Materials

L-carnitine and paraformaldehyde were purchased from Aladdin (Shanghai, China). Dexamethasone and coumarin 6 were purchased from Macklin (Shanghai, China). Labrafil ® M 1944 CS was kindly gifted from Gattefossé (Lyon, France). Tween 80 was purchased from BASF (Ludwigshafen, Germany). α-methyl-DL-tryptophan, primary antibody against ATB0,+ and OCTN2, Alexa Fluor 594 labeled secondary antibody, and rabbit anti-ATB0,+ antibody were purchased from Sigma-Aldrich (St. Louis, Missouri, USA). Rabbit anti-OCTN2 antibody was purchased from ABclonal (Woburn, USA). BCA kit, Cell Counting Kit-8 (CCK8), RAPI buffer, and DAPI (4’,6-diamidino-2-phenylindole) were purchased from Beyotime (Shanghai, China). Rhodamine B and endocytosis inhibitors including chlorpromazine, indomethacin, colchicine, and quercetin were purchased from Aladdin (Shanghai, China). DEX solution was prepared as follows: DEX was first dissolved in dimethyl sulfoxide (DMSO) and then the stock solution was diluted to a final DEX concentration of 0.5 mg/mL and DMSO concentration of 5% [24].

Human corneal epithelial cells (HCECs) were kindly gifted by Prof. Hongtao Song (Department of Pharmacy, 900 Hospital of the Joint Logistics Team, Fuzhou, China). HCECs were incubated in a DMEM medium including 10% FBS, 100 IU/mL penicillin, and 100 µg/mL streptomycin. The cells were incubated at 37 °C with 5% CO2.

New Zealand albino rabbits (2.5–3.0 kg) were purchased from the animal center of Nantong University. All the animal studies were approved by Nantong University Ethics Committee.

Synthesis and characterization of stearoyl L-carnitine

Stearoyl L-carnitine was synthesized as described previously [25]. Details are provided in the Supporting information.

Preparation and characterization of stearoyl L-carnitine-modified nanoemulsions (SC-NEs)Preparation of SC-NEs

SC-NEs were prepared by emulsion solvent evaporation method. Briefly, the oil phase was prepared by dissolving DEX with the required amount (3 mg), Labrafil ® M 1944 CS (50 mg), Tween 80 (50 mg), and different amounts of SC in ethanol under magnetic stirring at 75 °C. The ratios of SC to the amount of Labrafil ® M 1944 CS were set as 5%, 10%, 20%, 40%, and 80%, respectively. The deionized water (5 mL) was used as an aqueous phase by heated to 75 °C. The aqueous phase was added dropwise to the oil phase under magnetic stirring at 75 °C. The resultant primary emulsions were sonicated using a probe sonicator (Scientz, China). Subsequently, the obtained nanoemulsions were centrifuged to remove the unloaded drugs. Similarly, the unmodified NEs were prepared in the same as the aforementioned procedure except without the addition of SC.

Characterization of SC-NEsParticle size and zeta potential

The particle size and zeta potential were determined by a 90 Plus Particle Size Analyzer (Brookhaven, USA) [26]. The morphology of DEX NEs and DEX SC-NEs was visualized by a transmission electron microscope (TEM) (Hitachi, Japan) [27].

Encapsulation efficiency

The EE of DEX in NEs and SC-NEs was assayed by an ultrafiltration method. The unencapsulated drug was removed by ultrafiltration with an ultrafilter (MWCO: 5 kDa, Beijing Genosys). Briefly, 0.5 mL of nanoemulsions was added into an ultrafilter and centrifuged at 4000 rpm for 15 min. The DEX concentration in the filtrate was measured by HPLC. To assay the amount of drug in nanoemulsion suspensions, ethanol was placed to dissolve nanoemulsions followed by water sonication and centrifugation (13,000 rpm, 15 min). The DEX concentration in the filtrate and nanoemulsion suspensions was determined by high-performance liquid chromatography (HPLC). The HPLC analysis was performed on a Waters HPLC system consisting of a Waters 1525 binary pump, a Waters 2487 dual λ absorbance detector, and a Waters 717 plus autosampler and equipped with a Waters C18 column (5 μm, 0.46 × 25 cm). The mobile phase was composed of acetonitrile and water (40:60, v/v). The flow rate was 1.0 mL/min, and the detection wavelength was 240 nm [28]. The EE was calculated as follows.

$$\mathrm=\frac}}_}- }}_}}}}_}}\times 100\mathrm$$

Differential scanning calorimetry (DSC) and fourier transform infrared spectroscopy (FTIR)

The physical state of DEX in different formulations was analyzed by differential scanning calorimeter (Netzsch, Germany). After freeze-drying, the samples were weighed and scanned (50–300 °C) at a rate of 10 °C/min under nitrogen flow. The intermolecular interaction was analyzed using FTIR (Niclet is50, Thermo, USA). The spectrum was obtained using the KBr disk method in the range of 500 cm−1 to 4000 cm−1.

In vitro release

The in vitro DEX release behavior from DEX solution, NEs, and SC-NEs containing different amounts of SC (5, 10, 20, 40, 80%) was investigated by a dialysis method. Briefly, 0.5 mL of SC-NEs was dropped into a dialysis bag (MWCO: 8 kDa) and placed in 10 mL of simulated tear fluid at 35 ± 0.5 °C with continuous shaking (120 rpm). At the designated time point, an aliquot of the sample was taken out and replaced with a fresh medium. The released DEX content was determined by the HPLC method as described above.

In vitro stability

The in vitro stability of different formulations was studied at 4 °C. The drug content changes were determined at predetermined time points.

Expression of OCTN2 and ATB0, + in human corneal epithelial cells and corneal tissuesExpression of OCTN2 and ATB.0,+ in human corneal epithelial cells (HCECs)

The expression of OCTN2 and ATB0,+ in HCECs was observed by immunofluorescence. HCECs were seeded on a 12-mm coverslip in 24-well plat at 1 × 105 cells/well density. After cells grew to 90% confluence, the cells were fixed with 4% paraformaldehyde for 20 min and blocked with 3% BSA solutions for 30 min. Then, the primary antibody against ATB0,+ and OCTN2 was added to the cells and cultured for 12 h at 4 °C. Alexa Fluor 594 labeled secondary antibody was added to the cells and incubated for 60 min at 37 °C. The cell nucleus was stained with DAPI. Lastly, the cells were observed by fluorescence microscope (Carl Zeiss, Germany).

Expression of ATB0,+ and OCTN2 in rabbit cornea

The expression of ATB0,+ and OCTN2 in rabbit corneal tissue was also examined by immunofluorescence analysis. The rabbits were euthanized with air injection. The corneal tissues were excised and rinsed with ice-cold pH 7.4 phosphate-buffered saline. For immunofluorescence analysis, the corneal tissues were fixed with 4% paraformaldehyde and paraffin sections of 10 μm were prepared. The sections were blocked with 3% BSA solutions for 1 h. The samples were stained with rabbit anti-ATB0,+ antibody and rabbit anti-OCTN2 antibody for 12 h. Lastly, the corneal tissues were observed by fluorescence microscope (Carl Zeiss, Germany).

Cellular uptake and mechanism studies in HCECsCellular uptake of SC-NEs

HCECs were seeded in a 24-well plate at 1 × 105 cells/well density and cultured for 24 h. Fluorescence dye coumarin 6 (C6)-loaded NEs or SC-NEs were prepared as the same as the procedure in "Preparation and characterization of stearoyl L-carnitine-modified nanoemulsions (SC-NEs)" section except that DEX was replaced with C6. When the cells reach 90% confluence, C6 (2.5 μg/mL)-loaded NEs (C6 NEs) or SC-NEs (C6 SC-NEs) in NaCl buffer (140 mM NaCl, 25 mM Hepes/Tris, 5.4 mM KCl, 1.8 mM CaCl2, 0.8 mM MgSO4, and 5 mM glucose, pH 7.4) was added to the cells and cultured for 1 h. Next, the cells were rinsed with cold NaCl buffer 3 times and were solubilized by RAPI buffer for 60 min. The fluorescence intensity was measured using a multimode microplate reader (Bio Tek, USA) at λex = 460 nm and λem = 505 nm. Protein concentration was measured by a BCA kit.

For the fluorescence imaging study, HCECs were seeded on a 12-mm coverslip in a 24-well plate at 1 × 105 cells/well density. After cells reach 90% confluence, the medium was removed, and C6 SC-NEs in NaCl buffer were added and cultured for 1 h. The nucleus was stained by DAPI. The cell was observed under a Leica fluorescence microscope.

Cellular uptake mechanisms

To confirm the effect of ATB0,+ and OCTN2 in the uptake of SC-NEs, α-methyl-DL-tryptophan (α-MT, 2.5 mM) was selected as a specific inhibitor of ATB0,+, and L-carnitine (10 mM) was selected as a competitive inhibitor for both ATB0,+ and OCTN2.

To investigate Na+ and Cl− on the cellular uptake of SC-NEs, Na+ or Cl− free buffer was selected in this study. Na+ free buffer was prepared by substituting NaCl in the buffer with methylglucamine hydrochloride. The gluconic acid sodium salt, gluconic acid potassium salt, and gluconic acid calcium salt were used instead of NaCl, KCl, and CaCl2 for preparing Cl− free buffer.

To investigate the endocytosis pathway of SC-NEs, different endocytosis inhibitors (chlorpromazine: 50 μM, indomethacin: 100 μM, colchicine: 10 μM, quercetin: 10 μM) were added to the 24-well plate containing HCECs at 1 × 105 cells/well density. After treatment for 30 min, the medium was withdrawn and C6 SC-NEs were added to cells and cultured for 1 h. Cells without treatment were set as a control group.

Corneal permeation study

The in vivo corneal permeation of SC-NEs in rabbit eyes was evaluated by the fluorescence imaging technique. Briefly, C6 NEs and C6 10% SC-NEs (50 μL, 0.06 mg/mL) were instilled into the right eyes every 2 min for 5 times consecutively. At the predetermined time points (15 min and 60 min), the rabbits were euthanatized, and their corneas were removed for the frozen section. The cell nucleus was stained with DAPI, and the slide was observed by a Leica fluorescence microscope. The relative fluorescence intensity was quantified by Image J software.

The ex vivo corneal permeation of SC-NEs was studied using a vertical franz cell system (RYG-6B, Shanghai, China). Briefly, after New Zealand albino rabbits were euthanized, the corneas were isolated and washed with glutathione bicarbonate ringer's (GBR) solution. Then, the cornea was clipped in the middle of the donor chamber and the receptor chamber with the epithelial surface toward the donor chamber. 7 mL of glutathione bicarbonate GBR solution was added into the receptor chamber. 1 mL of DEX solution, DEX NEs and DEX 10% SC-NEs (0.1 mg/mL) was added into the donor chamber. The cell diameter is 1 cm. The water bath temperature was set at 35 ± 0.5 °C. At different time points, equal volume (20 μL) of sample was taken from the receptor chamber. The samples were diluted with methanol and drug concentration was measured by HPLC. The cumulative permeation (Q), permeability coefficient (Papp) and steady-state flux (Jss) were calculated using the following equations:

$$_=\frac_}(_+\frac_}\sum_^_)$$

where V0 is the volume of receptor chamber; V is the sampling volume; Cn is the drug concentration at the sampling time points; Ci is drug concentration at the previous time points; A is the effective diffusion area; C0 is the initial drug concentration;ΔQ/Δt is the slope of regression equation.

Ocular surface retention study

Fluorescence imaging was used to investigate the ocular surface retention of SC-NEs and Rhodamine B (RhB) was selected as a fluorescence probe [29]. RhB-loaded NEs (RhB NEs) or SC-NEs (RhB SC-NEs) were prepared as the same as the procedure in "Preparation and characterization of stearoyl L-carnitine-modified nanoemulsions (SC-NEs)" section except that DEX was replaced with RhB. Rabbits were topically administered with RhB NEs and RhB 10% SC-NEs (20 μL, 1 mg/mL). The Ocular surface fluorescence signals were captured by an in vivo fluorescence imaging system (Bruker, USA) at 1, 30, and 60 min, respectively.

In vivo pharmacokinetics

Rabbits were randomly divided into three groups and each group had three rabbits (n = 3). After topical administration with DEX solution, DEX NEs, and DEX 10% SC-NEs (50 μL, 0.5 mg/mL) into the lower conjunctival sac, the rabbits were euthanized with air injection at 10, 20, 40, 60, 90, and 120 min. After euthanized, an insulin needle was used to puncture the anterior chamber to collect aqueous humor and isolate the cornea tissue from the eyeball at each time point. The aqueous humor (0.15 mL) was mixed with 0.15 mL methanol via the vortex. The cornea tissues were homogenized with 0.5 mL methanol. Then, the aqueous humor samples and cornea homogenates were centrifuged (12,000 rpm, 10 min), and the DEX concentration in the supernatant was determined by the HPLC method as described above.

In vivo anti-inflammatory efficacy

The in vivo anti-inflammatory efficacy of DEX SC-NEs was investigated using the EIU model in rabbits. After establishing the EIU model by intravitreal injection of Lipopolysaccharide (0.1 μgE. O111:B4, Sigma-Aldrich), DEX solution, DEX NEs, and DEX 10% SC-NEs (50 μL, 0.5 mg/mL) were dripped into the inferior conjunctival sac (4 times over 24 h). The clinical signs were evaluated with a slip lamp examination. At 24 h after uveitis induction, the rabbits were killed and the iris-ciliary body was collected for mRNA measurement. Total RNA was isolated with TRIzol reagent (Invitrogen, CA), and reverse-transcribed to cDNA using HiScript® III RT SuperMix for qPCR (+ gDNA wiper) kit (Vazyme, China). The qRT-PCR was achieved by ChamQ Universal SYBR qPCR Master Mix (Vazyme, China). The primer sequences of monocyte chemoattractant protein 1 (MCP-1), Matrix metalloproteinases 1 (MMP-1), and vascular cell adhesion molecule 1(VCAM-1) are listed in Additional file 1: Table S3. GAPDH was used as a reference gene. In addition, the eyeballs were fixed in formaldehyde, embedded in paraffin, sectioned, and stained with hematoxylin and eosin (H&E) for histopathological examination.

Ocular safety assessmentIn vitro cytotoxicity test in HCECs.

The cytotoxicity assay was performed on HCECs. Briefly, HECEs (1 × 104 cells/well) were seeded into the wells of 96-well plates and incubated overnight. Then, DEX 10% SC-NEs were added to the wells. After 24 h incubation, CCK8 reagent was added to the wells and cultured for 1 h. The absorbance values were determined with a microplate reader (Bio Tek, Vermont, USA) at 490 nm.

In vivo ocular irritation in rabbit eyes

A modified Draize eye irritation test was used to investigate the ocular safety of SC-NEs in vivo [30]. First, the rabbit eyes were topically administered with DEX solution, DEX NEs, and DEX 10% SC-NEs (50 μL, 0.5 mg/mL). The clinic signs of the ocular anterior segment were assessed by slit lamp at 24 h after instillation. The corneal surface integrity was also evaluated by fluorescein staining. Finally, the rabbit was euthanatized by intravenous air injection, and the eyeball was taken out to prepare hematoxylin and eosin (H&E) pathological sections.

Statistical analysis

All values are presented as mean ± standard deviation (SD). The Student’s t-test was used to analyze the statistical difference between the data. The P-value < 0.05 was considered statistically significant.