Condition responsive nanoparticles for managing infection and inflammation in keratitis

Saad M. Ahsan and Ch. Mohan Rao *
Centre for Cellular and Molecular Biology (CCMB), Council of Scientific and Industrial Research (CSIR), Uppal Road, Hyderabad – 500 007, Telangana, India. E-mail:; Web: Fax: +91-40-27195633; Tel: +91-40-27195631

Received 7th February 2017 , Accepted 31st May 2017

First published on 5th June 2017

Keratitis is a major cause of avoidable visual impairment. About 30% of patients with fungal keratitis eventually become permanently blind in the developing world. Proteases, secreted by the pathogen and the host, damage the cornea before the infection is resolved. Treating keratitis is a challenge because both infection and inflammation need to be addressed. An additional challenge is to maintain a therapeutic dose at the corneal surface as blinking and tear film wash away the drugs, administered as eye drops. We have developed a nanoparticle-based drug delivery system that enhances the drug residence time by anchoring to the cornea, down-regulates inflammation and releases the antifungal drug: all in a condition-responsive manner. The expression of Toll-Like Receptors (TLR4) on the corneal epithelial cells increases in response to infection. We have conjugated anti-TLR4 antibodies on the surface of ketoconazole-encapsulated gelatin nanoparticles. The anti-TLR4 antibody not only facilitates binding of nanoparticles to the cornea, enhancing their residence time, but also reduces the levels of inflammatory cytokines. Host and fungal proteases degrade the gelatin nanoparticle, an alternative substrate for proteases, thereby reducing corneal damage and releasing the encapsulated drug, ketoconazole, proportional to the severity of infection. After testing the efficacy of the system with human corneal epithelial cells, we have extended our studies to a rat model of keratitis. The results show a significantly increased corneal retention, suppressed inflammation and resolution of infection in the infected eyes. We believe that this will be an excellent approach to manage keratitis as well as other topical ocular infections.


The eye is an important sensory organ in most living systems. Due to its superficial location and immune-privileged nature1 the eye, during the course of evolution, has developed various anatomical and physiological barriers to protect itself from any external damage.2 However, the same barriers come in the way of treatment when the eye is infected.3 Keratitis is one such situation, where the cornea is either infected (infectious keratitis)4–7 or damaged (sterile keratitis).8–10 Although largely restricted to the farming population in tropical countries,5,11 the occurrence of keratitis, in recent times, has been on the rise in the developed world as well.12 This is mainly due to an increase in the use of contact lenses, increased surgical interventions at the corneal surface and an indiscriminate use of steroids in ocular therapies.8–10,13–15 Treatment of infectious keratitis requires the administration of antimicrobial and/or anti-inflammatory drugs as eye drops for topical instillation. However, the continuous dilution and washing effect of the tear film results in less than 5% of the applied drug being effective post instillation.16,17 Treatment of infectious keratitis thus requires rapid and intense antibiotic therapy18 complemented with an anti-inflammatory drug to minimize inflammation and host- and pathogen-mediated damage to the corneal surface.19

Successful treatment for keratitis requires multiple challenges to be addressed in a single formulation. A fundamental requirement is to strike a balance in the drug dosage along with maintaining therapeutic concentrations for prolonged periods. Designing a stimuli-responsive system is speculated to eliminate multiple spikes and drops in drug concentrations along with the associated toxicity and lack of response respectively.

Nanotechnology offers a unique opportunity to develop therapeutic formulations that can simultaneously tackle both aspects of keratitis viz. elimination of infection and controlling inflammation. Systems utilized for ocular drug delivery fall into two major categories viz. film-based and particle-based systems. Film-based systems may include hydrogel lens,20,21 wafers,22 corneal shields,23etc.; while particle based systems may include nanosuspensions,24 nanoparticles,25 liposomes,26 micelles,27,28etc. Both these systems have their own benefits and drawbacks. However, a major disadvantage of these systems has been their non-specific nature and less control over drug release. Most nanocarriers release drug by Fickian diffusion which has a major disadvantage of being unregulated. Stimuli-responsive nanocarriers can overcome this drawback by providing an on-demand delivery of drugs.29 Despite several advances in stimuli-responsive nanocarriers, no such system has been developed for keratitis management.

In this context, the development of a nanostructure-based drug delivery system with an ability to sense the severity of an infection/inflammation and modulate drug release accordingly is extremely important. We demonstrate ketoconazole (ket) loaded gelatin nanostructures which are lysed by the action of proteases in the microenvironment of the infection, thereby leading to ket release and action. The surface of the nanostructure is conjugated with anti-TLR4 antibodies, which aid in infection/inflammation sensing, corneal adhesion and suppress inflammation. As TLR4 over-expression in the corneal epithelium is a direct consequence of infection/inflammation severity, the nanoparticle binding and its subsequent anti-inflammatory actions are thus modulated accordingly. The present approach demonstrates a simple yet smart and novel method to precisely control anti-microbial and anti-inflammatory activities of the administered agents over prolonged periods. The proof-of-concept study demonstrated here creates a paradigm for future studies in developing nanoparticle based targeted formulations for managing corneal infections and inflammation.



Gelatin, from bovine skin, lime-cured (Type B), with a bloom strength of 225, penicillin, streptomycin, insulin, epidermal growth factor from murine submaxillary gland (EGF), bovine serum albumin (BSA) and methyl-β-cyclodextrin were from Sigma-Aldrich (St Louis, MO, USA), acetone (HPLC grade) was from Spectrochem (Mumbai, India) and glutaraldehyde (25% aqueous solution) was from Calbiochem. Human corneal epithelial (HCE) cells were a gift from L.V. Prasad Eye Institute (Hyderabad, India). Aspergillus flavus (NCIM535) was obtained from CSIR-IMTECH (Chandigarh, India). A SuperScript® First-Strand Synthesis System for RT-PCR was from Invitrogen™ (Thermo Fisher Scientific, Waltham, MA, USA), glucose was from Thermo Scientific, tryptone and agar were from HiMedia (Mumbai, India), and PCR mix and primers were from Bioserve India (Hyderabad, India). Anti-TLR4 and anti-actin antibodies were from Abcam (Cambridge, UK), horseradish peroxidase (HRP) conjugated secondary anti-mouse and anti-rabbit antibodies were from PerkinElmer (Waltham, MA, USA). The FITC conjugated IgG antibody (goat anti-rabbit FITC) was from Santa Cruz and pronase was from Roche. Ketoconazole, ketamine hydrochloride and prednisolone were purchased from local vendors, eosin yellowish was purchased from Merck, and hematoxylin LR was purchased from Thomas Baker.

Drug dissolution studies

An excess amount of ketoconazole in 10 mL of deionized water was dissolved using 2.5 mM, 5 mM, 7.5 mM, 10 mM, 15 mM and 20 mM methyl-β-cyclodextrin (mβCD). The contents of each tube were stirred for 3 days until equilibrium. After equilibration, undissolved ketoconazole was separated by filtration with a 0.45 μm PVDF membrane syringe filter. The resulting solution was then analyzed for ketoconazole content by measuring the absorbance at 260 nm after dilution of samples using a Lambda Model UV-Visible spectrophotometer (PerkinElmer, Waltham, MA, USA). The phase solubility diagram was obtained by plotting the concentration of dissolved ket (mM) versus the concentration of mβCD added (mM). The stability constant Ks was calculated from the equation: Ks = slope/So(1 − slope), where slope is the value obtained by linear fitting of the phase solubility curve and So is the intrinsic solubility of ket in water in the absence of mβCD.30,31

FT-IR analysis of ket-mβCD complex

Infrared spectra were obtained at a resolution of 4 cm−1 on a Fourier transform instrument Bruker Alpha T, Bruker Optik, (Ettlingen, Germany), equipped with a single reflection diamond attenuated total reflectance accessory. Samples were dried on the surface of the diamond crystal element by evaporating the solvent. 64 scans were recorded at a scan speed 200 cm per minute and averaged.

Synthesis of gelatin nanoparticles

The double-desolvation method was used for nanoparticle synthesis, partially modified by the method described earlier.32–35 Briefly, a 5% w/w solution of gelatin type B (Bloom225) was rapidly desolvated by the addition of equal amounts of acetone. After precipitation of the high molecular weight gelatin fraction, the supernatant containing soluble low molecular weight gelatin was discarded. The sediment was redissolved in water under gentle heating (50 °C). The redissolved high molecular weight gelatin was snap frozen in liquid nitrogen and lyophilized. The lyophilized gelatin was stored at 4 °C until further use. 0.1 g of freeze-dried high molecular weight gelatin was then dissolved in 10 mL of deionised water under gentle heating (50 °C) to obtain a 1% solution. The solution was filtered through a 0.22 μm filter and the pH was adjusted to 3.5 with 0.1 N HCl. Nanoparticle formation was initiated by the drop-wise addition of acetone (second desolvation step) under continuous stirring (500 rpm). After addition of the desired amount of acetone (∼70%), 20 μl glutaraldehyde (25%) diluted in 1 mL acetone was added to the reaction mixture to cross-link nanoparticles. After stirring for 12 hours (h), the particles were purified by centrifugation at 48[thin space (1/6-em)]000g for 10 min and redispersed in acetone/water (30/70 mixture) thrice. The purified nanoparticles were stored as dispersion in deionized water at 4–8 °C or lyophilized.

Synthesis of ket loaded GNPs

For the synthesis of ket-loaded nanoparticles, gelatin was dissolved in 10 mL of the prepared ket-mβCD complex (0.1, 0.5, 1, 5 and 10 mM of mβCD concentration) solution (before the second desolvation step) to obtain a gelatin (type B) concentration of 1% w/v.36 The protocol for nanoparticle synthesis mentioned above was followed thereafter. Drug-loading of ket in gelatin nanoparticles was determined by resuspending and incubating a known amount of particles in methanol for 2–3 h under continuous stirring in a rotator. The suspension was centrifuged and the supernatant was analyzed spectrophotometrically at 260 nm. The drug loading efficiency and loading content were calculated by comparing the OD260 with the standard curve obtained by the spectrophotometric analysis of increasing concentrations of ket in methanol.

Preparation of anti-TLR4 conjugated GNPs

Antibody conjugation to the nanoparticles was carried out using carbodiimide chemistry. Briefly, nanoparticles were suspended in 0.1 M MES buffer (pH 6.0) containing 0.5 M NaCl. The surface carboxylate functional groups of the nanoparticles were activated with EDC (2 mM) and stabilized by NHS (5 mM) by incubation at room temperature for 2 h with vortexing. After incubation, the particles were washed thrice by centrifugation at 48[thin space (1/6-em)]000g for 10 min and redispersed in PBS (pH 7.4). For the conjugation of FITC-IgG, increasing amounts of the antibody (0, 0.02, 0.05, 0.1, 0.2, 0.3, 0.5, 1.0, 1.5 and 2.0 μg) were added per mg of EDC-activated nanoparticles suspended in PBS, pH 7.4. The samples were incubated for 24 h at 4 °C, washed with PBS and analyzed for antibody conjugation by flow cytometry using a BD FACSCalibur™ cell analyzer (Becton, Dickinson and Company, Franklin Lakes, NJ, USA). Particles were gated on forward and side scatter to exclude clumps, and 10[thin space (1/6-em)]000 gated events were recorded. Data analysis was performed by Cell Quest software provided by the supplier. For the conjugation of anti-TLR4, 2 μg per mL antibody solution were added drop-wise to the activated nanoparticles and vortexed at 4 °C for 24 h. The conjugated nanoparticles were separated by centrifugation (48[thin space (1/6-em)]000g, 10 min) and stored at 4 °C until further use.

Characterization of anti-TLR4-ket-mβCD-GNPs

Gelatin nanoparticles were resuspended in PBS (pH 7.4) and analyzed for their size and polydispersity in a Nanopartica nanoparticle analyzer system (Horiba Scientific, Kyoto, Japan) equipped with a diode-pumped solid-state (DPSS) laser of wavelength 532 nm and a temperature controller unit. For TEM analysis particles were resuspended in MQ water and placed on formvar/carbon coated - copper 300 mesh grids. Excess sample was removed by blotting with filter paper and the grid was air dried. The particle-bearing grids were counterstained with a 1% w/v uranyl acetate solution for 1 minute. The samples were examined with a JEM-2100 (M/S Jeol Limited, Tachikawa, Tokyo, Japan) transmission electron microscope (TEM) at 100 kV accelerating voltage.

For stability analysis, gelatin nanoparticles were synthesized as described previously. Post synthesis particles were filtered through a 0.45 μm syringe filter and stored at 4 °C. At various time intervals (once every week), aliquots were taken and diluted in phosphate buffered saline (PBS) (100 mM, pH 7.4). The particles were analyzed for size and zeta potential in a Nanopartica nanoparticle analyzer system (Horiba Scientific, Kyoto, Japan).

Minimum inhibitory concentration (MIC) of ket-mβCD-GNPs

MIC was determined by the protocol suggested by Espinel-Ingroff et al.37 Briefly, Aspergillus flavus was grown on Sabouraud dextrose agar (SDA) for 7 days at 35 °C. After 7 days, 1 mL of sterile PBS was added to the culture and the flask was gently vortexed. The spore suspension was taken out and placed in a sterile tube for a few minutes, allowing the heavy particles to settle. The upper homogeneous suspension was transferred to a sterile tube. The density of the spore suspensions was read at 560 nm and adjusted to optical densities (OD) in the range of 0.09 to 0.11 (80 to 82% transmittance). These suspensions were diluted 1[thin space (1/6-em)]:[thin space (1/6-em)]100 in the RPMI media (without sodium bicarbonate), which would correspond to 0.4 × 104 to 5 × 104 colony forming units (CFU) per mL. Ketoconazole was diluted and added to different wells of a 96 well plate at increasing concentrations. Similarly, the ket-mβCD complex and gelatin nanoparticles loaded with the ket-mβCD complex were added at various concentrations. The inoculum was incubated at 35 °C for 48 h after which the contents of each well were plated on SDA plates and incubated at 35 °C. Colonies were counted after 2–3 days. MIC values (100% growth inhibition) were calculated based on concentrations of the formulation that showed no colonies post incubation.

Human corneal epithelial (HCE) cell interaction studies

Human corneal epithelial (HCE) cells were cultured in DMEM and Ham's F10 (1[thin space (1/6-em)]:[thin space (1/6-em)]1), supplemented with 5% FBS, 10 ng per mL EGF, 5 μg per mL insulin, antibiotics (5 μg per mL penicillin & 6 μg per mL streptomycin) at 37 °C under 95% humidity and 5% CO2. Cells were grown in serum-free media for 24 h prior to the experiments. The cells were then treated with varying amounts of LPS (0 ng mL−1, 1 ng mL−1, 10 ng mL−1, 100 ng mL−1, 1 μg mL−1 and 10 μg mL−1) for 4 h. In a separate experiment, fungal hyphae antigens were isolated by the method described by Zhao et al.38 Briefly, Aspergillus flavus mycelium was obtained by growing the fungus in Sabouraud media. Mycelia were washed twice with PBS followed by sterilization in 70% ethanol-phosphate buffered saline and incubation at 4 °C for 24 h. The mycelia were then disrupted by using a homogenizer and dried. For cell culture studies, 0.2 μg mL−1 of Aspergillus flavus mycelia was added to HCE cell culture and incubated for 4 h. Post treatment, the cells were harvested by gentle scraping and whole cell protein was extracted by the incubation of cells in lysis buffer (50 mM Tris-Cl, pH 8.0, containing 1% Triton X-100, 1% sodium deoxycholate, 0.1% sodium dodecyl sulphate, 150 mM sodium chloride, 1× protease inhibitor, 1× phosphatase inhibitor and sodium orthovanadate) for 30 minutes on a rotator at 4 °C. After a brief sonication, the lysate was centrifuged (12[thin space (1/6-em)]000g for 20 minutes at 4 °C) to remove any aggregates. The protein in the supernatant was estimated by Bradford's method according to the instructions of the manufacturer. 25 μg of the total cell protein was loaded on to a 12% SDS-polyacrylamide electrophoresis gel. After electrophoresis, the protein was transferred to a Hybond C-extra nitrocellulose membrane using semi-dry transfer apparatus. The membrane was then placed in blocking solution (5% BSA in Tris Buffered Saline (TBS)) for 2 h with gentle shaking at room temperature. The primary antibody was diluted (in 1% BSA in TBS) according to the manufacturer's instructions and the membrane was incubated with the primary antibody overnight at 4 °C. The membrane was then washed with TBS containing 0.1% Tween-20 (TBST) thrice and incubated with a horseradish peroxidase (HRP) conjugated secondary antibody at room temperature for 1 h. The membrane was then washed with TBS containing 0.1% Tween-20 (TBST) thrice and developed by using a Vilber-Lourmat Chemiluminescence Imaging System (MArne-la-Valée Cedex 3, France) using Chemi-Capt software, after adding the HRP substrate. Densitometric analysis and comparative calculations were performed using ImageJ software (CSHL, NY, USA).

For GNP–HCE cell interaction studies, cells were treated with 100 ng per mL of LPS. FITC-labelled anti-TLR4 conjugated gelatin nanoparticles were added at a concentration of 200 μg mL−1 and incubated for 4 h. The cells were washed with PBS and counter-stained with DAPI. Images were acquired on a confocal microscope using a 63× objective lens on a Leica confocal microscope (TCS-SP8; Leica Microsystems, Wetzlar, Germany). Images were analyzed using Leica Application Suite AF software provided by the company. For flow cytometry analysis, FITC-labelled anti-TLR4-GNPs were added to HCE cells under culture conditions at a concentration of 200 μg mL−1 and incubated for 4 h. The cell uptake of FITC-labelled anti-TLR4-GNPs was studied by flow cytometry using a BD FACSCalibur™ cell analyzer (Becton, Dickinson and Company, Franklin Lakes, NJ, USA). The cells were gated on forward and side scatter to exclude cell debris, and 10[thin space (1/6-em)]000 gated events were recorded. Data analysis was performed using Cell Quest software provided by the supplier.

Suppression of inflammation by anti-TLR4-conjugated GNPs

For RT-PCR (reverse transcription-polymerase chain reaction) studies, cells were cultured in serum-free media and subsequently treated with LPS, LPS along with GNPs and LPS along with anti-TLR4-GNPs. Untreated cells served as normal controls. The cells were collected after 4 h of treatment and resuspended in TRIzol® (Thermo Fisher Scientific, Waltham, MA, USA). Alternatively the cells were treated with Aspergillus flavus (as obtained previously) mycelia, Aspergillus flavus mycelia along with GNPs and Aspergillus flavus mycelia along with anti-TLR4-GNPs. The total RNA isolated was given a DNase I (New England Biolabs (NEB), Ipswich, MA, USA) treatment according to the manufacturer's instructions. 1 μg of DNase I treated RNA was taken for cDNA synthesis using the SuperScript® First-Strand Synthesis System for RT-PCR kit as per instructions. For the quantitation of genes, 1 μl of cDNA was used as a template with the corresponding primers for the genes to be studied. The polymerase chain reaction (PCR) products were analyzed by running on a 2% agarose gel in Tris-Acetate-EDTA (TAE) buffer. The resulting bands were documented by using a Gene Genius Classic Gel Documentation System (Syngene, Cambridge, UK) using the GeneSnap software. Densitometric analysis and comparative calculations were performed using ImageJ software (CSHL, NY, USA). The primer sequences used for the PCR are GAPDH forward primer 5′ CAA TGC CTC CTG CAC CAC C 3′, reverse primer 5′ GGT GGC AGT GAT GGC ATG G 3′; IL8 forward primer 5′ TGC AGT TTT GCC AAG GAG TGC, reverse primer 5′ CAG ACA GAG CTC TCT TCC ATC 3′; TNFα forward primer 5′ GAG TGA CAA GCC TGT AGC CC 3′, reverse primer 5′ CCT TGA AGA GGA CCT GGG AG 3′ and MMP2 forward primer 5′ TCA GCC AGC ACC CTG GAG C 3′, reverse primer 5′ GCC AGG ATC CAT TTT CTT CTT CAC 3′.

Biocompatibility studies of gelatin nanoparticles

Human corneal epithelial (HCE) cells were cultured in 35 mm Petri-dishes. Gelatin nanoparticles were added at concentrations of 2, 4, 6, 8 and 10 mg mL−1 to the cells and incubated for 12 h. HCE cells treated with 200 μM H2O2 for 6 h served as a positive control. Cells were scraped and resuspended in phosphate buffered saline (PBS). Propidium iodide (PI) was added at a final concentration of 0.5 μg mL−1, and the samples were analyzed by flow cytometry using a BD FACSCalibur™ cell analyzer (Becton, Dickinson and Company, Franklin Lakes, NJ, USA). The cells were gated on forward and side scatter to exclude cell debris, and 10[thin space (1/6-em)]000 gated events were recorded. Data analysis was performed by using Cell Quest software provided by the supplier.

Protease responsive release of ket

100 mg GNPs were resuspended in PBS with varying concentrations of pronase (0 mg mL−1, 0.01 mg mL−1 & 0.1 mg mL−1) and packed in dialysis tubings with MWCO 8k to 12k. Samples were collected at varying time points and analyzed by using reverse-phase high performance liquid chromatography (RP-HPLC) for the ket content. ket was analyzed using an Agilent 1200 Series reverse-phase high performance liquid chromatography (RP-HPLC) system.39 Separation was carried out using an Agilent 300SB-C18 (5 μm, 4.5 × 250 mm) column. Isocratic elution was carried out with the mobile phase consisting of acetonitrile and 0.2% triethylamine with pH adjusted to 6.4 using phosphoric acid (48[thin space (1/6-em)]:[thin space (1/6-em)]52, v/v) at a flow rate of 1 mL min−1. The mobile phase was filtered under vacuum and degassed. Chromatographic separation was monitored at 260 nm. All the samples were analyzed at room temperature. Total run time for the analysis was 20 minutes.

In vivo studies

28 Wistar rats weighing 200 to 250 grams each were randomly distributed into 7 groups of four rats each. All rats were treated in accordance with the Association for Research in Vision and Ophthalmology (ARVO) statement for the use of animals in ophthalmic vision and research. The protocol was approved by the Institutional Animal Ethics Committee (IAEC). Rats were anaesthetized for the duration of the experiment with an intraperitoneal injection of ketamine hydrochloride. Rats were administered with a subcutaneous injection of prednisolone prior to inoculation of fungal spores in order to establish the infection.40 Rats in all groups received 20 μl of the spore suspension (106 CFU mL−1) in their right eye. The left eye in each rat served as an un-inoculated control. Post inoculation, the rats were examined daily for 7–8 days. Rats that developed keratitis (characterized by a white infiltrate under a slit-lamp biomicroscope) within 48 h post inoculation were considered for the testing of different formulations (2 μg equivalent dose of ket, given at an interval of 12 h each) as follows: Group 1: untreated, Group 2: administered with PBS, Group 3: administered with ket, Group 4: administered with ket-mβCD, Group 5: administered with empty GNPs, Group 6: administered with ket-mβCD-GNPS, and Group 7: administered with anti-TLR4-ket-mβCD-GNPs.

Clinical examinations

Following the inoculation of Aspergillus suspensions into the rat corneas, the eye was examined using a slit-lamp biomicroscope until 7 days for size and depth of infiltrate. The size of infiltrate was expressed as mm (diameter). The corneal disease was graded as described by Guo et al. (Table 3).41

Corneal binding of nanoparticles

Rats were infected with Aspergillus flavus spores as mentioned earlier. On day 1 or 2 post-infection, rats that showed a clinical score of approximately 1 were administered with either FITC-labelled GNPs or FITC-labelled anti-TLR4-GNPs. The rats were sacrificed by asphyxiation 4, 8 and 12 h post administration of the nanoparticle suspension. The eyeballs were collected, snap frozen in liquid nitrogen and stored at −80 °C until further use. Eyeballs were embedded in optimal cutting temperature (OCT) compound tissue freezing medium (Leica Biosystems, Wetzlar, Germany) and cryo-sectioned to obtain 10 μm thick sections on positively charged slides. The sections were kept in isopropanol for 10 min followed by washing with phosphate-buffered saline twice. The sections were counter-stained with DAPI and images were acquired on a confocal microscope using a 63× objective lens on a Leica confocal microscope (TCS-SP8; Leica Microsystems, Wetzlar, Germany). Images were analyzed by using Leica Application Suite AF software provided by the company.

Microbiological evaluation of rat eyes

Rats were sacrificed by asphyxiation after the entire duration of the experiment. Eyeballs were excised from the eye socket under sterile conditions. The entire eyeball was inoculated on a SDA plate and incubated at 35 °C for 7 days. The plates were examined daily for 7 days for the growth of Aspergillus flavus.

Histopathological evaluation of rat eyes

Rats were sacrificed by asphyxiation after the entire duration of the experiment. Eyeballs were excised from the eye socket and fixed in 10% formalin. The cornea was isolated and embedded in paraffin wax for sectioning. The paraffin sections were subjected to staining by haematoxylin and eosin (H&E) stain for the visualization of inflammatory cells and tissue integrity.

Statistical analysis

An unpaired, two-tailed Student's t-test was used to determine statistical significance. Data were considered significant at p < 0.05. All values were expressed as means ± S.D. (standard deviation) of at least three independent experiments. Data analysis was performed in Microsoft Excel and OriginPro 8 software from OriginLab.

Results and discussion

Nanoparticle synthesis and characterization

Gelatin nanoparticle synthesis was carried out by the double desolvation method.32,33 We have shown, in our recent study, that the pH and acetone concentrations govern the matrix density of the particles which in turn affects nanoparticle degradation rates and drug release.35 For optimal nanoparticle preparation and ensuring the maximum response to protease-based degradation, synthesis was carried out with high molecular weight gelatin fractions obtained after the first desolvation with pH adjusted to 3.5. Acetone was added to a final concentration of 75% to induce precipitation and nanoparticle formation.35 To obtain high drug loading efficiencies, ket was complexed with methyl-β-cyclodextrin (mβCD).36 Complexation with mβCD increases the aqueous solubility of hydrophobic drugs.42,43 The phase-solubility diagram for the ket-mβCD complex (Fig. 1b) was found to be linear (AL type of curve) up to a mβCD concentration of 10 mM. Further increasing the mβCD concentration resulted in a positive deviation (AP type of curve, data not shown). The stability constant for a 1[thin space (1/6-em)]:[thin space (1/6-em)]1 complex (K1:1) was found to be 3.3 × 103 M−1 (calculated by the Higuchi–Connors equation).44 FTIR of the ket-mβCD complex suggested the inclusion of ket in the hydrophobic cavity of the mβCD molecule. The complex shows the characteristic peaks of mβCD at 1642 cm−1, 1453 cm−1, 1331 cm−1, 1194 cm−1, 1154 cm−1, 1082 cm−1, 1037 cm−1 and 965 cm−1 (Fig. 1a). However, the characteristic peaks of ket at 1585 cm−1, 1510 cm−1 (C–C and N–H stretchings), 1458 cm−1 (C–C and C–H stretchings), and 1291 cm−1 and 1231 cm−1 (C–N and C–O stretchings) were not observed for the complex, due to the decrease in the corresponding bond vibrations of ket post inclusion in the cyclodextrin cavity. Such a loss in an IR signal of the drug upon inclusion in the cyclodextrin cavity has been reported earlier.45 An optimum drug loading efficiency (DLE) of 41.9 ± 2.4% and a drug loading content (DLC) of 3.7 ± 0.4% (Fig. 1c), as determined by UV-VIS spectroscopy, were obtained by utilizing 1 mM ket-mβCD complex for a 1% gelatin solution.
image file: c7nr00922d-f1.tif
Fig. 1 (a) FTIR of mβCD (black), ket (red) and ket-mβCD complex (blue). (b) Phase-solubility diagram for the ket-mβCD complex. All values are expressed as mean ± SD (n = 3). (c) Drug-loading efficiency (DLE) and drug-loading content (DLC) of ket-mβCD-GNPs with mβCD concentrations of 0.1, 0.5, 1, 5 and 10 mM. Bars represent mean ± SD (n = 3). (d) TEM micrograph of anti-TLR4-mβCD-GNPs. TEM images show a uniform size and a narrow size distribution of the nanoparticles.

For optimization of antibody conjugation, an FITC-labelled IgG antibody, in increasing concentrations, was conjugated to the nanoparticles and the conjugation efficiency was determined by counting fluorescent (FITC-IgG conjugated) particles through flow cytometry.46 The maximal amount of the antibody that conjugated with 1 mg GNPs was found to be 2 μg (ESI Fig. S1). For all further experiments, 2 μg anti-TLR4 antibody was used per mg GNP. The rationale behind using the anti-TLR4 antibody in the present study was to target the over-expressed TLR4 in the corneal epithelium and provide an increased retention of the GNPs on the corneal surface. Furthermore, anti-TLR4 conjugation is also expected to block the subsequent TLR4 mediated inflammatory pathways, thereby minimizing inflammation-mediated host damage. The conjugation of the anti-TLR4 antibody increased the particle size to 155.2 ± 3.5 nm compared to 141.7 ± 1.4 nm for bare GNPs (Table 1). The increase in size could be attributed to the surface bound antibody molecules.47 The particle size, as measured by TEM, was found to be in the range of 80 to 100 nm before conjugation which did not show a significant increase after antibody conjugation (Fig. 1d). The particle size obtained by TEM is the projected area of the diameter of the dehydrated particles and is usually smaller than the hydrodynamic radii obtained by DLS. The small difference (∼9%) between the hydrodynamic radii of the bare and antibody conjugated GNPs was not reflected in the TEM studies. The stability of the ket-mβCD-anti-TLR4-GNPs, as studied by size and zeta potential measurements, revealed no change in size or surface charge upon storage after filtration through a 0.45 μm filter even after 10 weeks at 4 °C (ESI Fig. S2). The nanoparticles synthesised were also found to be compatible with HCE cells even at high concentrations viz. 10 mg mL−1 (ESI Fig. S3).

Table 1 Physico-chemical characterization of GNPs
Nanoparticle formulation Size distribution (nm) Zeta potential (mV) Polydispersity index (PdI)
GNPs 141.7 ± 1.4 −43 ± 1.25 0.181 ± 0.045
Anti-TLR4-GNPs 155.2 ± 3.5 −41.47 ± 0.75 0.302 ± 0.024

LPS-induced nanoparticle binding to human corneal epithelial (HCE) cells

Over-expression of Pattern Recognition Receptors (PRRs) such as Toll-like receptors (TLRs) is a host strategy for mounting an inflammatory response.48 LPS has been shown to induce an inflammatory response through TLR4 as well as other Toll-like receptors (TLRs).49 Furthermore, TLR4 has been reported to be involved in the inflammatory pathways mediated by both fungal and bacterial Pathogen Associated Molecular Patterns (PAMPs).38 In the present study lipopolysaccharide (LPS) was used as a model ligand to induce TLR4 over-expression and generate an inflammatory response in Human Corneal Epithelial (HCE) cell cultures.50 Our results, as shown in Fig. 2a and quantified in Fig. 2b, show that the expression of TLR4 increases with increasing concentrations of LPS. Apart from LPS, the exposure of HCE cells to Aspergillus flavus mycelia was also found to increase the expression of TLR4 (ESI Fig. S4a).
image file: c7nr00922d-f2.tif
Fig. 2 (a) Western blot analysis of TLR4 over-expression in Human Corneal Epithelial (HCE) cells upon treatment with increasing concentrations of lipopolysaccharide (LPS) (0 to 10 μg). (b) Histogram representation of fold increase in TLR4 expression upon treatment with increasing concentrations of LPS (0 to 10 μg). Actin was used as an endogenous control. (c) Flow cytometry analysis of Human Corneal Epithelial (HCE) cells treated with FITC labeled GNPs and anti-TLR4-GNPs in the presence and absence of LPS. Nanoparticles were labeled with FITC and incubated with HCE cells in the presence or absence LPS for 4 h. Cells were washed with PBS and analyzed on a flow cytometer for nanoparticles (M1 represents cell population without fluorescence; M2 represents cell population with associated fluorescence). (d) Histogram representation of cell associated fluorescence as obtained by flow cytometry. (e) Cellular interaction of nanoparticles. Confocal microscopy images of Human Corneal Epithelial (HCE) cells treated with GNPs, anti-TLR4-GNPs (i) and anti-TLR4-GNPs with LPS (ii). Nanoparticles were labeled with FITC and added to HCE cell cultures and incubated for 4 h. Cells were washed with PBS, fixed with 4% formaldehyde, counter-stained with DAPI and imaged using a Leica confocal microscope (TCS-SP8; Leica Microsystems, Wetzlar, Germany).

To demonstrate the ability of the anti-TLR4-conjugated GNPs to bind HCE cells, FITC labeled GNPs and anti-TLR4-GNPs (200 μg mL−1) were incubated with HCE cells with or without LPS for 2 h at 37 °C. The cells were washed and quantified for fluorescence by flow cytometry (Fig. 2c and d). Our results suggest that the conjugation of the anti-TLR4 antibody to GNPs leads to an enhancement in nanoparticle adhesion with the HCE cells. The anti-TLR4-GNP adhesion to the HCE cells was found to increase further upon LPS treatment due to an increased surface expression of TLR4.

The cells were then visualized by using a confocal laser microscope to determine particle adhesion and uptake. Human Corneal Epithelial (HCE) cells were treated with FITC labeled GNPs, anti-TLR4-GNPs and anti-TLR4-GNPs + LPS (Fig. 2e). As evident from Fig. 2e, cell adhesion was low when the HCE cells were treated with bare GNPs. The surface adhesion was found to be higher for anti-TLR4 conjugated GNPs (Fig. 2e(i)) which further increased with LPS treatment (Fig. 2e(ii)).

Large particle sizes (above 100 nm) and high hydrophilicity have been reported to inhibit cellular uptake of polymeric nanoparticles.51 Therefore, the low cellular uptake of the gelatin nanoparticles could be attributed to the larger nanoparticle sizes (approximately 150 nm) used in the present study and the high hydrophilicity of gelatin.

Bare GNPs showed some interaction with HCE cells (Fig. 2c–e) which could be attributed to the RGD domains present in gelatin.52,53 The RGD motif has been previously employed for targeting the ocular surface.54 A charge-based interaction can, however, be ruled out due to the negative surface charge of the nanoparticles and the epithelial cell surface. The zeta potential of gelatin nanoparticles as determined at pH 7.4 (in 10 mM phosphate buffer) and pH 6.2 (in 10 mM MES buffer) was found to be around −41.47 ± 0.75 mV and −35.7 ± 4.5 mV respectively (Table 1).

The ability of the anti-TLR4-GNPs to efficiently bind the over-expressed TLR4 in HCE cells has been used as an advantage in the present study for increasing corneal adhesion thereby providing an increased retention time on the corneal surface. Since the expression of TLRs is directly related to the severity of the infection/inflammation, the binding of anti-TLR4-GNPs to the corneal surface would be determined by the infection/inflammation severity. Such a system is expected to provide a better control over drug dosage in keratitis management.

Suppression of inflammatory cytokines

Having established the ability of the anti-TLR4-GNPs to bind to the HCE cell surface, we evaluated the efficacy of the anti-TLR4-GNPs and bare GNPs to suppress the inflammatory response induced by LPS. LPS is known to induce inflammation through TLR4 along with other TLRs.49 In the present study, LPS stimulated HCE cells were treated with GNPs or anti-TLR4-GNPs and the mRNA levels of interleukin-8 (IL8) and Tumor Necrosis Factor α (TNFα) were monitored by RT-PCR. IL8 and TNFα are pro-inflammatory cytokines expressed upon TLR4 activation by LPS and are involved in recruiting host effector cells for mounting an inflammatory response.55 As shown in Fig. 3a, the mRNA levels of IL8 and TNFα were found to be down-regulated upon anti-TLR4-GNP treatment. We also measured the mRNA levels of matrix metalloproteinase 2 (MMP2), which plays an important role in tissue remodeling and neutrophil infiltration. MMP2 is also known to be activated by LPS through multiple TLRs56–58 including TLR4.59 As shown in Fig. 3a, the mRNA levels of MMP2 decreased significantly upon treatment of cells with anti-TLR4-GNPs. In comparison with anti-TLR4-GNPs, bare GNPs did not cause any significant suppression in the mRNA levels of the genes studied. The slight reduction observed in the expression could, however, be due to the non-specific binding of the GNPs to the HCE cell surface (shown in Fig. 2c–e), and non-specific blocking of LPS binding sites. The suppression of inflammatory cytokines by anti-TLR4-GNPs was also observed in the case of HCE cells exposed to Aspergillus flavus mycelia (ESI Fig. S4b).
image file: c7nr00922d-f3.tif
Fig. 3 (a) Fold change in the mRNA levels of TNF-α, IL-8 and MMP2 in Human Corneal Epithelial (HCE) cells treated with LPS, GNPs + LPS and anti-TLR4-GNPs + LPS. Untreated cells served as the control. Bars represent mean ± SD (n = 3) (*p < 0.05, **p < 0.01). (b) Protease dependent ket release profile from GNPs: representative kinetics of ket release from GNPs incubated with different concentrations of pronase (black curve: 0 mg mL−1 pronase, red curve: 0.01 mg mL−1 pronase and blue curve: 0.1 mg mL−1 pronase) up to 72 h at 37 °C. All values are expressed as mean ± SD (n = 3). (c) Growth percent of Aspergillus flavus at different concentrations of ket (black), ket-mβCD complex (red) and ket-mβCD-GNPs (blue). (The minimum concentration at which no growth was observed is referred to as MIC.) All values are expressed as mean ± SD (n = 3).

Protease-induced ket release from nanoparticles

A major advantage of using nanoparticles for drug delivery is the localization and modulation of the drug concentration at the site of action. While the former can be achieved by targeting the nanoparticles, the latter can be tuned by controlling the release of drug from the nanoparticle matrix.

In the present study, the over-expression of proteases (both host and pathogen-based)40 in the infection/inflammatory microenvironment has been exploited to regulate ket release from GNPs. To demonstrate the protease-based ket release from GNPs, ket-mβCD-GNPs were incubated in phosphate-buffered saline (PBS) with and without pronase and the drug release was subsequently analyzed by RP-HPLC. As shown in Fig. 3b, ket release continues until 12 h (cumulative ket release at 12 h ∼88%) with 0.1 mg mL−1 pronase incubation (Fig. 3b, blue curve). However, with the utilization of 0.01 mg mL−1 pronase (Fig. 3b, red curve), the ket release rate was significantly lowered and continued up to 36 h (cumulative ket release at 36 h ∼79%). In contrast, only ∼8% of ket was released in the absence of pronase (Fig. 3b, black curve) during the time course of the study (72 h). The drug release was found to be largely dependent on the protease concentration utilized, due to the degradation of gelatin nanoparticles by proteases, as demonstrated previously by our group.35 The high protease activity in the infection/inflammation microenvironment40 is thus expected to degrade the GNPs and provide a stimuli-responsive, site-specific drug release. Moreover, the nanoparticles also act as alternative substrates for the proteases and dilute their deleterious effect on the inflammatory milieu thereby minimizing damage to the corneal tissue.

Anti-fungal activity of nanoparticles

To determine the anti-fungal efficacy of the nanoparticles, ket, ket-mβCD complex and ket-mβCD-GNPs were incubated with 106 CFU of Aspergillus flavus spores for 24 h and subsequently plated onto Sabouraud dextrose agar plates. Colonies were counted 48 to 72 h post plating. The MIC values are reported as the minimum concentration of formulation required to completely inhibit the growth of Aspergillus flavus. The MIC value for ket was found to be around 2 ± 0.2 μg mL−1 (Fig. 3c, black curve), which was comparable to the values reported earlier.60 However, the ket-mβCD complex showed a higher MIC value compared to ket alone (Fig. 3c, red curve). The higher MIC value of ket-mβCD (0.4 to 0.45 mM corresponding to 35.4 ± 2.5 μg mL−1) could be due to the complexation with mβCD. It is possible that the ket molecules preferentially remain buried in the cyclodextrin core and hence are less available for the required action. The in vivo efficacy of the ket-mβCD complex is expected to be better than ket alone due to its high aqueous solubility which may increase the pre-corneal (tear film) concentration of ket. ket due to its hydrophobic nature may thereafter partition into the non-polar lipid phase of the cellular membranes, thus increasing the ket levels in the corneal tissue and aqueous humor.43 The MIC of the ket-mβCD-GNP (6 to 10 mg mL−1 corresponding to 28 ± 7 μg mL−1) (Fig. 3c, blue curve) was slightly lower compared to that of the ket-mβCD complex. Our results, summarized in Table 2, indicate that the encapsulated drug retains its efficacy post-release from the GNPs.
Table 2 Minimum Inhibitory Concentration (MIC) of ket, ket-mβCD and ket-mβCD-GNPs for Aspergillus flavus
Formulation MIC (μg mL−1)
ket 2 ± 0.2
ket-mβCD 35.4 ± 2.5
ket-mβCD-GNPs 28 ± 7

Table 3 Visual scoring system for murine fungal keratitis41
  Grade 1 Grade 2 Grade 3 Grade 4
Area of corneal opacity 1%–25% 26%–50% 51%–75% 76–100%
Density of corneal opacity Slight cloudiness, outline of iris and pupil discernible Cloudy, but outline of iris and pupil remain visible Cloudy, opacity not uniform Uniform opacity
Surface regularity Slight surface irregularity Rough surface, some swelling Significant swelling, crater or serious descemetocele formation Perforation or descemetocele

In vivo corneal adhesion of nanoparticles

To determine the ability of anti-TLR4-GNPs to bind the corneal epithelium, FITC labeled GNPs with and without anti-TLR4 conjugation were administered topically to rats infected with Aspergillus flavus. Nanoparticle binding was monitored at different time points by collecting eyeballs and imaging the corneal tissue sections to detect NP-associated fluorescence. An uninfected healthy cornea comprises of a multilayered epithelial layer and a stromal layer with relatively fewer cells. An infected cornea, however, shows cellular infiltration in the stromal layer and a loss in epithelial integrity as shown in Fig. 4a (DAPI panel).
image file: c7nr00922d-f4.tif
Fig. 4 (a) Nanoparticle binding to infected corneas: FITC conjugated GNPs and anti-TLR4-GNPs were administered to infected corneas, eye balls were collected 4 h, 8 h and 12 h post nanoparticle administration, corneal tissues were isolated and 10 μm sections were obtained. Sections were counter-stained with DAPI and visualised using a Leica confocal microscope (TCS-SP8; Leica Microsystems, Wetzlar, Germany). Bare GNPs were not observed at the 4 h time point. In contrast, anti-TLR4-GNPs were observed at the 4 h and 8 h time points. (b) Clinical manifestation and scoring of Aspergillus flavus infected rat corneas treated with different formulations of nanoparticles. Corneal ulcers were evident in groups 1, 2, 3, 4 and 5. Group 6 showed a comparatively mild infiltration and less opacity which was further reduced in group 7 (n = 4). (c) Slit lamp examination and clinical scoring of an ocular disease response was performed at day 1, 3, 5 and 7 p.i. on each infected cornea. (Group 1 (black), group 2 (red), group 3 (blue), group 4 (green), group 5 (cyan), group 6 (purple) and group 7 (orange).)

Corneal sections obtained from rat eyeballs infected and administered with FITC labeled anti-TLR4-GNPs showed fluorescence at 4 h and 8 h post administration. Fluorescence was not observed in the corneal sections obtained 12 h post administration (Fig. 4a). The absence of fluorescence in the GNP-treated eyes, even 4 h (Fig. 4a) post administration, suggests the inability of bare GNPs to be retained on the corneal surface for prolonged periods. Our results suggest that anti-TLR4 conjugation significantly increases the residence time of the nanoparticles on the corneal surface in the infected corneas compared to bare GNPs.

The dosage frequency for the subsequent experiments was thus fixed at once every 12 h.

Clinical evaluation

After demonstration of the effectiveness of anti-TLR4 conjugation in increasing the residence time of GNPs on the corneal surface, the infected corneas were administered with various formulations at an interval of 12 h each, to determine their efficacy in managing the infection and inflammation. The right eyes of rats were inoculated with 20 μl of Aspergillus flavus spore suspension (106 CFU mL−1) to induce infection, while the left eyes served as an un-inoculated control. Infected eyes from rats in all groups showed fungal infiltration within the first 24–48 h post infection. The infection was characterized by corneal ulceration and cloudiness. Rats that developed keratitis (characterized by a white infiltrate under a slit-lamp biomicroscope) within 48 h post inoculation were divided into the following groups: Group 1: untreated, Group 2: administered with PBS, Group 3: administered with ket, Group 4: administered with ket-mβCD, Group 5: administered with empty GNPs, Group 6: administered with ket-mβCD-GNPS, and Group 7: administered with anti-TLR4-ket-mβCD-GNPs. The infection progressed rapidly in rats from group 1, 2, 3, 4 and 5 (Fig. 4b), as characterized by an increase in the depth of the fungal infiltrate, corneal cloudiness and epithelial damage with mucous discharge. However, the progression of infection in rats of groups 6 and 7 (Fig. 4b) was gradual with mild ulceration and fungal infiltration. While the infection regressed in rats of group 7 as followed over a period of 7 days, rats of group 6 still showed mild fungal infiltration and corneal ulceration on day 7. The regression of infection in rats of group 7 can be attributed to the increased corneal residence time of the nanoparticles, and the prolonged effect of anti-TLR4 and the loaded drug-ket in managing both the infection and inflammation. The lower efficacy of ket-mβCD-GNPs compared to anti-TLR4-ket-mβCD-GNPs should be due to the lower retention of the bare nanoparticles on the corneal surface in addition to the absence of the anti-inflammatory activity of anti-TLR4. Fig. 4c shows the clinical scores of rat corneas, administered with various formulations, as assessed by using a slit lamp microscope over a period of 7 days.41 The clinical scoring of the progression of infection (Fig. 4c) clearly brings out the efficacy of the anti-TLR4-ket-mβCD-GNPs in managing the infection.

Histopathological evaluation

To assess neutrophil infiltration, H&E stained corneal sections were imaged using a bright field microscope. As mentioned earlier, a healthy cornea is characterized by a multi-layered epithelium followed by a stromal layer which is sparsely populated by keratocytes (Fig. 5a, first panel). However, once infected, the scleral layer is populated with red blood cells (RBCs) and neutrophils, appearing as a result of neovascularization and inflammation in the corneal tissue accompanied by edema. The isolation and sectioning of corneas from the infected eyes in rats from groups 1, 2, 3, 4 and 5 was not possible due to the difficulty in isolating and sectioning of the corneal tissues which were found to be extensively damaged by day 7. Shown in Fig. 5a (second panel) is a representative histology image of rats in group 1, at day 3 post infection. It shows extensive infiltration of neutrophils and RBCs and edema. Rats in group 6 also showed a significant infiltration of neutrophils and RBCs due to an increased inflammatory response. However, the density of the infiltration was found to be less than that of the untreated control at day 3 post infection. The reduction in neutrophil infiltration and neovascularization is due to a decrease in the fungal load and hence a lower inflammatory response (Fig. 5a, third panel).
image file: c7nr00922d-f5.tif
Fig. 5 (a) Histopathology and PMN infiltration in corneas treated with different formulations. Panel 1 shows a normal cornea with no infiltration. Panel 2 shows a severe and overwhelming cellular infiltration in untreated corneas (group 1) at day 3 p.i. Panel 3 shows cellular infiltration in ket-mβCD-GNP treated corneas. Panel 4 shows preserved morphological integrity of the cornea with moderate infiltration in anti-TLR4-ket-mβCD-GNP treated corneas at day 7 p.i. Magnification: ×200 (a–d). (b) Sabouraud dextrose agar plates revealing the growth of Aspergillus flavus from rat eye balls obtained from uninfected rats, groups 1, 2, 3, 4, 5, 6 and 7.

The fourth panel in Fig. 5a is a representative corneal histopathological image of rats in group 7. A significant decrease in infiltration of neutrophils and red blood cells (RBCs) was observed. The effect is attributable to the blocking of the TLR4 receptors and the shutting down of the subsequent inflammatory pathway by the nanoparticles (anti-TLR4-ket-mβCD-GNPs) along with a decreased fungal load.

Microbiological evaluation

To demonstrate the reduction in the pathogen load upon nanoparticle administration, the eyeballs from the treated and control rats were excised under sterile conditions and cultured on Sabouraud dextrose agar plates with pH adjusted to 5.6 to minimize bacterial growth. Uninfected eyeballs (Fig. 5b) did not reveal any growth of Aspergillus flavus over the time course of the study. However, the eyeballs obtained from groups that showed active infection viz. groups 1, 2, 3, 4 and 5 (Fig. 5b) showed a significant growth of Aspergillus flavus.

The eyeballs obtained from group 6 showed a slight reduction in the size of the Aspergillus flavus colony formed whereas the eyeballs obtained from group 7 showed no growth of Aspergillus flavus and are comparable to the uninfected control.

Our in vivo results highlight the importance of the anti-TLR4 antibodies in corneal adhesion and suppression of inflammation. An increased corneal adhesion thereby prolongs the drug residence time in the corneal tissue resulting in a better management of infection and inflammation even at a low dosage frequency.


We developed a condition-responsive, smart nanoparticle-based drug delivery system for keratitis. Protein nanoparticles decorated with TLR4 antibodies increase the residence time of the drug on the cornea, thereby significantly reducing the frequency of dosing. TLR4 antibodies bind to their receptors, blocking the downstream TLR4 signaling pathway and suppress inflammation. Protein-based nanoparticles offer an alternative substrate for host and pathogen-associated proteolytic enzymes, thereby reducing the corneal damage and scar formation. Proteolytic degradation of the nanoparticle leads to the release of the drug, which then acts on the pathogen. The corneal binding and drug release are proportional to the severity of infection: more severe the infection, greater is the expression of TLR4 on the surface of corneal epithelium, resulting in an increased binding of the anti-TLR4-GNPs. Similarly, more severe the infection, higher is the proteolytic activity leading to an increased degradation of GNPs releasing the drug. We have tested this nanoparticle system for corneal binding and the effect on inflammation and infection. Our rat model studies confirm the efficacy of these particles. To the best of our knowledge, this work reports, for the first time, a strategy for targeting an over-expressed ocular surface receptor for increasing the residence time of a nanoparticle-based formulation. We trust this will be an excellent approach to manage keratitis. In addition, this approach may be of use to treat other topical infections as well.

Conflict of interest

The authors have filed a patent for the nanoparticle-based system reported in the study.


This work was supported by the Department of Science and Technology (DST), Govt. of India funded project “Nanotechnology for biomedical applications”. SMA acknowledges Council for Scientific and Industrial Research (CSIR) for financial support. CMR acknowledges the Department of Science and Technology, India for Sir J. C. Bose National Fellowship. The authors thank Mr T. Avinash Raj for help in histology experiments and Mr A. Harikrishna for help in TEM image acquisition.


  1. J. W. Streilein, Nat. Rev. Immunol., 2003, 3, 879–889 CrossRef CAS PubMed.
  2. E. Knop and N. Knop, Chem. Immunol. Allergy, 2007, 92, 36–49 CAS.
  3. R. Gaudana, H. K. Ananthula, A. Parenky and A. K. Mitra, AAPS J, 2010, 12, 348–360 CrossRef CAS PubMed.
  4. T. P. O'Brien, Eye, 2003, 17, 957–974 CrossRef PubMed.
  5. M. Srinivasan, J. Mascarenhas and C. N. Prashanth, Indian J. Ophthalmol., 2008, 56, 203–207 CrossRef CAS PubMed.
  6. P. A. Thomas, Eye, 2003, 17, 852–862 CrossRef CAS PubMed.
  7. L. Xie, W. Zhong, W. Shi and S. Sun, Ophthalmology, 2006, 113, 1943–1948 CrossRef PubMed.
  8. R. I. Angunawela, F. Arnalich-Montiel and B. D. Allan, J. Cataract Refractive Surg., 2009, 35, 606–607 CrossRef PubMed.
  9. F. Stapleton, L. Keay, I. Jalbert and N. Cole, Optom. Vis. Sci., 2007, 84, 257–272 CrossRef PubMed.
  10. S. MacRae, D. C. Macaluso and L. F. Rich, J. Cataract Refractive Surg., 1999, 25, 1679–1681 CrossRef CAS PubMed.
  11. M. Srinivasan, Curr. Opin. Ophthalmol., 2004, 15, 321–327 CrossRef CAS PubMed.
  12. D. S. Lam, E. Houang, D. S. Fan, D. Lyon, D. Seal and E. Wong, Eye, 2002, 16, 608–618 CrossRef CAS PubMed.
  13. J. K. Dart, F. Stapleton and D. Minassian, Lancet, 1991, 338, 650–653 CrossRef CAS.
  14. T. Cumurcu, P. Firat, E. Ozsoy, M. Cavdar and Y. Yakupogullari, Clinics, 2011, 66, 1509–1510 CrossRef PubMed.
  15. S. A. Iyer, S. S. Tuli and R. C. Wagoner, Eye Contact Lens, 2006, 32, 267–271 Search PubMed.
  16. A. Urtti, Adv. Drug Delivery Rev., 2006, 58, 1131–1135 CrossRef CAS PubMed.
  17. U. B. Kompella, R. S. Kadam and V. H. Lee, Ther. Deliv., 2010, 1, 435–456 CrossRef CAS PubMed.
  18. N. S. Gokhale, Indian J. Ophthalmol., 2008, 56, 215–220 CrossRef PubMed.
  19. K. R. Wilhelmus, Ophthalmology, 2002, 109, 835–842 CrossRef PubMed.
  20. H. J. Kim, K. Zhang, L. Moore and D. Ho, ACS Nano, 2014, 8, 2998–3005 CrossRef CAS PubMed.
  21. J. F. Huang, J. Zhong, G. P. Chen, Z. T. Lin, Y. Deng, Y. L. Liu, P. Y. Cao, B. Wang, Y. Wei, T. Wu, J. Yuan and G. B. Jiang, ACS Nano, 2016, 10, 6464–6473 CrossRef CAS PubMed.
  22. X. Yuan, D. C. Marcano, C. S. Shin, X. Hua, L. C. Isenhart, S. C. Pflugfelder and G. Acharya, ACS Nano, 2015, 9, 1749–1758 CrossRef CAS PubMed.
  23. J. Silbiger and G. A. Stern, Ophthalmology, 1992, 99, 889–892 CrossRef CAS PubMed.
  24. S. Shi, Z. Zhang, Z. Luo, J. Yu, R. Liang, X. Li and H. Chen, Sci. Rep., 2015, 5, 11337 CrossRef CAS PubMed.
  25. Z. G. Konat, L. Contreras-Ruiz, J. E. Parraga, A. Lopez-Garcia, B. R. Romero, Y. Diebold, B. Seijo and A. Sanchez, Mol. Pharm., 2011, 8, 1783–1788 CrossRef PubMed.
  26. Y. Diebold, M. Jarrin, V. Saez, E. L. Carvalho, M. Orea, M. Calonge, B. Seijo and M. J. Alonso, Biomaterials, 2007, 28, 1553–1564 CrossRef CAS PubMed.
  27. R. D. Vaishya, V. Khurana, S. Patel and A. K. Mitra, Wiley Interdiscip. Rev.: Nanomed. Nanobiotechnol., 2014, 6, 422–437 CrossRef CAS PubMed.
  28. G. Prosperi-Porta, S. Kedzior, B. Muirhead and H. Sheardown, Biomacromolecules, 2016, 17, 1449–1457 CrossRef CAS PubMed.
  29. S. Mura, J. Nicolas and P. Couvreur, Nat. Mater., 2013, 12, 991–1003 CrossRef CAS PubMed.
  30. T. Higuchi and A. Connors, Adv. Anal. Chem. Instrum., 1965, 4, 117–212 CAS.
  31. M. Másson, B. V. Sigurdardóttir, K. Matthíasson and T. Loftsson, Chem. Pharm. Bull., 2005, 53, 958–964 CrossRef.
  32. C. Wong, T. Stylianopoulos, J. Cui, J. Martin, V. P. Chauhan, W. Jiang, Z. Popovic, R. K. Jain, M. G. Bawendi and D. Fukumura, Proc. Natl. Acad. Sci. U. S. A., 2011, 108, 2426–2431 CrossRef CAS PubMed.
  33. C. J. Coester, K. Langer, B. H. van and J. Kreuter, J. Microencapsulation, 2000, 17, 187–193 CrossRef CAS PubMed.
  34. S. M. Ahsan and C. M. Rao, Int. J. Biol. Macromol., 2017, 95, 1126–1134 CrossRef CAS PubMed.
  35. S. M. Ahsan and C. M. Rao, Int. J. Nanomed., 2017, 12, 795–808 CrossRef PubMed.
  36. J. Vandervoort and A. Ludwig, Eur. J. Pharm. Biopharm., 2004, 57, 251–261 CrossRef CAS PubMed.
  37. A. Espinel-Ingroff, V. Chaturvedi, A. Fothergill and M. Rinaldi, J. Clin. Microbiol., 2002, 40, 3776–3781 CrossRef CAS PubMed.
  38. J. Zhao and X. Y. Wu, Curr. Eye Res., 2008, 33, 447–454 CrossRef CAS PubMed.
  39. V. K. Venishetty, N. Parikh, R. Sistla, F. J. Ahmed and P. V. Diwan, J. Chromatogr. Sci., 2011, 49, 136–141 CAS.
  40. U. Gopinathan, T. Ramakrishna, M. Willcox, C. M. Rao, D. Balasubramanian, A. Kulkarni, G. K. Vemuganti and G. N. Rao, Exp. Eye Res., 2001, 72, 433–442 CrossRef CAS PubMed.
  41. H. Guo, J. Gao and X. Wu, Immunol. Cell Biol., 2012, 90, 352–357 CrossRef CAS PubMed.
  42. E. Albers and B. W. Muller, Crit. Rev. Ther. Drug Carrier Syst., 1995, 12, 311–337 CrossRef CAS PubMed.
  43. J. Zhang, L. Wang, C. Gao, L. Zhang and H. Xia, J. Ocul. Pharmacol. Ther., 2008, 24, 501–506 CrossRef CAS PubMed.
  44. M. M. Doile, K. A. Fortunato, I. C. Schmucker, S. K. Schucko, M. A. Silva and P. O. Rodrigues, AAPS PharmSciTech, 2008, 9, 314–321 CrossRef CAS PubMed.
  45. C. Jianbin, C. Liang, X. Hao and M. Dongpin, Spectrochim. Acta, Part A, 2002, 58, 2809–2815 CrossRef.
  46. E. V. van Gaal, G. Spierenburg, W. E. Hennink, D. J. Crommelin and E. Mastrobattista, J. Controlled Release, 2010, 141, 328–338 CrossRef CAS PubMed.
  47. M. Reth, Nat. Immunol., 2013, 14, 765–767 CrossRef CAS PubMed.
  48. R. L. Redfern and A. M. McDermott, Exp. Eye Res., 2010, 90, 679–687 CrossRef CAS PubMed.
  49. A. C. Johnson, F. P. Heinzel, E. Diaconu, Y. Sun, A. G. Hise, D. Golenbock, J. H. Lass and E. Pearlman, Invest. Ophthalmol. Visual Sci., 2005, 46, 589–595 Search PubMed.
  50. S. M. Leal Jr., S. Cowden, Y. C. Hsia, M. A. Ghannoum, M. Momany and E. Pearlman, PLoS Pathog., 2010, 6, e1000976 Search PubMed.
  51. Y. Bai, H. Xing, P. Wu, X. Feng, K. Hwang, J. M. Lee, X. Y. Phang, Y. Lu and S. C. Zimmerman, ACS Nano, 2015, 9, 10227–10236 CrossRef CAS PubMed.
  52. A. V. Taubenberger, M. A. Woodruff, H. Bai, D. J. Muller and D. W. Hutmacher, Biomaterials, 2010, 31, 2827–2835 CrossRef CAS PubMed.
  53. D. Gullberg, K. R. Gehlsen, D. C. Turner, K. Ahlen, L. S. Zijenah, M. J. Barnes and K. Rubin, EMBO J., 1992, 11, 3865–3873 CAS.
  54. J. Lu, M. Shi and M. S. Shoichet, Bioconjugate Chem., 2009, 20, 87–94 CrossRef CAS PubMed.
  55. J. E. Oakes, C. A. Monteiro, C. L. Cubitt and R. N. Lausch, J. Virol., 1993, 67, 4777–4784 CAS.
  56. L. Li, F. W. Cheng, F. Wang, B. Jia, X. Luo and S. Q. Zhang, Mol. Cell. Biochem., 2014, 389, 43–49 CrossRef CAS PubMed.
  57. T. Saber, D. J. Veale, E. Balogh, J. McCormick, S. NicAnUltaigh, M. Connolly and U. Fearon, PLoS One, 2011, 6, e23540 CAS.
  58. M. Ruan, Z. Zhang, S. Li, M. Yan, S. Liu, W. Yang, L. Wang and C. Zhang, PLoS One, 2014, 9, e92748 Search PubMed.
  59. W. Li, Q. Xu, Y. Deng, Z. Yang, S. Xing, X. Zhao, P. Zhu, X. Wang, Z. He and Y. Gao, Lab. Invest., 2015, 95, 635–647 CrossRef CAS PubMed.
  60. A. Espinel-Ingroff, K. Dawson, M. Pfaller, E. Anaissie, B. Breslin, D. Dixon, A. Fothergill, V. Paetznick, J. Peter, M. Rinaldi and T. Walsh, Antimicrob. Agents Chemother., 1995, 39, 314–319 CrossRef CAS PubMed.


Electronic supplementary information (ESI) available: Fig. S1: Flow cytometry quantification of FITC-IgG conjugation to GNPs. Fig. S2: Stability profile of gelatin nanoparticles. Fig. S3: Biocompatibility of gelatin nanoparticles. Fig. S4: Aspergillus flavus induced TLR4 expression in human corneal epithelial (HCE) cells and anti-TLR4-GNP mediated suppression of inflammatory cytokines. See DOI: 10.1039/C7NR00922D

This journal is © The Royal Society of Chemistry 2017