endocytosis, the intracellular pathway taken, and the cells’ response to the NPs. Interestingly, the cationic NPs showed approximately 2-fold greater uptake than anionic NPs in HeLa cells. In this study based on semiquantitative experiments, cationic NPs incubated at the apical PM also showed increased uptake compared to anionic NPs, but not to the extent seen in the HeLa cells,11 probably due to the low
440 Biomacromolecules, Vol. 9, No. 2, 2008
Figure 3. Quantitative evaluation of NP interactions with MDCK cells. MDCK cells were incubated for the indicated time periods with either cationic (solid line, diamonds) or anionic (dashed line, squares) NPs. Subsequently, cells were either harvested and analyzed by FACS (A) or washed and measured in a fluorescent plate reader (Fluostar, BMG) (B). (A) Cationic NPs show a moderate increase in fluorescence intensity compared to anionic NPs. (B) Cationic NPs show 2-fold more uptake than anionic NPs. The difference in the extent of the increase in uptake with the FACS analysis (moderate, A) and live cell measurement by plate reader (prominent, B) is most likely due to the trypsin treatment in the former, which eliminates many of the PMassociated NPs through trypsinization of PM proteins.
rate of clathrin-mediated endocytosis operating at the apical surface of MDCK cells 16 Mechanism of PEG-PLA NP Internalization into Polarized MDCK Cells. The endocytic machineries of the apical PM of epithelia affect NPs’ internalization. Several endocytic mechanisms operate at the apical PM of MDCK cells. These include the clathrin-mediated pathway17,38,39 and clathrin independent pathways that depend on PKC and Rho-family GTPase.40,41 Specific to MDCK cells, caveolae are present on the BL PM and are absent from the apical PM.42 Antibody crosslinking of specific proteins may induce the appearance of caveolae at the apical PM. However, they do not increase membrane internalization to any great extent.43 Therefore, the
Harush-Frenkel et al.
low rate of endocytosis at the apical PM of MDCK cells is mainly due to the low rate of maturation and scission of the clathrin-mediated machinery. NP endocytosis and delivery to the endosome was evidenced by the accumulation of punctate intracellular staining, suggesting that these NPs utilize the clathrin-mediated endocytosis pathway. To examine whether NPs utilize this endocytic pathway at the apical PM, we expressed well-characterized dominant negative mutant polypeptides of dynamin and clathrin that have been shown to inhibit the clathrin-mediated pathway at the apical PM of epithelia.38,44 MDCK cells were infected with recombinant adenoviruses encoding for tagged dynamin (HA tag) and clathrin hub (T7 tag). To confirm expression of the different mutant constructs, cell lysates were resolved by SDS-PAGE. A reacted Western blot (Figure 4F) with either anti-HA or antiT7 antibodies revealed that both proteins are expressed in these cells. Expression of the dominant negative dynamin I K44A (for 18 h) and clathrin heavy chain mutant lacking its N-terminal domain (clathrin hub, for 24 h) followed by incubation for 60 min with both cationic and anionic NP formulations on the apical PM resulted in markedly reduced endocytosis of NPs (Figure 4A). Specifically, expression of the dynamin mutant resulted in loss of NP staining within the endosomal system while preserving NP staining along the lateral PM. Expression of the dynamin mutant is known to cause the accumulation of membrane proteins mainly on the PM, and we therefore hypothesized that a large fraction of the NP binding factors are endogenous membrane proteins that bind the NPs at the apical PM and subsequently transcytose them to the lateral PM. This is supported by the redistribution of NPs to the lateral PM in cells expressing dynamin I mutant (Figure 4A-C). In Figure 4B,C, cells accumulating high levels of NPs also revealed circular organelles whose surfaces were decorated with NPs: this can be explained by the appearance of binding factors on the PM of these cells. By expressing the dynamin I mutant, 80% of endocytosis was inhibited at both PMs, resulting in redistribution of many membrane proteins from endosomal compartments to either apical or BL PMs and, as a result, NP binding sites are artificially increased on the PMs.44 This brings about the increased formation of circular organelles decorated by NPs in proximity to the PM. To quantitate NP uptake by the clathrin-mediated machinery, we inhibited this endocytic system by expressing the dynamin and clathrin mutant constructs and measured uptake of NPs over a period of 60 min. In Figure 4E, we observe higher association to, and uptake of, the cationic NPs compared to their anionic counterparts. In addition, expression of these endocytosis mutants dramatically reduced the uptake of cationic NPs (65% inhibition), indicating that these NPs internalize mainly through the clathrin-mediated pathway. Inhibition of anionic NP uptake was slightly less (54% inhibition), implying that in addition to internalization through the clathrin-mediated pathway, they may utilize other dynamin-independent internalization pathways. The results clearly indicate that both NP formulations associate with binding factors that endocytose through the clathrin-mediated pathway. Our observation that dynamin and clathrin dominant negative mutants inhibit both NP formulations indicates that their main endocytic entry takes place through the clathrin-mediated pathway (because, in MDCK cells, caveolae do not appear on the apical PM). Inhibition of this pathway results in the stimulation of an alternative entry pathway known as macropinocytosis: because macropinosome-like structures were observed within the cells as megacircular peripheral organelles
Charged Nanoparticle Uptake by Polarized Epithelial Cells
Biomacromolecules, Vol. 9, No. 2, 2008 441
Figure 4. NPs applied to the apical PM endocytose through a clathrin-dynamin-mediated pathway. MDCK cells were grown for 3 days to enable epithelial polarization. Control cells and cells expressing the dominant negative dynamin I K44A mutant and dominant negative clathrin hub mutant were used to monitor endocytosis of NPs applied to the apical PM. Cells were incubated with cationic (left) and anionic (right) NPs for 60 min. Subsequently, cells were fixed and processed for confocal microscopy (A-D). Both dynamin and clathrin mutants inhibit NP uptake, although the inhibitory effect is more prominent for cationic NPs (A); bar, 10 µM. (B-D) Enlarged images of cells from A showing anionic NP uptake (B,C) and cationic NP uptake (D); bar, 2 µM. (B-C) Despite inhibition of clathrin-mediated apical endocytosis, anionic NPs accumulate at the lateral PM and in circular organelles resembling macropinosomes. (D) Cells expressing dominant negative clathrin hub accumulate cationic NPs at the lateral PM in a punctate fashion, suggesting clathrin-coated pits. Arrows indicate NPs at the lateral PM; arrowheads indicate the circular organelles resembling macropinosomes. Ten images of each time point, each containing approximately 25-30 cells were taken. (E) Fluorescence intensity of cells expressing different endocytic mutants and incubated with cationic and anionic NPs. Experiment performed in triplicates, 3 times; t test performed and p value indicated in graph. *0.01 < p < 0.05; **p < 0.01. (F) Cell extracts of control cells and cells expressing dynamin 1 K44A mutant and clathrin hub mutant were separated by SDS-PAGE followed by Western blotting with anti-HA (dynamin) and anti-T7 (clathrin hub) antibodies. Western blot shows that cells express the relevant mutants.
that are intensely decorated with NPs (Figure 4B).45–47 Our results differ from the previous observation that in primary cultured rabbit conjunctival epithelial cells, PLGA NP uptake may take place in part via a clathrin-mediated process and endocytosis is most likely of the adsorptive type.33,34 The different cell type used, PEGylation, different doses, and different durations of incubation may explain this discrepancy. It should be noted that our results are in agreement with several other studies reporting clathrin-mediated endocytosis of layered double-hydroxide NPs into MNNG/HOS cells,48 mesoporous NPs into human mesenchymal stem cells,49 and chitosan NPs into Caco-2 and A549 cells.50,51 In addition, our results are in accordance with Panyam et al. who reported that the uptake of PLGA NPs in vascular smooth muscle cells is significantly reduced following inhibition of the clathrin-mediated pathway, but not of the caveola-dependent pathway.10 Cationic NP internalization occurs through the clathrin-mediated pathway in HeLa cells11 and at the apical PM of polarized MDCK cells. The higher inhibitory effect of the clathrin and dynamin
mutant constructs in HeLa cells relative to the apical PM result probably from the overall low rate of clathrin-mediated endocytosis at the apical PM relative to that in HeLa cells. This could explain the reduced advantage of cationic over anionic NPs in apical internalization into MDCK cells.17,38 The NPs’ ability to follow the transcytotic pathway may open the way for efficient delivery of drugs that have previously been repelled by the epithelial monolayer. In Figure 2, we provide clear evidence that internalized NPs from the apical PM transcytose and are delivered to the lateral PM, indicating their ability to follow the transcytotic pathway. Moreover, inhibition of endocytosis by clathrin and dynamin mutants that inhibit clathrin-mediated endocytosis at both the BL and apical PMs did not inhibit transcytosis but slightly restricted the number of NPs at the lateral PM (Figure 2). This suggests that the presumed NP carrier was preferentially transcytosed and trapped at the lateral PM due to inhibition of endocytosis or high affinity of the NPs to the lateral PM. Support for this phenomenon comes from work on NPs in Caco-2 cells that showed kinetics of transcytosis similar to those found here
442 Biomacromolecules, Vol. 9, No. 2, 2008
Harush-Frenkel et al.
Figure 5. Intracellular accumulation of NP formulations. (A) MDCK cells incubated for 60 min with cationic and anionic NPs, fixed, blocked with 3% BSA, permeabilized, and then incubated with P-58 mouse monoclonal antibody for 2 h, followed by incubation with RRX-conjugated secondary antibody for 2 h. Merged NP fluorescence (green) and P58 fluorescence (red) shows partial overlap of NPs and lateral PM in yellow staining. (B) MDCK cells incubated for 60 min with cationic and anionic NPs and LysoTracker were fixed and processed for confocal microscopy. Merged NP fluorescence (green), and LysoTracker fluorescence (red) shows overlap of NPs and LysoTracker yellow punctate staining indicates colocalization between NPs and lysosomes. Interference contrast image of the cells shown in gray. While anionic NPs are partly localized to lysosomes, cationic NPs are excluded from the lysosomes. Bar, 2 µM.
(Figure 2A).8 Similar studies supporting preferential uptake and transcytosis of cationic NPs have been performed on endothelial cell monolayers. To assay the NPs’ potential to cross the blood-brain barrier.52 These studies showed that the cationic BSA NPs were transcytosed several-fold more effectively than the noncationic BSA-NPs.2,53 Anionic, But Not Cationic NPs are Also Targeted to the Degradative Lysosomal Route. To visualize NPs association with lateral plasma membrane of MDCK cells, immunofluorescence experiments were performed. In Figure 5A, it can be observed that the anti P58 monoclonal antibody clearly stains the entire BL PM in red while the visualized NPs are in green. The merging of both images revealed partial colocalization between NPs formulation and BL PM marker (yellow), indicating partial delivery of NPs (both cationic and anionic) to the lateral PM, while a fraction remains intracellular at the BL side of the cell. The possible effect of NP surface charge on their sorting during intracellular transport was evaluated. For the purpose of differentiating between NPs targeted to the degradative lysosomal pathway and hence exposed to a harsh cellular environment that may reduce their half-life and NPs targeted to other pathways within the cell; we coincubated MDCK cells with the NP formulations and LysoTracker, a lysosomal fluorescent tracer that binds to membranes under acidic conditions. As shown in Figure 5B, cationic NPs were rarely seen colocalized with the lysosomal marker. Our findings are further supported by published results showing that the anionic surface charge of PLGA NPs applied to human nonpolarized arterial smooth muscle cells changes to a cationic charge during their transport along the endolysosomal pathway, resulting in their escape into the cytosol and indicating that the charge may have a critical
effect on the endocytic pathway within the cell.10 Thus, a large fraction of the anionic NPs were colocalized with the marker. These results indicate that the negative charge incorporated on the NPs greatly affects their intracellular route and targets them to the degradative pathway. Possible Biofate of Cationic and Anionic NPs in MDCK Cells. On the basis of our study, an illustrated model of NPs membrane transport within MDCK cells is depicted in Figure 6. NPs applied to the apical PM of MDCK cells are mostly targeted to the clathrin endocytic machinery (Figure 6A1), with a fraction internalized through a dynamin- and clathrin-independent process, like the macropinocytic pathway (Figure 6A2). The positive charge contributes to the enhanced localization of NPs on the PM into clathrin-coated pits. Part of the internalized anionic and cationic NPs follow an endocytic pathway, while the remaining NP fraction is assumed to recycle back to the PM by an exocytotic process similar to the phenomena observed by Panyam et al. for PLGA NPs.9 Importantly, a significant amount of cationic NPs and a fraction of the anionic NPs transcytose and accumulate at the lateral PM (Figure 6C). Moreover, during their intracellular passage, cationic NPs avoid the lysosomal degradative pathway, while a fraction of the anionic NPs follows this route (Figures 4, 6D).
Conclusions In this study, we show that the charge exposed at the surface of NPs may significantly affect their internalization machinery and intracellular route in polarized epithelial MDCK cells. When NPs applied to the apical PM of these cells, they are mostly targeted to the clathrin endocytic machinery. The positive charge stimulates
Charged Nanoparticle Uptake by Polarized Epithelial Cells
Figure 6. Cationic and anionic NPs follow different intracellular pathways in polarized epithelial MDCK cells (redrawn with permission from ref 28). NPs applied to the apical PM of polarized epithelial cells associate with PM components and further internalize, mostly via the clathrin-mediated machinery (A1) and a fraction by a macropinosomallike, clathrin-dynamin-independent machinery (A2). Both routes of entry come together at the apical early endosomes (B) from which the NPs are sorted to either the common endosome and further to the lateral PM (C) or to the late endosome and further to the lysosome (D). At the apical early endosome, a significant fraction of the anionic NPs (red line) are either targeted to the degradative lysosomal pathway (D) or transcytosed to the lateral PM (C). On the other hand, cationic NPs (blue line) sorted at the apical early endosomes are either retained within the endosomal system or transcytosed to the lateral PM (C), avoiding the lysosomal pathway.
movement of the NP on the PM into clathrin-coated pits. A fraction of NP internalized through a dynamin- and clathrin-independent process, macropinocytic like pathway. Overall, our results suggest that cationic NPs may be promising drugs carriers aimed for release at the lateral PM and in the blood stream due to their transcytotic pathway and their absence in lysosomes. Acknowledgment. This research was supported in part by the Israel Science Foundation (grant nos. 1318/04 and 622/04) to Y.A. and S.B., respectively. Y.A. and S.B. are affiliated with the David R. Bloom Center for Pharmacy at the Hebrew University.
References and Notes (1) Dong, Y.; Feng, S. S. Biomaterials 2005, 26, 6068–6076. (2) Lu, W.; Tan, Y. Z.; Hu, K. L.; Jiang, X. G. Int. J. Pharm. 2005, 295, 247–260. (3) Gao, X.; Tao, W.; Lu, W.; Zhang, Q.; Zhang, Y.; Jiang, X.; Fu, S. Biomaterials 2006, 27, 3482–3490. (4) Vila, A.; Sanchez, A.; Evora, C.; Soriano, I.; Vila Jato, J. L.; Alonso, M. J. J. Aerosol. Med. 2004, 17, 174–185. (5) Olivier, J. C. NeuroRx 2005, 2, 108–119. (6) Chavanpatil, M. D.; Khdair, A.; Panyam, J. J. Nanosci. Nanotechnol. 2006, 6, 2651–2663. (7) Fiegel, J.; Ehrhardt, C.; Schaefer, U. F.; Lehr, C. M.; Hanes, J. Pharm. Res. 2003, 20, 788–796. (8) Behrens, I.; Pena, A. I.; Alonso, M. J.; Kissel, T. Pharm. Res. 2002, 19, 1185–1193. (9) Panyam, J.; Labhasetwar, V. Pharm. Res. 2003, 20, 212–220. (10) Panyam, J.; Zhou, W. Z.; Prabha, S.; Sahoo, S. K.; Labhasetwar, V. FASEB J. 2002, 16, 1217–1226.
Biomacromolecules, Vol. 9, No. 2, 2008 443 (11) Harush-Frenkel, O.; Debotton, N.; Benita, S.; Altschuler, Y. Biochem. Biophys. Res. Commun. 2007, 353, 26–32. (12) Mao, S.; Germershaus, O.; Fischer, D.; Linn, T.; Schnepf, R.; Kissel, T. Pharm. Res. 2005, 22, 2058–2068. (13) Jung, T.; Kamm, W.; Breitenbach, A.; Kaiserling, E.; Xiao, J. X.; Kissel, T. Eur. J. Pharm. Biopharm. 2000, 50, 147–160. (14) Labhasetwar, V.; Song, C.; Humphrey, W.; Shebuski, R.; Levy, R. J. J. Pharm. Sci. 1998, 87, 1229–1234. (15) Rabinovich-Guilatt, L.; Couvreur, P.; Lambert, G.; Dubernet, C. J. Drug Targeting 2004, 12, 623–633. (16) von Bonsdorff, C. H.; Fuller, S. D.; Simons, K. EMBO J. 1985, 4, 2781–2792. (17) Naim, H. Y.; Dodds, D. T.; Brewer, C. B.; Roth, M. G. J. Cell Biol. 1995, 129, 1241–1250. (18) Altschuler, Y.; Hodson, C.; Milgram, S. L. Curr. Opin. Cell Biol. 2003, 15, 423–429. (19) Mostov, K. E.; Verges, M.; Altschuler, Y. Curr. Opin. Cell Biol. 2000, 12, 483–490. (20) Hyman, T.; Shmuel, M.; Altschuler, Y. Mol. Biol. Cell 2006, 17, 427– 437. (21) Marsh, M.; Helenius, A. Cell 2006, 124, 729–740. (22) Conner, S. D.; Schmid, S. L. Nature 2003, 422, 37–44. (23) Fessi, H.; Puisieux, F.; Devissaguet, J. P.; Ammoury, N.; Benita, S. Int. J. Pharm. 1989, 55, R1–R4. (24) Panyam, J.; Labhasetwar, V. Mol. Pharm. 2004, 1, 77–84. (25) Musumeci, T.; Ventura, C. A.; Giannone, I.; Ruozi, B.; Montenegro, L.; Pignatello, R.; Puglisi, G. Int. J. Pharm. 2006, 325, 172–179. (26) Devalapally, H.; Shenoy, D.; Little, S.; Langer, R.; Amiji, M. Cancer Chemother. Pharmacol. 2007, 59, 477–484. (27) Dailey, L. A.; Kleemann, E.; Wittmar, M.; Gessler, T.; Schmehl, T.; Roberts, C.; Seeger, W.; Kissel, T. Pharm. Res. 2003, 20, 2011–2020. (28) Shmuel, M.; Nodel-Berner, E.; Hyman, T.; Rouvinski, A.; Altschuler, Y. Mol. Biol. Cell 2007, 18, 1570–1585. (29) Mosmann, T. J. Immunol. Methods 1983, 65, 55–63. (30) Win, K. Y.; Feng, S. S. Biomaterials 2005, 26, 2713–2722. (31) Rodriguez-Boulan, E.; Kreitzer, G.; Musch, A. Nat. ReV. Mol. Cell Biol. 2005, 6, 233–247. (32) Fessi, H.; Puisieux, F.; Devissaguet, J.-P.; Ammoury, N.; Benita, S. Int. J. Pharm. 1989, 55, R1–R4. (33) Qaddoumi, M. G.; Ueda, H.; Yang, J.; Davda, J.; Labhasetwar, V.; Lee, V. H. Pharm. Res. 2004, 21, 641–648. (34) Qaddoumi, M. G.; Gukasyan, H. J.; Davda, J.; Labhasetwar, V.; Kim, K. J.; Lee, V. H. Mol. Vision 2003, 9, 559–568. (35) Davda, J.; Labhasetwar, V. Int. J. Pharm. 2002, 233, 51–59. (36) Mellman, I. Annu. ReV. Cell. DeV. Biol. 1996, 12, 575–625. (37) Kim, H. R.; Gil, S.; Andrieux, K.; Nicolas, V.; Appel, M.; Chacun, H.; Desmaele, D.; Taran, F.; Georgin, D.; Couvreur, P. Cell. Mol. Life Sci. 2007, 64, 356–364. (38) Altschuler, Y.; Liu, S.-H.; Katz, L.; Tang, K.; Hardy, S.; Brodsky, F.; Mostov, K. J. Cell Biol. 1999, 147, 7–12. (39) Shmuel, M.; Santy, L. C.; Frank, S.; Avrahami, D.; Casanova, J. E.; Altschuler, Y. J. Biol. Chem. 2006, 281, 13300–13308. (40) Holm, P. K.; Eker, P.; Sandvig, K.; van Deurs, B. Exp. Cell Res. 1995, 217, 157–168. (41) Garred, O.; Rodal, S. K.; van Deurs, B.; Sandvig, K. Traffic 2001, 2, 26–36. (42) Vogel, U.; Sandvig, K.; van Deurs, B. J. Cell Sci. 1998, 111 (Pt 6), 825–832. (43) Verkade, P.; Harder, T.; Lafont, F.; Simons, K. J. Cell Biol. 2000, 148, 727–739. (44) Altschuler, Y.; Barbas, S. M.; Terlecky, L. J.; Tang, K.; Hardy, S.; Mostov, K. E.; Schmid, S. L. J. Cell Biol. 1998, 143, 1871–1881. (45) Wadia, J. S.; Stan, R. V.; Dowdy, S. F. Nat. Med. 2004, 10, 310–315. (46) Meier, O.; Boucke, K.; Hammer, S. V.; Keller, S.; Stidwill, R. P.; Hemmi, S.; Greber, U. F. J. Cell Biol. 2002, 158, 1119–1131. (47) Pelkmans, L.; Helenius, A. Curr. Opin. Cell Biol. 2003, 15, 414–422. (48) Oh, J. M.; Choi, S. J.; Kim, S. T.; Choy, J. H. Bioconjugate Chem. 2006, 17, 1411–1417. (49) Huang, D. M.; Hung, Y.; Ko, B. S.; Hsu, S. C.; Chen, W. H.; Chien, C. L.; Tsai, C. P.; Kuo, C. T.; Kang, J. C.; Yang, C. S.; Mou, C. Y.; Chen, Y. C. FASEB J. 2005, 19, 2014–2016. (50) Ma, Z.; Lim, L. Y. Pharm. Res. 2003, 20, 1812–1819. (51) Huang, M.; Ma, Z.; Khor, E.; Lim, L. Y. Pharm. Res. 2002, 19, 1488– 1494. (52) Kreuter, J. J. Nanosci. Nanotechnol. 2004, 4, 484–488. (53) Chuang, V. T.; Kragh-Hansen, U.; Otagiri, M. Pharm. Res. 2002, 19, 569–577.
BM700535P