Gallic acid grafting effect on delivery performance and antiglaucoma efficacy of antioxidant-functionalized intracameral pilocarpine carriers

Acta Biomaterialia xxx (2016) xxx–xxx

Contents lists available at ScienceDirect

Acta Biomaterialia journal homepage: www.elsevier.com/locate/actabiomat

Full length article

Gallic acid grafting effect on delivery performance and antiglaucoma efficacy of antioxidant-functionalized intracameral pilocarpine carriers Shih-Feng Chou a,b,1, Li-Jyuan Luo a,1, Jui-Yang Lai a,c,d,e,⇑ a

Institute of Biochemical and Biomedical Engineering, Chang Gung University, Taoyuan, Taiwan 33302, Republic of China Department of Bioengineering, University of Washington, Seattle, WA 98195-5061, USA c Biomedical Engineering Research Center, Chang Gung University, Taoyuan, Taiwan 33302, Republic of China d Molecular Medicine Research Center, Chang Gung University, Taoyuan, Taiwan 33302, Republic of China e Center for Tissue Engineering, Chang Gung Memorial Hospital, Taoyuan, Taiwan 33305, Republic of China b

a r t i c l e

i n f o

Article history: Received 13 December 2015 Received in revised form 21 April 2016 Accepted 25 April 2016 Available online xxxx Keywords: Therapeutic delivery carrier Intracameral pilocarpine administration Antiglaucoma effects Antioxidant-functionalized biomaterial Grafting amount of gallic acid

a b s t r a c t Functionalization of therapeutic carrier biomaterials can potentially provide additional benefits in drug delivery for disease treatment. Given that this modification determines final therapeutic efficacy of drug carriers, here, we investigate systematically the role of grafting amount of antioxidant gallic acid (GA) onto GN in situ gelling copolymers made of biodegradable gelatin and thermo-responsive poly(Nisopropylacrylamide) for intracameral delivery of pilocarpine in antiglaucoma treatment. As expected, increasing redox reaction time increased total antioxidant activities and free radical scavenging abilities of synthesized carrier biomaterials. The hydrophilic nature of antioxidant molecules strongly affected physicochemical properties of carrier materials with varying GA grafting amounts, thereby dictating in vitro release behaviors and mechanisms of pilocarpine. In vitro oxidative stress challenges revealed that biocompatible carriers with high GA content alleviated lens epithelial cell damage and reduced reactive oxygen species. Intraocular pressure and pupil diameter in glaucomatous rabbits showed correlations with GA-mediated release of pilocarpine. Additionally, enhanced pharmacological treatment effects prevented corneal endothelial cell loss during disease progression. Increasing GA content increased total antioxidant level and decreased nitrite level in the aqueous humor, suggesting a much improved antioxidant status in glaucomatous eyes. This work significantly highlights the dependence of physicochemical properties, drug release behaviors, and bioactivities on intrinsic antioxidant capacities of therapeutic carrier biomaterials for glaucoma treatment. Statement of Significance Development of injectable biodegradable polymer depots and functionalization of carrier biomaterials with antioxidant can potentially provide benefits such as improved bioavailability, controlled release pattern, and increased therapeutic effect in intracameral pilocarpine administration for glaucoma treatment. For the first time, this study demonstrated that the biodegradable in situ gelling copolymers can incorporate different levels of antioxidant gallic acid to tailor the structure-property-function relationship of the intracameral drug delivery system. The systematic evaluation fully verified the dependence of phase transition, degradation behavior, drug release mechanism, and antiglaucoma efficacy on intrinsic antioxidant capacities of carrier biomaterials. The report highlights the significant role of grafting amount of gallic acid in optimizing performance of antioxidant-functionalized polymer therapeutics as new drug delivery platforms in disease treatment. Ó 2016 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.

1. Introduction ⇑ Corresponding author at: Institute of Biochemical and Biomedical Engineering, Chang Gung University, Taoyuan, Taiwan 33302, Republic of China. E-mail address: [email protected] (J.-Y. Lai). 1 Equal contribution to first author.

It is known that high intraocular pressure (IOP) strongly involved in glaucomatous development and progression [1,2]. In addition, studies have described the effects of excessive oxidative

http://dx.doi.org/10.1016/j.actbio.2016.04.035 1742-7061/Ó 2016 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.

Please cite this article in press as: S.-F. Chou et al., Gallic acid grafting effect on delivery performance and antiglaucoma efficacy of antioxidant-functionalized intracameral pilocarpine carriers, Acta Biomater. (2016), http://dx.doi.org/10.1016/j.actbio.2016.04.035

2

S.-F. Chou et al. / Acta Biomaterialia xxx (2016) xxx–xxx

stress on high IOP as a result of trabecular meshwork dysfunction leading to abnormal outflow of aqueous humor [3]. Currently, medical treatments for glaucoma patients involve in the use of prescribed drugs in the form of eye drops and gels to lower IOP. However, the need for drug molecules to diffuse and penetrate through the protective barrier of cornea often reduces their bioavailability and efficacy in ocular tissues, and thus frequent applications are associated with these types of formulations. Due to this, injections of drug-containing hydrogels appear to be an alternative approach to achieve improved effectiveness in glaucoma treatment. Specifically, we have demonstrated the synthesis of in situ gelling GN copolymeric hydrogels using biodegradable gelatin and thermoresponsive poly(N-isopropylacrylamide) (PNIPAAm) for intracameral delivery of pilocarpine [4]. Furthermore, we compared differences in the molecular weights of PNIPAAm as a result of the physicochemical changes in GN copolymers and their effects in drug delivery characteristics [5]. Moreover, in situ gelling GN copolymers were further grafted with gallic acid (GA) molecules to provide additional antioxidant activities [6]. Supported by our previous studies, this work further examine the effects of different grafting amounts of GA onto GN copolymers on the physicochemical properties, drug release characteristics, and intracameral treatment of various antioxidative in situ gelling GNGA biomaterials. Polyphenolics, including GA, have molecular structures consisting of aromatic rings and hydroxyl groups, which account for their outstanding antiviral, antibacterial, anti-inflammatory and antioxidant properties [7–10]. Traditionally, antioxidant small molecules were entrapped in the polymer carriers whereas the intrinsic properties of the matrix dictate the release rate of antioxidant [11,12]. Recently, the idea of grafting antioxidants onto biodegradable polymers has drawn many attentions since it functionalizes drug carriers. Theoretically, grafting antioxidants onto the backbone of polymeric structure allows a relatively high level of total antioxidant activity. Furthermore, continuous antioxidative stress ability can be achieved since the release of antioxidant depends on the degradation of the matrix, which is rather slow as compared to the diffusion-controlled mechanism in conventional encapsulation methods. Owing to these advantages, Yang et al. grafted antioxidative citric acid onto copolymers of poly(ethylene glycol) and PNIPAAm followed by the entrapping of chemokine [13]. Their results suggested that grafting of citric acid affected the physicochemical properties of the polymeric matrix, sustained the release of chemokine, and increased free radical scavenging ability and bioactivities of the drug carrier. In a similar study, van Lith et al. reported the grafting of citric acid and ascorbic acid onto 1,8octanediol with excellent antioxidative ability in free radical scavenging, iron chelation, and inhibition of lipid peroxidation [14]. Of particular importance, studies in oxidative challenges using 200 lM hydrogen peroxide suggested that the drug carriers protected cells from attacks generated by reactive oxygen species (ROS) while antioxidative ability remained at a high level after degradation of the matrix materials. However, both studies only considered the changes in intrinsic antioxidative properties from grafting antioxidants onto biodegradable polymers rather than comparing the effects of different grafting amounts of antioxidant on their antioxidative abilities. Therefore, this work addresses the unknowns of different grafting amounts of antioxidant on drug release and bioactivities of the drug carriers. Typical grafting procedures of GA onto polymer backbones involve in the use of a redox pair from ascorbic acid (AA) and hydrogen peroxide as the radical initiator [15,16]. It is a one step process with a relatively high yield. In addition, using AA redox reaction as a radical initiator allows the instantaneous functionalization of the polymer backbones without the production of toxic chemicals, which can seriously hinder the biocompatibility of the drug carriers. Furthermore, the hydroxyl groups on the aromatic ring of GA

are preserved leading to strong antioxidant activities of GA for a prolonged time. Senevirathne et al. grafted GA onto chitosan using AA redox reaction as a radical initiator where the grafting amounts of GA were adjusted by the molar ratio of chitosan to GA [17]. These authors demonstrated that the antioxidant ability of GA-g-chitosan carriers against ROS increased with increasing grafting amounts of GA. Isßıklan et al. demonstrated itaconic acid-grafted sodium alginate using cerium ammonium nitrate/nitric acid as a redox reaction initiator [18]. In addition, they studied the parameters in the redox reactions explicitly and suggested that a 5-h reaction time (with all other parameters set at optimal values) reached the highest grafting yield and efficiency. Others have reported the correlations between redox reaction times and grafting amounts of antioxidants onto polymer backbones using different polymer-redox systems [19–21]. These earlier findings motivate us to study the dependence of grafting amounts of GA onto GN copolymers on the AA redox reaction time. More importantly, we use AA redox reaction time to control the antioxidative performance of GNGA carrier materials designed for intracameral drug delivery application. In current study, GNGA carrier materials were synthesized from three redox reaction times (i.e., 30 min, 90 min, and 180 min). This variation allows us to obtain in situ gelling carriers with different total antioxidant activities and free radical scavenging abilities. Since GA is a hydrophilic small molecule, increasing the grafting amount of GA (by increasing redox reaction time) alters the physicochemical characteristics of the GNGA carriers. In addition, we hypothesize that the amount of GA present in the GNGA biomaterials has a significant impact on their antiglaucomatous effects as well as their bioactivities as a drug carrier vehicle. Furthermore, pilocarpine, oldest and most frequently used medication to treat glaucoma, is encapsulated in GNGA injections to study their release rates associated with the grafting amount of the GA. However, it is noteworthy that the ability to lower oxidative stress in the anterior chamber from drug-containing polymer injections was independent of the presence of pilocarpine based on our previous results. By contrast, oxidative stress in the anterior chamber correlated highly with functionalization of the antioxidant GA molecules. Therefore, considering that pilocarpine and GA contribute to separate effects in the clinical responses of glaucoma, current research efforts mainly address on the effects of various grafting amounts of GA onto GN on the development of antiglaucomatous biomaterial carriers. 2. Materials and methods 2.1. Materials Gelatin (type A; 300 Bloom), GA, AA, hydrogen peroxide, sulfuric acid, sodium phosphate, ammonium molybdate, 2,20 -diphenyl1-picrylhydrazyl (DPPH), matrix metalloproteinase-2 (MMP-2, EC 3.4.24.24), pilocarpine nitrate, and a-chymotrypsin were purchased from Sigma-Aldrich (St. Louis, MO, USA). Before use, NIPAAm (Acros Organics, Geel, Belgium) was purified by recrystallization from n-hexane. Deionized water used was purified with a Milli-Q system (Millipore, Bedford, MA, USA). Balanced salt solution (BSS, pH 7.4) was obtained from Alcon Laboratories (Fort Worth, TX, USA). Phosphate-buffered saline (PBS, pH 7.4) was acquired from Biochrom (Berlin, Germany). All the other chemicals were of reagent grade and used as received without further purification. 2.2. Synthesis of GA-functionalized gelatin-g-PNIPAAm (GNGA) Detailed materials used in this work along with synthesis of GA-functionalized gelatin-g-PNIPAAm (GNGA) were described

Please cite this article in press as: S.-F. Chou et al., Gallic acid grafting effect on delivery performance and antiglaucoma efficacy of antioxidant-functionalized intracameral pilocarpine carriers, Acta Biomater. (2016), http://dx.doi.org/10.1016/j.actbio.2016.04.035

S.-F. Chou et al. / Acta Biomaterialia xxx (2016) xxx–xxx

previously [4–6]. Briefly, GN carriers comprised backbones of Bloom 300 aminated gelatin with PNIPAAm end-caps at the carboxylic terminals using carbodiimide-coupling chemistry [4]. Specifically, the grafting efficiency of GN samples was 18.6% at 0.36 feed molar ratio of NH2/COOH. Later, an aqueous solution was obtained by dissolving 0.5 g of GN into 50 ml of deionized water followed by adding 0.25 g AA and 1 ml of hydrogen peroxide to the solution while maintaining at 25 °C for 30, 90, and 180 min. At the end of the each time point, 60 mg of GA was added to the solution and allowed the mixture to agitate for 24 h. After completion, the solution was exhaustively dialyzed for 4 days in 4 °C deionized water to remove unreacted components followed by lyophilization at 50 °C and 0.08 mbar for 3 days. Here, the GN samples reacted with AA and hydrogen peroxide at 25 °C for 30 min was designated as M30. 2.3. Determination of total antioxidant activity and scavenging ability against DPPH radical Total antioxidant activities in GNGA samples were examined using phosphomolybdenum assays. Briefly, 10 mg of GNGA polymers were dissolved in 0.3 ml of deionized water followed by mixing the solution with 1.2 ml of phosphomolybdenum assay reagent. After mixing, samples were maintained at 95 °C for 90 min followed by measurements of absorbance at 695 nm using an UV-vis spectrophotometer (Thermo Scientific, Waltham, MA, USA) in room temperature against a blank sample, containing 1.2 ml of the same reagent with 0.3 ml of deionized water. Total antioxidant activities of GNGA carriers were described as equivalent antioxidant concentrations (n = 6). Free radical scavenging activities of GNGA samples were evaluated using DPPH assays. Briefly, 250 mg of GNGA samples were dissolved in 12.5 ml of deionized water followed by the addition of 12.5 ml ethanol solution containing 1 mg of DPPH radical. After incubation for 30 min at 25 °C, the absorbance of GNGA samples was measured by an UV-vis spectrophotometer (Thermo Scientific) at 517 nm. Calculations of DPPH scavenging activity (%) was based on ((A0  A1)/A0)  100, where A0 is the absorbance of blank DPPH solution at the same reaction conditions in the absence of synthesized polymers, and A1 is the absorbance of DPPH solution in the presence of polymer samples (n = 6). 2.4. Determination of water content, phase transition temperature, and degradability For water content measurements, 1.5 ml of GNGA solutions (10% w/v) were prepared at 25 °C followed by a 10 min gelation process at 34 °C. Later, GNGA hydrogels were dried in vacuum until a constant weight (Wi) equal to the initial weight of GNGA powder (150 mg) used for the preparation of polymer solutions. The swollen gel weight (Ws) was measured by immersing samples for 4 h in 34 °C BSS (10 ml) on a reciprocal rotary shaker (50 rpm). Water content (%) of the GNGA samples was calculated as ((Ws  Wi)/ Ws)  100 (n = 5) [22]. Thermal properties of GNGA solutions (10% (w/v) in BSS) were measured by a DSC 2010 differential scanning calorimeter (TA Instruments, New Castle, DE, USA). DSC experiments were carried by hermetically sealing of 8 mg GNGA solutions after conditioning in room temperature for 1 h. Samples were scanned from 25 to 45 °C using a temperature ramp rate of 3 °C/min purging with nitrogen gas. Lower critical solution temperatures (LCST) were measured from the onset of the endothermic peak (n = 4) [5]. Degradation measurements were performed on GNGA hydrogels from the same gelling method described earlier. The initial weight (Wi) of GNAG hydrogels after drying to constant weight (150 mg) in vacuum was recorded. Samples were then immersed

3

in 10 ml of MMP-2 containing BSS at 34 °C for 2 weeks. The MMP-2 concentration was 50 ng/ml determined by the level of gelatinase present in aqueous humor of glaucomatous eyes [23]. Later, samples were collected and dried to determine the weight after degradation (Wd). Percentage weight loss (%) was calculated as ((Wi  Wd)/Wi)  100 (n = 4). 2.5. In vitro drug release studies To determine the percentage drug encapsulations, 2% (w/v) of pilocarpine nitrate was added in 10% w/v GNGA solutions at room temperature. The same gel formation procedures described earlier was used [6]. Later, drug-loaded hydrogels were redissolved at 25 °C in an empty vial for high performance liquid chromatography (HPLC) measurements. A L-2400 UV detector and a Mightysil RP-18 column (Kanto Chemical, Tokyo, Japan) were employed. Detailed HPLC procedures, including pilocarpine standard and spiked controls, were described previously [4–6]. Drug encapsulation efficiency was determined from entrapped pilocarpine nitrate in the GNGA hydrogels during gelation process as compared to the initial drug feeding in the solutions (n = 4). Drug release studies were performed similarly as the degradation studies earlier using 50 ng/ml of MMP-2. Release buffer was collected at predetermined time points and analyzed by HPLC. Pilocarpine concentrations, after being released from GNGA hydrogels, were calculated using a calibration curve (n = 4). The cumulative release percentage of drug at each time point was determined by dividing the amount of the averaged released pilocarpine nitrate by the total amount of the loaded pilocarpine nitrate and multiplied by 100. 2.6. In vitro biocompatibility studies In vitro biocompatibility of human lens epithelial (HLE-B3; ATCC No: CRL-11421) cell lines were tested using previously described methods [4,24]. Briefly, cell viability was performed using a Live/Dead Viability/Cytotoxicity Kit (Molecular Probes, Eugene, OR, USA) followed by observations under fluorescence microscopy (Axiovert 200M; Carl Zeiss, Oberkochen, Germany) [25]. The living cells were identified through signals of green fluorescence from the intracellular esterase activities due to cleavage of calcein AM. By contrast, a red fluorescence was produced via binding of EthD-1 to the nucleic acids in dead cells with damaged cell membranes. Cell proliferation was determined by using WST-1 assay (Roche Diagnostics, Indianapolis, IN, USA) [26]. Cell staining was achieved by adding 100 ll of WST-1 reagent to the cultures followed by incubation for 4 h at 37 °C in a 5% CO2 incubator. The optical density (OD) value at 450 nm was recorded using a Multiskan Spectrum Microplate Spectrophotometer (ThermoLabsystems, Vantaa, Finland) (n = 4). Detailed procedures of the IL-6 measurements, indicated by the messenger RNA (mRNA) levels of HLE-B3 cells, were described previously (n = 3) [4,27]. 2.7. Measurement of antioxidant activity against oxidative stress The hydrogen peroxide-induced oxidative stress of HLE cell model was used according to our previously published methods [6]. The HLE-B3 cells with a density of 5  104 cells/well were seeded in 24-well plates and incubated with 150 ll of sterile GNGA solutions (10% w/v) for 24 h. Then, the cell cultures were treated with a further incubation of 24 h in medium containing 200 lM hydrogen peroxide. For comparison purpose, the cells were exposed to hydrogen peroxide of 0 lM (Control group) and 200 lM (HP group) for 24 h following 24 h of incubation in the absence of the polymer carrier materials.

Please cite this article in press as: S.-F. Chou et al., Gallic acid grafting effect on delivery performance and antiglaucoma efficacy of antioxidant-functionalized intracameral pilocarpine carriers, Acta Biomater. (2016), http://dx.doi.org/10.1016/j.actbio.2016.04.035

4

S.-F. Chou et al. / Acta Biomaterialia xxx (2016) xxx–xxx

A phase-contrast microscopy (Nikon, Melville, NY, USA) was used for cell morphology observations while WST-1 assays using the similar procedures described earlier were used for determination of cell viability. Results were expressed as relative metabolic activity comparing to the mean OD of the control groups (n = 4). Intracellular accumulation of ROS was measured by oxidative conversion of cell-permeable 20 ,70 -dichlorodihydrofluorescein diacetate (DCFH-DA) (Molecular Probes) to fluorescent 20 ,70 -dichlor ofluorescein (DCF). The HLE cells in the culture wells were incubated with 10 lM DCFH-DA solutions at 37 °C for 1 h. Then, the cells were washed three times with PBS. The DCF fluorescence imaging (Ex. 488 nm; Em. 525 nm) was acquired with a fluorescence microscope (Carl Zeiss). Furthermore, the fluorescence reading was done with a multi-mode microplate reader (BioTek Instruments, Winooski, VT, USA) to detect the difference in the fluorescence intensity (n = 4). 2.8. Animal studies All animal procedures were approved by the Institutional Review Board and were conducted in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. Twenty-four adult New Zealand white rabbits (National Laboratory Animal Breeding and Research Center, Taipei, Taiwan, ROC), weighing 3.0–3.5 kg and 16–20 weeks of age, were used for this study. Animals were healthy and free of clinically observable ocular surface disease. Surgical operation was performed in the single eye of animals, with the normal fellow eye. Experimental glaucoma model was obtained in accordance with our previous established model [4]. In the three test groups (M30, M90, and M180) of animals (6 rabbits/group), the glaucomatous rabbits received intracameral injections of 50 ll of a mixture containing pilocarpine nitrate (2% w/v) and GNGA solutions (10% w/v). Without treatment with any polymers and drugs, the remaining 6 rabbits with experimental glaucoma served as a control group (Ctrl). The glaucomatous rabbits were anesthetized intramuscularly followed by delivery of 50 ll of 2% (w/v) pilocarpine and 10% (w/v) GNGA solutions to the anterior chamber, using a 30-gauge needle. IOP measurements were performed by using a Schiotz tonometer (AMANN Ophthalmic Instruments, Liptingen, Germany) and five readings were averaged [4,28]. Measurements of pupil diameter were performed using a pupillary diameter gauge (Smith and Newphew Pharmaceuticals, Essex, UK) after acclimatization in a room with constant lighting [5]. Each data point represents the average value of four measurements. Finally, a specular microscopy (Topcon Optical, Tokyo, Japan) was used for rabbits’ corneal endothelial cell morphology observation and three cell densities were averaged [29]. After euthanization of animals with CO2 gas, total antioxidant and nitrite levels from aqueous humor of each rabbit eye were evaluated through biochemical analyses (n = 6) [30,31]. 2.9. Statistical analyses Unless specified, data were described as mean ± standard deviation. One-way analyses of variance (ANOVA) were performed for comparative studies at an acceptable significance level of P < 0.05. 3. Results and discussion 3.1. Determination of total antioxidant activity and scavenging ability against DPPH radical Fig. 1a displays images of GNGA biomaterials in phosphomolybdenum assay, where greenish solid complex (phosphate-Mo (V))

can be clearly seen due to promotion of Mo (VI) to Mo (V) by GA [32]. From the images, increasing redox reaction time results in a higher amount of sedimentation. Quantitative analyses, shown in Fig. 1b, suggest that total antioxidant activities are 0.12 ± 0.03 mg/g, 0.33 ± 0.06 mg/g, and 1.42 ± 0.08 mg/g for M30, M90, and M180 groups, respectively (P < 0.05). Free radical scavenging ability of GNGA biomaterials against DPPH assay, shown in Fig. 1c, reveals decolorization of stable DPPH radical in the presence of different levels of antioxidants [33,34]. Quantitative analyses by means of measuring the amount of purple chromogen radicals remaining in the solution, shown in Fig. 1d, suggest a 29.2 ± 2.4%, 38.9 ± 3.2%, and 79.1 ± 3.3% inhibition of the DPPH radicals for M30, M90, and M180 groups, respectively (P < 0.05). Our data are in agreement with Spizzirri et al., who reported a 66 ± 3% inhibition of DPPH radicals from gelatin-GA conjugates using a redox reaction time of 120 min [35]. Overall, total antioxidant activity and free radical scavenging ability against DPPH depend on the grafting amount of antioxidant GA molecules in GNGA biomaterials (also see the Supporting Information). 3.2. Determination of water content, phase transition temperature, and degradability The water content of GNGA biomaterials, shown in Fig. 2a, suggests that increasing redox reaction time increases the overall water content in the samples and decreases their corresponding hydrophobicity. For example, the water content for M30 groups is 56.3 ± 1.2% whereas M90 and M180 groups have a water content of 61.0 ± 1.8% and 75.5 ± 3.1%, respectively (P < 0.05). LCST from M30, M90, and M180 groups, shown in Fig. 2b, reveals that increasing redox reaction time increases the LCST from 27.5 ± 0.1 °C, 29.0 ± 0.3 °C, to 31.3 ± 0.1 °C for M30, M90, and M180, respectively (P < 0.05). Fig. 2c displays the weight loss of GNGA samples prepared at various redox reaction times. Initially, weight losses of 7.1 ± 0.9%, 10.0 ± 0.7%, and 17.4 ± 1.5% at 8 h are observed for M30, M90, and M180 groups, respectively (P < 0.05). The weight loss of various GNGA samples increases significantly after 1 day and continues to increase to 31.3 ± 0.8%, 41.6 ± 1.4%, and 49.1 ± 1.2% at 14 days for M30, M90, and M180 groups, respectively (P < 0.05). These investigations indicate significant physicochemical differences between the three GNGA groups, suggesting that grafting amount of GA determines water content, phase transition temperatures, and degradation of carrier materials (also see the Supporting Information). 3.3. In vitro drug release studies Here, we evaluate the encapsulation efficiency of pilocarpine (loaded at 2% w/v) for various GNGA biomaterials in BSS at 34 °C. According to our previous study, drug encapsulation efficiency reached 62.4 ± 2.0% using only GN carriers [4]. Fig. 3a shows the effect of different grafting amount of GA onto GN on the encapsulation efficiency of pilocarpine nitrate. Our results suggest that drug encapsulation efficiencies are 63.5 ± 1.4%, 71.9 ± 2.0%, and 73.6 ± 1.2% for M30, M90, and M180 groups, respectively (P > 0.05). The significances in drug encapsulation efficiency indicate that increasing amounts of GA grafted to GN increased drug loading during formation of temperature triggered gels (also see the Supporting Information). In vitro drug release studies were performed in sink condition in physiological medium containing MMP-2 at 34 °C over a time course of 14 days. The desirable therapeutic level of pilocarpine was reported in the range of 10–33 lg/ml in order to decrease the elevated IOP [36]. Results of time-course changes in pilocarpine concentrations are shown in Fig. 3b. Here, we categorize

Please cite this article in press as: S.-F. Chou et al., Gallic acid grafting effect on delivery performance and antiglaucoma efficacy of antioxidant-functionalized intracameral pilocarpine carriers, Acta Biomater. (2016), http://dx.doi.org/10.1016/j.actbio.2016.04.035

S.-F. Chou et al. / Acta Biomaterialia xxx (2016) xxx–xxx

5

Fig. 1. (a) Photographs of the reaction of phosphomolybdenum reagent with various GNGA samples. (b) Total antioxidant activities of various GNGA samples were analyzed by UV-Visible spectrophotometry. Results are expressed as antioxidant equivalent concentration. Values are mean ± SD (n = 6). *P < 0.05 vs all groups. (c) Photographs of the reaction of DPPH reagent with various GNGA samples. (d) DPPH scavenging activities of various GNGA samples were analyzed by UV-Visible spectrophotometry. Results are expressed as percentage inhibition of the DPPH radical. Values are mean ± SD (n = 6). *P < 0.05 vs all groups.

the release kinetics of pilocarpine into four stages and they are described as below. Stage one – initial burst release: Significant higher drug concentrations at the initial 0.5 h are detected for all three groups of GNGA samples as compared to their later time points. These observations suggest a burst release behavior perhaps due to pilocarpine localized at the surface of GNGA carriers. Drug concentrations for M30, and M90, and M180 at 0.5 h are 257.3 ± 2.8 lg/ml, 266.3 ± 3.8 lg/ml, and 301.4 ± 5.5 lg/ml, respectively. Higher initial concentrations of released pilocarpine found in M90 and M180 groups are attributed to the increase of drug encapsulation efficiency (Fig. 3a) resulting in a higher amount of surface drug than the M30 groups. In addition, water content measurements (Fig. 2a) suggest that M180 groups are more hydrophilic than M90 followed by M30. This effect yields a faster initial burst and a more rapid release profile of the M180 groups than the other GNGA materials. Stage two – effect of diffusion: After the initial burst release, drug concentrations for M30, M90, and M180 groups significantly decrease at 1 h (P < 0.05). From 1 to 4 h, concentrations of pilocarpine remain relatively stable and are above 10 lg/ml for all GNGA groups. However, there appears to be statistical differences in the level of released drug among the three groups at any given time point in this stage where the pilocarpine concentration for M180 groups is higher than M90 followed by M30 (P < 0.05). The steady release of drug along with grafting amount of GA dependence suggests a diffusion-dominated release phase. In addition, polymer swelling due to differences in hydrophobicity is expected to dominate current diffusion-controlled stage. As mentioned previously, increasing amount of GA grafted onto GN results in the decrease of overall hydrophobicity of the GNGA materials (Fig. 2a). This may allow the GNGA to swell more, which facilitates the diffusion of drug molecules from carrier interior due to a loose polymeric network. In general, our explanation on stage two release associated with the effect of polymer swelling is in accordance with Haseeb et al., who demonstrated a stimuli-responsive release behavior by controlling the dynamic swelling of a polysaccharide-based material [37].

Stage three – combined diffusion and degradation: After stage two, drug concentrations increase again significantly for M90 and M180 groups at 8 h while a significant increase in drug concentration is observed for M30 at 1 day (P < 0.05). The sudden increases of pilocarpine concentrations suggest a different release kinetic pattern as compared to the previous diffusion-controlled release mechanism found in stage two. One possible explanation for this phenomenon is associated with biodegradation of the carrier materials that accelerates drug release resulting in the increase of pilocarpine concentrations at various time points. In addition, the present results reveal that M180 and M90 groups achieve a 10% weight loss at 8 h whereas M30 groups receive a similar weight loss at 1 day (Fig. 2c). Similarly, drug concentrations increase at 8 h for M180 and M90 groups and at 1 day for M30 groups, which suggests that a 10% weight loss of the GNGA materials significantly adds more drugs through degradation to the already established pilocarpine concentration based on the diffusion-controlled release. Therefore, initiation of the degradation in delivery carriers facilitates the release of encapsulated drug in this stage. Stage four – effect of polymer degradation: After stage three, the level of released drug slowly decreases to a time point where pilocarpine concentrations are significantly lower than the previous time point (P < 0.05). The decrease in drug concentration can be observed from Fig. 3b at 3 days for both M90 and M180 groups and 7 days for M30 groups. Since the additional effect of carrier degradation accelerates the drug release from stage three, the remaining pilocarpine in GNGA materials decreases significantly resulting in the sudden decrease and/or depletion of drug concentration (P < 0.05). In addition, Hedberg et al. reported the release behavior of biodegradable carriers and suggested that degradation of the polymer could contribute to the increase of release rates at later stage [38]. Furthermore, as mentioned previously, M90 and M180 samples reach a 20% weight loss at 3 days while a similar weight loss is achieved by M30 at 7 days (Fig. 2c). This observation may account for the further decrease of drug concentration in release behavior. Therefore, in stage four, GNGA degradation dominates the release of pilocarpine and significantly decreases drug

Please cite this article in press as: S.-F. Chou et al., Gallic acid grafting effect on delivery performance and antiglaucoma efficacy of antioxidant-functionalized intracameral pilocarpine carriers, Acta Biomater. (2016), http://dx.doi.org/10.1016/j.actbio.2016.04.035

6

S.-F. Chou et al. / Acta Biomaterialia xxx (2016) xxx–xxx

Fig. 2. (a) Equilibrium water content of various GNGA samples. Values are mean ± SD (n = 5). *P < 0.05 vs all groups. (b) DSC thermograms of GNGA samples in BSS (10% w/v). Each LCST data point represents the average of four different values. *P < 0.05 vs all groups. (c) Time-course of weight loss of GNGA samples after incubation at 34 °C in BSS containing MMP-2. An asterisk indicates statistically significant differences (*P < 0.05; n = 4) for the mean value of weight loss compared to value at previous time point. #P < 0.05 vs all groups (compared only within each time point group). Incubation time point: hour (h); day (d).

concentration under the therapeutic level, especially for the M180 groups. Our delivery carriers demonstrate a tunable release characteristic of drug depending on the grafting amount of GA onto GN copolymers. Here, cumulative release curves of M30, M90, and M180 samples containing pilocarpine are shown in Fig. 3c. All samples exhibit an initial burst of pilocarpine up to 40% of the cumulative release due to surface drug. In addition, increasing the amount of GA in carrier samples increases the cumulative release of pilocarpine from GNGA samples at any given time. For example,

Fig. 3. (a) Drug encapsulation efficiency of various GNGA samples. Values are mean ± SD (n = 4). *P < 0.05 vs all groups; #P < 0.05 vs M30 groups. (b) Time-course of the concentration of pilocarpine released from GNGA samples at 34 °C in BSS containing MMP-2. An asterisk indicates statistically significant differences (*P < 0.05; n = 4) for the mean value of the pilocarpine concentration compared to the value at the previous time point. #P < 0.05 vs all groups (compared only within each time point group). Incubation time point: hour (h); day (d). (c) Cumulative release percentage.

the cumulative release at 1 day is 87% for M180 samples whereas M90 and M30 samples receive 66% and 62% cumulative releases at the same time point, respectively. Furthermore, M180 reaches a 95% cumulative release after 2 days and maintains at this level throughout the release study. Finally, M90 and M30 appear to exhibit a zero order release behavior showing a gradual increase of the cumulative release during the study. Given that 2% (w/v) of pilocarpine was mixed in the GNGA carriers and drug loading was between 63.5 ± 1.4% and 73.6 ± 1.2% for all groups (Fig. 3a), the effect of initial burst was minimized due to a relatively low

Please cite this article in press as: S.-F. Chou et al., Gallic acid grafting effect on delivery performance and antiglaucoma efficacy of antioxidant-functionalized intracameral pilocarpine carriers, Acta Biomater. (2016), http://dx.doi.org/10.1016/j.actbio.2016.04.035

S.-F. Chou et al. / Acta Biomaterialia xxx (2016) xxx–xxx

7

drug content in the materials. For example, the burst release behavior of pilocarpine was demonstrated by Hsiue et al. using a drug loading from 5 wt% to 30 wt% in poly(2-hydroxyethyl methacrylate) films [36]. Their observations showed that 5 wt% of drug loading achieved a 45% cumulative release at 0.5 h while increasing pilocarpine loading to 30 wt% increased the cumulative release to 75% at the same time point. The underlying mechanism of accelerated initial burst release rate due to increasing drug concentration is related to the increasing hydration behavior from additional hydrophilic pilocarpine. In accordance with these findings, our data demonstrate increasing release rates caused by the increase in hydrophilicity of GNGA carriers (Fig. 2a). Overall, in vitro release studies of pilocarpine reveal a concentration dependence of released pilocarpine on the grafting amount of GA onto GN and release mechanisms are associated with diffusion of the drug as well as degradation of the GNGA materials. 3.4. In vitro biocompatibility studies In vitro biocompatibility studies of GNGA carrier materials were performed using human lens epithelial cell cultures. Cytotoxicity data were obtained from live/dead bioassays where green and red fluorescence indicated live and dead cells, respectively. Fig. 4a shows representative fluorescence images of HLE-B3 cells after 2 days of incubation with M30, M90, M180, and the control groups (without any contact of the GNGA biomaterials). Cells from the control groups appear to be healthy and actively proliferate as indicated by prominent green fluorescence with a minimal amount of red fluorescent signal. Similarly to the control groups, HLE-B3 cells maintain their cell viability by fluorescing strong green signals after exposing to M30, M90, and M180 samples, indicating that the drug delivery carriers are biocompatible with the investigated cells and cause no cell damage. In addition, the mitochondrial dehydrogenase activity of HLE-B3 cells after a 2-day exposure to various GNGA biomaterials was further quantitatively measured by means of WST-1 assays. Fig. 4b shows the OD values of M30, M90, and M180 groups as compared to the controls. Results indicate that the control cultures receive an OD of 0.52 ± 0.04 while the OD values for M30, M90, and M180 groups are 0.55 ± 0.06, 0.51 ± 0.03, and 0.52 ± 0.06, respectively and without any statistical significance (P > 0.05). This finding is in accordance with the observations made in Fig. 4a where cells are viable in the presence of test materials under growth conditions. As a result, GNGA biomaterials have little to negligible effects on the cellular proliferative capacity during the course of study. Furthermore, we performed pro-inflammatory gene expression for HLE-B3 cells exposing to various GNGA biomaterials after 2 days. Fig. 4c shows quantitative measurements of real-time RT-PCR on IL-6 mRNA levels of M30, M90, and M180 groups as compared to the controls (set as 100%). The IL-6 expression levels in M30, M90, and M180 groups are 106.1 ± 8.2%, 103.8 ± 10.0%, and 108.0 ± 4.7%, respectively. These values show no statistical difference (P > 0.05), indicating that the test materials have a minimal effect on cellular inflammation. Overall, our results suggest that all studied GNGA carriers are biocompatible in vitro with human lens epithelial cells (also see the Supporting Information).

Fig. 4. (a) Cell viability of HLE-B3 cell cultures was determined by staining with Live/Dead Viability/Cytotoxicity Kit in which live cells fluoresce green and dead cells fluoresce red. Fluorescence images of cells after a 2-day exposure to the GNGA samples. Control: without test materials. Scale bars: 50 lm. (b) The OD value at 450 nm for HLE-B3 cells exposed to the GNGA samples for 2 days. Control: without test materials. Values are mean ± SD (n = 4). (c) Gene expression of IL-6 in HLE-B3 cells incubated with the GNGA samples for 2 days, measured by real-time RT-PCR. Normalization was done by using GAPDH. Data in the experimental groups are percentages relative to that of control groups (without materials). Values are mean ± SD (n = 3).

3.5. Measurement of antioxidant activity against oxidative stress As demonstrated in Fig. 1, our GNGA biomaterials exhibit promising antioxidant activities and abilities to inhibit DPPH free radicals that depend on the grafting amount of GA onto GN. Here, we further examine the antioxidant activity of GNGA carrier materials with various grafting amounts of GA using an in vitro oxidative stress challenge model, involving in cultivation of HLE-B3 cells in a hydrogen peroxide containing environment (200 lM)

[6]. Fig. 5a shows representative optical microscopic images of HLE-B3 cells cultured in various conditions including the presence of hydrogen peroxide and GNGA biomaterials after 2 days. As seen from the images, cells in the control group are healthy and exhibit a typical lens epithelial morphology without the addition of GNGA biomaterials and hydrogen peroxide. By contrast, cells in the HP groups show shrinkage and partial detachment from the culture substrate confirming the toxic effect of hydrogen peroxide on the

Please cite this article in press as: S.-F. Chou et al., Gallic acid grafting effect on delivery performance and antiglaucoma efficacy of antioxidant-functionalized intracameral pilocarpine carriers, Acta Biomater. (2016), http://dx.doi.org/10.1016/j.actbio.2016.04.035

8

S.-F. Chou et al. / Acta Biomaterialia xxx (2016) xxx–xxx

Fig. 5. Effect of polymer carrier materials on H2O2-induced (a, b) cell viability and (c, d) intracellular ROS. (a) Representative phase-contrast micrographs of the HLE-B3 cells after incubation with various GNGA samples M30, M90, and M180 for 24 h and further exposure to H2O2 for 24 h. The cells exposed to 0 (Control group) or 200 (HP group) lM H2O2 for 24 h following 24 h of incubation in the absence of the polymers were used for comparison. Scale bars: 100 lm. (b) Cell viability was measured by the WST-1 assay. Results of metabolic activity were expressed as percentage of control groups. Values are mean ± standard deviation (n = 4). *P < 0.05 vs all groups. (c) Representative fluorescent images of the HLE-B3 cells after incubation with various GNGA samples M30, M90, and M180 for 24 h and further exposure to H2O2 for 24 h. The cells exposed to 0 (Control group) or 200 (HP group) lM H2O2 for 24 h following 24 h of incubation in the absence of the polymers were used for comparison. Scale bars: 50 lm. (d) Intracellular levels of ROS were measured by the fluorescence intensity of DCFH-DA, with a microplate reader. Quantification results were the mean of four independent experiments.

excess oxidative stress in HLE-B3 cell morphology. Furthermore, there appears to be a difference for cells cultured in hydrogen peroxide with the presence of antioxidative GNGA biomaterials. For example, cells in M30 groups show less extent of cell shrinkage and death as compared to the HP groups. In addition, cell morphology of the M180 groups is comparable to the control groups suggesting that increasing the amount of GA grafted on the GN copolymers inhibits the effect from hydrogen peroxide-generated oxidative stress. Quantitative analysis for HLE-B3 cell viability was performed using WST-1 assay. Fig. 5b shows the mitochondrial dehydrogenase activity from various cell cultures as compared to the control groups (100%). The metabolic activity level in the HP groups is 47.9 ± 3.3%, indicating that hydrogen peroxide is a strong oxidant that generates excess oxidative stress on cells. Choudhary et al. studied the effect of hydrogen peroxide on the viability of HLEB3 cells and reported a 50% of cell viability corresponding to 200 lM of hydrogen peroxide in the culture [39]. In the attempt of establishing the baseline for the HP groups on cell viability, our results are in agreement with their reported value. By contrast, cells exposed to GNGA followed by hydrogen peroxide treatment show an increase in metabolic activity. For example, M30 groups receive a metabolic activity of 58.6 ± 2.7% whereas increasing the amount of GA attached to GN significantly increases the activity level to 70.9 ± 2.1% and 85.2 ± 3.5% for M90 and M180, respectively (P < 0.05). The effects of GA concentrations on cell viability after exposure to an organic hydroperoxide were previously demonstrated where a 1.2-fold increase in cell viability was associated with a 10-fold increase in GA molar ratio during grafting process

[17]. In addition, increasing GA concentrations from 1 lM to 100 lM significantly increased viability of human micro-vascular endothelial cells after exposing to hydrogen peroxide (500 lM) [40]. Our results are in agreement with these earlier findings suggesting that GNGA biomaterials mitigate oxidative damage on cells caused by hydrogen peroxide. In addition, we report the dependence of antioxidant activity on the grafting amount of GA onto GNGA carrier materials. In parallel to oxidative stress-induced cell death, generations of ROS in an oxidative environment can often lead to cell death [41,42]. Here, we examine the effects of different grafting amounts of GA in antioxidative GNGA biomaterials on ROS generation. Fig. 5c shows fluorescence images of ROS generation in various HLE-B3 cell cultures containing hydrogen peroxide (200 lM) and GNGA samples. As seen from the images, the control groups exhibit a minimal green fluorescence, indicating a low level of ROS generation. By contrast, in the HP groups, a high level of ROS generation is observed. In addition, M30 groups receive a high level of ROS generation while increasing the grafting amount of GA in GNGA decreases the green fluorescence, indicating the reduction of ROS generation. It appears that the ROS generation in M180 groups is comparable to the control groups suggesting an excellent ability in antioxidant activity. Quantitative analyses of the fluorescence intensities are shown in Fig. 5d. For all groups examined, the intensity profiles are very similar. However, the intensities at near 525 nm wavelength suggest that the order of ranking in all groups is HP > M30 > M90 > M180 > Control. This finding is supported by the DPPH assays where M180 has a better free radical scavenging

Please cite this article in press as: S.-F. Chou et al., Gallic acid grafting effect on delivery performance and antiglaucoma efficacy of antioxidant-functionalized intracameral pilocarpine carriers, Acta Biomater. (2016), http://dx.doi.org/10.1016/j.actbio.2016.04.035

S.-F. Chou et al. / Acta Biomaterialia xxx (2016) xxx–xxx

capacity than M90 followed by M30 (Fig. 1c and d). In addition, phenolic antioxidants, such as GA, exhibited excellent free radical scavenging abilities and metal chelating properties [43]. Therefore, phenolic antioxidants were able to suppress cell signaling pathways and down-regulate gene expression, and consequently, lower the effect of ROS generation on cell behaviors. Our findings are in agreement with this mechanistic explanation in the level of ROS generation using different GNGA biomaterials. In general, in vitro antioxidant studies against oxidative stress caused by hydrogen peroxide suggest that different levels of antioxidant activities can be reached by varying the grafting amount of GA onto GN copolymers. 3.6. Animal studies 3.6.1. IOP and pupil diameter Fig. 6a shows the IOP of rabbits, expressed as differences between treated and normal (i.e., baseline) eye, over 14 days of investigation. All rabbits have IOP near baseline before induction of experimental glaucoma whereas glaucomatous rabbits receive an average IOP of 22.5 ± 1.3 mmHg after 4 weeks of achymotrypsin injection, indicating a successful induction of experimental glaucoma. The IOP in the control rabbits (glaucomatous animals without the injection of polymer and drug) maintain at

Fig. 6. Measurements of (a) IOP and (b) pupil diameter after intracameral injection of various GNGA polymer solutions (M30, M90, and M180) containing pilocarpine in rabbits with glaucoma (GL). Glaucomatous animals receiving no polymer and drug served as control groups (Ctrl). Asterisks indicate statistically significant differences (*P < 0.05; **P < 0.005; n = 6) as compared with the baseline (a) IOP and (b) pupil diameter values. Follow-up time point: preoperation (Pre); hour (h); day (d).

9

an elevated level during the follow-up examinations and gradually increase to 27.9 ± 1.5 mmHg at the end of experiments. By contrast, the IOP significantly decrease to 10.3 ± 1.0 mmHg, 8.1 ± 0.8 mmHg, and 5.9 ± 1.5 mmHg after 4 h for rabbits subjecting to treatment from M30, M90, and M180 groups, respectively (P < 0.05). It is noteworthy that M180 groups achieve a lower IOP at early time points than M30 and M90 groups due to a higher initial drug loading and fast release of pilocarpine within the examination period of 2 days (Fig. 3). Generally, IOP of M30, M90, and M180 reach to a minimal at around 1 day where M30 and M90 maintain at low IOP up to 14 days. Due to depletion of the drug as a result of fast release from M180 groups, IOP start to increase at day 3 and reach to 13.8 ± 1.5 mmHg at 14 days. By contrast, M30 and M90 sustain pilocarpine release up to 14 days within the therapeutic level (Fig. 3), and thus, the IOP for both groups remain at a low level. Our results suggest the dependence of IOP on the GA-mediated drug loading and release rates of M30, M90, and M180 groups. In addition, IOP profiles from in vivo experiments generally follow our in vitro drug release results where decreasing of the IOP in each group is closely related to the 4stage release mechanism of pilocarpine involving diffusion and swelling/degradation of GNGA biomaterials. Pilocarpine regulates IOP due to its ability in constriction of iris/ciliary body leading to pupillary contraction, which further enlarges the trabecular meshwork to drain excess aqueous humor away from the anterior chamber. Fig. 6b shows the profiles of decrease in pupil diameter due to pharmacological action of released pilocarpine from M30, M90, and M180 groups as compared to the Ctrl groups over 14 days. As expected, no significant changes are found in the reduction of pupil diameter for Ctrl groups due to the absence of drug treatments. By contrast, M30, M90, and M180 groups achieve significant decreases in pupil diameter of 2.4 ± 0.3 mm, 3.0 ± 0.2 mm, 3.6 ± 0.2 mm at 4 h, respectively (P < 0.05). Moreover, further reductions of pupil diameter are noted at 8 h followed by plateaus in the decrease of pupil diameter up to 2 days for M30, M90, and M180 groups. At early time points (up to 2 days), M180 groups achieve a greater extent of decrease in pupil diameter than M90 and M30 groups due to a higher drug loading and faster release rate of pilocarpine. However, a significant increase in pupil diameter for the M180 groups from day 2 to day 3 indicates that the release of pilocarpine has been slowed and/or depleted. The decrease in pupil diameter for M180 groups is 1.2 ± 0.3 mm at day 3 and gradually increases to 0.3 ± 0.3 mm at the end of the study. This observation is supported by our in vitro release data where a significant decrease in drug concentration for M180 groups from day 2 to day 3 is noted (Fig. 3b). Apart from M180 groups, M30 and M90 show similar profiles of decrease in pupil diameter over time. For example, the reduction of pupil diameter starts to decrease for M30 groups and M90 groups from day 3 to 1.3 ± 0.2 mm and 2.5 ± 0.3 mm at the end of the study, respectively. The differences in the reduction of pupil diameter between M30 and M90 groups can be explained by our drug release data (Fig. 3b) where M90 groups have a higher pilocarpine concentration than M30 groups over the course of the study. In general, we show that the decrease in pupil diameter is highly related to the pilocarpine released from biomaterial carriers with varying amounts of GA molecules. Sarchahi et al. studied the effects of IOP and pupil diameter using eye drops containing 2% pilocarpine in an in vivo animal model, and their results suggested the IOP decreased from 15.8 ± 1.8 mmHg to 12.6 ± 1.1 mmHg whereas the pupil diameter decreased from 8.4 ± 0.8 mm to 5.8 ± 0.8 mm at a maximum treatment of 5 days [44]. Comparing to their data, the present results suggest a more effective treatment using the same loading of pilocarpine via in situ gelling of GNGA biomaterials. This example clearly demonstrates the difference on the effectiveness in glau-

Please cite this article in press as: S.-F. Chou et al., Gallic acid grafting effect on delivery performance and antiglaucoma efficacy of antioxidant-functionalized intracameral pilocarpine carriers, Acta Biomater. (2016), http://dx.doi.org/10.1016/j.actbio.2016.04.035

10

S.-F. Chou et al. / Acta Biomaterialia xxx (2016) xxx–xxx

coma treatment using a same drug at a same loading but via different dosage forms and intake routes. Furthermore, Kao et al. compared the release of pilocarpine and its corresponding effect on decrease in pupil diameter using different drug delivery systems, in the forms of eye drops, gels, liposomes, and nanoparticles [45]. At the same dosage concentration (0.1% w/v), the order of ranking in the reduction of pupil diameter was nanoparticles > liposomes > gels > eye drops similar to their corresponding in vitro release rates. However, it is noteworthy that the decrease in pupil diameter returned to zero after 1 day of treatment even with nanoparticles as the drug delivery vehicle. The high surface area and low drug loading are the major drawbacks for particle delivery system including nanoparticles and liposomes. In addition, the gel method was less promising overall perhaps due to the requirement for drug to penetrate through the cornea and the possibility of dilution from outside fluids. In another study, Hsiue et al. loaded 20 wt % pilocarpine into poly(2-hydroxyethyl methacrylate) films and suggested that the IOP decreased from 25.4 mmHg to 15 mmHg after 5 h and was maintained for 2 days whereas the pupil diameter decreased from 11 mm to 8 mm during the same time frame [36]. However, the effects from drug potency may occur at such a high loading, which the authors failed to report. Interestingly, the loading of pilocarpine was increased by 10-fold to receive a similar magnitude of effects on IOP and decrease in pupil diameter for films as compared to our in situ gelling systems. Therefore, we demonstrated a promising drug delivery system using GNGA in situ gelling biomaterials carrying minimal loading of pilocarpine with maximum effects in the decrease of IOP and pupil diameter. 3.6.2. Corneal endothelial cell morphology and density A potential effect from oxidative stress and elevated IOP during the progression and development of glaucoma is the change in morphology and density of corneal endothelial cells. Normal corneal endothelial cells on Descemet’s membrane typically exhibit a hexagonal morphology. As seen from Fig. 7a, normal corneal endothelium before induction of experimental glaucoma (Pre groups) shows typical hexagonal cellular morphology whereas endothelial monolayer after glaucoma induction for 4 weeks (GL groups) exhibits an irregular cell shape. In addition, cells subjecting to the induction of experimental glaucoma without any treatment for 14 days (Ctrl groups) appear to exhibit less hexagonal cell morphology than the GL groups. By contrast, M30 and M90 groups show a clear hexagonal cell shape after 14 days, indicating the presence of healthy cells due to antioxidative GNGA biomaterials loaded with pilocarpine. However, it is noteworthy that corneal endothelium from M180 groups exhibits an irregular cell structure perhaps due to the depletion of drug at day 3 (Fig. 3b) followed by an elevated IOP up to 14 days (Fig. 6a). Fig. 7b shows the quantitative measurement of corneal endothelial cell density from images acquired by specular microscopy. The cell density from the Pre groups is 3389 ± 76 cells/ mm2 as compared to the significantly lower cell density from the GL groups at 3011 ± 109 cells/mm2 (P < 0.05), indicating a successful induction of experimental glaucoma. Furthermore, the cell density of the Ctrl groups significantly decreases to 2410 ± 88 cells/ mm2 as compared to the GL groups (P < 0.05) suggesting that glaucoma without treatment can lead to further endothelial cell loss overtime. In such case, a high IOP may account for the significant decrease in cell counts of the Ctrl groups (Fig. 6a). On the other hand, when injecting with antioxidative GNGA biomaterials loaded with pilocarpine, cell densities increase to 2778 ± 132 cells/mm2, 3030 ± 101 cells/mm2, and 2531 ± 105 cells/mm2 for M30, M90, and M180 groups, respectively. Interestingly, M90 and M30 groups have significantly higher cell counts than Ctrl and M180 groups (P < 0.05). This observation can be explained by data from drug release study and IOP measurement, which M90 and M30 groups

Fig. 7. (a) Representative specular microscopic images of corneal endothelium at preoperation (Pre) and in experimental rabbit glaucoma (GL) eyes 2 weeks after intracameral injection of various GNGA polymer solutions (M30, M90, and M180) containing pilocarpine. Glaucomatous animals receiving no polymer and drug served as control groups (Ctrl). (b) Preoperative and postoperative specular microscopy measurements of corneal endothelial cell density. Values are mean ± standard deviation (n = 6). *P < 0.05 vs all groups; #P < 0.05 vs Pre, Ctrl, and M180 groups; +P < 0.05 vs Pre, GL, M30, and M90 groups; ^P < 0.05 vs Pre, Ctrl, M90, and M180 groups; P < 0.05 vs Pre, Ctrl, M30, and M180 groups.

have a higher concentration of pilocarpine (above therapeutic level) than M180 groups (Fig. 3b) leading to a lower IOP than M180 and Ctrl groups (Fig. 6a) after 14 days. Overall, our results suggest that antioxidative GNGA biomaterials loaded with pilocarpine maintain the morphology and density of corneal endothelial cells. The effects of IOP on corneal endothelial cell size, density, and morphology were reported by Ollivier et al., who used an argon

Please cite this article in press as: S.-F. Chou et al., Gallic acid grafting effect on delivery performance and antiglaucoma efficacy of antioxidant-functionalized intracameral pilocarpine carriers, Acta Biomater. (2016), http://dx.doi.org/10.1016/j.actbio.2016.04.035

S.-F. Chou et al. / Acta Biomaterialia xxx (2016) xxx–xxx

11

laser to induce ocular hypertension and elevate IOP [46]. According to these authors, cell morphology changed while an increase in average cell size from 252.4 lm2 to 408.7 lm2 and a decrease in average cell density from 3990.2 cells/mm2 to 2601.7 cells/mm2 were associated with an increase in average IOP from 14.8 mmHg to 38.3 mmHg. In addition, Cho et al. suggested a decrease in corneal endothelial cell density for patients who had glaucoma as well as those with ocular hypertension (high IOP, above 21 mmHg) without glaucoma [47]. Furthermore, our previous study demonstrated that a loss of corneal endothelial cells was attributed to the persistent elevation of IOP in untreated rabbits [4]. These early reports demonstrated the effect of IOP on corneal endothelium. In addition, our antioxidative GNGA biomaterials effectively reduce the IOP (Fig. 6a) and result in the preservation of cell morphology and density. Specifically, drug release data and IOP results suggest that M30 groups are at the therapeutic level to alleviate glaucoma but exhibiting less pharmacological responses than M90 groups. In addition, M90 groups appear to be the ideal GNGA biomaterials among all due to their strong antioxidant activities and efficacious drug release profiles. Finally, M180 groups have the highest amount of antioxidative GA grafted onto GN; however, fast release and depletion of drug in 2 days suggest the elevated high IOP is still the leading cause of changes in cell morphology and decreases in cell density. Overall, we correlate the IOP and corneal endothelial cell morphology and density after treatment using antioxidative GNGA biomaterials loaded with pilocarpine. Our results indicate that treatment with M90 groups is better than M30 groups followed by M180 groups in terms of the preservation of corneal endothelial integrity. 3.6.3. Total antioxidant level and nitrite level Fig. 8a shows the total antioxidant level in the aqueous humor from Ctrl groups (subjecting to induction of experimental glaucoma) and GNGA groups after 14 days. In the Ctrl groups, the antioxidant level is 0.13 ± 0.05 mM, which is significantly lower than that of a normal eye (0.98 ± 0.17 mM) found in our previous study [6]. In addition, after injection of drug-loaded GNGA biomaterials into glaucomatous rabbits, the total antioxidant level increases with increasing amounts of GA molecules grafted to GN. For example, the total antioxidant levels for M30, M90, and M180 are 0.42 ± 0.04 mM, 0.57 ± 0.08 mM, and 0.86 ± 0.09 mM, respectively (P < 0.05). It is known that antioxidants in anterior chamber play an important role in the regulation of oxidative stress by balancing the free radical species. Richer et al. reported the concentrations of several major antioxidants found in mammalian aqueous humor, including cysteine, ascorbic acid, glutathione, uric acid, and tyrosine [48]. In accordance with their results, our findings suggest a high level of total antioxidant in glaucomatous rabbits after receiving drug-containing GNGA injections due to the presence of antioxidant GA molecules. Izzotti et al. evaluated the sensitivity of several anterior chamber tissues to oxidative stress and suggested that the trabecular meshwork is the most susceptible tissue to oxidative stress in progression of glaucoma [49]. In addition, these authors reported significant increases in oxidative DNA damage and DNA fragmentation from the trabecular meshwork cells (1.9-fold and 1.4-fold, respectively) by hydrogen peroxide. Therefore, reduction of oxidative stress may be able to aid to the successful treatment of glaucoma while our antioxidative GNGA biomaterial carriers demonstrate the capability of increasing the total antioxidant level in aqueous humor. In addition to total antioxidant level, Fig. 8b shows the nitrite level in the aqueous humor from Ctrl groups and GNGA groups after 14 days. The nitrite level for the Ctrl groups is 60.4 ± 4.0 lM whereas the nitrite level in normal eyes were reported at 9.4 ± 1.2 lM, previously [6]. For glaucomatous rabbits receiving drug-containing GNGA injections, the nitrite levels for M30, M90,

Fig. 8. The levels of (a) total antioxidant and (b) nitrite in the aqueous humor of experimental rabbit glaucoma eyes 2 weeks after intracameral injection of various GNGA polymer solutions (M30, M90, and M180) containing pilocarpine. Glaucomatous animals receiving no polymer and drug served as control groups (Ctrl). Values are mean ± standard deviation (n = 6). *P < 0.05 vs all groups.

and M180 are 40.2 ± 2.3 lM, 34.3 ± 2.1 lM, and 23.9 ± 3.7 lM, respectively (P < 0.05). Of particular significance, nitrite level determines diverse ocular effects on vasodilation. In addition, preclinical models and clinical studies reported the role of nitrite levels in mediating IOP [50]. Haefliger et al. described the role of nitrites in regulating IOP and suggested that the fundamental mechanism was associated with the intrinsic contractile elements in the trabecular meshwork regulated by the level of nitrites [51]. In accordance with these explanations, our results suggest that GNGA biomaterials lower nitrite level and reduce IOP (Fig. 6a). More importantly, nitrite level in the aqueous humor appears to depend on the amount of GA grafted onto GN copolymers. Overall, we reported the antioxidative in situ gelling biomaterials exhibiting excellent performance in up-regulating the total antioxidant level and down-regulating the nitrite level. The present data will also be beneficial to a more in-depth exploration of GAfunctionalized polymeric carriers in the improvement in glaucoma-related antioxidant status of anterior chamber. Because of combining the advantages of both biodegradability and thermo-responsiveness of polymers, biodegradable thermogels have attracted much attention for their potential applications in the field of injectable biomaterial carriers for sustained drug delivery [52–54]. By means of intracameral pilocarpine administration using antioxidative GNGA hydrogels, an injection depot formulation is found to be able to be applied for further development of more efficient glaucoma therapeutic strategies. 4. Conclusions In this work, we reported the synthesis of antioxidative GNGA in situ gelling drug carriers for antiglaucomatous treatment.

Please cite this article in press as: S.-F. Chou et al., Gallic acid grafting effect on delivery performance and antiglaucoma efficacy of antioxidant-functionalized intracameral pilocarpine carriers, Acta Biomater. (2016), http://dx.doi.org/10.1016/j.actbio.2016.04.035

12

S.-F. Chou et al. / Acta Biomaterialia xxx (2016) xxx–xxx

Specifically, GA was grafted onto GN copolymers at various redox reaction times using a redox pair of AA and hydrogen peroxide as a radical initiator. By increasing the redox reaction time, total antioxidant activities and free radical scavenging abilities against DPPH radicals increased as a result of higher grafting amount of GA onto GN copolymers. Due to hydrophilic nature of GA molecules, physicochemical properties of GNGA hydrogels showed different water content, phase transition temperatures, and degradability directly correlated with grafting amounts of GA on the carrier materials. In addition, pilocarpine encapsulation depended on the grafting amount of GA while in vitro drug release studies suggested a 4-stage release mechanism, dominated by initial burst, diffusion, and degradation. Our results suggested that M90 groups delivered pilocarpine at a steady and continuous rate within the therapeutic level while higher than M30 groups over 14 days. HLE-B3 cell viability assays and pro-inflammatory gene expression suggested that GNGA carriers, regardless of the grafting amount of GA, were biocompatible and caused no inflammatory response to the investigated cells. Furthermore, in vitro oxidative stress challenges using 200 lM of hydrogen peroxide on HLE-B3 cells revealed that GNGA carriers alleviated cell damage from oxidative stress and generation of ROS. Again, M90 groups provided an adequate antioxidative protection to the investigated cells even though the grafting amount of GA was lower than M180 groups. After experimentally induction of glaucomatous rabbits, injections of pilocarpine-loaded GNGA biomaterials to the anterior chamber showed a decreased IOP and pupil diameter closely related to the in vitro drug release behavior. At the end of the in vivo work, observations on the corneal endothelial cells suggested the preservation of hexagonal cell morphology with increased cell density, indicating the mitigation of experimental glaucoma from drug-loaded GNGA biomaterials. Lastly, biochemical analyses of the aqueous humor reveled that increasing grafting amounts of GA in GNGA carriers increased total antioxidant level and decreased the overall nitrite level. Overall, we demonstrated the potential use of antioxidative GNGA hydrogels for glaucoma treatment. Specifically, while it was obvious that increasing grafting amount of GA onto GN copolymers enhanced overall antioxidant activities, our work reflected the facts that grafting antioxidant small molecules facilitated the changes in physicochemical properties, drug release behaviors, and bioactivities of the carrier materials. Of equal importance, this study implied the significance on grafting amounts of antioxidant small molecules to the optimal performance of the in situ gelling drug carrier materials. Disclosures The authors have no conflicts of interest relevant to this article. Acknowledgements This work was supported by grant NHRI-EX104-10311EC from the National Health Research Institutes of Taiwan. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.actbio.2016.04. 035. References [1] R.D. Fechtner, R.N. Weinreb, Mechanisms of optic nerve damage in primary open angle glaucoma, Surv. Ophthalmol. 39 (1994) 23–42.

[2] M.C. Leske, A. Heijl, M. Hussein, B. Bengtsson, L. Hyman, E. Komaroff, Early Manifest Glaucoma Trial Group, Factors for glaucoma progression and the effect of treatment: the early manifest glaucoma trial, Arch. Ophthalmol. 121 (2003) 48–56. [3] S.C. Saccà, A. Izzotti, P. Rossi, C. Traverso, Glaucomatous outflow pathway and oxidative stress, Exp. Eye Res. 84 (2007) 389–399. [4] J.-Y. Lai, A.-C. Hsieh, A gelatin-g-poly(N-isopropylacrylamide) biodegradable in situ gelling delivery system for the intracameral administration of pilocarpine, Biomaterials 33 (2012) 2372–2387. [5] J.-Y. Lai, Biodegradable in situ gelling delivery systems containing pilocarpine as new antiglaucoma formulations: effect of a mercaptoacetic acid/Nisopropylacrylamide molar ratio, Drug Des. Dev. Ther. 7 (2013) 1273–1285. [6] J.-Y. Lai, L.-J. Luo, Antioxidant gallic acid-functionalized biodegradable in situ gelling copolymers for cytoprotective antiglaucoma drug delivery systems, Biomacromolecules 16 (2015) 2950–2963. [7] K. Wolfe, X. Wu, R.H. Liu, Antioxidant activity of apple peels, J. Agric. Food Chem. 51 (2003) 609–614. [8] M.-S. Kang, J.-S. Oh, I.-C. Kang, S.-J. Hong, C.-H. Choi, Inhibitory effect of methyl gallate and gallic acid on oral bacteria, J. Microbiol. 46 (2008) 744–750. [9] J.M. Kratz, C.R. Andrighetti-Fröhner, P.C. Leal, R.J. Nunes, R.A. Yunes, E. Trybala, T. Bergström, C.R.M. Barardi, C.M.O. Simões, Evaluation of anti-HSV-2 activity of gallic acid and pentyl gallate, Biol. Pharm. Bull. 31 (2008) 903–907. [10] S.-H. Kim, C.-D. Jun, K. Suk, B.-J. Choi, H. Lim, S. Park, S.H. Lee, H.-Y. Shin, D.-K. Kim, T.-Y. Shin, Gallic acid inhibits histamine release and pro-inflammatory cytokine production in mast cells, Toxicol. Sci. 91 (2006) 123–131. [11] L. Tavano, R. Muzzalupo, N. Picci, B. de Cindio, Co-encapsulation of antioxidants into niosomal carriers: gastrointestinal release studies for nutraceutical applications, Colloid Surf. B Biointerfaces 114 (2014) 82–88. [12] I.S. Raja, N.N. Fathima, A gelatin based antioxidant enriched biomaterial by grafting and saturation: towards sustained drug delivery from antioxidant matrix, Colloid Surf. B Biointerfaces 128 (2015) 537–543. [13] J. Yang, R. van Lith, K. Baler, R.A. Hoshi, G.A. Ameer, A thermoresponsive biodegradable polymer with intrinsic antioxidant properties, Biomacromolecules 15 (2014) 3942–3952. [14] R. van Lith, E.K. Gregory, J. Yang, M.R. Kibbe, G.A. Ameer, Engineering biodegradable polyester elastomers with antioxidant properties to attenuate oxidative stress in tissues, Biomaterials 35 (2014) 8113–8122. [15] M. Curcio, F. Puoci, F. Iemma, O.I. Parisi, G. Cirillo, U.G. Spizzirri, N. Picci, Covalent insertion of antioxidant molecules on chitosan by a free radical grafting procedure, J. Agric. Food Chem. 57 (2009) 5933–5938. [16] F. Iemma, F. Puoci, M. Curcio, O.I. Parisi, G. Cirillo, U.G. Spizzirri, N. Picci, Ferulic acid as a comonomer in the synthesis of a novel polymeric chain with biological properties, J. Appl. Polym. Sci. 115 (2010) 784–789. [17] M. Senevirathne, Y.-J. Jeon, Y.-T. Kim, P.-J. Park, W.-K. Jung, C.-B. Ahn, J.-Y. Je, Prevention of oxidative stress in Chang liver cells by gallic acid-graftedchitosans, Carbohydr. Polym. 87 (2012) 876–880. [18] N. Isßıklan, F. Kursßun, Synthesis and characterization of graft copolymer of sodium alginate and poly(itaconic acid) by the redox system, Polym. Bull. 70 (2013) 1065–1084. [19] B.S. Kaith, A.S. Singha, S. Kumar, S. Kalia, Mercerization of flax fiber improves the mechanical properties of fiber-reinforced composites, Int. J. Polym. Mater. Polym. Biomater. 57 (2008) 54–72. [20] V.K. Thakur, A.S. Singha, M.K. Thakur, Graft copolymerization of methyl acrylate onto cellulosic biofibers: synthesis, characterization and applications, J. Polym. Environ. 20 (2012) 164–174. [21] W. Xie, P. Xu, Q. Liu, J. Xue, Graft-copolymerization of methylacrylic acid onto hydroxypropyl chitosan, Polym. Bull. 49 (2002) 47–54. [22] J.-Y. Lai, Biocompatibility of chemically cross-linked gelatin hydrogels for ophthalmic use, J. Mater. Sci. Mater. Med. 21 (2010) 1899–1911. [23] J.-Y. Lai, Y.-T. Li, Functional assessment of cross-linked porous gelatin hydrogels for bioengineered cell sheet carriers, Biomacromolecules 11 (2010) 1387–1397. [24] J.-Y. Lai, P.-L. Lu, K.-H. Chen, Y. Tabata, G.-H. Hsiue, Effect of charge and molecular weight on the functionality of gelatin carriers for corneal endothelial cell therapy, Biomacromolecules 7 (2006) 1836–1844. [25] J.-Y. Lai, S.-J. Lue, H.-Y. Cheng, D.H.-K. Ma, Effect of matrix nanostructure on the functionality of carbodiimide cross-linked amniotic membranes as limbal epithelial cell scaffolds, J. Biomed. Nanotechnol. 9 (2013) 2048–2062. [26] J.-Y. Lai, Biofunctionalization of gelatin microcarrier with oxidized hyaluronic acid for corneal keratocyte cultivation, Colloid Surf. B Biointerfaces 122 (2014) 277–286. [27] J.-Y. Lai, Relationship between structure and cytocompatibility of divinyl sulfone cross-linked hyaluronic acid, Carbohydr. Polym. 101 (2014) 203–212. [28] J.-Y. Lai, T.-P. Wang, Y.-T. Li, I.-H. Tu, Synthesis, characterization and ocular biocompatibility of potential keratoprosthetic hydrogels based on photopolymerized poly(2-hydroxyethyl methacrylate)-co-poly(acrylic acid), J. Mater. Chem. 22 (2012) 1812–1823. [29] J.-Y. Lai, Effect of chemical composition on corneal tissue response to photopolymerized materials comprising 2-hydroxyethyl methacrylate and acrylic acid, Mater. Sci. Eng. C 34 (2014) 334–340. [30] D. Koracevic, G. Koracevic, V. Djordjevic, S. Andrejevic, V. Cosic, Method for the measurement of antioxidant activity in human fluids, J. Clin. Pathol. 54 (2001) 356–361. [31] L.C. Green, D.A. Wagner, J. Glogowski, P.L. Skipper, J.S. Wishnok, S.R. Tannenbaum, Analysis of nitrate, nitrite and [15N]nitrate in biological fluids, Anal. Biochem. 126 (1982) 131–138.

Please cite this article in press as: S.-F. Chou et al., Gallic acid grafting effect on delivery performance and antiglaucoma efficacy of antioxidant-functionalized intracameral pilocarpine carriers, Acta Biomater. (2016), http://dx.doi.org/10.1016/j.actbio.2016.04.035

S.-F. Chou et al. / Acta Biomaterialia xxx (2016) xxx–xxx [32] A.Y. Loo, K. Jain, I. Darah, Antioxidant and radical scavenging activities of the pyroligneous acid from a mangrove plant, Rhizophora apiculata, Food Chem. 104 (2007) 300–307. [33] U. Bae, S. Park, S. Jung, B. Park, S. Chae, Hypoglycemic effects of aqueous persimmon leaf extract in a murine model of diabetes, Mol. Med. Rep. 12 (2015) 2547–2554. [34] K.P. Suja, A. Jayalekshmy, C. Arumughan, Free radical scavenging behavior of antioxidant compounds of sesame (Sesamum indicum L.) in DPPH system, J. Agric. Food Chem. 52 (2004) 912–915. [35] U.G. Spizzirri, F. Iemma, F. Puoci, G. Cirillo, M. Curcio, O.I. Parisi, N. Picci, Synthesis of antioxidant polymers by grafting of gallic acid and catechin on gelatin, Biomacromolecules 10 (2009) 1923–1930. [36] G.-H. Hsiue, J.-A. Guu, C.-C. Cheng, Poly (2-hydroxyethyl methacrylate) film as a drug delivery system for pilocarpine, Biomaterials 22 (2001) 1763–1769. [37] M.T. Haseeb, M.A. Hussain, S.H. Yuk, S. Bashir, M. Nauman, Polysaccharides based superabsorbent hydrogel from Linseed: dynamic swelling, stimuli responsive on–off switching and drug release, Carbohydr. Polym. 136 (2016) 750–756. [38] E.L. Hedberg, C.K. Shih, L.A. Solchaga, A.I. Caplan, A.G. Mikos, Controlled release of hyaluronan oligomers from biodegradable polymeric microparticle carriers, J. Control. Release 100 (2004) 257–266. [39] S. Choudhary, W. Zhang, F. Zhou, G.A. Campbell, L.L. Chan, E.B. Thompson, N.H. Ansari, Cellular lipid peroxidation end-products induce apoptosis in human lens epithelial cells, Free Radic. Biol. Med. 32 (2002) 360–369. [40] L. Zhang, Z.-Z. Ma, Y.-Y. Che, N. Li, P.-F. Tu, Protective effect of a new amide compound from Pu-erh tea on human micro-vascular endothelial cell against cytotoxicity induced by hydrogen peroxide, Fitoterapia 82 (2011) 267–271. [41] E. Siemieniuk, L. Kolodziejczyk, E. Skrzydlewska, Oxidative modifications of rat liver cell components during fasciola hepatica infection, Toxicol. Mech. Methods 18 (2008) 519–524. [42] S.K. Tusi, L. Khalaj, G. Ashabi, M. Kiaei, F. Khodagholi, Alginate oligosaccharide protects against endoplasmic reticulum- and mitochondrial-mediated apoptotic cell death and oxidative stress, Biomaterials 32 (2011) 5438–5458. [43] M.A. Soobrattee, V.S. Neergheen, A. Luximon-Ramma, O.I. Aruoma, T. Bahorun, Phenolics as potential antioxidant therapeutic agents: mechanism and actions, Mutat. Res. Fundam. Mol. Mech. Mutagen. 579 (2005) 200–213.

13

[44] A.A. Sarchahi, N. Abbasi, M.A. Gholipour, Effects of an unfixed combination of latanoprost and pilocarpine on the intraocular pressure and pupil size of normal dogs: effects of latanoprost and pilocarpine on IOP of dogs, Vet. Ophthalmol. 15 (2012) 64–70. [45] H.-J. Kao, Y.-L. Lo, H.-R. Lin, S.-P. Yu, Characterization of pilocarpine-loaded chitosan/Carbopol nanoparticles, J. Pharm. Pharmacol. 58 (2006) 179–186. [46] F.J. Ollivier, D.E. Brooks, A.M. Komaromy, M.E. Kallberg, S.E. Andrew, H.L. Sapp, M.B. Sherwood, W.W. Dawson, Corneal thickness and endothelial cell density measured by non-contact specular microscopy and pachymetry in Rhesus macaques (Macaca mulatta) with laser-induced ocular hypertension, Exp. Eye Res. 76 (2003) 671–677. [47] S.W. Cho, J.M. Kim, C.Y. Choi, K.H. Park, Changes in corneal endothelial cell density in patients with normal-tension glaucoma, Jpn. J. Ophthalmol. 53 (2009) 569–573. [48] S.P. Richer, R.C. Rose, Water soluble antioxidants in mammalian aqueous humor: interaction with UV B and hydrogen peroxide, Vision Res. 38 (1998) 2881–2888. [49] A. Izzotti, S.C. Saccà, M. Longobardi, C. Cartiglia, Sensitivity of ocular anterior chamber tissues to oxidative damage and its relevance to the pathogenesis of glaucoma, Invest. Ophthalmol. Vis. Sci. 50 (2009) 5251–5258. [50] M.E. Cavet, J.L. Vittitow, F. Impagnatiello, E. Ongini, E. Bastia, Nitric oxide (NO): an emerging target for the treatment of glaucoma, Invest. Ophthalmol. Vis. Sci. 55 (2014) 5005–5015. [51] I.O. Haefliger, E. Dettmann, R. Liu, P. Meyer, C. Prünte, J. Messerli, J. Flammer, Potential role of nitric oxide and endothelin in the pathogenesis of glaucoma, Surv. Ophthalmol. 43 (1999) S51–S58. [52] W. Shen, J. Luan, L. Cao, J. Sun, L. Yu, J. Ding, Thermogelling polymer-platinum (IV) conjugates for long-term delivery of cisplatin, Biomacromolecules 16 (2015) 105–115. [53] K. Lei, W. Shen, L. Cao, L. Yu, J. Ding, An injectable thermogel with high radiopacity, Chem. Commun. 51 (2015) 6080–6083. [54] L. Zhang, W. Shen, J. Luan, D. Yang, G. Wei, L. Yu, W. Lu, J. Ding, Sustained intravitreal delivery of dexamethasone using an injectable and biodegradable thermogel, Acta Biomater. 23 (2015) 271–281.

Please cite this article in press as: S.-F. Chou et al., Gallic acid grafting effect on delivery performance and antiglaucoma efficacy of antioxidant-functionalized intracameral pilocarpine carriers, Acta Biomater. (2016), http://dx.doi.org/10.1016/j.actbio.2016.04.035