Nonionic Polymeric Micelles for Oral Gene Delivery In Vivo
SHWU-FEN CHANG,1 HAN-YI CHANG,2 YAW-CHONG TONG,2 SY-HANN CHEN,3FEI-CHIN HSAIO,2 SHAO-CHUN LU,4 and JIAHORNG LIAW2
ABSTRACT
The main aim of this study was to investigate the feasibility of using nonionic polymeric micelles of poly(eth-ylene oxide)–poly(proppoly(eth-ylene oxide)–poly(ethpoly(eth-ylene oxide) (PEO-PPO-PEO) as a carrier for oral DNA delivery in vivo. The size and appearance of DNA/PEO-PPO-PEO polymeric micelles were examined, respectively, by dynamic light scattering and atomic force microscopy, and their z potential was measured. Expression of the delivered lacZ gene in various tissues of nude mice was assessed qualitatively by
5-bromo-4-chloro-3-indolyl-b-D-galactopyranoside staining of sections and quantitatively by measuring enzyme activity in tissue extracts, using the substrate of b-galactosidase, chlorophenol red-b-D-galactopyranoside. In addition, the types of cells expressing the lacZ gene in the duodenum were identified by histological analysis. DNA/PEO-PPO-PEO poly-meric micelles are a single population of rounded micelles with a mean diameter of 170 nm and a z potential of 24.3 mV. Duodenal penetration of DNA/PEO-PPO-PEO polymeric micelles was evaluated in vitro by cal-culating the apparent permeability coefficient. The results showed a dose-independent penetration rate of (5.75 6 0.37) 3 1025cm/sec at low DNA concentrations (0.026–0.26 mg/ml), but a decrease to (2.89 6 0.37) 3 1025cm/sec at a concentration of 1.3 mg/ml. Furthermore, when 10 mM RGD peptide or 10 mM EDTA was administered before and concurrent with the administration of DNA/PEO-PPO-PEO polymeric micelles, trans-port was inhibited ([0.95 6 0.57] 3 1025 cm/sec) by blocking endocytosis or enhanced ([29.8 6 5.7] 3 1025 cm/sec) by opening tight junctions, respectively. After oral administration of six doses at 8-hr intervals, the highest expression of transferred gene lacZ was seen 48 hr after administration of the first dose, with gene expression detected in the villi, crypts, and goblet cells of the duodenum and in the crypt cells of the stom-ach. Reporter gene activity was seen in the duodenum, stomach, and liver. Activity was also seen in the brain and testis when mice were administered 10 mM EDTA before and concurrent with DNA/PEO-PPO-PEO poly-meric micelle administration. lacZ mRNA was detected in these five organs and in the blood by reverse tran-scription-polymerase chain reaction. Taken together, these results show efficient, stable gene transfer can be achieved in mice by oral delivery of PEO-PPO-PEO polymeric micelles.
481 OVERVIEW SUMMARY
Gene delivery through the gastrointestinal tract not only has many potential applications, but is also less invasive and more easily performed. PEO-PPO-PEO polymeric micelles were investigated to determine the feasibility of oral gene delivery and their usefulness as vectors for gene delivery in vivo. The stomach, duodenum, liver, and cir-culating blood are the primary targets for gene transfer by oral gene delivery using these nonionic polymeric mi-celles. Furthermore, gene transfer could be enhanced by
administration of 10 mM EDTA before the administra-tion of DNA/PEO-PPO-PEO polymeric micelles. Our re-sults further advance the promising therapeutic usage of PEO-PPO-PEO polymeric micelles for gene transfer by oral delivery.
INTRODUCTION
G
ENE DELIVERY through the gastrointestinal (GI) tract notonly has many potential applications, such as the
comple-1Graduate Institute of Cell and Molecular Biology, School of Medicine, Taipei Medical University, Taipei 110, Taiwan. 2School of Pharmacy, Taipei Medical University, Taipei 110, Taiwan.
3Research and Development Division, Precision Instrument Development Center, Hsinchu 30077, Taiwan.
mentation of single gene disorders (e.g., cystic fibrosis), but also has the advantages of ease of administration, a large sur-face area for targeting, and the accessibility to stem cells lo-cated in the GI tract (Alton et al., 1993). Nanotechnology and nanomaterials are being developed as advanced approaches to overcome some of the problems associated with oral viral gene delivery (Ledley, 1996; Croyle et al., 1998a; Mao et al., 2001; Cheng et al., 2003). Limitations in the use of nanoparticles as oral gene delivery carriers are that they must be absorbed from the GI tract and they must overcome the physiological condi-tions in the GI tract, such as extreme pH, enzyme activity, and GI motility. However, factors affecting the absorption of poly-meric nanoparticles in GI membranes are still not well under-stood. In particular, studies on the relationship between gene delivery and the physicochemical properties of the polymeric carrier, such as the stability of the complexes, their interaction with biomembranes, and the possible entry pathways of these polymeric complexes, are needed to reveal the application po-tential of these polymers (Florence, 1997; Kabanov et al., 2002). In addition, biocompatibility and biodegradability of the polymer complexes are another two factors requiring more con-sideration. New biomaterials able to overcome all these prob-lems would be of great interest for designing nanoparticulate carriers for mucosal oral gene delivery.
A nonionic PEO-PPO-PEO triblock copolymer has proved useful in the medical, pharmaceutical, and cosmetic fields (Saski, 1968; Guy and Hadgraft, 1987; Stolnik et al., 1995; Liaw and Lin, 2000; Kabanov et al., 2002). The PEO-PPO-PEO copolymer is resistant to adsorption by serum proteins, liver uptake, and enzyme degradation (Yalin et al., 1997; Ka-banov et al., 2002). Drug formulations containing these copolymer complexes were found to be more stable in the blood stream (Schmolka, 1991; Kabanov et al., 2002). Miyazaki et al. (1995) found that this type of nanopolymeric micelle showed good potential as a drug delivery system be-cause of the simplicity of micelle preparation and the ease of drug incorporation into the micelles. Further advantages of the nonionic PEO-PPO-PEO triblock copolymer are its increased drug-loading capacity, sustained systemic release, and in-creased half-life (Liaw and Lin, 2000). We have reported on the efficacy of gene delivery of topical eye drops formulations using DNA/PEO-PPO-PEO polymeric micelles in the pres-ence of interfering factors, such as tear mucin and degrada-tion enzymes (Liaw et al., 2001). Using these hydrophilic mi-celles, gene expression was increased by 28 and 38% in the eyes of nude mice and rabbits, respectively. Gene expression was further enhanced up to 70% by pretreatment with EDTA or cytochalasin B, which are thought to alter tight junctions in the cornea (Rojanasakul et al., 1990; Liaw and Robinson, 1992; Liaw et al., 2001). The aim of the present study was to explore the ability of nonionic PEO-PPO-PEO polymeric mi-celles to protect and deliver plasmid DNA orally in vivo. We assessed the expression of a reporter gene in GI tissues and examined gene distribution to other organs through the blood circulation. The in vitro duodenal penetration of nanosized DNA/PEO-PPO-PEO polymeric micelles was quantified with Hoechst H33258 DNA-staining dye. EDTA, a tight junction-opening enhancer (Rojanasakul et al., 1990; Liaw et al., 2001), and RGD peptide, a receptor-mediated endocytotic inhibitor (Croyle et al., 1998b), were used to modify in vitro gene trans-fer in the duodenum.
MATERIALS AND METHODS
Materials
PEO-PPO-PEO copolymer, with an average molecular mass of 8400 Da, was obtained from BASF (Ludwigshafen, Ger-many). EDTA and RGD peptide (arginine-glycine-aspartic acid) were obtained from Sigma (St. Louis, MO). RGE peptide (arginine-glycine-glutamic acid) was obtained from GIBCO-BRL (Rockville, MD). All other chemicals were of analytical reagent grade and were used without further purification.
Animals
Male nude mice (BALB/c-nu), 6 to 8 weeks old, were used for in vivo and in vitro oral delivery studies and were purchased from the National Laboratory Animal Breeding and Research Center (Taipei, Taiwan). They were maintained under specific pathogen-free conditions.
Plasmid DNA
The pCMV-lacZ plasmid, carrying the lacZ gene encoding
b-galactosidase (b-Gal) under the control of the cy-tomegalovirus (CMV) promoter, was used as the transferred gene. Plasmid DNA was amplified in Escherichia coli host strain TG-1 and purified by equilibrium centrifugation on a CsCl–ethidium bromide gradient (Macgregor and Caskey, 1989). Purity was determined with a QIAprep Spin column un-der endotoxin-free conditions. Stability of the plasmid DNA prepared was determined by electrophoresis on an agarose gel followed by ethidium bromide staining. DNA concentration was measured by ultraviolet (UV) absorption at 260 nm. Stability of plasmid DNA in DNA/PEO-PPO-PEO polymeric micelle formulations was determined as described previously by elec-trophoresis immediately after preparation, after 2 days of stor-age at room temperature, and after 3 freeze–thaw cycles (Liaw
et al., 2001).
Preparation of DNA/PEO-PPO-PEO polymeric micelles
All DNA/PEO-PPO-PEO polymeric micelles were freshly prepared on a weight percentage basis as described previously (Liaw et al., 2001). Polymeric micelles were formed with 6% (w/w) PEO-PPO-PEO in water. A pyrene fluorescence probe was used to determine the formation of micelles as previously reported (Liaw and Lin, 2000). Various concentrations of plas-mid DNA were gently mixed with PEO-PPO-PEO polymeric micelles in a vial for 2 hr at 25°C.
Size and z potential of DNA/PEO-PPO-PEO polymeric micelles
The size and z potential of complexes in mixtures contain-ing plasmid DNA (0.26 mg/ml) and a 6% PEO-PPO-PEO poly-mer solution were measured and compared with mixtures con-taining plasmid DNA (0.26 mg/ml) or 6% PEO-PPO-PEO polymer solution alone. The average particle size and z poten-tial of the polymeric micelles were determined by quasielastic laser dynamic light scattering (DLS) (Zetasizer 3000; Malvern Instruments, Malvern, UK), using an assumed refractive index ratio of 1.33 and a viscosity of 0.88 (Liaw et al., 2001). The
sampling time for each sample was 10 msec and the experi-mental duration was 100 sec. All measurements were performed at 25°C at a measurement angle of 90°.
Atomic force microscopy
Ten microliters of DNA/PEO-PPO-PEO polymeric micelles was placed on a mica surface with no further treatment. The AFM (NanoScope III; Digital Instruments/Veeco Metrology Group, Santa Barbara, CA) was operated in constant tapping mode, as described in the previous study (Liaw et al., 1999). The cantilevers were standard NanoProbe silicon single-crystal levers (125 mm); the constant force mode was used with a typ-ical scan frequency of 5 Hz. A scanner with a 1-mm scanning range was used, and all images were collected within a 1 3 1 or 3 3 3 mm2area. Unless otherwise stated, all images shown
were subjected only to the normal image processing of level-ing.
In vitro duodenal penetration studies
For the in vitro studies, nude mice were killed by cervical dislocation and upper duodenal sections, from the pylorus to 1 cm distal to the ligament of Treitz, were retrieved. Duodenal tissues were gently rinsed three times in warm phosphate-buffered saline (PBS) and then placed in an in vitro vertical dif-fusion apparatus (Liaw and Lin, 2000). A tissue surface area of 0.09 cm2was exposed to the donor and receiver compartments
of the diffusion apparatus, containing 0.5 and 6 ml of PBS, re-spectively. DNA/PEO-PPO-PEO polymeric micelles and other reagents were added to the donor compartment, and 0.5-ml ples were taken from the receiver compartment at various sam-pling times; the volume in the receiver compartment was main-tained by the addition of 0.5 ml of prewarmed PBS. The concentration of DNA penetrating through the duodenum in the presence or absence of 10 mM RGD, 10 mM RGE, or 10 mM EDTA was assayed with Hoechst H33258 dye (Molecular Probes, Eugene, OR). A F4500 fluorescence spectrophotome-ter (Hitachi, Tokyo, Japan) was set at a fixed excitation wave-length of 352 nm and fluorescence was determined at an emis-sion wavelength of 460 nm. The fluorescence of penetrating plasmids in the tissues was compared with a standard curve generated using plasmid DNA at 0.1 to 500 mg/ml with H33258 dye (Jong et al., 1997). The apparent permeability coefficient (P) was calculated according to the following equation as de-scribed previously (Liaw et al., 1999): P 5 (dC/dt)V/A 3 C0,
where V(dC/dt) is the steady state rate of DNA appearing in the receiver chamber after the initial lag time, C0is the initial
plas-mid concentration in the donor chamber, and A is the area of duodenal tissue exposed (0.09 cm2). Data from all experiments
were pooled to determine the mean and standard error of the mean. Analysis of variance (ANOVA) using Dunnett’s multi-ple comparison tests with a 95% confidence level determined the significance of differences between each group of experi-ments.
Oral gene transfer in vivo
For the in vivo studies, nude mice were fasted for 24 hr be-fore the experiments, but were allowed free access to water. Formulations were administered with a stomach feeding nee-dle for mice (KN-342; Natume Seisakusho, ). Six doses of each
polymeric micelle formulation (150 ml), formed from plasmid (0.26 mg/ml) and 6% (w/v) PEO-PPO-PEO, were administered at 8-hr intervals (6 A.M., 2 P.M., and 10 P.M.). Mice receiving
only plasmid DNA or polymeric micelles served as control groups. To evaluate gene transfer in vivo, mice were killed by cervical dislocation at set time points and all major organs and tissues were removed and processed immediately for individ-ual analysis. The duodenum was dissected from the pylorus to 1 cm distal to the ligament of Treitz. All tissues were gently rinsed three times in warm PBS, pH 7.4. In transfection en-hancer studies, 150 ml of enen-hancer (10 mM EDTA) was ministered orally 10 min before and concurrent with the ad-ministration of each dose of DNA/PEO-PPO-PEO polymeric micelles.
b-Galactosidase enzyme histochemistry
Organs and tissues were immersed for 5 min in ice-cold fix-ation solution (2% paraformaldehyde, 0.2% glutaraldehyde, and 5 mM EGTA in 100 mM sodium phosphate buffer, pH 7.6) and then rinsed in ice-cold PBS containing 5 mM EGTA. EGTA was added to these solutions to abolish endogenous b-Gal ac-tivity (Rosenberg et al., 1992; Sferra et al., 1997; Weiss et al., 1999). Two adjacent 1-cm sections were taken from the entire length of the duodenum and separated in paired series. One se-ries of sections was frozen in cryoprotective medium (O.C.T.; Sakura Finetek U.S.A., Torrance, CA) and the other series of sections was processed for en bloc b-Gal staining. Sections from brain, liver, stomach, and testis were paired and processed in the same way.
Cryosections (10 mm) of the O.C.T.-embedded organs were fixed for 5 min at 4°C in 4% paraformaldehyde and 0.2% glu-taraldehyde in 100 mM phosphate buffer, pH 7.4, washed twice in PBS, and stained for 24 hr at 37°C in 100 mM so-dium phosphate buffer, pH 7.4, containing 1.3 mM MgCl2, 3
mM K3Fe(CN)6, 3 mM K4(CN)6, and a 1-mg/ml
concentra-tion of the substrate 5-bromo-4-chloro-3-indolyl-b-D
-galtopyranoside (X-Gal; GIBCO-BRL, Grand Island, NY), ac-cording to the method of Oshima et al. (1998). Tissues to be stained en bloc were fixed for 90 min at 4°C in 4% parafor-maldehyde and then incubated for 24 hr at 37°C with PBS containing 10 mM K4Fe(CN)6, 10 mM K3Fe(CN)6, 0.01%
so-dium deoxycholate, 0.02% Nonidet P-40 (NP-40), 2 mM MgCl2, and X-Gal (1 mg/ml). The lacZ gene was considered
to be expressed when the tissue was blue-green in color un-der an operating microscope at magnifications of 340 and 3100, unless specifically mentioned otherwise. In addition, cryosections (10 mm) were stained with hematoxylin–eosin for pathological assessment.
Preparation of tissue extracts and determination of transgene expression
b-Gal expression was quantified with the enzyme substrate chlorophenol red-b-D-galactopyranoside (CPRG, 1 mg/ml;
Gene Therapy Systems, San Diego, CA). Induced color devel-opment was measured at 580 nm as previously described (Liaw
et al., 2001). Total tissue protein was measured with a DC
pro-tein assay reagent kit (Bio-Rad, Hercules, CA) and used to nor-malize the b-Gal activity for each sample. Statistical compar-isons were determined by ANOVA (Dunnett’s multiple comparison tests) with a 95% confidence level.
Polymerase chain reaction analysis of
lacZ gene mRNA
Forty-eight hours after the first oral dosing of DNA/PEO-PPO-PEO polymeric micelles, total RNA was extracted from the duodenum, stomach, blood, brain, liver, and testis with TRI-zol reagent (Invitrogen Life Technologies, Carlsbad, CA) ac-cording to the manufacturer’s instructions. cDNA was prepared with SuperScript II reverse transcriptase (Invitrogen Life Tech-nologies) and stored at 220°C. A 249-bp segment of the lacZ gene was amplified by polymerase chain reaction (PCR) as de-scribed below. Sequences of the primer pair for amplification of the transferred b-Gal-encoding gene were 59-CTA CAC CAA CGT AAC CTA TCC C-39 and 59-TCC TCC GGC GCG TAA AAA TGC G-39 (Hanazono et al., 1999). The level of housekeeping gene b-actin expression was analyzed and used to demonstrate the presence of the same amount of total cDNA in each RNA sample. Sequences of the primer pair for ampli-fication of a 659-bp segment of the b-actin gene were 59-CTA GAA GCA TTG CGG TGG ACG ATG GAG GG-39 and 59-TGA CGG GGT CAC CCA CAC TGT GCC CAT CTA-39
(Hanazono et al., 1999). PCRs included 32 cycles of denatura-tion for 1 min at 94°C, annealing for 1 min at 56°C, and ex-tension for 2 min at 72°C. PCR products were loaded onto a 0.8% acrylamide gel for electrophoresis and then the DNA bands were visualized under UV light. Quantification of band intensities was performed with a Kodak EDAS290 Analysis system (Kodak Scientific Imaging Systems, New Haven, CT).
RESULTS
Characterization of DNA/PEO-PPO-PEO polymeric micelles
Using DLS for particle size measurement, PEO-PPO-PEO polymers at a concentration of 6% (w/w) were found to exist as a single population of particles with a mean diameter of 165 6 19 nm (Table 1). In the presence of PEO-PPO-PEO polymers, plasmid DNA carrying the lacZ gene condensed from a size of 403 6 27 to 170 6 20 nm; the condensed DNA size is similar to that of polymeric micelles alone (Table 1). AFM was used to vi-TABLE1. SIZE ANDzPOTENTIAL OFDNA/PEO-PPO-PEO POLYMERIC MICELLESa
Size AFM z potential
Formulation (nm) (nm) (mV)
Plasmid DNAb 403 6 27 — .220 6 1.3
PEO-PPO-PEO polymeric micellesc 165 6 19 160 (100–200)d 24.5 6 2.3
DNA/PEO-PPO-PEO polymeric micellese 170 6 20 170 (88–236)d0 24.3 6 1.7 aResults are expressed as the mean and standard deviation for five experiments. bPlasmid DNA (0.26 mg/ml).
cPEO-PPO-PEO polymers (6%).
dValues in parentheses represent the range of particle sizes measured by AFM. ePlasmid DNA (0.26 mg/ml) plus 6% PEO-PPO-PEO polymeric micelles.
FIG. 1. Size distribution of DNA/PEO-PPO-PEO polymeric micelles. (a) AFM images of 10 ml of PEO-PPO-PEO polymeric
micelles (6%) with plasmid solution (0.26 mg/ml) on a mica surface (the size section analysis of particles 1, 2, and 3 are 164, 236, and 88 nm, respectively); (b) AFM images of plasmid alone.
sualize the morphology of DNA/PEO-PPO-PEO polymeric mi-celles, and the results showed that the complexes readily attached and remained bounded to mica and could be imaged with an AFM tip (Fig. 1a). DNA/PEO-PPO-PEO polymeric micelles appear to exist as a single population of round micelles with diameters rang-ing from 88 to 236 nm and an average diameter of 170 nm (Table 1). Images of plasmid DNA alone on mica are shown in Fig. 1b. The measured surface charge was 220 6 1.3 mV for plasmid DNA alone, 24.5 6 2.3 mV for PEO-PPO-PEO polymeric mi-celles alone, and 24.3 6 1.7 mV for DNA/PEO-PPO-PEO poly-meric micelles (Table 1).
In vitro nude mouse duodenal penetration of
DNA/PEO-PPO-PEO polymeric micelles
To evaluate mechanisms involved in the penetration of DNA/PEO-PPO-PEO polymeric micelles through duodenal tis-sues, in vitro penetration studies were performed. Apparent per-meability coefficients (P) for DNA/PEO-PPO-PEO polymeric micelles formulated with various concentrations of plasmid DNA are shown in Table 2. No significant differences were seen for formulated micelles with plasmid DNA at 0.026, 0.13, or 0.26 mg/ml. However, the P value decreased when the DNA TABLE2. EFFECT OFDNA CONCENTRATION WITH 6% PEO-PPO-PEO POLYMERIC MICELLES ON
In Vitro DUODENAL PERMEABILITY IN NUDEMICEa
Apparent permeability coefficient (P)
Treatment (31025cm/sec)b
DNA concentration with PEO-PPO-PEO
0.026 mg/ml 05.12 6 0.98
0.13 mg/ml 05.52 6 1.31
0.26 mg/ml 05.75 6 0.37
1.30 mg/ml 02.89 6 0.37c
10 mM RGD 1 DNA (0.26 mg/ml) with PEO-PPO-PEOd 00.95 6 0.57c
10 mM RGE 1 DNA (0.26 mg/ml) with PEO-PPO-PEOd 05.60 6 0.62
10 mM EDTA 1 DNA (0.26 mg/ml) with PEO-PPO-PEOe 29.81 6 5.73f an 5 9.
bValues are expressed as means 6 SEM.
cDenotes a statistically significant decrease (p , 0.01) compared with DNA (0.26 mg/ml) with
polymeric micelles.
dRGD or RGE (10 mM) was added to the donor compartment 10 min before addition of DNA
(0.26 mg/ml) with polymeric micelles.
eEDTA (10 mM) was added to the donor compartment 10 min before addition of DNA (0.26
mg/ml) with polymeric micelles.
fDenotes a statistically significant increase (p , 0.01) compared with a DNA concentration of
0.26 mg/ml.
TABLE3. b-GALACTOSIDASE ACTIVITYAFTER ORALADMINISTRATION OFREPORTERVECTOR WITH PEO-PPO-PEO POLYMERIC MICELLES INNUDE MICEa
b-Gal activityb
Treatment (mU/mg protein)
Polymeric micelles alone 5.05 6 0.26 (n 5 12)c
DNA alone 5.89 6 0.55 (n 5 12)c
DNA/polymeric micellesa
One dose (48 hr) 5.89 6 0.35 (n 5 12) Three doses (three times per day for 1 day, 12 hr) 6.59 6 0.18 (n 5 12)d
Three doses (three times per day for 1 day, 48 hr) 7.10 6 0.35 (n 5 12)d
Six doses (three times per day for 2 days, 48 hr) 7.90 6 0.50 (n 5 12)d
Six doses (three times per day for 2 days, 72 hr) 6.20 6 0.48 (n 5 12)d,e
Six doses (three times per day for 2 days, 96 hr) 5.78 6 0.18 (n 5 12) 10 mM EDTA 1 DNA/polymeric micelles 9.00 6 1.15 (n 5 10)d
aSix- to eight-week-old BALB/c nude mice were administered one, three, or six doses of
plas-mid (0.26 mg/ml) plus 6% polymeric micelles over the indicated time period (days); then, at the indicated time (hr) after the start of treatment, the duodenum was removed and homogenized, and its b-Gal activity was measured.
bValues are expressed as means 6 SEM. Numbers in parentheses are the numbers of specimens
examined.
cNo significant difference between polymeric micelles alone and DNA alone (p . 0.1). dSignificant (p , 0.05) increase compared with DNA-only mice.
concentration was 1.30 mg/ml, indicating mediated and satu-rated penetration processes for the DNA/PEO-PPO-PEO poly-meric micelles formulation. P values of highly penetrating com-pounds, such as phenol red, in the GI membrane are on the order of 10–5cm/sec, whereas the P value of insulin is lower,
on the order of 10–7cm/sec (Schilling and Mitra, 1990). These
data suggest that the penetration of DNA/PEO-PPO-PEO poly-meric micelles across the GI membrane was relatively high.
To investigate the subsequent internalization of DNA/PEO-PPO-PEO polymeric micelles after entry into the duodenal membrane, RGD peptide was used to block the clathrin-coated vesicle pathway (Croyle et al., 1998b). As shown in Table 2, pretreatment with 10 mM RGD peptide for 10 min before and concurrent with polymeric micelle administration resulted in a significant 6-fold decrease in the P value of DNA/PEO-PPO-PEO polymeric micelles. However, administration of a 10 mM concentration of the analog RGE, in which aspartic acid was replaced by glutamic acid, in place of RGD, did not affect trans-port in the duodenum. In contrast, the P value increased by about 5- to 6-fold when the tight junction-opening reagent EDTA was used before and concurrent with polymeric micelle administration.
In vivo oral gene transfer
An intubation procedure was used to administer DNA/PEO-PPO-PEO polymeric micelles directly to the stomach of nude mice and to assess quantitatively the efficiency of gene trans-fer. b-Gal activity in the duodenum with various numbers of doses (one, three, or six) after administration of DNA (0.26
mg/ml) with PEO-PPO-PEO polymeric micelles was measured 48 hr after the first dose. In mice administered three doses at 8-hr intervals, we detected significant enhancement of trans-gene expression in duodenal tissue both 12 and 48 hr after ini-tial dosing (Table 3). However, in mice administered only a sin-gle dose, no significant transgene expression in the duodenum was seen 48 hr postadministration (Table 3). Results showed that transfer of the lacZ gene could be achieved in the duode-num, and that b-Gal activity in the duodenum was most sig-nificantly increased (by 32%) when six doses were adminis-tered, compared with the control groups. There was no significant difference between the two control groups receiving six doses of either plasmid DNA alone or polymeric micelles alone (p . 0.1), further indicating that the detected b-Gal ac-tivity stated above was from plasmid DNA transferred by the polymeric carrier. In addition, we investigated the duration of
lacZ expression in the duodenum after administration of six
doses of DNA/PEO-PPO-PEO polymeric micelles by measur-ing b-Gal activity 48, 72, and 96 hr postadministration of the first dose. Transgene expression at 72 hr was lower than that detected at 48 hr and returned to background levels by 96 hr.
To examine the effect of EDTA on copolymer complex-medi-ated gene transfer, 150-ml doses of 10 mM EDTA were ad-ministered 10 min before and concurrent with administration of the micelle formulation for each of the six doses of DNA/PEO-PPO-PEO polymeric micelles. Transgene expression at 48 hr was significantly increased (by 52%) compared with that in mice administered DNA alone, as shown in Table 3.
Six doses were administered at 8-hr intervals (at 6 A.M.,
2 P.M., and 10 P.M.), and expression of the delivered lacZ gene
was evaluated 48 hr after administration of the first dose, us-ing histological tissue sections with b-Gal substrate or whole-mount staining with X-Gal. In mice administered DNA/PEO-PPO-PEO polymeric micelles orally, a blue–green color, indicating expression of exogenous b-Gal, was detected mainly in the epithelium of the duodenal villi (Fig. 2D–F). Most of the transgene expression was localized in the duodenal epithelium (Fig. 2G–K); the intensity of staining was much stronger in the villi and the crypts (Fig. 2J and K). EGTA treatment was used to block endogenous b-Gal activity (Rosenberg et al., 1992; Sferra et al., 1997; Weiss et al., 1999) (see b-Galactosidase En-zyme Histochemistry, above), and no background staining was seen in control tissues (Fig. 2A–C and L–O). In addition, no signs of toxicity or inflammatory response to the delivered poly-meric micelles were evident by microscopy in any duodenal specimens examined during the entire study period; this obser-vation was supported by analysis of hematoxylin and eosin-stained tissue sections (Fig. 2M and N).
Biodistribution of b-Gal in other tissues
In this study, a delivery regimen of six doses administered at 8-hr intervals was used, because transgene expression in duodenal tissues was highest with this dosing regimen. Rep-resentative examples of X-Gal staining of other major tissues of the nude mouse after oral gene delivery using DNA/PEO-PPO-PEO polymeric micelles are presented in Fig. 3, and a summary of the quantitative data for the amount of gene trans-fection is given in Table 4. b-Gal activity in the stomach ex-tract showed a significant increase from 0.46 6 0.09 mU/mg protein in the control group to 0.53 6 0.06 mU/mg protein in the group receiving DNA/PEO-PPO-PEO polymeric micelles (p , 0.05), and a further increase to 0.66 6 0.05 mU/mg pro-tein when using EDTA as a penetration enhancer. A similar pattern of increased b-Gal activity was seen in the liver. Sig-nificant increases in transgene expression were seen in the brain and the testis (95 and 21%, respectively) only after EDTA treatment. Compared with the evenly distributed pat-tern of b-Gal expression in the duodenum, the small, local-ized areas of blue–green color in the stomach, brain, and testis indicated only low amounts of b-galactosidase activity in these tissues. No increases in b-Gal activity were observed
FIG. 2. Whole-mount mucosa of duodenal tissue analysis (A–F) and histological analysis of the duodenum (G–O) of nude
mice 48 hr after the start of oral delivery of six 0.26-mg/ml doses of plasmid/PEO-PPO-PEO polymeric micelles (PPM) admin-istered at 8-hr intervals [(A–C) delivery plasmid alone; (D–K) plasmid/PPM; (L) delivery plasmid alone; (O) delivery polymeric micelles (PM) alone]. Note the speckled blue–green staining corresponding to exogenous b-Gal activity in the duodenal tissues (villi [V] and crypt cells [cr] are indicated by asterisks and arrows, respectively). Light micrographs of 20-mm O.C.T. cryosec-tions of duodenum counterstained with eosin (E) (I–K) or hematoxylin–eosin (H/E) (M and N). No inflammatory reaction was noted after DNA/PPM delivery. Original magnification is indicated in parentheses in each panel.
with or without EDTA treatments in other tissues, such as the heart, lung, spleen, and kidney.
Reverse transcription-polymerase chain reaction for RNA analysis
Transcription of the lacZ transgene was further analyzed by reverse transcription-polymerase chain reaction (RT-PCR) of total duodenal RNA extracts from mice administered six doses of DNA/PEO-PPO-PEO polymeric micelles; b-actin expression served as internal control. Total RNA was extracted from duo-denal samples 48 hr after administration of the first dose and a 249-bp DNA fragment of the lacZ gene amplified by RT-PCR is shown in Fig. 4. Interestingly, lacZ mRNA was detectable in samples from blood, brain, stomach, liver, and testis tissues at 48 hr with or without EDTA pretreatment (Fig. 4). No PCR product was detected when using cDNA from tissues of the plasmid DNA-treated control group. Direct PCR analysis of the RNA samples did not produce the 249-bp band, showing that DNA contamination of RNA extracts was not responsible for the appearance of the 249-bp band (data not shown).
DISCUSSION
It is generally believed that certain features of the polymeric carrier system, including size, surface charge, and hydrophilic-ity of the carriers, affect intestinal absorption characteristics (Jong et al., 1997). Jani et al. (1990) observed that particles with mean diameters of 50–100 nm showed a higher uptake in rat intestine than did larger particles. Rejman et al. (2004) also revealed that the mechanism by which the beads were inter-nalized, and their subsequent intracellular routing, were strongly dependent on the particle size of the carrier system. Internalization of microspheres with a diameter less than 200 nm involved clathrin-coated pits. With increasing size, a shift to a mechanism that relied on caveola-mediated internalization became apparent; caveola-mediated internalization is the pre-dominant pathway of entry for particles with a size of 500 nm. In our study, the diameter of DNA/PEO-PPO-PEO polymeric micelles ranged from 88 to 236 nm, which is comparable to that reported by Jeong and Park (2001) for DNA/poly(D,L
-lac-tic-co-glycolic acid) copolymer complexes, as determined by AFM. The condensation of DNA from 400 to 170 nm on in-teraction with the polymeric micelles indicates that this formu-lation may generate more compact particles for endocytosis, thus making it more resistant to enzymatic digestion and to the pH of the GI tract, as well as more sensitive to influence by junctional penetration enhancer.
Particles with less hydrophobic surfaces show higher uptake in the GI tract than do particles with more hydrophobic sur-faces (Norris and Sinko, 1997). These authors found that, in contrast to a more hydrophilic polymeric carrier system, hy-drophobic poly(styrene) beads showed poor mucus penetration. The hydrophilic polymeric micelles we used in this study, which had almost no z potential (24 mV), enhanced gene transfer in
vivo. These surface properties may partially explain the ability
of EDTA to increase the ability of these hydrophilic, nanosized particles to penetrate the duodenal epithelium and distribute to other tissues (Tables 2–4), because such particles would be dis-tributed into the extracellular space and be able to pass around cells. This process would be aided by the opening of paracel-lular junctions. Furthermore, it has been hypothesized that the addition of hydrophilic PEO polymers to liposome or polymers can prevent complexes from binding to protein/serum or from being degraded by digestive enzymes (Kwon et al., 1994; Torchilin, 2001). Mathiowitz et al. (1997) and Luessen et al. (1996) showed that surface charge seems to be a prerequisite structural factor for colloidal carriers, leading to significant bioadhesion to the mucosa of the small intestine. However, the mechanisms of gene transfer by PEO-PPO-PEO polymeric mi-celles under physiological conditions seen in the GI tract are still under examination in our laboratory.
b-Gal activity in duodenal extracts increased 32% over con-trols when using the PEO-PPO-PEO polymeric micelles trans-fer system. Foreman et al. (1998) obtained a similar enhance-ment of 20% after direct injection of the b-Gal gene into the duodenum. However, Cheng et al. (1997) demonstrated an ap-proximately 9-fold increase in transfection when using an ade-noviral vector to deliver the b-Gal gene into the duodenum; this higher transfection efficiency could be partially due to their combination use of gentamicin, ranitidine, prednisolone, and glucagons before inserting the oral feeding tube, which reduced the bacterial burden, reduced secretion of acid fluid from stom-ach, lowered the intestinal mucous barrier, and slowed bowel motility in the duodenum. We found unacceptable levels of b-Gal background activities in the jejunum, the ileum, and the colon of our experimental mice even after EGTA incubation. This endogenous b-Gal activity in the intestine has been ob-served in other studies (Sferra et al., 1997; Foreman et al., 1998; Weiss et al., 1999). A study by Foreman et al. (1998) indicated that potential sources of this background activity included mam-malian lysosomal enzymes, the digestive enzyme lactase, and
b-Gal produced by enteric bacteria. We therefore focused our study on duodenal sections of the GI tract.
PEO-PPO-PEO copolymers have long been used as compo-nents of various pharmaceutical formulations to enhance the oral bioavailability of drugs (Saski, 1968; Schubert and Wretlin,
FIG. 3. (A–D) Whole-mount and (E–G) histological analysis of brain, stomach, liver, and testis 48 hr after the start of oral
de-livery of six 0.26-mg/ml doses of plasmid/PEO-PPO-PEO polymeric micelles administered at 8-hr intervals. [(B and D) plasmid alone or polymeric micelles alone; (E) stained with hematoxylin/eosin (H/E); (G) counterstained with eosin (E); (A, C, F, and
G) treatment with 10 mM EDTA before and during DNA/PEO-PPO-PEO polymeric micelle delivery]. Note the speckled
blue–green staining corresponding to X-Gal activity (brain in the cortex area [co]; stomach in the chief cells of the gastric mu-cosa [m]; liver in hepatocytes [h]; testis in Leydig’s cells [l] are indicated by arrows). No inflammatory reaction was noted af-ter DNA/PEO-PPO-PEO polymeric micelle delivery. Original magnification is indicated in parentheses in each panel. Whole mount original magnification in (A–D), 310.
1983; Damage et al., 1997; Yang et al., 1999). However, their effects varied when different concentrations of the copolymer were used. First, at concentrations below the critical micelle concentration (CMC), the copolymers perturbed the plasma membrane and inhibited the P-glycoprotein efflux system in Caco-2 cell monolayers, resulting in a significant enhancement of absorption and permeability of drugs (Batrakova et al., 1999). Second, at concentrations above the CMC, block copolymers self-assembled into micelles and increased the P value of var-ious compounds in Caco-2 cell monolayers to 0.2–3.5 3 10–5
cm/sec (Batrakova et al., 1999). This range of P values is sim-ilar to that seen in our study (,10–5cm/sec). Taken together,
these results suggest that this copolymer system should be use-ful in increasing the oral bioavailability of plasmid DNA in the duodenum.
Improved transfection efficiency was achieved both in vivo and in vitro by the use of EDTA, which is known to open tight junctions of the duodenum and increase paracellular transport,
allowing hydrophilic polymeric micelles containing plasmid DNA to be distributed to deeper levels of tissues (Zhang et al., 2003). This is similar to our previous results in which we en-hanced transfection of ocular tissues by using EDTA and cy-tochalasin B (Rojanasakul et al., 1990; Liaw and Robinson, 1992; Liaw et al., 2001). However, it is also known that in-creased intracellular calcium levels lead to an inin-creased rate of uptake of complexes (Yuan et al., 2001), so the mechanism of enhancement by EDTA is not totally clear. We hypothesize that EDTA causes an increase in intracellular calcium levels, trig-gering endocytosis and the opening of tight junctions, thus al-lowing the complexes to penetrate into deeper tissues and dis-tribute to other organs. However, our data also showed that, after treatment of the duodenum in vitro with RGD peptide be-fore and concurrent with polymeric micelle administration, gene transfer efficiency using DNA/PEO-PPO-PEO polymeric mi-celles was reduced by 84%, implying the entry of particles into cells via receptor-mediated endocytosis. Croyle et al. (1998b) TABLE 4. b-GALACTOSIDSEACTIVITY IN OTHERMAJOR TISSUES OFNUDE MICEAFTERORALADMINISTRATION OF
REPORTERVECTOR WITHPEO-PPO-PEO POLYMERICMICELLES
b-Gal activitya(mU/mg protein)
Organ Plasmid control Oral deliveryb EDTA 1 oral deliveryc
Stomach 0.46 6 0.09 (n 5 7) 0.53 6 0.06d(n 5 7) 0.66 6 0.05d(n 5 7) Liver 0.37 6 0.03 (n 5 6) 0.45 6 0.04d(n 5 6) 0.42 6 0.04d(n 5 6) Brain 0.27 6 0.07 (n 5 8) 0.33 6 0.01 (n 5 8) 0.52 6 0.11d(n 5 8) Testis 0.63 6 0.03 (n 5 7) 0.66 6 0.05 (n 5 8) 0.76 6 0.06d(n 5 8) Heart 0.15 6 0.03 (n 5 8) 0.15 6 0.02 (n 5 8) 0.16 6 0.02 (n 5 8) Lung 0.40 6 0.04 (n 5 8) 0.38 6 0.03 (n 5 8) 0.44 6 0.02 (n 5 6) Spleen 1.33 6 0.18 (n 5 8) 1.33 6 0.18 (n 5 8) 1.39 6 0.12 (n 5 8) Kidney 1.23 6 0.16 (n 5 7) 1.19 6 0.12 (n 5 7) 1.20 6 0.10 (n 5 7)
aValues are expressed as means 6 SEM. Numbers in parentheses are the numbers of specimens examined.
bSix- to eight-week-old BALB/c nude mice received six doses of plasmid (0.26 mg/ml) with micelles and then, 48 hr after the
start of treatment, organs were removed and homogenized, and b-gal activity was measured.
cMice were treated with 10 mM EDTA before and during treatment, as described in footnote b.
dSignificant (p , 0.05) increase in activity compared with that in the same organ in plasmid-only control mice.
FIG. 4. RT-PCR analysis of lacZ gene transcription. Total RNA was prepared from duodenum (D), blood (B), brain (Br),
stom-ach (S), liver (L), and testis (T) of animals administered plasmid alone (1) or DNA/polymeric micelles (2), or treated with 10 mM EDTA before and during DNA/polymeric micelle administration (3). The samples were amplified using a 249-bp sequence of the lacZ gene and a 659-bp sequence of the b-actin gene. A typical direct PCR analysis of the RNA samples (lane 0) did not result in production of the 249-bp band.
reported that pretreatment of Caco-2 cells with 3.5 mM RGD peptide reduced adenovirus-mediated gene transfer by 80%. Rejman et al. (2004) also demonstrated that internalization of microspheres with a diameter less than 200 nm involves mainly a clathrin-coated pit-mediated mechanism. It has been demon-strated that PEO polymeric micelles were taken up into cells via a receptor-mediated endocytotic mechanism that can be abolished by low temperatures (Batrakova et al., 1999; Zhang
et al., 2003). Thus, taken together, our data suggest that the
transport of PEO-PPO-PEO copolymer micelles into tissues may be through endocytosis and could be enhanced by the open-ing of tight junctions.
We found that X-Gal staining of epithelial cells was local-ized to the cytoplasm of intestinal enterocytes. Under higher magnification, an apical duodenal localization of exogenous b-Gal activity was observed in the region of the villi. The find-ing that the majority of goblet cells showed positive stainfind-ing, also seen in the crypts, where stem cells are located, is similar to other reports ( Lau et al., 1995; Cheng et al., 2003) using a viral delivery system. In contrast to retroviral vectors (Lau et
al., 1995), plasmid-containing polymeric micelles cannot insert
their DNA into host chromosomal DNA; hence the crypt stem cells of the villi may not be able to express the delivered b-Gal reporter gene because of loss of episomal genetic material af-ter continuous cell division. Almost all enaf-terocytes in the crypt–villus axis were b-Gal positive, which is consistent with the migration of crypt cells to the villi by 48 hr (Clatworthy and Subramanian, 2001), suggesting that crypt cells are the pos-sible target cells for plasmid DNA delivered by polymeric mi-celles. However, we cannot rule out the possibility that the X-Gal-stained epithelial cells in the villi were cells derived from crypt stem cells harboring the transferred gene delivered by polymeric micelles.
After oral administration of DNA/PEO-PPO-PEO polymeric micelles, b-Gal activity in the duodenum was maximal at 48 hr and returned to background levels by 96 hr. This finding fur-ther suggests that rapid turnover of intestinal epithelial cells limits the time that the micelles remain in the tissue and thus the expression of the delivered gene. However, it is unclear whether the time when reporter gene expression started to de-crease was affected by mucus, enzymes, pH, and bacterial com-ponents in the duodenum, resulting in a decrease in transfec-tion efficiency. Furthermore, because of the extremely low gene transfer efficiency at time points beyond 96 hr, it was difficult to tell whether the trace signal observed was indeed due to the gene product. Possible explanations for this transient gene ex-pression could be degradation of the delivered gene, shutdown of gene expression, or loss of transfected cells.
It is interesting that transfection was detected in the circu-lating blood 48 hr after oral delivery of polymeric micelles. Fur-thermore, we detected positive RT-PCR signals and protein ex-pression in the liver, stomach, brain, and testis. These results are similar to those obtained in studies on the tissue biodistri-bution of genes transferred by adenoviral vectors after local in
vivo gene transfer to the arterial wall, which showed that, after
transfection, X-Gal staining of blood cells was increased by 1.8%, whereas that of the liver and testis was increased by 0.7 and 0.06%, respectively (Hiltunen et al., 2000; Zhang et al., 2003). In addition, after a single intravenous injection of
plas-mid packaged in the interior of PEO liposomes, Shi et al. (2001) detected expression of the exogenous gene in the brain, liver, and spleen that lasted for at least 6 days. Kabanov et al. (2002) reported that these PEO-PPO-PEO formulations not only re-mained stable at concentrations above the CMC in the blood stream, but also showed 3-fold higher stability in the brain and better oral bioavailability for various drugs. This further em-phasizes that it should be possible to locally transfect the stom-ach and duodenum and other major organs through the blood circulation, using oral gene delivery with PEO-PPO-PEO poly-meric micelles.
In summary, the present study demonstrates the feasibility of gene transfer into duodenal epithelial cells, using orally ad-ministered PEO-PPO-PEO polymeric micelles containing the
lacZ gene. Expression of the b-Gal gene in the duodenal
ep-ithelium peaked at 48 hr and fell to background levels by 96 hr after polymeric micelle administration. Use of EDTA and RGD peptide to modify the transfection efficiency of DNA/PEO-PPO-PEO polymeric micelles suggested that the transfection mechanism involved endocytosis and was enhanced by the opening of tight junctions in the duodenal tissues and by aug-menting the paracellular pathway, allowing distribution to deeper tissues and more distant organs.
ACKNOWLEDGMENTS
This work was supported by grants from the National Sci-ence Council of Taiwan (NSC92-2320-B038-035) and Taipei Medical University (SKH-TMU-92-25).
REFERENCES
ALTON, E.W., MIDDLETON, P.G., CAPLEN, N.J., SMITH, S.N., STEEL, D.M., MUNKONGE, F.M., JEFFERY, P.K., GEDDES, D.M., HART, S.L., WILLIAMSON, R., et al. (1993). Non-invasive liposome-mediated gene delivery can correct the ion transport defect in cystic fibrosis mutant mice. Nat. Genet. 5, 135–142.
BATRAKOVA, E.V., LI, S., MILLER, D.W., and KABANOV, A.V. (1999). Pluronic P85 increases permeability of a broad spectrum of drugs in polarized BBMEC and Caco-2 cell monolayers. Pharm. Res. 16, 1366–1372.
CHENG, D.Y., KOLLS, J.K., LEI, D., and NOEL, R.A. (1997). In vivo and in vitro gene transfer and expression in rat intestinal epithelial cells by E1-deleted adenoviral vector. Hum. Gene Ther. 8, 755–764. CHENG, X., MING, X., and CROYLE, M.A. (2003). PEGylated ade-noviruses for gene delivery to the intestinal epithelium by the oral route. Pharm. Res. 20, 1444–1451.
CLATWORTHY, J.P., and SUBRAMANIAN, V. (2001). Stem cells and the regulation of proliferation, differentiation and patterning in the intestinal epithelium: Emerging insights from gene expression patterns, transgenic and gene ablation studies. Mech. Dev. 101, 3–9. CROYLE, M.A., ANDERSON, D.J., ROESSLER, B.J., and AMIDON, G.L. (1998a). Development of a highly efficient purification process for recombinant adenoviral vectors for oral gene delivery. Pharm. Dev. Technol. 3, 365–372.
CROYLE, M.A., WALTER, E., JANICH, S., ROESSLER, B.J., and AMIDON, G.L. (1998b). Role of integrin expression in adenovirus-mediated gene delivery to the intestinal epithelium. Hum. Gene Ther. 9, 561–573.
DAMAGE, C., VRANXKX, H., BALSCHMIDT, P., and COU-VREUR, P. (1997). Poly(alkyl cyanoacrylate) nanospheres for oral administration of insulin. J. Pharm. Sci. 86, 1403–1407.
FLORENCE, A.T. (1997). The oral absorption of micro- and nanopar-ticles: Neither exceptional nor unusual. Pharm. Res. 14, 259– 266.
FOREMAN, P.K., WAINWRIGHT, M.J., ALICKE, B., KOVESDI, I., WICKHAM, T.J., SMITH, J.G., MEIER-DAVIS, S., FIX, J.A., DADDONA, P., GARDNER, P., and HUANG, M.T. (1998). Ade-novirus-mediated transduction of intestinal cells in vivo. Hum. Gene Ther. 9, 1313–1321.
GUY, R.H., and HADGRAFT, J. (1987). Transdermal drug delivery: A perspective. J. Control. Release 4, 237–251.
HANAZONO, Y., BROWN, K.E., HANDA, A., METZGER, M.E., HEIM, D., KURTZMAN, G.J., DONAHUE, R.E., and DUNBAR, C.E. (1999). In vivo marking of rhesus monkey lymphocytes by adeno-associated viral vectors: Direct comparison with retroviral vectors. Gene Ther. 94, 2263–2270.
HILTUNEN, M.O., TURUNEN, M.P., TURUNEN, A., RISSANEN, T.T., LAITNEN, M., KOSMA, V., and YLA-HERTTUALA, S. (2000). Biodistribution of adenoviral vector to nontarget tissues after local in vivo gene transfer to arterial wall using intravascu-lar and periadventitial gene delivery methods. FASEB J. 14, 2230–2236.
JANI, P.U., HALBERT, G.W., LANGRIDGE, J., and FLORENCE, A.T. (1990). Nanoparticle uptake by the rat gastrointestinal mucosa: Quantitation and particle size dependency. J. Pharm. Pharmacol. 42, 821–826.
JEONG, J.H., and PARK, T.G. (2001). Novel polymer–DNA hybrid polymeric micelles composed of hydrophobic poly(D,L
-lactic-co-gly-colic acid) and hydrophilic oligonucleotides. Bioconjug. Chem. 12, 917–923.
JONG, Y.S., JACOB, J.S., YIP, K.P., GARDNER, G., SEITELMAN, E., WHITNEY, M., MONTGOMERY, S., and MATHIOWITZ, E. (1997). Controlled release of plasmid DNA. J. Control. Release 47, 123–134.
KABANOV, A.V., BATRAKOVA, E.V., and ALAKHOV, V.Y. (2002). Pluronic block copolymers as novel polymer therapeutics for drug and gene delivery. J. Control. Release 82, 189–212. KWON, G., SUWA, S., YOKOYAMA, M., OKANO, T., SAKURAI,
Y., and KATAOKA, K. (1994). Enhanced tumor accumulation and prolonged circulation times of micelle-forming poly(ethylene oxide-aspartate) block copolymer–Adriamycin conjugates. J. Control. Re-lease 29, 17–23.
LAU, C., SORIANO, H.E., LEDLEY, F.D., FINEGOLD, J.M., WOLFE, J.H., BIRKENMEIER, E.H., and HENNING, S.J. (1995). Retroviral gene transfer into the intestinal epithelium. Hum. Gene Ther. 6, 1145–1151.
LEDLEY, F.D. (1996). Pharmaceutical approach to somatic gene ther-apy. Pharm. Res. 13, 1595–1614.
LIAW, J., and LIN, Y.C. (2000). Evaluation of poly(ethylene oxide)–poly(propylene oxide)–poly(ethylene oxide) gels as a re-lease vehicle for percutaneous fentanyl. J. Control. Rere-lease 68, 273–282.
LIAW, J., and ROBINSON, J.R. (1992). The effect of polyethylene glycol molecular weight on corneal transport and the related influ-ence of penetration enhancers. Int. J. Pharm. 88, 125–140. LIAW, J., AOYAGI, T., KATAOKA, K., SAKURAI, Y., and OKANO,
T. (1999). Permeation of PEO-PBLA-FITC polymeric micelles in aortic endothelial cells. Pharm. Res. 16, 213–220.
LIAW, J., CHANG, S.F., and HSIAO, F.C. (2001). In vivo gene de-livery into ocular tissues by eye drops of poly(ethylene oxide)– poly(propylene oxide)–poly(ethylene oxide) polymeric micelles. Gene Ther. 8, 999–1004.
LUESSEN, H.L., DE LEEUW, B.J., LANGEMEYER, M.W., DE
BOER, A.B., VERHOEF, J.C., and JUNGINGER, H.E. (1996). Mu-coadhesive polymers in perioral peptide drug delivery. VI. Carbomer and chitosan improve the intestinal absorption of the peptide drug buserelin in vivo. Pharm. Res. 13, 1668–1672.
MACGREGOR, G.R., and CASKEY, C.T. (1989). Construction of plasmids that express E. coli-galactosidase in mammalian cells. Nu-cleic Acids Res. 17, 2365–2369.
MAO, H.Q., KRISHNENDU, R., TROUNG-LE, V.L., JANES, K.A., LIN, K.Y., WANG, Y., AUGUST, T., and LEONG, K.W. (2001). Chitosan–DNA nanoparticles as gene carriers: Synthesis characteri-zation and transfection efficiency. J. Control. Release 70, 399–421. MATHIOWITZ, E., JACOB, J.S., JONG, Y.S., CARINO, G.P., CHICHERING, D.E., CHATURVEDI, P., SANTOS, C.A., VIJA-YARAGHAVAN, K., MONTGOMERY, S., BASSETT, M., and MORRELL, C. (1997). Biologically erodable microspheres as po-tential oral drug delivery systems. Nature 386, 410–414.
MIYAZAKI, S., TOBIYAMA, T., TAKADA, M., and ATTWOOD, D. (1995). Percutaneous absorption of indomethacin from pluronic F127 gels in rats. J. Pharm. Pharmacol. 47, 455–457.
NORRIS, D.A., and SINKO, P.J. (1997). Effect of size, surface charge, and hydrophobicity on the translocation of polystyrene mi-crospheres through gastrointestinal mucin. J. Appl. Polym. Sci. 63, 1481–1492.
OSHIMA, Y., SAKAMOTO, T., YAMANAKA, I., and NISHI, T. (1998). Targeted gene transfer to corneal endothelium in vivo by electric pulse. Gene Ther. 5, 1347–1354.
REJMAN, J., OBERLE, V., ZUHORN, I.S., and HOEKSTRA, D. (2004). Size-dependent internalization of particles via the pathways of clathrin- and caveolae-mediated endocytosis. Biochem. J. 377, 159–169.
ROJANASAKUL, Y., LIAW, J., and ROBINSON, J.R. (1990). Mech-anisms of action of some penetration enhancers in the cornea: Laser scanning confocal microscopic and electrophysiology studies. Int. J. Pharm. 66, 131–142.
ROSENBERG, W., BREAKEFIELD, X.O., DEANTONIO, C., and ISACSON, O. (1992). Authentic and artifactual detection of the E. coli lacZgene product in the rat brain by histochemical methods. Mol. Brain Res. 16, 311–315.
SASKI, W. (1968). Effect of a nonionic surface-active polymer on pas-sage of hydrocortisone across rat intestine in vitro. J. Pharm. Sci. 57, 836–838.
SCHILLING, J.R., and MITRA, A.K. (1990). Intestinal mucosal trans-port of insulin. Int. J. Pharm. 62, 53–64.
SCHMOLKA, I.R. (1991). Poloxamers in the pharmaceutical industry. In Polymers for Controlled Drug Delivery. Tarcha, P.J., ed. (CRC Press, Boca Raton, FL) pp. 189–214.
SCHUBERT, O., and WRETLIN, A. (1961). Intravenous infusion of fat emulsions, phosphatides and emulsifying agents. Acta Chir. Scand. 1(Suppl. 278).
SFERRA, T.J., MCNELLY, D., and JOHNSON, P.R. (1997). Gene transfer to the intestinal tract: A new approach using selective in-jection of the superior mesenteric artery. Hum. Gene Ther. 8, 681–687.
SHI, N., BOADO, R.J., and PARDRIDGE, W.M. (2001). Receptor-mediated gene targeting to tissues in vivo following intravenous ad-ministration of pegylated immunoliposomes. Pharm. Res. 18, 1091–1095.
STOLNIK, S., ILLUM, L., and DAVIS, S.S. (1995). Long circulat-ing microparticle drug carriers. Adv. Drug Deliv. Rev. 16, 195– 214.
TORCHILIN, V.P. (2001). Structure and design of polymeric sur-factant-based drug delivery systems. J. Control. Release 73, 137–172.
WEISS, D.J., LIGGITT, D., and CLARK, J.K. (1999). Histochemical discrimination of endogenous mammalian galactosidase activity
from that resulting from LacZ gene expression. Histochem. J. 31, 231–236.
YALIN, M., ONER, F., ONER, L., and HINCAL, A.A. (1997). Prepa-ration and properties of a stable intravenous lorazepam emulsion. J. Clin. Pharm. Ther. 22, 39–44.
YANG, S., ZHU, J., LU, Y., LIANG, B., and YANG, C. (1999). Body distribution of camptothecin solid lipid nanoparticles after oral ad-ministration. Pharm. Res. 16, 751–756.
YUAN, A., SIU, C.H., and CHIA, C.P. (2001). Calcium requirement for efficient phagocytosis by Dictyostelium discoideum. Cell Calcium 29, 229–238.
ZHANG, X., SWAYER, G.J., DONG, X., QIU, Y., COLLINS, L., and FABRE, J.W. (2003). The in vivo use of chloroquine to promote non-viral gene delivery to the liver via the portal vein and bile duct. J. Gene Med. 5, 209–218.
Address reprint requests to:
Dr. Jiahorng Liaw School of Pharmacy Taipei Medical University 250 Wu Hsing Street Taipei 110, Taiwan E-mail: [email protected]
Received for publication October 27, 2003; accepted after re-vision March 16, 2004.
