Effects of surface hydrophilic properties of PEG-based mucus-penetrating nanostructured lipid carriers on oral drug delivery

Abstract

Nano-structured lipid carriers (NLCs) have been widely used for oral drug delivery due to their remarkable biocompatibility and biodegradability. In addition to formulation and particle size, particle surface properties are important for overcoming gastrointestinal (GI) barriers to oral drug delivery by lipid-based nanoparticles. In this study, we describe polyethylene glycol (PEG)-based mucus-penetrating NLCs and we report the effect of surface hydrophilic properties on their oral delivery across the GI mucus and intestinal epithelium cells in vivo and in vitro. The solvent evaporation method was used to prepare PEG-modified NLCs (pNLCs) with three different levels of polyethylene glycol (100) monostearate (S100) as the hydrophilic coating material. The stability and cytotoxicity of pNLCs, as well as their uptake by Caco-2 cells and transport across Caco-2 monolayers and mucus-secreting Caco-2/HT29 coculture monolayers, were investigated to confirm the optimal PEG level for oral delivery. In situ single-pass perfusion experiments, in vivo imaging and in vivo oral pharmacokinetic studies were also performed in rodents to evaluate the oral absorption of pNLCs with different hydrophilic surface properties. Our results showed that the hydration layer thickness of the pNLCs increased with the increasing S100 content. NLC containing 20% S100 (pNLC-20%) showed the highest uptake by Caco-2 cells and the most effective permeation through Caco-2/HT29 coculture monolayers compared with unmodified NLC, pNLC-40% and pNLC-80%. In addition, pNLC-20% was the most stable during the process of oral administration and uptake into the blood circulation, as judged by fluorescence resonance energy transfer (FRET). Moreover, in situ single-pass perfusion experiments, ligated intestinal loops tests, in vivo imaging and pharmacokinetics studies also demonstrated that pNLC-20% most rapidly penetrated the intestinal epithelium and had the highest AUC and C_max of all the NLCs. In conclusion, optimization of surface hydrophilic properties is essential for enhancing the absorption of drugs delivered by oral administration of lipid-based nanocarriers.

---

1 Introduction

The oral route is the most widely used and strongly preferred for drug administration owing to the highly absorptive properties of the intestinal epithelium and the high surface area of the villi.^1,2 Nanostructured lipid carriers (NLCs) have been widely used for oral drug delivery due to their remarkable biocompatibility and biodegradability.^3,4 The controlled nanostructure formed by solid and liquid lipids in NLCs provides sufficient space to accommodate drug molecules in the matrix, resulting in maximum drug-loading capacity, stability and ability to control the release of encapsulated substances.^5–7

However, in order to achieve therapeutic levels, orally administered NLCs must overcome the rapid secretion and shedding of gastrointestinal (GI) tract mucus, and then effectively penetrate the intestinal epithelia or membranes of target cells.⁸ Moreover, the final NLC product is typically a hydrophobic particle, whether the poorly soluble drug is micronized into a suspension or encapsulated within conventional lipid nanoparticles. Therefore, the highly protective mucus layer of the GI tract will present a particularly formidable barrier to NLC diffusion.^9,10

In order to enhance the bioavailability of cargo therapeutics and overcome the high viscoelasticity and adhesivity of mucus traps, mucoadhesive particles (MAPs) and mucus-penetrating particles (MPPs) have recently been developed. MAPs depend on modified cationic materials to adhere to the intestinal mucus, thus prolonging their retention time. MPPs use hydrophilic nonionic long chain polymers, such as polyethylene glycol (PEG), to rapidly penetrate the mucus layer.^8,11 However, Katharina Maisel et al. reports that compared with MPP, the distribution of MAP strongly sticks to the epithelial surface due to the obstacle of mucus barrier. It results in aggregation and limited distribution throughout the GI tract after oral administration. MPP can penetrate into mucus in the deeply in-folded surfaces to evenly coat the entire epithelial surface, which is likely to provide improved oral drug delivery.^9,12

It is particularly important to enhance the mucus-penetrating properties of MPPs, which will result in more uniform distribution and improved retention of particles on the mucosal surface.¹³ PEG is an electrically neutral and hydrophilic material routinely used in pharmaceutics to improve systemic circulation and minimize opsonization following intravenous injection of nanoparticles.¹⁴ Moreover, PEG is an FDA-approved nontoxic material which endows engineered nanoparticles with a mucus penetrating coating.^15–17 Densely PEGylated nanoparticles exhibit rapid penetration of highly viscoelastic mucus secretions.¹⁸ PEG-modified surfaces serves as “mucus-inert” interfaces that minimize adhesive interaction between nanoparticles and mucus constituents, and facilitate their diffusion through low viscosity pores in the mucus matrix.^12,16 However, in most cases, the uncharged hydrophilic surface may have a negative effect on the interaction between nanoparticles and the cell membrane, which may decrease their uptake and transport by intestinal epithelial cells.^19,20 There is an urgent and critical need to clearly understand the effect of the surface hydrophilic properties of MPPs on gastrointestinal absorption, in order to overcome the obstacles of both the mucus layer and the epithelium.

Herein, we investigated the effect of surface hydrophilic properties on oral drug delivery by NLCs. NLCs were prepared with three different levels of surface hydrophilicity to evaluate the importance of surface hydrophilic properties on oral drug delivery in vitro and in vivo (shown in Fig. 1). Cell uptake, permeation and pharmacokinetic studies confirmed that NLC with 20% PEG modification (pNLC-20%) showed maximum enhancement of oral drug delivery compared with unmodified NLC, pNLC-40% and pNLC-80%.

Fig. 1 Schematic presentation showing penetration of the intestinal epithelium by PEG-based mucus-penetrating nanostructured lipid carriers with different surface hydrophilic properties. (1) Cell uptake by caveolae and clathrin-mediated pathways; (2) a higher efficient penetration of the viscoelastic mucus layers and intestinal epithelium of the GI tract of pNLC-20% than NLC, pNLC-40% and pNLC-80%.

2 Materials and methods

2.1 Materials and animals

Phosphatidylcholine was purchased from Taiwei Biotechnology Co, Ltd. (Shanghai, China). S100, PEG molecular weight 4000 g mol⁻¹ and 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) were purchased from Sigma-Aldrich (Milwaukee, WI, USA). Labrafac CC was a kind gift from the Gattefosse office (Shanghai, China). Coumarin-6 (C6) and trilaurin were purchased from TCI (Japan). BCA Protein Assay Kit was from Thermo Scientific Pierce (Rockford, USA). Chlorpromazine was purchased from Sigma-Aldrich Co. (Shanghai, China). Fetal bovine serum (FBS), DMEM/HG medium, penicillin, streptomycin and hanks balanced salt solution (HBSS) were purchased from Hyclone (USA). DMSO, nystatin, amiloride and methyl-β-cyclodextrin were purchased from Aladdin (Shanghai, China). 4′,6-Diamidino-2-phenylindole (DAPI), 3,3′-dioctadecyloxacarbocyanine perchlorate (DiO) and 1,1′-dioctadecyl-3,3,3′,3′-tetramethylindocarbocyanine perchlorate (DiI) were purchased from Beyotime (Shanghai, China). The near-infrared dye 1,1′-dioctadecyl-3,3,3′,3′-tetramethylindotricarbocyanine (DiR) was obtained from Beijing Fanbo Science and Technology Co., Ltd. (Beijing, China). All other chemicals and reagents were of analytical grade.

Sprague-Dawley (SD) rats (weighing 180–220 g) and male ICR mice were purchased from the Experimental Animal Centre of Nantong University (Nantong, China). All animal experiments were conducted in full compliance with local, national, ethical and regulatory principles with the approval of the Institutional Animal Care and Use Committee at China Pharmaceutical University.

2.2 Cell culture

Caco-2 cells and HT 29 cells were both obtained from the cell bank of the Chinese Academy of Sciences. Cells were maintained in Dulbecco's modified Eagle's minimal essential medium (DMEM) containing 10% (v/v) fetal bovine serum (FBS), 1% penicillin/streptomycin (100 IU mL⁻¹) and 1% (v/v) non-essential amino acids. Cells were cultured at 37 °C in a humidified atmosphere with 5% CO₂. The cells were harvested with 0.05% trypsin and rinsed. The resulting cell suspension was used in the following experiments.

2.3 Preparation and characterization of C6-loaded NLC and pNLCs

The solvent evaporation method²¹ was utilized to prepare C6-loaded plain and S-100-modified NLCs. C6, phosphatidylcholine (PC), Labrafac CC and trilaurin were dissolved in ethanol/methylene dichloride (8 : 2, v/v), and S100 with a total content of 0, 20%, 40% and 80% was also added into the mixture. The organic solvent was removed under reduced pressure at 37 °C for 12 h. The dried lipid film was hydrated with pH 7.4 isotonic phosphate buffer (PBS) at 37 °C to give a lipid suspension. The suspension was filtered through a 0.45 μm cellulose nitrate membrane after ultrasonication to obtain NLC. NLCs with a total content of 20%, 40% and 80% S100 were termed as pNLC-20%, pNLC-40%, and pNLC-80% respectively. The ultrasonication was subsequently performed as following to obtain NLCs with equal levels of particle size: ultrasonic intensity was 100 W for pNLC-40%, 200 W for NLC and 300 W for pNLC-20% and pNLC-80%. As for ultrasonic frequency, each ultrasonic cycle contains 1 s for ultrasonication and 2 s for interval. The final frequency of ultrasonic cycles were 180 times of cycle for NLCs, 360 times of cycle for pNLC-20%, 40 times of cycle for pNLC-40% and 30 times of cycle for pNLC-80%. NLCs loaded with DiO/DiI were also prepared using the same method. Particle size, polydispersity index and zeta potential of C6-loaded NLC and pNLCs were determined with a Brookhaven Instruments-ZetaPlus (Brookhaven, USA). Morphology analysis of C6-loaded NLC and pNLCs was performed by atomic force microscopy imaging with tapping mode (AFM, SPA3800N, SEIKO, Japan) and transmission electron microscopy (TEM, H-600, Hitachi, Japan). The encapsulation efficiency (EE) of NLC and pNLCs was determined as previously reported.⁵

2.4 Effect of different surface hydrophilic properties on fixed aqueous layer thickness (FALT) of NLC and pNLCs

According to the Gouy–Chapman theory,²² the zeta potential ψ(L) as the electrostatic potential at the position of the slipping plane L was expressed as eqn (1)

ln ψ(L) = ln A − kL (1)

where k was the Debye–Huckel parameter, which could be calculated by √(C)/0.3 (with C as the molality of electrolyte), and A was a constant. By using NaCl as the electrolyte, with concentrations of 0, 0.01, 0.05 and 0.1 M, the equation of zeta potential against k was calculated, and the slope L represented the thickness of the fixed aqueous layer in nm.

2.5 Evaluation of the stability in vitro and in vivo

FRET technology was applied to certify the stability of plain NLC and S100-modified NLCs. Through the energy transfer from an excited fluorophore (donor) to a nearby light-absorbing molecule (acceptor), FRET provides information on the proximity of two fluorescent molecules within a 1–10 nm range. All NLCs were labeled with DiO and DiI as a FRET pair, and changes in the FRET signals were monitored to detect variation in the integrity of the nanoparticles. NPs were suspended in isotonic PBS, Krebs–Ringer (K–R) solution and simulated gastric fluid (0.1 M HCl). Samples were taken at certain times and fluorescence was detected at the excitation wavelength of 460 nm using a fluorescence spectrophotometer (RF-5301, Shimadzu). The FRET ratio (FR) was calculated according to eqn (2):

FR (%) = F_r/(F_r + F_g) × 100% (2)

where F_r and F_g were the fluorescence intensity at 564 nm and 504 nm, respectively.

FRET NLC or pNLCs were administered to SD rats by oral gavage. Blood samples were collected at certain time points and centrifuged at 12 000 rpm for 10 min. 100 μL of plasma supernatant was diluted to 1 mL with PBS. The fluorescence spectra of the samples were detected at the excitation wavelength of 460 nm using the fluorescence spectrophotometer. The FRET ratio of each nanoparticle was determined at each time point.

2.6 In vitro release of C6-loaded NLC and pNLCs

The in vitro drug release of C6 from NLC and pNLCs was analyzed using a dialysis method.²³ 0.1 M isotonic PBS (pH 7.4) was used to simulate the cellular environment and the artificial intestinal fluid K–R solution was used as the dissolution medium. 1.0 mL of C6-loaded NLC or pNLCs (equivalent to 1000 ng of C6) was placed in a dialysis bag (MWCO of 12 kDa). The receptor compartment was filled with 80 mL dissolution medium at 37 ± 0.5 °C with stirring at 250 rpm. At predetermined time points, 1.0 mL of the dissolution medium was withdrawn and replaced with the same volume of pre-warmed fresh dialysis medium. The samples were then centrifuged at 12 000 rpm for 10 min and the supernatant was analyzed using an HPLC (Shimadzu Scientific Instrument Inc. Japan) system with an RF-10AXL Shimadzu fluorescence detector, Ex_{465 nm}/Em_{502 nm}. The mobile phase consisted of methanol : water (96 : 4) at a flow rate of 1 mL min⁻¹. The stationary phase, a Dikma Diamonsil C18 column (150 mm × 4.6 mm, 5 μm) was kept at 35 °C. The release experiment was carried out in triplicate.⁴

2.7 Cytotoxicity of NLC and pNLCs

The cytotoxicity of NLC and pNLCs in Caco-2 cells was assessed by MTT assay. 1 × 10⁵ cells per well were seeded in 96-well plates (Costar, USA). After 24 h of culture, the culture medium was removed and replaced with a series of concentrations of NLC and pNLCs (n = 3 for each group). After 24 h incubation at 37 °C, 20 μL of MTT solution (5 mg mL⁻¹) was added to each well and incubated for an additional 4 h. The supernatant was removed, and then the formed formazan salts were dissolved by adding 150 μL of DMSO and shaking the plates for a moment. Absorbance was measured at 570 nm using an ELISA reader. The UV absorbance intensity was measured at 570 nm with a microplate reader (Thermo Electron Corporation, USA). Cell viability was expressed as the percentage of A_{570 nm} of the study group relative to that of the control group.

2.8 Cellular uptake

1 × 10⁵ Caco-2 cells per well were seeded in 24-well plates and incubated at 37 °C for 24 h. To investigate time-dependent uptake, HBSS was replaced with 100 ng mL⁻¹ of C6-loaded NLC and pNLCs, and then incubated for 0.5 h, 1 h, 1.5 h and 2 h. To investigate concentration-dependent uptake, cells were incubated for 2 h with C6-loaded NLC and pNLCs at concentrations ranging from 25 ng mL⁻¹ to 100 ng mL⁻¹ (based on C6 concentration). To reveal the possible uptake mechanism of each nanoparticle by Caco-2 cells, cells were pre-incubated with the following endocytic inhibitors at 37 °C for 30 min: (1) sodium azide (0.1%, w/v), a cell energy metabolism inhibitor; (2) chlorpromazine (10 μg mL⁻¹), a clathrin-mediated endocytosis inhibitor; (3) nystatin (30 μg mL⁻¹), a caveolae-mediated endocytosis inhibitor; (4) amiloride (100 μM), a macropinocytosis inhibitor; (5) methyl-β-cyclodextrin (MβCD) (10 mM), an inhibitor of de novo synthesis of cholesterol which blocks the caveolae and clathrin-mediated pathways by depleting cholesterol.²⁴ The inhibitors were all used at concentrations that are non-toxic to cells. After the pre-treatment step, the uptake study was carried out in the presence of the endocytic inhibitor with a C6 concentration of 100 ng mL⁻¹. Subsequently, the cells were washed three times with ice-cold isotonic PBS, and the fluorescence intensity of C6 in the cells was analyzed by flow cytometry. The relative uptake index (RUI) was calculated according to the following eqn (3):

RUI = F_S/F_c × 100% (3)

where F_S was the fluorescence of C6-loaded NLC and pNLCs treated with various kinds of uptake inhibitor and F_c was the fluorescence of the control.

Confocal laser scanning microscopy (CLSM) was also used to assess cell uptake. Caco-2 cells were washed with PBS three times and fixed with 4% paraformaldehyde for 10 minutes. Following staining of cell nuclei with DAPI for 15 minutes, cellular uptake of NLC or pNLCs was visualized by CLSM (Leica TCS SP5, Heidelberg, Germany). In addition, the FRET technique was used to investigate the integrity of nanocarriers in Caco-2 cells. Caco-2 cells (1 × 10⁵ cells per well) were seeded into 24-well plates and incubated for 24 h, then incubated with FRET NLC or pNLCs for 2 h. Afterwards, the culture medium was removed and the cells were washed three times in PBS. After incubation with fresh medium for certain times, the cells were fixed and stained with DAPI as described above. The FRET signals of the nanoparticles were detected at the excitation wavelength of 460 nm using CLSM.

2.9 Transport studies

Transport assays of C6-loaded NLC and pNLCs across Caco-2 and Caco-2/HT-29 (75 : 25) coculture cell monolayers was performed using Transwell inserts® to evaluate oral absorption in vitro. The culture medium was removed and the cell monolayer was washed with HBSS on both apical (AP) and basolateral sides before treatment. 200 μL of test solution (100 ng mL⁻¹, based on C6 concentration) was added to the AP side while 800 μL of fresh HBSS at 37 °C was added to the BL side to estimate the AP to BL transport. 100 μL of solution was taken from the BL side at 0.5, 1, 1.5, 2 h after addition of the test solution and replaced with fresh HBSS at 37 °C. C6-loaded NLC and pNLCs were detected by HPLC. The apparent permeability coefficient (P_app) was calculated from the measurement of the transfer rate of C6 across the cells from the upper to lower compartments of the Transwell inserts according to eqn (4):

P_app (cm s⁻¹) = dQ/A × C₀ × dt (4)

where dQ/dt (ng s⁻¹) was the drug permeation rate, A was the surface area of the polycarbonate membrane (0.33 cm²) and C₀ (100 ng mL⁻¹) was the initial concentration.

2.10 Intestinal absorption of NLC and pNLCs by in situ permeability studies²⁵

Male SD rats weighing approximately 180–230 g were fasted for 24 h before the experiment with free access to water. The rats were anesthetized with an intraperitoneal injection of 20% urethane solution. The abdomen was cut open with a midline incision (2–3 cm), and the duodenum, jejunum, ileum and colon were collected. Using a peristaltic pump, the intestinal segments were flushed and perfused with 37 °C drug perfusion solution at a flow rate of 0.2 mL min⁻¹ for 30 min. The time when the first drop of drug perfusion solution flowed from the perfusion tube was set to zero. For the whole 90 min perfusion period, samples were collected every 15 min and the flow rate was kept stable at 0.2 mL min⁻¹. Phenol red acted as a nonabsorbable marker in the perfusion solution to correct for any appreciable influence of the secretion or absorption of water during the experiment. After a single-pass perfusion experiment, C6 in the collected perfusion solution samples was analyzed by HPLC as described above. Meanwhile, the phenol red concentration of the samples was determined using a Shimadzu LC-10AT HPLC system (Kyoto, Japan). The Diamonsil C18 column (150 mm × 4.6 mm, 5 μm) was kept at 40 °C, the mobile phase was a mixture of methanol: double distilled water (50 : 50, v/v) containing 0.025 mol L⁻¹ KH₂PO₄. The flow rate was set at 1.0 mL min⁻¹. The injection volume was 20 μL and the absorbance of the effluent was monitored at 415 nm. P_eff of the C6-loaded NLC and pNLCs in intestinal segments was calculated using the following eqn (5):where Q was the flow rate (mL min⁻¹), r and L were the radius and length of the perfused intestinal segment (cm), and C_in and C_out were the concentrations of C6 in the perfusion solution and collected samples respectively. PR_in and PR_out were the concentrations of phenol red in the perfusion solution and collected samples respectively.

2.11 Validation of in situ single-pass perfusion experiment

The in vivo uptake of C6-loaded NLC and pNLCs was evaluated using the ligated intestinal loops model.²⁶ The rats were anesthetized with an intraperitoneal injection of 20% urethane solution. From the opened abdominal incision (2–3 cm), 2 cm sections of each of the duodenum, jejunum, ileum and colon from the intestine were washed with K–R culture solution (37 °C). Equal amounts of C6-loaded NLC and pNLCs (100 ng mL⁻¹, based on C6 concentration, 0.2 mL) were injected into each loop, and the loops were ligated. After 2 h, each loop was cut out and washed in K–R culture solution. Subsequently, each loop was placed into 4% paraformaldehyde and stained with DAPI. Paraffin sections of the tissues were visualized using a fluorescence microscope (Olympus, Japan).

2.12 In vivo imaging and biodistribution analysis

NLC or pNLCs loaded with DiR were administered by oral gavage to mice at same dose. Images were taken at 2, 4 and 8 h after administration using an in vivo imaging system. To verify the biodistribution of nanoparticles in different organs, mice were sacrificed at each time point to obtain different organs for imaging. The organs were separated and washed three times with PBS. Dye accumulation and retention in organs was imaged and analyzed with Carestream Molecular Imaging Software V 5.3.5. The wavelength of DiR was fixed at 720 nm for excitation and 790 nm for emission.

2.13 Pharmacokinetic studies

Eighteen male rats (180–220 g) were used to evaluate the effect of different surface hydrophilic properties on the pharmacokinetics of NLC and pNLCs. NLC and pNLCs were administered to rats (n = 6) in a single dose of 1.4 mg kg⁻¹ by oral gavage. After oral administration and blood collection, blood samples (0.2 mL) were centrifuged at 12 000 rpm and a double volume of acetonitrile and methanol (1 : 1, v/v) was added to the supernatant plasma for deproteinization. The mixture was vortexed for 10 min and then centrifuged at 12 000 rpm for 10 min, and the C6 concentration in the blood was measured by HPLC. The plasma C6 concentration was plotted versus time and the pharmacokinetic parameters were analyzed using a single compartmental model.

2.14 Statistical analysis

Data are expressed as means of three separate experiments. Results are given as mean ± S.D. Differences between groups were assessed using unpaired two-tailed Student's t tests. In all cases, a p-value < 0.05 was considered statistically significant and a p-value < 0.01 was considered highly statistically significant.

3 Results and discussion

3.1 Preparation and characterization of C6-loaded NLC and pNLCs

S100 with polyethylene glycol molecular weight 4000 g mol⁻¹ was an amphiphilic polymeric derivative of hydrophilic PEG modified by attaching a hydrophobic moiety, it could be easily incorporated into the lipid core of NLC with the hydrophilic PEG chains on their surface. In PEG monostearate family, there were S100, S55 and S40. The number of codes meant how many –CH₂CH₂O– units in polymers. S100 could reach the best long cycle effect than S55 or S40 because its molecular weight of PEG segment was longer than the other two.²⁷ In the present study, C6-loaded NLC and pNLCs were prepared by the solvent evaporation method. The average particle size, PDI, zeta potential and EE (%) of C6-loaded NLC and pNLCs are summarized in Table 1. The results of particle size and PDI indicated that satisfactory homogeneity of NLCs. Meanwhile, the particle size of the NLC and pNLCs could be kept uniform by changing the ultrasound time and intensity. The EEs of the nanoparticles were all above 93%, indicating that most of the C6 was trapped within the nanoparticles. The morphology of NLC and pNLCs was observed by TEM and AFM (Fig. 2). The results showed that most of the nanoparticles appear as uniform spherical particles. The particle sizes were similar to those obtained by dynamic light scattering.

Table 1 Physicochemical properties of NLC and pNLCs. Data are presented as mean ± S.D. (n = 3)

Group	Size (nm)	PDI	Zeta (mV)	EE (%)
NLC	136.4 ± 1.3	0.124 ± 0.023	−59.00 ± 0.92	93.48 ± 0.78
pNLC-20%	130.5 ± 0.3	0.132 ± 0.023	−53.96 ± 0.82	98.91 ± 0.66
pNLC-40%	130.8 ± 1.8	0.213 ± 0.001	−58.34 ± 0.75	96.52 ± 0.45
pNLC-80%	133.3 ± 1.0	0.303 ± 0.003	−50.47 ± 0.64	96.44 ± 0.87

---

Fig. 2 TEM and AFM images of NLC (A), pNLC-20% (B), pNLC-40% (C), and pNLC-80% (D).

3.2 Fixed aqueous layer thickness (FALT) of NLC and pNLCs

As shown in Table 2, the FALT increased as the S100 level in the NLC formulation increased from 20% to 80% of the total content. The results are consistent with a paper reporting that the FALT is related to the PEG molecular weight and the amount of PEG used.²⁸ The FALT measurements suggest that the surface of our hydrophilic PEG-modified pNLCs is located somewhere between the “mushroom” and the “brush”. The results indicate that most of the PEG chains are slightly constricted and are dense enough to ensure that there are no gaps or spaces on the particle surface.²⁹

Table 2 FALT for NLC and pNLCs

Group	Equation	R^2	FALT (nm)
NLC	Y = −0.3917x + 4.0923	R^2 = 0.9889	0.39
pNLC-20%	Y = −1.4738x + 3.9349	R^2 = 0.9629	1.47
pNLC-40%	Y = −2.3147x + 3.9155	R^2 = 0.9743	2.31
pNLC-80%	Y = −2.9235x + 3.8178	R^2 = 0.9396	2.92

---

3.3 FRET determination of in vitro and in vivo stability

The FRET technique (shown in Fig. 3A) was used to estimate the integrity of NLC and pNLCs. The FRET pair contained DiO (460 Ex/504 Em) and DiI (495 Ex/564 Em). If the distance was less than 10 nm of two fluorescence pairs (indicating the complete of NLCs), the emitted energy of the donor dye DiO would transfer to the acceptor dye DiI because of the overlap between emission of DiO and excitation of DiI. Consequently, the fluorescence of emission by DiI at 564 nm would be observed when DiO was emitted at 460 nm, while fluorescence of DiO was disappear. Upon disassembly of NLCs, the FRET pair was separated and the normal fluorescence of the DiO was restored, whereas the DiI stopped fluorescing. As shown in Fig. 3B, the FRET nanoparticles showed an apparent emission from DiI at 564 nm after DiO was excited at 460 nm, suggesting that the integrity of the nanoparticles enabled energy transfer from the DiO donor to the DiI acceptor. It is worth noting that the stability and integrity of the nanoparticles was maintained following treatment with PBS, K–R solution and gastric fluid (0.1 M HCl) as shown in Fig. 3C. To analyze the stability in vivo, the FRET nanoparticles were orally administered to rats, and changes in the fluorescent signals in the plasma were monitored. Fig. 3D shows that the FRET signal was maintained up to 8 h, and hydrophilic PEG-modified pNLC-20% showed the highest level of integrity following oral administration and absorption into the blood circulation. The FRET ratio of pNLC-20% is similar to the initial FRET ratio value of 55% for all nanoparticles. 8 h after oral administration, the FRET ratio of pNLC-20% was calculated to be 52.3%, which is higher than that of NLC (42.7%). Therefore, 20% PEG-modified NLCs might be the stability of formulation in oral absorption.

Fig. 3 The cross-over region of the FRET donor DiO (emission) and the FRET acceptor DiI (excitation) (A). Fluorescence spectra, obtained using an excitation wavelength of 460 nm, of NLC and pNLCs loaded with the DiO/DiI fluorescent pair (B). FRET analysis of the stability and integrity of nanoparticles treated with PBS, K–R solution and HCl (C). FRET analysis of the integrity of nanoparticles in blood plasma following oral administration to rats (D).

3.4 In vitro release of C6-loaded NLC and pNLCs

In vitro drug release of C6-loaded NLC and pNLCs is shown in Fig. 4A and B. In order to simulate the cell environment and the in vivo biological environment, the in vitro release experiments were performed in isotonic PBS (pH = 7.4) and artificial intestinal fluid K–R solution as releasing medium to imitate in vivo physiological and intestinal environment, respectively. The K–R solution contained 0.2 g L⁻¹ calcium chloride, 0.22 g L⁻¹ magnesium chloride, 7.8 g L⁻¹ sodium chloride, 1.37 g L⁻¹ sodium bicarbonate, 0.22 g L⁻¹ sodium dihydrogen phosphate, 0.35 g L⁻¹ potassium chloride and 1.4 g L⁻¹ D-glucose. The drug release behavior was shown to be similar in NLC and hydrophilic PEG-modified pNLCs with same particle size (p > 0.05), which may be attributed to the nearly uniform distribution of C6 in each NLC.³⁰ More importantly, the cumulative leakage of C6 from the four NLCs was less than 10% in isotonic PBS and K–R solution after 24 h. The results indicated that PEG modification would be considered to negligibly interfere with the drug release. Therefore, C6-loaded NLCs could maintain relative stability in the blood circulation and gastrointestinal tract, which avoided the rapid leakage of C6 before reaching the targeted site.

Fig. 4 In vitro release profiles of C6-loaded NLC and pNLCs in PBS (A) and K–R solution (B) (mean ± S.D., n = 3). In vivo cytotoxicity of Caco-2 cell co-cultured with blank (non-loaded) NLC and pNLCs (C), and with C6-loaded NLC and pNLCs (D).

3.5 Cytotoxicity of NLC and pNLCs

The cytotoxicity of blank (non-loaded) and C6-loaded NLC and hydrophilic PEG-modified pNLCs to the Caco-2 cell line after treatment for 24 h was assessed by MTT assay, as shown in Fig. 4C and D. Slight cytotoxicity of both blank NLCs was observed at the concentration of NLCs ranging from 200 μg mL⁻¹ to 3000 μg mL⁻¹ and the cell survival viability was more than 80% when the concentration of NLC and pNLCs was 1000 μg mL⁻¹ which was used for further experiments. Similarly, the results of C6-loaded NLC and pNLCs demonstrated a high cell survival viability against Caco-2 cells, which further validated the high entrapment efficiency and good stability of drug.

3.6 Cellular uptake study

In order to investigate the intracellular integrity of NLC and hydrophilic PEG-modified pNLCs, Caco-2 cells were incubated with nanoparticles containing the DiO and DiI fluorescent pair. After the cells were incubated with the FRET nanocarriers for 6 h, a significant DiI signal could still be observed in all treatment groups (Fig. 5). This suggests that all the nanoparticles maintained their structural integrality during the process of entering the intestinal cells.

Fig. 5 FRET analysis of the intracellular integrity of NLC or pNLCs loaded with the DiO/DiI pair in Caco-2 cells. Images were taken with a laser scanning confocal microscope.

The effect of drug concentration on the uptake of NLC and hydrophilic PEG-modified pNLCs by Caco-2 cell is shown in Fig. 6A. When Caco-2 cells were incubated with C6-loaded nanoparticles for 2 h at 37 °C, the cellular uptake of both NLC and pNLCs increased with increasing C6 concentration, which indicated that the uptake of NLC and pNLCs was concentration-dependent. Fig. 6B illustrates that the nanoparticles were internalized by Caco-2 cells within 0.5 h and the uptake was time-dependent. Moreover, at each time point except 0.5 h, pNLC-20% exhibited higher uptake efficiency than the other nanoparticles. The uptake efficiency of pNLC-20% was about 1.20-fold that of NLC and 1.11-fold that of pNLC-80% nm at 2 h. The cellular uptake of pNLC-20% was significantly higher than that of NLC and pNLC-80% (p < 0.05, vs. NLCs and pNLC-80%) as the concentration increased from 25 ng mL⁻¹ to 200 ng mL⁻¹. As seen in CLSM images (Fig. 6C), the cytoplasmic fluorescence intensity was highest in Caco-2 cells treated with pNLC-20% for 2 h at a concentration of 100 ng mL⁻¹. This is consistent with the fluorescence intensity quantification. The results indicate that 20% PEG modification of NLCs could effective at enhancing cellular uptake.

Fig. 6 Uptake of C6-loaded NLC and pNLCs by Caco-2 cells at 37 °C (A) with different concentrations of C6 (n = 3); (B) with 100 ng mL⁻¹ of C6 for different time intervals (n = 3). (C) Laser scanning confocal images of Caco-2 cells after incubation with NLC and pNLCs (equivalent to 100 ng mL⁻¹ of C6) at 37 °C for 2 h.

3.7 In vitro cell uptake mechanism study

To further investigate the endocytosis pathways of NLC and hydrophilic PEG-modified pNLCs, Caco-2 cells were treated with several endocytosis inhibitors and the cellular uptake of the nanoparticles was evaluated. Compared with the control, the uptake of NLCs was significantly reduced by all the inhibitors (p < 0.05) except for sodium azide (Fig. 7A). The cellular uptake of pNLC-20% was similarly reduced by MβCD, nystatin and amiloride (EIPA) (p < 0.05). It is worth mentioning that after the cells were treated in advance with MβCD, which inhibits the caveolae and clathrin-mediated pathways by depleting cholesterol, the cellular uptake of NLC, pNLC-20%, pNLC-40% and pNLC-80% was decreased to 16.08%, 15.76%, 14.16 and 15.84% respectively (p < 0.01 vs. control). This demonstrates that endogenous cholesterol is involved in the internalization of all NLC and pNLCs. In the presence of chlorpromazine, which is known to inhibit the clathrin-associated pathway, the cellular uptake of NLC was significantly lower than pNLC-20% (p < 0.05). In the case of amiloride (EIPA), which is known to inhibit the macropinocytosis pathway, the uptake of NLCs was lower than pNLC-20% (p < 0.05) and pNLC-80% (p < 0.01), and the uptake of pNLC-20% and pNLC-40% was lower than pNLC-80% (p < 0.05). However, it should be taken into account that the different internalization processes act in parallel, and thus the different endocytic pathways might tend to compensate each other.³¹

Fig. 7 Relative uptake efficiency of NLC and pNLCs by Caco-2 cells treated with various cellular uptake inhibitors (n = 3). ^#p < 0.05 compared with the control at 37 °C, ^##p < 0.01 compared with the control at 37 °C (A). Apparent permeability coefficient (P_app) of Caco-2 cell monolayers (B) and Caco-2/HT-29 coculture cell monolayers (C) treated with NLC and pNLCs (n = 3).

3.8 Transport studies

Caco-2 monolayers with transepithelial electrical resistance (TEER) higher than 500 U cm² and Caco-2/HT-29 (75 : 25) coculture cell monolayers with TEER higher than 250 U cm² can be used to mimic intestinal epithelial transport. In Caco-2 cell monolayer transport assays, as shown in Fig. 7B, the largest amount of C6 transport was facilitated by NLC and pNLC-20%. When the amount of PEG increased to 40% and 80%, the apparent permeability coefficient (P_app) decreased markedly, which might be due to the increased hydrophilicity of these nanoparticles.

Fig. 7C shows that in Caco-2/HT29 cell monolayers, which incorporate mucus-secreting cells and can better simulate the intestinal epithelium, the P_app of pNLC-20% was the highest, and was significantly increased compared with NLC (p < 0.05) and pNLC-80% (p < 0.01). These results can be attributed to PEG-modified NLCs possessing rapid mucus-penetrating transport properties, which allows them to cross the intestinal epithelium, while NLC are trapped by the highly viscoelastic and adhesive gastrointestinal mucosa.¹⁶ Furthermore, the P_app of pNLC-40% and pNLC-80% is lower than that of pNLC-20%, indicating that suitable PEG modification and surface hydrophilic properties are also essential for the diffusion of nanoparticles through the mucus layer.

3.9 Evaluation of the intestinal absorption of NLC and pNLCs by the in situ single-pass intestine perfusion method

The in situ single-pass intestine perfusion method was used to assess the absorption of NLC and pNLCs by the intestine of Sprague-Dawley rats. Fig. 8A shows that the apparent permeability coefficient (P_eff) of duodenum for all nanoparticles was higher than the other parts of the intestine. In duodenum, P_eff of pNLC-20% was 2.04-, 1.80- and 1.76-fold higher than NLC, pNLC-40% and pNLC-80%, respectively (p < 0.01). The significantly higher P_eff of pNLC-20% was also found in jejunum and ileum. For colon, the PEG-modified pNLCs with higher surface hydrophilicity (pNLC-40% and pNLC-80%) had an enhanced P_eff, which might be useful for colon administration. All the results above could be explained by the fact that pNLC-20% diffused nearly unimpeded through the mucus barrier, whereas NLC were trapped. It is worth noting that densely PEGylated NLCs might easily cross the mucus barrier, but may then be blocked by the intestinal epithelium due to their highly hydrophilic properties. The above data are also consistent with the permeation study using Caco-2/HT29 cell monolayers. It should be noted that the presence of mucus-secreting cells was essential for the in vitro evaluation of delivery systems.

Fig. 8 Influence of NLC and pNLCs on in situ intestinal effective permeability of C6; values represent mean ± S.D. (n = 3) (A). Distribution of NLC and pNLCs in small intestine. Blue fluorescence indicates the nuclei of small intestinal cells, and green fluorescence indicates the nanoparticles (B). Graph showing plasma concentration of C6 within 24 h of oral administration of NLC and pNLCs to rats (n = 6) (C).

3.10 Validation of the in situ single-pass perfusion experiment by the ligated intestinal loops model in vivo

In order to visualize the absorption of NLC and pNLCs by villi, an in vivo uptake study was carried out using double fluorescent labeling with DAPI-stained nuclei (blue) and C6-loaded NLC and pNLCs (green). As shown in Fig. 8B, C6 fluorescence was strongest in the intestinal epithelium of pNLC-20%-treated duodenum, jejunum and ileum. This suggests that pNLC-20% is most effectively absorbed by the small intestine, which is consistent with the results of the single-pass perfusion study.

3.11 In vivo imaging and biodistribution

In order to monitor the different nanoparticles in the body following oral administration in mice, the in vivo biodistribution at certain time points was recorded with an in vivo imaging system (Fig. 9). pNLCs-20% exhibited the strongest fluorescence intensity in tissues 8 h post-injection, showing that pNLC-20% is quickly absorbed into the blood circulation by the oral route and then travel to various tissues and organs. The results also demonstrate the good bioavailability of pNLC-20%.

Fig. 9 In vivo biodistribution of DiR in mice monitored by in vivo imaging after oral administration of NLC or pNLCs.

3.12 Pharmacokinetic study

In order to investigate the effect of different surface hydrophilic properties on oral absorption of pNLCs, pharmacokinetic studies were performed in male Sprague-Dawley rats following a single oral dose (1.4 mg kg⁻¹). The plasma C6 concentration–time profiles are shown in Fig. 8C. The mean pharmacokinetic parameters of C6-loaded NLC or pNLCs are presented in Table 3. As shown in Fig. 8C and Table 3, pNLC-20% was absorbed into the blood circulation with the largest AUC_{0–12 h}. The C_max of pNLC-20% and pNLC-40% were 1.57-fold and 1.28-fold higher than NLC, respectively. This indicates that the uptake of pNLCs by the GI tract is greater than NLC. Based on the above data, pNLC-20% is the most suitable PEG-modified candidate for highly efficient penetration of viscoelastic mucus layers and the intestinal epithelium of the GI tract.

Table 3 Pharmacokinetic parameters of C6 after oral administration of NLC and pNLCs (mean ± S.D., n = 6). *p < 0.05 and **p < 0.01 vs. NLC

Group	AUC0–12 h (ng h mL^−1)	Cmax (ng mL^−1)	Tmax (h)	MRT (h)
NLC	21.11 ± 0.93	1.85 ± 0.18	0.42 ± 0.29	31.09 ± 0.52
pNLC-20%	25.08 ± 2.12*	3.03 ± 0.58**	0.50 ± 0.10	16.90 ± 2.07*
pNLC-40%	21.82 ± 0.15	2.24 ± 0.20*	0.42 ± 0.14	21.62 ± 1.10*
pNLC-80%	22.07 ± 2.36	2.31 ± 0.84	0.42 ± 0.15	22.49 ± 7.05*

---

4 Conclusions

To verify the importance of particle surface properties for overcoming gastrointestinal (GI) barriers to oral drug delivery by lipid-based nanoparticles, we designed and prepared C6-loaded NLC and pNLCs with different surface hydrophilic properties as a result of PEG-modification. We reported the effect of surface hydrophilic properties on their oral delivery across the GI mucus and intestinal epithelium cells in vivo and in vitro. In vitro drug release studies revealed that the leakage of C6 from the four nanoparticles was less than 10% in pH 7.4 isotonic PBS and K–R solution, C6-loaded NLCs could maintain relative stability in the blood circulation and gastrointestinal tract, which avoided the rapid leakage of C6 before reaching the targeted site. FRET study showed that pNLC-20% owned the highest level of integrity following oral administration and absorption into the blood circulation. In vitro cell uptake studies showed that pNLC-20% was taken up more effectively than NLC and the other pNLCs by Caco-2 cells. Moreover, transport studies showed that Caco-2/HT-29 coculture cell monolayers were more permeable to pNLC-20% than to the other three nanoparticles. In situ single pass perfusion studies showed that P_eff of pNLC-20% was significantly higher than the other three nanoparticles in duodenum, jejunum and ileum. This result was also confirmed by the ligated intestinal loops model in vivo. pNLC-20% exhibited the strongest fluorescence intensity in tissues 8 h post-injection by in vivo imaging. In addition, the pharmacokinetic study indicated that C_max of pNLC-20% was the highest of all the groups. Together, the results showed that the oral bioavailability increased after PEG-modification. It was necessary to get the optimal PEG density. pNLC-20% might be the superior level of PEG-modification for highly efficient penetration of the viscoelastic mucus layers and intestinal epithelium of the GI tract following oral administration.

Conflict of interest

The authors declare no competing financial interest.

Acknowledgements

This work was financially supported by the National Natural Science Foundation of China (No. 81373983 and 81573377), Natural Science Foundation (No. 20141352) and Six Talent Peaks Project (No. SWYY-011) of Jiangsu Province, Project Program of State Key Laboratory of Natural Medicines, China Pharmaceutical University (No. SKLNMZZJQ201603, SKLNMKF201608) and the Fundamental Research Funds for the Central Universities (No. 2016PT067).

References

1. A. C. Silva, A. Kumar, W. Wild, D. Ferreira, D. Santos and B. Forbes, Int. J. Pharm., 2012, 436, 798–805.
2. M. Schenk and C. Mueller, Best Pract. Res., Clin. Gastroenterol., 2008, 22, 391–409.
3. N. S. Ranpise, S. S. Korabu and V. N. Ghodake, Colloids Surf., B, 2014, 116, 81–87.
4. L. Hongying, X. Yanyu, N. Jiangxiu, C. Xi and P. Qineng, J. Nanosci. Nanotechnol., 2011, 11, 8547–8555.
5. Y. Shi, Z. Su, S. Li, Y. Chen, X. Chen, Y. Xiao, M. Sun, Q. Ping and L. Zong, Mol. Pharm., 2013, 10, 2479–2489.
6. A. Saupe, S. A. Wissing, A. Lenk, C. Schmidt and R. H. Müller, Bio-Med. Mater. Eng., 2005, 15, 393–402.
7. S. Doktorovova, E. B. Souto and A. M. Silva, Eur. J. Pharm. Biopharm., 2014, 87, 1–18.
8. L. M. Ensign, R. Cone and J. Hanes, Adv. Drug Delivery Rev., 2012, 64, 557–570.
9. K. Maisel, L. Ensign, M. Reddy, R. Cone and J. Hanes, J. Controlled Release, 2015, 197, 48–57.
10. A. W. Larhed, P. Artursson and E. Björk, Pharm. Res., 1998, 15, 66–71.
11. H. Takeuchi, H. Yamamoto, T. Niwa, T. Hino and Y. Kawashima, Pharm. Res., 1996, 13, 896–901.
12. S. K. Lai, Y. Y. Wang and J. Hanes, Adv. Drug Delivery Rev., 2009, 61, 158–171.
13. E. Roger, F. Lagarce, E. Garcion and J. P. Benoit, Nanomedicine, 2010, 5, 287–306.
14. J. M. Harris and R. B. Chess, Nat. Rev. Drug Discovery, 2003, 2, 214–221.
15. S. K. Lai, Y. Y. Wang, K. Hida, R. Cone and J. Hanes, Proc. Natl. Acad. Sci. U. S. A., 2010, 107, 598–603.
16. Y. Y. Wang, S. K. Lai, S. Jung Soo, P. Ama, C. Richard and H. Justin, Angew. Chem., Int. Ed., 2007, 47, 9726–9729.
17. Q. Xu, N. J. Boylan, S. Cai, B. Miao, H. Patel and J. Hanes, J. Controlled Release, 2013, 170, 279–286.
18. S. K. Lai, J. S. Suk, A. Pace, Y. Y. Wang, M. Yang, O. Mert, J. Chen, J. Kim and J. Hanes, Biomaterials, 2011, 32, 6285–6290.
19. K. Knop, R. Hoogenboom, D. Fischer and U. S. Schubert, Angew. Chem., Int. Ed., 2010, 49, 6288–6308.
20. E. Oh, J. B. Delehanty, K. E. Sapsford, K. Susumu, R. Goswami, J. B. Blanco-Canosa, P. E. Dawson, J. Granek, M. Shoff, Q. Zhang, P. L. Goering, A. Huston and I. L. Medintz, ACS Nano, 2011, 5, 6434–6448.
21. Z. Su, J. Niu, Y. Xiao, Q. Ping, M. Sun, A. Huang, W. You, X. Sang and D. Yuan, Mol. Pharm., 2011, 8, 1641–1651.
22. Y. Sadzuka, A. Nakade, R. Hirama, A. Miyagishima, Y. Nozawa, S. Hirota and T. Sonobe, Int. J. Pharm., 2002, 238, 171–180.
23. Y. Liu, S. Gao, S. Asghar, G. Liu, J. Song, W. Xuan, Q. Ping, C. Zhang and Y. Xiao, Int. J. Biol. Macromol., 2014, 72C, 1391–1401.
24. S. K. Rodal, G. Skretting, O. Garred, F. Vilhardt, D. B. Van and K. Sandvig, Mol. Biol. Cell, 1999, 10, 961–974.
25. F. Gao, Z. Zhang, H. Bu, Y. Huang, Z. Gao, J. Shen, C. Zhao and Y. Li, J. Controlled Release, 2011, 149, 168–174.
26. H. Yuan, C. Y. Chen, G. H. Chai, Y. Z. Du and F. Q. Hu, Mol. Pharm., 2013, 10, 1865–1873.
27. M. Garcia-Fuentes, D. Torres, M. Martin-Pastor and M. J. Alonso, Langmuir, 2004, 20, 8839–8845.
28. K. Shimada, A. Miyagishima, Y. Sadzuka, Y. Nozawa, Y. Mochizuki, H. Ohshima and S. Hirota, J. Drug Targeting, 1995, 3, 283–289.
29. D. E. O. Iii and N. A. Peppas, Int. J. Pharm., 2006, 307, 93–102.
30. R. H. Müller, M. Radtke and S. A. Wissing, Int. J. Pharm., 2002, 242, 121–128.
31. J. E. Schnitzer, J. Liu and P. Oh, J. Biol. Chem., 1995, 270, 14399–14404.

---