- Email: [email protected]
Electrospun nanofibers – A promising solid in-situ gelling alternative for ocular drug delivery
Journal Pre-proofs Research paper Electrospun nanofibers – a promising solid in-situ gelling alternative for ocular drug delivery Benedikt Göttel, Juliana Martins de Souza e Silva, Cristine Santos de Oliveira, Frank Syrowatka, Miltiadis Fiorentzis, Anja Viestenz, Arne Viestenz, Karsten Mäder PII: DOI: Reference:
S0939-6411(19)31312-8 https://doi.org/10.1016/j.ejpb.2019.11.012 EJPB 13189
To appear in:
European Journal of Pharmaceutics and Biopharmaceutics
Received Date: Revised Date: Accepted Date:
28 June 2019 4 November 2019 14 November 2019
Please cite this article as: B. Göttel, J. Martins de Souza e Silva, C. Santos de Oliveira, F. Syrowatka, M. Fiorentzis, A. Viestenz, A. Viestenz, K. Mäder, Electrospun nanofibers – a promising solid in-situ gelling alternative for ocular drug delivery, European Journal of Pharmaceutics and Biopharmaceutics (2019), doi: https://doi.org/10.1016/j.ejpb.2019.11.012
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
© 2019 Elsevier B.V. All rights reserved.
Electrospun nanofibers – a promising alternative for ocular drug delivery
solid
in-situ
gelling
Benedikt Göttel1, Juliana Martins de Souza e Silva2, Cristine Santos de Oliveira2, Frank Syrowatka3, Miltiadis Fiorentzis4, Anja Viestenz4, Arne Viestenz4, Karsten Mäder1 1Institute
of Pharmacy, Martin Luther University Halle-Wittenberg, Wolfgang-Langenbeck-Str. 4, 06120 Halle (Saale), Germany 2Institute of Physics, Martin Luther University Halle-Wittenberg, Heinrich-Damerow-Str. 4, 06120 Halle (Saale), Germany 3The Interdisciplinary Center of Materials Science, Martin Luther University Halle-Wittenberg, Heinrich-DamerowStr. 4, 06120 Halle (Saale), Germany 4Department of Ophthalmology, Martin Luther University Halle-Wittenberg, University Medicine Halle, ErnstGrube-Str. 40, 06120 Halle (Saale), Germany Corresponding author: [email protected]
Abstract A serious problem of the treatment of eye diseases is the very short residence time of the drug. The majority of the drug is cleared within few seconds due to the poor capability of the eye to accommodate additional liquids. We developed a new ocular drug delivery system, which is applied in dry form and forms immediately a gel after administration. The system is based on gellan gum / pullulan electrospun nanofibers. The rheological behavior of the spinning solution was investigated followed by further characterization of the in situ formed gel. Three-dimensional X-ray imaging with nanometric resolution (nano-CT) and electron scanning microscopy were used for a detailed characterization of the diameter and alignment of the fibers. A high porosity (87.5 ± 0.5%) and pore interconnectivity (99%) was found. To ensure a good fit to the eye anatomy, the prepared fibers were shaped into curved geometries. Additionally, a new innovative moistening chamber for the in vitro determination of the ocular residence time in porcine eyes was developed which mimics the liquid turnover. A clear prolongation of the fluorescein residence time compared to conventional eye drops was achieved with the application of the curved nanofiber in situ gelling mat. In summary, the developed in situ gelling system with adapted geometry is a promising alternative system for ocular drug delivery. Keywords Gellan Gum; electrospinning; In situ gelling; nanofibers; Pullulan; nano-CT; ocular drug delivery; ocular residence time; eye
Introduction Poor bioavailability is the major problem of ocular drug delivery. It results from the small volume, the fast turnover of the tear film and the presence of several physiological barriers which drug molecules have to pass [1,2]. The majority of the applied dose does not reach the desired location, and therefore, low efficacy in the eye and even systemic side effects result [2,3]. Formulation approaches to overcome the challenges include drug delivery systems with increased viscosity [4,5], liposomes [6], nanoparticles [7], hydrogels [8] and lyophilisates [9]. In situ forming systems are promising for ocular drug delivery, because they combine the advantages of low viscous liquids (easy application) and semi-solid dosage forms (longer residence time). Postulated triggers for in situ gelation include temperature [10], pH-shifts [11] or ionic interactions [12]. However, only few products based on ionic gelation have reached the market (e.g. Timoptic XE™). Solid ocular drug delivery systems do currently play only a minor role. However, they have several potential benefits compared to liquids, including a better storage stability (e.g. decreased hydrolysis) and a higher ocular bioavailability due to decreased clearance. In a preferred scenario, the solid DDS dissolves quickly and completely in the tear film and forms a transparent film with a prolonged residence time. Because drug uptake into the eye is mainly diffusion controlled, increased drug concentration gradients will translate into a higher bioavailability. A solid system must rapidly dissolve to a transparent film to avoid an impact of the patient’s vision. Rapid dissolution requires as porous structure which can be obtained by lyophilisation [13–16] or electrospinning [17,18]. The first (1) and main goal of our work was the development of a new solid ocular drug delivery system by electrospinning with optimized properties. Further goals include (2) a detailed characterization of the nanofiber structure and (3) the development of an in vitro test to characterize the formulation performance against liquid drainage. Because of the proven biocompatibility and good performance of gellan gum eye drops (e.g. Gelrite™), we selected gellan gum as an in situ forming polymer. Gellan gum based eye drops prolonged the activity therapeutic effects brinzolamide [19]. Gellan gum eye drops are well tolerated and increase the residence time of trypan blue [20] and the bioavailability of natamycin transfersomes [20]. Comparing different liquid ion gelling polymer systems in vitro and in vivo, gellan gum and carrageenan performed better than HPMC, alginate and chitosan [21,22]. Gellan gum is a negatively charged tetrasaccharide produced by the bacteria Pseudomonas elodea, which consists of two units of β-D-glucose, one of β-D-glucuronate and one of α-L-rhamnose [23,24]. Remarkably, gellan gum is able to gel not only with divalent cations, but also both with monovalent cations [25,26]. However, simple exposure of bulk gellan gum to buffer at body temperature does not result in dissolution and gel formation, but in lump formation. We hypothesized that we will overcome this dissolution problem with the high surface area of electrospun nonwovens. Pilot experiments to spin gellan gum solutions were not successful. The poor electrospinning properties of gellan gum were also observed by other groups [27]. Therefore, Pullulan was selected as a second polymer to assist the spinning process. Pullulan is a linear polysaccharide synthetized by Aureobasidium pullulans and consists of three glucose units, which are connected as maltotriose [28,29]. Pullulan is a nontoxic biopolymer [30] with excellent electrospinning properties [31,32]. The eye curvature makes the application of solid dosage forms more difficult than the instillation of eye drops. A drug delivery system with plane geometry is easy to make, but it has only limited local contact with the curved eye surface. To enable a good fit of the drug
delivery system on the surface of the eye, we aimed to develop solid DDS which mimics the eye curvature. To achieve our second goal, the characterization of the fiber dimension and structure, we applied environmental scanning electron microscopy (ESEM) and X-ray imaging with nanometric resolution (nano-CT). For the nano-CT measurements, we used phase contrast based methods for imaging to avoid the use of contrast agents which might cause different sample characteristics. The third goal was the development of a suitable in vitro model to quantify the ocular residence time. Existing models, like the tear-turnover apparatus with multilayered human immortalized cornea epithelial cells [33], allow the simulation of the tear flow, but the horizontal arrangement does not mirror the gravitation forces, which are present in a realistic scenario. Due to the lack of appropriate models to measure the ocular residence time in vitro, we decided to develop a novel 3D-printed moistening chamber that enables experiments with porcine eyes under physiological conditions. Multispectral fluorescence imaging was applied to measure the local distribution and to quantify the fluorescence signals.
1. Materials and Methods The descriptions of the nano-CT experiments and the measurements of the in vitro residence time are part of the supplement. 1.1. Materials Pullulan was a gift from Nagase GmbH® (Duesseldorf, Germany). The low acetylated gellun gum Kelcogel® GG-LA was bought from CP-Kelco® (Atlanta, USA). Nanofibers were prepared by electrospinning with the Electrospinning Starter Kit of Spraybase® (Maynooth, Ireland). Fluorescein sodium was purchased from Carl Roth GmbH+Co.KG (Karlsruhe, Germany). Simulated tear fluid (STF) was prepared with 6.8 g NaCl, 2.2 g NaHCO3; 0.084 g CaCl.2H2O, 1.4 g KCl and 1 L of double distilled water [34]; pH was adjusted to 7.4 with 0.1 M HCl. 1.2. Nanofiber preparation Mixtures of gellan gum with distilled water were heated up to 80 °C in a water bath until the gellan gum was completely hydrated. For the blended spinning solution, pullulan was added to the hot gellan gum solution. Afterwards, the mixture was treated by vortex mixing until a clear solution was obtained. To remove air bubbles, the spinning solution was centrifuged for 10 min at 1000 rpm. One mL of the spinning solution was filled in a syringe and placed in a pumping system with a flow rate of 0.75 mL/h. The needle-collector-distance was adjusted to 10 cm, the voltage was increased immediately until a stable Taylor cone was achieved. For the fiber formation, a voltage of 12-13 kV was necessary. All experiments were performed by a 22 G needle. Different spinning compositions were carried out. For the fiber preparation 15%, 17.5%, 20% and 22.5% pullulan were spun. The blended fibers were spun from 0.225% gellan gum solution with 15%, 17.5%, 20% and 22.5% pullulan. 1.3. Morphology and size distribution of electrospun fibers A light microscope Axiolab (Carl Zeiss, Jena, Germany) connected to a camera system OLYMPUS UC 30 was used to get a first impression on the structure and dimensions of the electrospun sheets. The fibers were also analyzed with a Philips ESEM XL 30 FEG electron microscope (ESEM). To gain information about the fiber size distribution, a Gaseous
Secondary Electron Detector (GSE) was used. Afterwards, the diameter of the fibers was estimated by the IC Measure® software. 1.4. Rheological investigation of the spinning solution The rheological measurements of the polymer solutions were carried out in a Kinexus lab+® (Malvern, Kassel, Germany). All solutions are expressed in w/v %. Pure pullulan solutions with 4%, 8%, 10% 15% 17.5% and 20% and pure gellan gum solutions with 0.225% and 0.5% in double distilled water were measured. In addition, solutions containing 0.225% gellan gum and 15% or 20% pullulan were analyzed. All solutions were prepared as described above (1.2). To gain a deeper insight of the influence of the viscosity on the spinning process, all measurements were performed at 20 °C and a shear rate of 1-1000 s-1 by a 60 mm cone geometry. 1.5. Rheological measurements of the pullulan-gellan gum fibers For the rheological experiments, a Kinexus lab+® (Malvern, Kassel, Germany) was used. All experiments were performed by plate-plate modus and a gap of 0.5 mm at 34 °C in the oscillation mode. The upper rotation plate had a diameter of 20 mm. The gels were prepared by mixing 100 mg of fibers with 250 µl simulated tear fluid (STF) or double distilled water directly at the rheometer. All measurements were performed within the linear viscoelastic region (LVR). To determine the viscoelastic region, an amplitude sweep was performed, first at a frequency of 1 Hz and a deformation of 0.01-100%. Thereafter, a frequency sweep was performed within the LVR at a deformation of 0.4% and over a range of 0.1-20 Hz. 1.6. Curvature formation A defined curvature of the sample was achieved by forming the spun fibers with a 3D printed matrix. Samples with 1.5 cm diameter and a mean weight of 4.3 mg ± 0.52 mg were punched-out from the prepared mesh. The 1.5 cm disks were formed between the components of the matrix. The punched out disks were placed at the bottom part of the matrix. Thereafter, the upper shell was joined to the matrix bottom. For the lens formation, the matrix was compressed moderately to obtain the required curvature. After 30 seconds of holding pressure, the lens was removed from the matrix. For determination of the curvature, the height h of the lenses was measured after formation in the spherical cap. The curvature determination was performed with the DesignSpark Mechanical 2.0 CAD software. 2. Results and Discussion 2.1. Electrospinning process and fiber morphology For the preparation of gellan gum fibers, our first attempt was to spin the pure gellan gum solution. We observed that the stream of pure gellan gum from the needle was discontinuous without forming fibers, as droplets detached from the needle tip without a steady stream of polymer. In an attempt to improve the fiber formation, we decided to use a second hydrophilic and fast dissolving polymer with good spinnable properties, such as pullulan. Different amounts of pure pullulan were tested as co-polymer to achieve fiber formation. Amounts equal to 10% and 12.5% m/V led to the formation of beaded fibers. Fibers with more homogeneous thickness were obtained when the co-polymer concentration was increased to 15% w/V or more. Thereafter, gellan gum was blended in different ratios with 15 % pullulan (or higher). Amounts between 0.3% and 0.7% of gellan gum fibers were used, but the spinning process was unstable. Droplets still detached from the Taylor cone and dropped down to the spun fibers without drying. After adjustment of the gellan gum content, the
blended solutions of 0.225% gellan gum and 15% pullulan (or higher) resulted in the formation of fibers without instabilities during the spinning process. Figure 1 displays several regions of solution compositions and their product characteristics.
Figure 1. 3D diagram of spinning solution compositions and the characteristics of the obtained product.
ESEM images of the electrospun products at different pullulan-gellan gum ratios (Figure 2 AC) showed that the fibers are smooth with homogenous surface without any spherical beads. The thickness of the fibers (Figure 2 D-E) were in a range from 225 to 450 nm for the 15% pullulan/0.225% gellan gum system; for the 17.5% pullulan/0.225% gellan gum, the thickness varied between 300-525 nm and for the highest pullulan concentration, a thickness range of 250-475 nm was obtained. The size distributions (Figure 2 D-E) illustrate that the thickness increased by increasing the pullulan concentration. The increase of the polymer concentration resulted in larger fiber diameters, in agreement the literature, that reports this same behavior for poly(ethylene oxide) and pullulan [35,36].
Figure 2. A-C: ESEM images of electrospun nanofibers from different pullulan–gellan gum ratios. D-F: Size distribution of different pullulan–gellan gum ratios.
2.2. Rheological investigation of the spinning solution The viscosity of different pullulan solutions (4, 8, 10, 15, 17.5 and 20%) dissolved in double distilled water was measured and the viscosity rises by increasing the amount of polymer (Figure 3 A). The viscosities of the 4%, 8% 15% and 20% pullulan solutions are 0.01; 0.06; 0.5 and 1.65 Pas at a shear rate of 1 s-1, respectively. Pullulan solutions with lower concentration, like 4%, behave almost like a Newtonian fluid, but the rheological behavior clearly changed by increasing the polymer content and shear thinning was observed. The Increasing the shear rate from 1 s-1 to 1000 s-1 caused a decrease of the viscosity to 36.4% for the 20% (w/w) pullulan solutions (from 1.65 Pas at 1 s-1 to 0.6 to Pas at 1000 s-1). With increasing shear rates, the 15% pullulan solution dropped from 0.5 Pas to 0.27 Pas, which corresponds to 54% of the initial viscosity. The shear thinning effect that becomes more significant by increasing the polymer concentration can be explained by the number of random coils of pullulan. With a higher number of polymer coils, the extent of entanglement and the interaction of the polymer chains with each other increases. Pullulan solutions of higher concentrations (≥ 15%) show a Newtonian plateau at lower shear rates, which can be explained by disruption of entanglements and an immediately replacement of the chain entanglements by the molecular motion [37]. If the rate of disruption by shear forces exceeds the entanglement replacement, the viscosity decreases.
Figure 3. A) Shear rate dependence of viscosity η for different pullulan solutions at 20 °C. The rheological data of 4%, 8% and 10% pullulan are magnified in the inset. B) Shear rate dependence of viscosity η for different pullulan, gellan gum and pullulan-gellan gum mixtures at 20 °C
To assess the impact of the presence of gellan gum in the spinning solution, pullulan and gellan gum were mixed at various ratios. The solution with 0.225% gellan gum has a viscosity of 0.03 Pas at 1 s-1 without any shear thinning behavior (Figure 3B). Higher amounts of gellan gum resulted in a higher shear thinning effect with an increased viscosity of 0.5 Pas at the lowest and 0.02 Pas at the highest shear rate. Presumably, the electrostatic repulsion of the carboxylic groups of the gellan gum induce lower interactions between the gellan gum polymeric chains. Therefore, the viscosity reduction takes place immediately after the treatment with shear forces. The applied shear forces at the beginning are sufficient to overcome the polymer chain interactions. It was not possible to obtain uniform fibers with pure gellan gum, probably because of the low extent of entanglement during the spinning process and its anionic nature [31]. For a blended pullulan–gellan gum solution, an increase of the viscosity in comparison to the pure pullulan solution is observed and a significant change of the curve progression with a higher shear thinning effect is found. The viscosity of 20% pullulan is 3.5 Pas and of 0.225% gellan gum 20% pullulan is 5.1 Pas at 1 s-1, thus, the viscosity is influenced by the addition of rather small amounts of gellan gum. Initiated by the polymer blend the plateau disappeared at the beginning, instead a shear thinning process appears immediately after the start of the measurement. During the spinning process of pure gellan gum, droplets detached from the Taylor cone at the needle tip, in a process induced by the electrostatic repulsion of the carboxylic groups [38]. The addition of 10% and 12.5% pullulan to the gellan gum solution causes an increase in the number of chain entanglements. The extent of entanglements correlates with the ability of fiber formation [39]. The interaction and overlapping of the molecular chains prevent the polymer stream from breaking into droplets. Beaded fibers were obtained as the amount of interaction was higher, but was still not sufficient to prevent the breakage of the polymer stream [39]. During the electrospinning process, the polymer solution is processed in regions with different shear rates. At first the polymer solution has to be pumped across the tube into the needle. After reaching the needle tip, the Taylor cone formation takes place. Thereby, high shear rates of 100-1000 s-1 affect the polymer solution, so that the solution can be stretched into fibers [31,40].
Figure 4. Viscosity η dependence of pullulan solutions in water at different concentrations at various shear rates at 20 °C.
In Figure 4, the viscosity of different pullulan solutions at various shear rates is presented. From 0% to 10% of pullulan, no significant differences in the progression of the curve at various shear rates are obtained. At pullulan concentrations below 10%, the extent of shear thinning is very low and the viscosity is independent from the applied shear rate. At higher pullulan concentrations, the shear thinning becomes more important. For concentrations between 10% and 15%, the viscosity (1000 s-1) passes a critical value, so that the polymer concentration allows fiber formation. The critical viscosity for fiber formation is between 0.1 Pas and 0.27 Pas at 1000 s-1. 2.3. Rheological measurements of the pullulan-gellan gum fibers The first step in rheological measurement was the determination of the linear viscoelastic region (LVR) by an amplitude sweep. The storage modulus G’, loss modulus G’’ and complex modulus G* did not change significantly during the deformation process below 1% deformation stress for all samples in the presence and absence of cations (Fig. S2Error! Reference source not found.). Deformation higher than 1% resulted in a significant change of G` and G* for the pullulan-gellan gum fibers gelled by STF. Hence, we concluded that a deformation of 0.4% was appropriate for the following frequency sweeps. The data of the frequency sweep are shown in Fig. S3Error! Reference source not found.. G`, G`` and G* increased by increasing the frequency. The elastic modulus G` of pure pullulan and pullulan-gellan gum fibers gelled with double distilled water is much lower in comparison with the G` of pullulan-gellan gum fibers gelled with STF (Fig. S3-A). This suggests that the elastic amount of the blended fibers is much higher. The loss angle δ of the pure pullulan fibers showed a constant decrease by increasing frequency from 70° to 40° (Fig. S3-D). Low values of the loss angle (20°-30°), which indicate gel formation [41] were observed for pullulan-gellan gum mixtures in STF media, but not in distilled water. Gel formation in gellan gum solutions prepared from gellan bulk material occurs only after heating up to 80 °C. In contrast, no heating step is necessary for the electrospun material, which immediately dissolves and gels after contact with STF. Figure 5 depicts the complex viscosity η* of the measured samples. All samples, except the pullulan-gellan gum fibers with STF showed the same curve progression. Gellan gum
blended fibers started at η* of 1939 Pas, η* decreased after increasing the frequency. The pullulan-gellan gum gel showed a high extent of shear thinning, all other samples showed a much lower effect. Shear thinning behavior is desired to avoid irritancy by high mechanical resistance during blinking of the eye.
Figure 5. Frequency dependence of complex viscosity η* of pure pullulan and pullulan-gellan gum blended fibers in double distilled water or simulated tear fluid at 34 °C.
2.4. Curvature formation For a prolonged residence time and high therapeutic efficacy, it is important to improve the contact between the drug delivery system and the cornea surface. A curved geometry of the fibers is required to achieve an optimal distribution of the drug delivery system on the ocular surface. As a guidance, the dimensions of porcine corneas were taken from literature data with a horizontal diameter of 14.9 mm and a vertical diameter of 12.4 mm [42,43]. Therefore, we prepared flat lenses with 1.5 cm diameter after electrospinning (Figure 6C). The lenses were formed by the matrix shown in Figure 6A,B. The electrospun 20% pullulan 0.225% gellan gum fibers shaped into lens shows lens height h of 2.54 mm ± 0.38 mm (Figure 6E-F). The curvature forming process reduced the lens diameter by 1 mm to 1.4 cm. Thereafter, the curvature of the formed lens was determined by CAD software DesignSpark Mechanical 2.0. The software was used to reconstruct the lens dimensions with the measured values and allowed the curvature determination at 10.92 mm. The lens will be used for human application. In contrast to the porcine eyes, the human horizontal cornea curvature is reported to range between 7.10 and 8.75 mm [44]. We are able to adjust the curvature to the anatomical conditions and this allows a very good adaptation of the ocular system, for human and animal use. These results emphasizes the simplicity of curvature formation of the electrospun product without the addition of crosslinking agents, like radicals or photo-initiating molecules for polymerization [45]. Furthermore, the use of water as a biocompatible spinning solvent makes any purification steps unnecessary.
Figure 6.: A) 3D image of the lens forming matrix (side view); B) 3D view of the forming matrix; C) 1.5 cm diameter flat lens image of mesh immediately after preparation; D) Side view of 1.4 cm diameter lens image after curvature formation; E) Microscopic picture of the lens (side view); F) Lens construction, scaled with the measured dimensions.
2.5.
Three-dimensional X-ray microscopy (nano-CT)
Nano-CT phase-contrast imaging enabled the visualization of the fibers distribution in the samples and their orientation in the space without the addition of any chemicals to increase the image contrast. Images of virtual cuts of the pullulan-gellan gum fiber-lens show that the fibers have an anisotropic orientation and do not show any beads (Figure 7).
Figure 7. Virtual cuts of an exemplary sample of fiber composed of 20% pullulan-0.225% gellan gum imaged with nano-CT. In A), the volume after a single cut is shown, in B) the fibers are pseudo-colored in green, and the virtual cut shown in blue illustrates the binarization of the image for the calculations performed further. In C), the entire volume imaged is shown. Scale bars: 10 μm.
The high porosity and pores interconnectivity is important to enable a fast water penetration and dissolution. A three-dimensional reconstruction of the samples and further image processing enabled the estimation of the porosity of the samples to be equal to 87.5 ± 0.5%, and the interconnectivity of pores to be equal to 99%. The porous structure of the fibers is further illustrated in the video of the supplement data, that allows visualizing the inner structure of the sample, the orientation of the fibers, the fibers thickness and the pores interconnected in all directions. The specific surface area was estimated from the 3D nano-CT datasets and is equal to 7.9 µ𝑚2 µ𝑚3
after curvature formation. The average diameter of the fibers is 317 nm and their
thickness distribution follows a Gaussian curve (Fig. 8Error! Reference source not found.).
Figure 8. Fiber diameter distribution calculated from the nano-CT images of the nanofibers (20% pullulan0.225% gellan gum).
In comparison to the averaged diameter measured by ESEM (398 ± 44 nm), the diameter of the fibers estimated after processing the X-ray 3D dataset was slightly lower (317 ± 42 nm). The size distribution images from nano-CT measurements are displayed in Figure 9.
Figure 9. 3D X-Ray images of the PL-GG nanofibers: A) Grayscale, B) Fiber size distribution image with scale bar.
The differences between the average diameter estimated by 3D X-ray imaging and by ESEM result most likely from differences between the 2D (ESEM) and 3D (nano-CT) imaging methods used. Both techniques have the limitation of providing images of only one small fraction of the entire sample. The ESEM only provides a 2D image of the surface of the sample, and the average diameter calculated was an estimation of the diameter of only 100 fibers. The nano-CT is a more time-consuming technique, as it involves a longer imaging procedure (roughly 9 h long for each 20% pullulan-0.225% gellan gum sample) and a dataset post-processing procedure that can take up to a similar amount of time. However, nano-CT provides a 3D dataset that enables the estimation of the average diameter of the fibers in all three directions, and take into account the variability of the diameter along every single fiber.
In that case, nano-CT produces a more statistically relevant quantification of the variable calculated. Nano-CT produces 3D images of the fibers and enables the estimation of a few parameters in a volumetric dataset. It has the limitation of producing a dataset that is small (in the μm range) when compared to the total size of the fiber-lens produced (in the cm range). However, the consistency of the results obtained for fiber diameter and porosity for three different samples with the same composition support the data obtained, showing that phasecontrast nano-CT is a suitable technique for pullulan fiber imaging. 2.6.
Ocular residence time
In Figure 10, the fluorescence images of fluorescein sodium drops and dye loaded fibers are displayed after different spray applications onto the porcine corneas. Time point zero (t0) shows the images immediately after instillation without any spray application. The conventional eye drops (Figure 10, t0 left) showed a spherical distribution in the center of the cornea, while the fibers showed a more homogenous distribution over the whole cornea surface (Figure 10, t0 right). The fibers showed higher fluorescence intensities at the lower ocular sections, due to the vertical experimental setup caused by the gravitational forces impact the ocular distribution. Over time, the gel formed by the fiber showed an area without any fluorescence along the spray intervals (Figure 10, arrow). This area increased in size with further spray applications, though after t8, a fluorescence signal caused by the fibers can still be detected. In contrast, the fluorescence intensity of the fluorescein drops changes considerably after just one spray application. Just few spots of dye outside of the cornea center were detected, whereas the fibers showed further steady cornea contact. With further spray applications, the signal intensity of the drops decreased very rapidly until no signal was detected after t6. Figure 11 shows kinetics of the fluorescence signal. For eye drops, only 15% signal intensity remained after one spray application. In contrast, the fibers were able to preserve 90% of their original signal intensity. After four applications, the eye drop signal decreased below 5%, but the dissolved fibers had still 65% of their initial intensity. After the last application, the signal of the drops decreased to 2.5%; the signal intensity from the dissolved fibers was still high (40%). These data support a prolonged in vitro residence time caused by the higher viscosity of the fiber formed gel in comparison to eye drops. In addition, compared to the eye drops, the fluorescein was much more homogenously distributed on the ocular surface, because of the fibers lens structure.
Figure 10. Overlay of regular (photo) and fluorescence images of porcine eyes treated with fluorescein sodium eye drops (5 µg/ml) (left), and with pullulan-gellan gum nano-fiber lens (0.0001% fluorescein sodium) (right) after different spray applications.
Figure 11. Fluorescence intensity of sodium fluorescein drops (5 µg/ml) and dye loaded pullulan-gellan gum fibers (0.0001%) during different spray application at the cornea surface.
3. Conclusion We developed a new solid in situ gelling system for the treatment of topical ocular diseases. Nanofibers with high porosity and narrow size distribution were produced by electrospinning of pullulan-gellan gum solutions. Nano-CT imaging enabled the noninvasive and detailed 3D visualization of the inner structure of the nanofibers. Rheological experiments showed that the gelling behavior still existed after the spinning process. We improved the application of the fiber-lens in the eye by creating a curvature with a 3D printed matrix. The system could be easily handled and applied. A new 3D printed model for the in vitro determination of the ocular residence time at porcine eyes was developed. After application, a rapid dissolution and a homogeneous distribution on the porcine cornea combined with a prolonged ocular residence time was observed. Therefore, the new solid in situ gelling system is an attractive alternative to existing ocular drug delivery systems. Acknowledgment The authors thank the material support from Nagase (Europe) GmbH, Duesseldorf, Germany, and DFG support (project grants MA1648/12-2 and WE4051/21-1).
Literature [1]
J.L. Mishima, S., Gasset, A., Klyce, S.D., and Baum, Determination of tear volume and tear flow, [Association for Research in Vision and Ophthalmology, etc.], 1966.
[2]
D.E. Everitt, J. Avorn, Systemic Effects of Medications Used To Treat Glaucoma, Ann. Intern. Med. 112 (1990) 120. doi:10.7326/0003-4819-112-2-120.
[3]
T. Kaila, L. Salminen, R. Huuponen, Systemic absorption of topically applied ocular timolol, J. Ocul. Pharmacol. 1 (1985) 79–83. doi:10.1089/jop.1985.1.79.
[4]
Vincent Hon-Kin Li, J.R. Robinson, Solution viscosity effects on the ocular disposition of cromolyn sodium in the albino rabbit, Int. J. Pharm. 53 (1989) 219–225. doi:10.1016/03785173(89)90315-3.
[5]
C.A. Adler, D.M. Maurice, M.E. Paterson, The effect of viscosity of the vehicle on the penetration of fluorescein into the human eye, Exp. Eye Res. 11 (1971) 34–42. doi:10.1016/S0014-4835(71)80062-3.
[6]
S.. Law, K.. Huang, C.. Chiang, Acyclovir-containing liposomes for potential ocular delivery: Corneal penetration and absorption, J. Control. Release. 63 (2000) 135–140. doi:10.1016/S0168-3659(99)00192-3.
[7]
P. Aksungur, M. Demirbilek, E.B. Denkbaş, J. Vandervoort, A. Ludwig, N. Ünlü, Development and characterization of Cyclosporine A loaded nanoparticles for ocular drug delivery: Cellular toxicity, uptake, and kinetic studies, J. Control. Release. (2011). doi:10.1016/j.jconrel.2011.01.010.
[8]
Y. Luo, K.R. Kirker, G.D. Prestwich, Cross-linked hyaluronic acid hydrogel films: new biomaterials for drug delivery, J. Control. Release. 69 (2000) 169–184. doi:10.1016/S01683659(00)00300-X.
[9]
S. Dinslage, M. Diestelhorst, A. Weichselbaum, R. Süverkrüp, Lyophilisates for drug delivery in ophthalmology: pharmacokinetics of fluorescein in the human anterior segment., Br. J. Ophthalmol. 86 (2002) 1114–7.
[10]
L. Mayol, F. Quaglia, A. Borzacchiello, L. Ambrosio, M.I. La Rotonda, A novel poloxamers/hyaluronic acid in situ forming hydrogel for drug delivery: Rheological, mucoadhesive and in vitro release properties, Eur. J. Pharm. Biopharm. 70 (2008) 199–206. doi:10.1016/J.EJPB.2008.04.025.
[11]
H. Ibrahim, C. Bindschaedler, E. Doelker, P. Buri, R. Gurny, Concept and development of ophthalmic pseudo-latexes triggered by pH, Int. J. Pharm. 77 (1991) 211–219. doi:10.1016/0378-5173(91)90319-J.
[12]
A. Rozier, C. Mazuel, J. Grove, B. Plazonnet, Gelrite®: A novel, ion-activated, in-situ gelling polymer for ophthalmic vehicles. Effect on bioavailability of timolol, Int. J. Pharm. 57 (1989) 163–168. doi:10.1016/0378-5173(89)90305-0.
[13]
K. Abduljalil, M. Diestelhorst, O. Doroshyenko, A. Lux, A. Steinfeld, S. Dinslage, R. Süverkrüp, U. Fuhr, Modelling ocular pharmacokinetics of fluorescein administered as lyophilisate or conventional eye drops, Eur. J. Clin. Pharmacol. 64 (2008) 521–529. doi:10.1007/s00228-0070457-3.
[14]
H. Refai, R. Tag, Development and characterization of sponge-like acyclovir ocular minitablets,
Drug Deliv. 18 (2011) 38–45. doi:10.3109/10717544.2010.509364. [15]
R.M. Moosa, Y.E. Choonara, L.C. Du Toit, P. Kumar, L.K. Tomar, C. Tyagi, T.R. Carmichael, V. Pillay, In vitro pharmaceutical characterization and statistical optimization of a novel topically applied instantly-soluble solid eye drop matrix, Pharm. Dev. Technol. 20 (2015) 854–862. doi:10.3109/10837450.2014.930488.
[16]
O. Saher, D.M. Ghorab, N.M. Mursi, Levofloxacin hemihydrate ocular semi-sponges for topical treatment of bacterial conjunctivitis: Formulation and in-vitro/in-vivo characterization, J. Drug Deliv. Sci. Technol. 31 (2016) 22–34. doi:10.1016/j.jddst.2015.11.004.
[17]
G. Di Prima, M. Licciardi, F. Carfì Pavia, A.I. Lo Monte, G. Cavallaro, G. Giammona, Microfibrillar polymeric ocular inserts for triamcinolone acetonide delivery, Int. J. Pharm. 567 (2019) 118459. https://www.sciencedirect.com/science/article/pii/S0378517319304934?via%3Dihub (accessed October 3, 2019).
[18]
Gagandeep, T. Garg, B. Malik, G. Rath, A.K. Goyal, Development and characterization of nanofiber patch for the treatment of glaucoma, Eur. J. Pharm. Sci. 53 (2014) 10–16. doi:10.1016/J.EJPS.2013.11.016.
[19]
J. Sun, Z. Zhou, A novel ocular delivery of brinzolamide based on gellan gum: In vitro and in vivo evaluation, Drug Des. Devel. Ther. 12 (2018) 383–389. doi:10.2147/DDDT.S153405.
[20]
A. Fernández-Ferreiro, M. González Barcia, M. Gil-Martínez, A. Vieites-Prado, I. Lema, B. Argibay, J. Blanco Méndez, M.J. Lamas, F.J. Otero-Espinar, In vitro and in vivo ocular safety and eye surface permanence determination by direct and Magnetic Resonance Imaging of ion-sensitive hydrogels based on gellan gum and kappa-carrageenan, Eur. J. Pharm. Biopharm. 94 (2015) 342–351. doi:10.1016/j.ejpb.2015.06.003.
[21]
I.D. Rupenthal, C.R. Green, R.G. Alany, Comparison of ion-activated in situ gelling systems for ocular drug delivery. Part 1: Physicochemical characterisation and in vitro release, Int. J. Pharm. 411 (2011) 69–77. doi:10.1016/j.ijpharm.2011.03.042.
[22]
I.D. Rupenthal, C.R. Green, R.G. Alany, Comparison of ion-activated in situ gelling systems for ocular drug delivery. Part 2: Precorneal retention and in vivo pharmacodynamic study, Int. J. Pharm. 411 (2011) 78–85. doi:10.1016/j.ijpharm.2011.03.043.
[23]
P.-E. Jansson, B. Lindberg, P.A. Sandford, Structural studies of gellan gum, an extracellular polysaccharide elaborated by Pseudomonas elodea, Carbohydr. Res. 124 (1983) 135–139. doi:10.1016/0008-6215(83)88361-X.
[24]
E.R. Morris, K. Nishinari, M. Rinaudo, Gelation of gellan – A review, Food Hydrocoll. 28 (2012) 373–411. doi:10.1016/J.FOODHYD.2012.01.004.
[25]
H. Grasdalen, O. Smidsrød, Gelation of gellan gum, Carbohydr. Polym. 7 (1987) 371–393. doi:10.1016/0144-8617(87)90004-X.
[26]
H. Moritaka, H. Fukuba, K. Kumeno, N. Nakahama, K. Nishinari, Effect of monovalent and divalent cations on the rheological properties of gellan gels, Food Hydrocoll. 4 (1991) 495– 507. doi:10.1016/S0268-005X(09)80200-2.
[27]
M. Rostami, M. Ghorbani, M. Aman mohammadi, M. Delavar, M. Tabibiazar, S. Ramezani, Development of resveratrol loaded chitosan-gellan nanofiber as a novel gastrointestinal delivery system, Int. J. Biol. Macromol. 135 (2019) 698–705. doi:10.1016/j.ijbiomac.2019.05.187.
[28]
B.J. Catley, A. Ramsay, C. Servis, Observations on the structure of the fungal extracellular polysaccharide, pullulan, Carbohydr. Res. 153 (1986) 79–86. doi:10.1016/S00086215(00)90197-6.
[29]
H. Bender, J. Lehmann, K. Wallenfels, Pullulan, ein extracelluläres Glucan von Pullularia pullulans, Biochim. Biophys. Acta. 36 (1959) 309–316. doi:10.1016/0006-3002(59)90172-6.
[30]
T. Kimoto, T. Shibuya, S.S.-F. and C. Toxicology, undefined 1997, Safety studies of a novel starch, pullulan: chronic toxicity in rats and bacterial mutagenicity, Elsevier. (n.d.).
[31]
A.C. Stijnman, I. Bodnar, R. Hans Tromp, Electrospinning of food-grade polysaccharides, Food Hydrocoll. 25 (2011) 1393–1398. doi:10.1016/J.FOODHYD.2011.01.005.
[32]
M.R. Karim, H.W. Lee, R. Kim, B.C. Ji, J.W. Cho, T.W. Son, W. Oh, J.H. Yeum, Preparation and characterization of electrospun pullulan/montmorillonite nanofiber mats in aqueous solution, Carbohydr. Polym. 78 (2009) 336–342. doi:10.1016/J.CARBPOL.2009.04.024.
[33]
J. Li, S. Tian, Q. Tao, Y. Zhao, R. Gui, F. Yang, L. Zang, Y. Chen, Q. Ping, D. Hou, Montmorillonite/chitosan nanoparticles as a novel controlled-release topical ophthalmic delivery system for the treatment of glaucoma., Int. J. Nanomedicine. 13 (2018) 3975–3987. doi:10.2147/IJN.S162306.
[34]
M. Paulsson, H. Hägerström, K. Edsman, Rheological studies of the gelation of deacetylated gellan gum (Gelrite(®)) in physiological conditions, Eur. J. Pharm. Sci. 9 (1999) 99–105. doi:10.1016/S0928-0987(99)00051-2.
[35]
A. Torre-Muruzabal, L. Daelemans, G. Van Assche, K. De Clerck, H. Rahier, Creation of a nanovascular network by electrospun sacrificial nanofibers for self-healing applications and its effect on the flexural properties of the bulk material, Polym. Test. 54 (2016) 78–83. doi:10.1016/J.POLYMERTESTING.2016.06.026.
[36]
J.M. Deitzel, J. Kleinmeyer, D. Harris, N.C. Beck Tan, The effect of processing variables on the morphology of electrospun nanofibers and textiles, Polymer (Guildf). 42 (2001) 261–272. doi:10.1016/S0032-3861(00)00250-0.
[37]
E.R. Morris, A.N. Cutler, S.B. Ross-Murphy, D.A. Rees, J. Price, Concentration and shear rate dependence of viscosity in random coil polysaccharide solutions, Carbohydr. Polym. 1 (1981) 5–21. doi:10.1016/0144-8617(81)90011-4.
[38]
P. Vashisth, P.A. Pruthi, R.P. Singh, V. Pruthi, Process optimization for fabrication of gellan based electrospun nanofibers, Carbohydr. Polym. 109 (2014) 16–21. doi:10.1016/J.CARBPOL.2014.03.003.
[39]
S.L. Shenoy, W.D. Bates, H.L. Frisch, G.E. Wnek, Role of chain entanglements on fiber formation during electrospinning of polymer solutions: Good solvent, non-specific polymerpolymer interaction limit, Polymer (Guildf). 46 (2005) 3372–3384. doi:10.1016/j.polymer.2005.03.011.
[40]
T. Han, A.L. Yarin, D.H. Reneker, Viscoelastic electrospun jets: Initial stresses and elongational rheometry, Polymer (Guildf). 49 (2008) 1651–1658. doi:10.1016/J.POLYMER.2008.01.035.
[41]
J. Carlfors, K. Edsman, R. Petersson, K. Jörnving, Rheological evaluation of Gelrite® in situ gels for ophthalmic use, Eur. J. Pharm. Sci. 6 (1998) 113–119. doi:10.1016/S0928-0987(97)000742.
[42]
C. Faber, E. Scherfig, J.U. Prause, K.E. Sørensen, Corneal thickness in pigs measured by
ultrasound pachymetry in vivo, Scand. J. Lab. Anim. Sci. (2008). [43]
J. HEICHEL, F. WILHELM, K.S. KUNERT, T. HAMMER, Topographic Findings of the Porcine Cornea, Med. Hypothesis, Discov. Innov. Ophthalmol. 5 (2016) 125.
[44]
P.M. Kiely, G. Smith, L.G. Carney, Meridional variations of corneal shape, Am. J. Optom. Physiol. Opt. (1984). doi:10.1097/00006324-198410000-00001.
[45]
H. Hiratani, A. Fujiwara, Y. Tamiya, Y. Mizutani, C. Alvarez-Lorenzo, Ocular release of timolol from molecularly imprinted soft contact lenses, Biomaterials. 26 (2005) 1293–1298. doi:10.1016/J.BIOMATERIALS.2004.04.030.
S0939-6411(19)31312-8 https://doi.org/10.1016/j.ejpb.2019.11.012 EJPB 13189
To appear in:
European Journal of Pharmaceutics and Biopharmaceutics
Received Date: Revised Date: Accepted Date:
28 June 2019 4 November 2019 14 November 2019
Please cite this article as: B. Göttel, J. Martins de Souza e Silva, C. Santos de Oliveira, F. Syrowatka, M. Fiorentzis, A. Viestenz, A. Viestenz, K. Mäder, Electrospun nanofibers – a promising solid in-situ gelling alternative for ocular drug delivery, European Journal of Pharmaceutics and Biopharmaceutics (2019), doi: https://doi.org/10.1016/j.ejpb.2019.11.012
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
© 2019 Elsevier B.V. All rights reserved.
Electrospun nanofibers – a promising alternative for ocular drug delivery
solid
in-situ
gelling
Benedikt Göttel1, Juliana Martins de Souza e Silva2, Cristine Santos de Oliveira2, Frank Syrowatka3, Miltiadis Fiorentzis4, Anja Viestenz4, Arne Viestenz4, Karsten Mäder1 1Institute
of Pharmacy, Martin Luther University Halle-Wittenberg, Wolfgang-Langenbeck-Str. 4, 06120 Halle (Saale), Germany 2Institute of Physics, Martin Luther University Halle-Wittenberg, Heinrich-Damerow-Str. 4, 06120 Halle (Saale), Germany 3The Interdisciplinary Center of Materials Science, Martin Luther University Halle-Wittenberg, Heinrich-DamerowStr. 4, 06120 Halle (Saale), Germany 4Department of Ophthalmology, Martin Luther University Halle-Wittenberg, University Medicine Halle, ErnstGrube-Str. 40, 06120 Halle (Saale), Germany Corresponding author: [email protected]
Abstract A serious problem of the treatment of eye diseases is the very short residence time of the drug. The majority of the drug is cleared within few seconds due to the poor capability of the eye to accommodate additional liquids. We developed a new ocular drug delivery system, which is applied in dry form and forms immediately a gel after administration. The system is based on gellan gum / pullulan electrospun nanofibers. The rheological behavior of the spinning solution was investigated followed by further characterization of the in situ formed gel. Three-dimensional X-ray imaging with nanometric resolution (nano-CT) and electron scanning microscopy were used for a detailed characterization of the diameter and alignment of the fibers. A high porosity (87.5 ± 0.5%) and pore interconnectivity (99%) was found. To ensure a good fit to the eye anatomy, the prepared fibers were shaped into curved geometries. Additionally, a new innovative moistening chamber for the in vitro determination of the ocular residence time in porcine eyes was developed which mimics the liquid turnover. A clear prolongation of the fluorescein residence time compared to conventional eye drops was achieved with the application of the curved nanofiber in situ gelling mat. In summary, the developed in situ gelling system with adapted geometry is a promising alternative system for ocular drug delivery. Keywords Gellan Gum; electrospinning; In situ gelling; nanofibers; Pullulan; nano-CT; ocular drug delivery; ocular residence time; eye
Introduction Poor bioavailability is the major problem of ocular drug delivery. It results from the small volume, the fast turnover of the tear film and the presence of several physiological barriers which drug molecules have to pass [1,2]. The majority of the applied dose does not reach the desired location, and therefore, low efficacy in the eye and even systemic side effects result [2,3]. Formulation approaches to overcome the challenges include drug delivery systems with increased viscosity [4,5], liposomes [6], nanoparticles [7], hydrogels [8] and lyophilisates [9]. In situ forming systems are promising for ocular drug delivery, because they combine the advantages of low viscous liquids (easy application) and semi-solid dosage forms (longer residence time). Postulated triggers for in situ gelation include temperature [10], pH-shifts [11] or ionic interactions [12]. However, only few products based on ionic gelation have reached the market (e.g. Timoptic XE™). Solid ocular drug delivery systems do currently play only a minor role. However, they have several potential benefits compared to liquids, including a better storage stability (e.g. decreased hydrolysis) and a higher ocular bioavailability due to decreased clearance. In a preferred scenario, the solid DDS dissolves quickly and completely in the tear film and forms a transparent film with a prolonged residence time. Because drug uptake into the eye is mainly diffusion controlled, increased drug concentration gradients will translate into a higher bioavailability. A solid system must rapidly dissolve to a transparent film to avoid an impact of the patient’s vision. Rapid dissolution requires as porous structure which can be obtained by lyophilisation [13–16] or electrospinning [17,18]. The first (1) and main goal of our work was the development of a new solid ocular drug delivery system by electrospinning with optimized properties. Further goals include (2) a detailed characterization of the nanofiber structure and (3) the development of an in vitro test to characterize the formulation performance against liquid drainage. Because of the proven biocompatibility and good performance of gellan gum eye drops (e.g. Gelrite™), we selected gellan gum as an in situ forming polymer. Gellan gum based eye drops prolonged the activity therapeutic effects brinzolamide [19]. Gellan gum eye drops are well tolerated and increase the residence time of trypan blue [20] and the bioavailability of natamycin transfersomes [20]. Comparing different liquid ion gelling polymer systems in vitro and in vivo, gellan gum and carrageenan performed better than HPMC, alginate and chitosan [21,22]. Gellan gum is a negatively charged tetrasaccharide produced by the bacteria Pseudomonas elodea, which consists of two units of β-D-glucose, one of β-D-glucuronate and one of α-L-rhamnose [23,24]. Remarkably, gellan gum is able to gel not only with divalent cations, but also both with monovalent cations [25,26]. However, simple exposure of bulk gellan gum to buffer at body temperature does not result in dissolution and gel formation, but in lump formation. We hypothesized that we will overcome this dissolution problem with the high surface area of electrospun nonwovens. Pilot experiments to spin gellan gum solutions were not successful. The poor electrospinning properties of gellan gum were also observed by other groups [27]. Therefore, Pullulan was selected as a second polymer to assist the spinning process. Pullulan is a linear polysaccharide synthetized by Aureobasidium pullulans and consists of three glucose units, which are connected as maltotriose [28,29]. Pullulan is a nontoxic biopolymer [30] with excellent electrospinning properties [31,32]. The eye curvature makes the application of solid dosage forms more difficult than the instillation of eye drops. A drug delivery system with plane geometry is easy to make, but it has only limited local contact with the curved eye surface. To enable a good fit of the drug
delivery system on the surface of the eye, we aimed to develop solid DDS which mimics the eye curvature. To achieve our second goal, the characterization of the fiber dimension and structure, we applied environmental scanning electron microscopy (ESEM) and X-ray imaging with nanometric resolution (nano-CT). For the nano-CT measurements, we used phase contrast based methods for imaging to avoid the use of contrast agents which might cause different sample characteristics. The third goal was the development of a suitable in vitro model to quantify the ocular residence time. Existing models, like the tear-turnover apparatus with multilayered human immortalized cornea epithelial cells [33], allow the simulation of the tear flow, but the horizontal arrangement does not mirror the gravitation forces, which are present in a realistic scenario. Due to the lack of appropriate models to measure the ocular residence time in vitro, we decided to develop a novel 3D-printed moistening chamber that enables experiments with porcine eyes under physiological conditions. Multispectral fluorescence imaging was applied to measure the local distribution and to quantify the fluorescence signals.
1. Materials and Methods The descriptions of the nano-CT experiments and the measurements of the in vitro residence time are part of the supplement. 1.1. Materials Pullulan was a gift from Nagase GmbH® (Duesseldorf, Germany). The low acetylated gellun gum Kelcogel® GG-LA was bought from CP-Kelco® (Atlanta, USA). Nanofibers were prepared by electrospinning with the Electrospinning Starter Kit of Spraybase® (Maynooth, Ireland). Fluorescein sodium was purchased from Carl Roth GmbH+Co.KG (Karlsruhe, Germany). Simulated tear fluid (STF) was prepared with 6.8 g NaCl, 2.2 g NaHCO3; 0.084 g CaCl.2H2O, 1.4 g KCl and 1 L of double distilled water [34]; pH was adjusted to 7.4 with 0.1 M HCl. 1.2. Nanofiber preparation Mixtures of gellan gum with distilled water were heated up to 80 °C in a water bath until the gellan gum was completely hydrated. For the blended spinning solution, pullulan was added to the hot gellan gum solution. Afterwards, the mixture was treated by vortex mixing until a clear solution was obtained. To remove air bubbles, the spinning solution was centrifuged for 10 min at 1000 rpm. One mL of the spinning solution was filled in a syringe and placed in a pumping system with a flow rate of 0.75 mL/h. The needle-collector-distance was adjusted to 10 cm, the voltage was increased immediately until a stable Taylor cone was achieved. For the fiber formation, a voltage of 12-13 kV was necessary. All experiments were performed by a 22 G needle. Different spinning compositions were carried out. For the fiber preparation 15%, 17.5%, 20% and 22.5% pullulan were spun. The blended fibers were spun from 0.225% gellan gum solution with 15%, 17.5%, 20% and 22.5% pullulan. 1.3. Morphology and size distribution of electrospun fibers A light microscope Axiolab (Carl Zeiss, Jena, Germany) connected to a camera system OLYMPUS UC 30 was used to get a first impression on the structure and dimensions of the electrospun sheets. The fibers were also analyzed with a Philips ESEM XL 30 FEG electron microscope (ESEM). To gain information about the fiber size distribution, a Gaseous
Secondary Electron Detector (GSE) was used. Afterwards, the diameter of the fibers was estimated by the IC Measure® software. 1.4. Rheological investigation of the spinning solution The rheological measurements of the polymer solutions were carried out in a Kinexus lab+® (Malvern, Kassel, Germany). All solutions are expressed in w/v %. Pure pullulan solutions with 4%, 8%, 10% 15% 17.5% and 20% and pure gellan gum solutions with 0.225% and 0.5% in double distilled water were measured. In addition, solutions containing 0.225% gellan gum and 15% or 20% pullulan were analyzed. All solutions were prepared as described above (1.2). To gain a deeper insight of the influence of the viscosity on the spinning process, all measurements were performed at 20 °C and a shear rate of 1-1000 s-1 by a 60 mm cone geometry. 1.5. Rheological measurements of the pullulan-gellan gum fibers For the rheological experiments, a Kinexus lab+® (Malvern, Kassel, Germany) was used. All experiments were performed by plate-plate modus and a gap of 0.5 mm at 34 °C in the oscillation mode. The upper rotation plate had a diameter of 20 mm. The gels were prepared by mixing 100 mg of fibers with 250 µl simulated tear fluid (STF) or double distilled water directly at the rheometer. All measurements were performed within the linear viscoelastic region (LVR). To determine the viscoelastic region, an amplitude sweep was performed, first at a frequency of 1 Hz and a deformation of 0.01-100%. Thereafter, a frequency sweep was performed within the LVR at a deformation of 0.4% and over a range of 0.1-20 Hz. 1.6. Curvature formation A defined curvature of the sample was achieved by forming the spun fibers with a 3D printed matrix. Samples with 1.5 cm diameter and a mean weight of 4.3 mg ± 0.52 mg were punched-out from the prepared mesh. The 1.5 cm disks were formed between the components of the matrix. The punched out disks were placed at the bottom part of the matrix. Thereafter, the upper shell was joined to the matrix bottom. For the lens formation, the matrix was compressed moderately to obtain the required curvature. After 30 seconds of holding pressure, the lens was removed from the matrix. For determination of the curvature, the height h of the lenses was measured after formation in the spherical cap. The curvature determination was performed with the DesignSpark Mechanical 2.0 CAD software. 2. Results and Discussion 2.1. Electrospinning process and fiber morphology For the preparation of gellan gum fibers, our first attempt was to spin the pure gellan gum solution. We observed that the stream of pure gellan gum from the needle was discontinuous without forming fibers, as droplets detached from the needle tip without a steady stream of polymer. In an attempt to improve the fiber formation, we decided to use a second hydrophilic and fast dissolving polymer with good spinnable properties, such as pullulan. Different amounts of pure pullulan were tested as co-polymer to achieve fiber formation. Amounts equal to 10% and 12.5% m/V led to the formation of beaded fibers. Fibers with more homogeneous thickness were obtained when the co-polymer concentration was increased to 15% w/V or more. Thereafter, gellan gum was blended in different ratios with 15 % pullulan (or higher). Amounts between 0.3% and 0.7% of gellan gum fibers were used, but the spinning process was unstable. Droplets still detached from the Taylor cone and dropped down to the spun fibers without drying. After adjustment of the gellan gum content, the
blended solutions of 0.225% gellan gum and 15% pullulan (or higher) resulted in the formation of fibers without instabilities during the spinning process. Figure 1 displays several regions of solution compositions and their product characteristics.
Figure 1. 3D diagram of spinning solution compositions and the characteristics of the obtained product.
ESEM images of the electrospun products at different pullulan-gellan gum ratios (Figure 2 AC) showed that the fibers are smooth with homogenous surface without any spherical beads. The thickness of the fibers (Figure 2 D-E) were in a range from 225 to 450 nm for the 15% pullulan/0.225% gellan gum system; for the 17.5% pullulan/0.225% gellan gum, the thickness varied between 300-525 nm and for the highest pullulan concentration, a thickness range of 250-475 nm was obtained. The size distributions (Figure 2 D-E) illustrate that the thickness increased by increasing the pullulan concentration. The increase of the polymer concentration resulted in larger fiber diameters, in agreement the literature, that reports this same behavior for poly(ethylene oxide) and pullulan [35,36].
Figure 2. A-C: ESEM images of electrospun nanofibers from different pullulan–gellan gum ratios. D-F: Size distribution of different pullulan–gellan gum ratios.
2.2. Rheological investigation of the spinning solution The viscosity of different pullulan solutions (4, 8, 10, 15, 17.5 and 20%) dissolved in double distilled water was measured and the viscosity rises by increasing the amount of polymer (Figure 3 A). The viscosities of the 4%, 8% 15% and 20% pullulan solutions are 0.01; 0.06; 0.5 and 1.65 Pas at a shear rate of 1 s-1, respectively. Pullulan solutions with lower concentration, like 4%, behave almost like a Newtonian fluid, but the rheological behavior clearly changed by increasing the polymer content and shear thinning was observed. The Increasing the shear rate from 1 s-1 to 1000 s-1 caused a decrease of the viscosity to 36.4% for the 20% (w/w) pullulan solutions (from 1.65 Pas at 1 s-1 to 0.6 to Pas at 1000 s-1). With increasing shear rates, the 15% pullulan solution dropped from 0.5 Pas to 0.27 Pas, which corresponds to 54% of the initial viscosity. The shear thinning effect that becomes more significant by increasing the polymer concentration can be explained by the number of random coils of pullulan. With a higher number of polymer coils, the extent of entanglement and the interaction of the polymer chains with each other increases. Pullulan solutions of higher concentrations (≥ 15%) show a Newtonian plateau at lower shear rates, which can be explained by disruption of entanglements and an immediately replacement of the chain entanglements by the molecular motion [37]. If the rate of disruption by shear forces exceeds the entanglement replacement, the viscosity decreases.
Figure 3. A) Shear rate dependence of viscosity η for different pullulan solutions at 20 °C. The rheological data of 4%, 8% and 10% pullulan are magnified in the inset. B) Shear rate dependence of viscosity η for different pullulan, gellan gum and pullulan-gellan gum mixtures at 20 °C
To assess the impact of the presence of gellan gum in the spinning solution, pullulan and gellan gum were mixed at various ratios. The solution with 0.225% gellan gum has a viscosity of 0.03 Pas at 1 s-1 without any shear thinning behavior (Figure 3B). Higher amounts of gellan gum resulted in a higher shear thinning effect with an increased viscosity of 0.5 Pas at the lowest and 0.02 Pas at the highest shear rate. Presumably, the electrostatic repulsion of the carboxylic groups of the gellan gum induce lower interactions between the gellan gum polymeric chains. Therefore, the viscosity reduction takes place immediately after the treatment with shear forces. The applied shear forces at the beginning are sufficient to overcome the polymer chain interactions. It was not possible to obtain uniform fibers with pure gellan gum, probably because of the low extent of entanglement during the spinning process and its anionic nature [31]. For a blended pullulan–gellan gum solution, an increase of the viscosity in comparison to the pure pullulan solution is observed and a significant change of the curve progression with a higher shear thinning effect is found. The viscosity of 20% pullulan is 3.5 Pas and of 0.225% gellan gum 20% pullulan is 5.1 Pas at 1 s-1, thus, the viscosity is influenced by the addition of rather small amounts of gellan gum. Initiated by the polymer blend the plateau disappeared at the beginning, instead a shear thinning process appears immediately after the start of the measurement. During the spinning process of pure gellan gum, droplets detached from the Taylor cone at the needle tip, in a process induced by the electrostatic repulsion of the carboxylic groups [38]. The addition of 10% and 12.5% pullulan to the gellan gum solution causes an increase in the number of chain entanglements. The extent of entanglements correlates with the ability of fiber formation [39]. The interaction and overlapping of the molecular chains prevent the polymer stream from breaking into droplets. Beaded fibers were obtained as the amount of interaction was higher, but was still not sufficient to prevent the breakage of the polymer stream [39]. During the electrospinning process, the polymer solution is processed in regions with different shear rates. At first the polymer solution has to be pumped across the tube into the needle. After reaching the needle tip, the Taylor cone formation takes place. Thereby, high shear rates of 100-1000 s-1 affect the polymer solution, so that the solution can be stretched into fibers [31,40].
Figure 4. Viscosity η dependence of pullulan solutions in water at different concentrations at various shear rates at 20 °C.
In Figure 4, the viscosity of different pullulan solutions at various shear rates is presented. From 0% to 10% of pullulan, no significant differences in the progression of the curve at various shear rates are obtained. At pullulan concentrations below 10%, the extent of shear thinning is very low and the viscosity is independent from the applied shear rate. At higher pullulan concentrations, the shear thinning becomes more important. For concentrations between 10% and 15%, the viscosity (1000 s-1) passes a critical value, so that the polymer concentration allows fiber formation. The critical viscosity for fiber formation is between 0.1 Pas and 0.27 Pas at 1000 s-1. 2.3. Rheological measurements of the pullulan-gellan gum fibers The first step in rheological measurement was the determination of the linear viscoelastic region (LVR) by an amplitude sweep. The storage modulus G’, loss modulus G’’ and complex modulus G* did not change significantly during the deformation process below 1% deformation stress for all samples in the presence and absence of cations (Fig. S2Error! Reference source not found.). Deformation higher than 1% resulted in a significant change of G` and G* for the pullulan-gellan gum fibers gelled by STF. Hence, we concluded that a deformation of 0.4% was appropriate for the following frequency sweeps. The data of the frequency sweep are shown in Fig. S3Error! Reference source not found.. G`, G`` and G* increased by increasing the frequency. The elastic modulus G` of pure pullulan and pullulan-gellan gum fibers gelled with double distilled water is much lower in comparison with the G` of pullulan-gellan gum fibers gelled with STF (Fig. S3-A). This suggests that the elastic amount of the blended fibers is much higher. The loss angle δ of the pure pullulan fibers showed a constant decrease by increasing frequency from 70° to 40° (Fig. S3-D). Low values of the loss angle (20°-30°), which indicate gel formation [41] were observed for pullulan-gellan gum mixtures in STF media, but not in distilled water. Gel formation in gellan gum solutions prepared from gellan bulk material occurs only after heating up to 80 °C. In contrast, no heating step is necessary for the electrospun material, which immediately dissolves and gels after contact with STF. Figure 5 depicts the complex viscosity η* of the measured samples. All samples, except the pullulan-gellan gum fibers with STF showed the same curve progression. Gellan gum
blended fibers started at η* of 1939 Pas, η* decreased after increasing the frequency. The pullulan-gellan gum gel showed a high extent of shear thinning, all other samples showed a much lower effect. Shear thinning behavior is desired to avoid irritancy by high mechanical resistance during blinking of the eye.
Figure 5. Frequency dependence of complex viscosity η* of pure pullulan and pullulan-gellan gum blended fibers in double distilled water or simulated tear fluid at 34 °C.
2.4. Curvature formation For a prolonged residence time and high therapeutic efficacy, it is important to improve the contact between the drug delivery system and the cornea surface. A curved geometry of the fibers is required to achieve an optimal distribution of the drug delivery system on the ocular surface. As a guidance, the dimensions of porcine corneas were taken from literature data with a horizontal diameter of 14.9 mm and a vertical diameter of 12.4 mm [42,43]. Therefore, we prepared flat lenses with 1.5 cm diameter after electrospinning (Figure 6C). The lenses were formed by the matrix shown in Figure 6A,B. The electrospun 20% pullulan 0.225% gellan gum fibers shaped into lens shows lens height h of 2.54 mm ± 0.38 mm (Figure 6E-F). The curvature forming process reduced the lens diameter by 1 mm to 1.4 cm. Thereafter, the curvature of the formed lens was determined by CAD software DesignSpark Mechanical 2.0. The software was used to reconstruct the lens dimensions with the measured values and allowed the curvature determination at 10.92 mm. The lens will be used for human application. In contrast to the porcine eyes, the human horizontal cornea curvature is reported to range between 7.10 and 8.75 mm [44]. We are able to adjust the curvature to the anatomical conditions and this allows a very good adaptation of the ocular system, for human and animal use. These results emphasizes the simplicity of curvature formation of the electrospun product without the addition of crosslinking agents, like radicals or photo-initiating molecules for polymerization [45]. Furthermore, the use of water as a biocompatible spinning solvent makes any purification steps unnecessary.
Figure 6.: A) 3D image of the lens forming matrix (side view); B) 3D view of the forming matrix; C) 1.5 cm diameter flat lens image of mesh immediately after preparation; D) Side view of 1.4 cm diameter lens image after curvature formation; E) Microscopic picture of the lens (side view); F) Lens construction, scaled with the measured dimensions.
2.5.
Three-dimensional X-ray microscopy (nano-CT)
Nano-CT phase-contrast imaging enabled the visualization of the fibers distribution in the samples and their orientation in the space without the addition of any chemicals to increase the image contrast. Images of virtual cuts of the pullulan-gellan gum fiber-lens show that the fibers have an anisotropic orientation and do not show any beads (Figure 7).
Figure 7. Virtual cuts of an exemplary sample of fiber composed of 20% pullulan-0.225% gellan gum imaged with nano-CT. In A), the volume after a single cut is shown, in B) the fibers are pseudo-colored in green, and the virtual cut shown in blue illustrates the binarization of the image for the calculations performed further. In C), the entire volume imaged is shown. Scale bars: 10 μm.
The high porosity and pores interconnectivity is important to enable a fast water penetration and dissolution. A three-dimensional reconstruction of the samples and further image processing enabled the estimation of the porosity of the samples to be equal to 87.5 ± 0.5%, and the interconnectivity of pores to be equal to 99%. The porous structure of the fibers is further illustrated in the video of the supplement data, that allows visualizing the inner structure of the sample, the orientation of the fibers, the fibers thickness and the pores interconnected in all directions. The specific surface area was estimated from the 3D nano-CT datasets and is equal to 7.9 µ𝑚2 µ𝑚3
after curvature formation. The average diameter of the fibers is 317 nm and their
thickness distribution follows a Gaussian curve (Fig. 8Error! Reference source not found.).
Figure 8. Fiber diameter distribution calculated from the nano-CT images of the nanofibers (20% pullulan0.225% gellan gum).
In comparison to the averaged diameter measured by ESEM (398 ± 44 nm), the diameter of the fibers estimated after processing the X-ray 3D dataset was slightly lower (317 ± 42 nm). The size distribution images from nano-CT measurements are displayed in Figure 9.
Figure 9. 3D X-Ray images of the PL-GG nanofibers: A) Grayscale, B) Fiber size distribution image with scale bar.
The differences between the average diameter estimated by 3D X-ray imaging and by ESEM result most likely from differences between the 2D (ESEM) and 3D (nano-CT) imaging methods used. Both techniques have the limitation of providing images of only one small fraction of the entire sample. The ESEM only provides a 2D image of the surface of the sample, and the average diameter calculated was an estimation of the diameter of only 100 fibers. The nano-CT is a more time-consuming technique, as it involves a longer imaging procedure (roughly 9 h long for each 20% pullulan-0.225% gellan gum sample) and a dataset post-processing procedure that can take up to a similar amount of time. However, nano-CT provides a 3D dataset that enables the estimation of the average diameter of the fibers in all three directions, and take into account the variability of the diameter along every single fiber.
In that case, nano-CT produces a more statistically relevant quantification of the variable calculated. Nano-CT produces 3D images of the fibers and enables the estimation of a few parameters in a volumetric dataset. It has the limitation of producing a dataset that is small (in the μm range) when compared to the total size of the fiber-lens produced (in the cm range). However, the consistency of the results obtained for fiber diameter and porosity for three different samples with the same composition support the data obtained, showing that phasecontrast nano-CT is a suitable technique for pullulan fiber imaging. 2.6.
Ocular residence time
In Figure 10, the fluorescence images of fluorescein sodium drops and dye loaded fibers are displayed after different spray applications onto the porcine corneas. Time point zero (t0) shows the images immediately after instillation without any spray application. The conventional eye drops (Figure 10, t0 left) showed a spherical distribution in the center of the cornea, while the fibers showed a more homogenous distribution over the whole cornea surface (Figure 10, t0 right). The fibers showed higher fluorescence intensities at the lower ocular sections, due to the vertical experimental setup caused by the gravitational forces impact the ocular distribution. Over time, the gel formed by the fiber showed an area without any fluorescence along the spray intervals (Figure 10, arrow). This area increased in size with further spray applications, though after t8, a fluorescence signal caused by the fibers can still be detected. In contrast, the fluorescence intensity of the fluorescein drops changes considerably after just one spray application. Just few spots of dye outside of the cornea center were detected, whereas the fibers showed further steady cornea contact. With further spray applications, the signal intensity of the drops decreased very rapidly until no signal was detected after t6. Figure 11 shows kinetics of the fluorescence signal. For eye drops, only 15% signal intensity remained after one spray application. In contrast, the fibers were able to preserve 90% of their original signal intensity. After four applications, the eye drop signal decreased below 5%, but the dissolved fibers had still 65% of their initial intensity. After the last application, the signal of the drops decreased to 2.5%; the signal intensity from the dissolved fibers was still high (40%). These data support a prolonged in vitro residence time caused by the higher viscosity of the fiber formed gel in comparison to eye drops. In addition, compared to the eye drops, the fluorescein was much more homogenously distributed on the ocular surface, because of the fibers lens structure.
Figure 10. Overlay of regular (photo) and fluorescence images of porcine eyes treated with fluorescein sodium eye drops (5 µg/ml) (left), and with pullulan-gellan gum nano-fiber lens (0.0001% fluorescein sodium) (right) after different spray applications.
Figure 11. Fluorescence intensity of sodium fluorescein drops (5 µg/ml) and dye loaded pullulan-gellan gum fibers (0.0001%) during different spray application at the cornea surface.
3. Conclusion We developed a new solid in situ gelling system for the treatment of topical ocular diseases. Nanofibers with high porosity and narrow size distribution were produced by electrospinning of pullulan-gellan gum solutions. Nano-CT imaging enabled the noninvasive and detailed 3D visualization of the inner structure of the nanofibers. Rheological experiments showed that the gelling behavior still existed after the spinning process. We improved the application of the fiber-lens in the eye by creating a curvature with a 3D printed matrix. The system could be easily handled and applied. A new 3D printed model for the in vitro determination of the ocular residence time at porcine eyes was developed. After application, a rapid dissolution and a homogeneous distribution on the porcine cornea combined with a prolonged ocular residence time was observed. Therefore, the new solid in situ gelling system is an attractive alternative to existing ocular drug delivery systems. Acknowledgment The authors thank the material support from Nagase (Europe) GmbH, Duesseldorf, Germany, and DFG support (project grants MA1648/12-2 and WE4051/21-1).
Literature [1]
J.L. Mishima, S., Gasset, A., Klyce, S.D., and Baum, Determination of tear volume and tear flow, [Association for Research in Vision and Ophthalmology, etc.], 1966.
[2]
D.E. Everitt, J. Avorn, Systemic Effects of Medications Used To Treat Glaucoma, Ann. Intern. Med. 112 (1990) 120. doi:10.7326/0003-4819-112-2-120.
[3]
T. Kaila, L. Salminen, R. Huuponen, Systemic absorption of topically applied ocular timolol, J. Ocul. Pharmacol. 1 (1985) 79–83. doi:10.1089/jop.1985.1.79.
[4]
Vincent Hon-Kin Li, J.R. Robinson, Solution viscosity effects on the ocular disposition of cromolyn sodium in the albino rabbit, Int. J. Pharm. 53 (1989) 219–225. doi:10.1016/03785173(89)90315-3.
[5]
C.A. Adler, D.M. Maurice, M.E. Paterson, The effect of viscosity of the vehicle on the penetration of fluorescein into the human eye, Exp. Eye Res. 11 (1971) 34–42. doi:10.1016/S0014-4835(71)80062-3.
[6]
S.. Law, K.. Huang, C.. Chiang, Acyclovir-containing liposomes for potential ocular delivery: Corneal penetration and absorption, J. Control. Release. 63 (2000) 135–140. doi:10.1016/S0168-3659(99)00192-3.
[7]
P. Aksungur, M. Demirbilek, E.B. Denkbaş, J. Vandervoort, A. Ludwig, N. Ünlü, Development and characterization of Cyclosporine A loaded nanoparticles for ocular drug delivery: Cellular toxicity, uptake, and kinetic studies, J. Control. Release. (2011). doi:10.1016/j.jconrel.2011.01.010.
[8]
Y. Luo, K.R. Kirker, G.D. Prestwich, Cross-linked hyaluronic acid hydrogel films: new biomaterials for drug delivery, J. Control. Release. 69 (2000) 169–184. doi:10.1016/S01683659(00)00300-X.
[9]
S. Dinslage, M. Diestelhorst, A. Weichselbaum, R. Süverkrüp, Lyophilisates for drug delivery in ophthalmology: pharmacokinetics of fluorescein in the human anterior segment., Br. J. Ophthalmol. 86 (2002) 1114–7.
[10]
L. Mayol, F. Quaglia, A. Borzacchiello, L. Ambrosio, M.I. La Rotonda, A novel poloxamers/hyaluronic acid in situ forming hydrogel for drug delivery: Rheological, mucoadhesive and in vitro release properties, Eur. J. Pharm. Biopharm. 70 (2008) 199–206. doi:10.1016/J.EJPB.2008.04.025.
[11]
H. Ibrahim, C. Bindschaedler, E. Doelker, P. Buri, R. Gurny, Concept and development of ophthalmic pseudo-latexes triggered by pH, Int. J. Pharm. 77 (1991) 211–219. doi:10.1016/0378-5173(91)90319-J.
[12]
A. Rozier, C. Mazuel, J. Grove, B. Plazonnet, Gelrite®: A novel, ion-activated, in-situ gelling polymer for ophthalmic vehicles. Effect on bioavailability of timolol, Int. J. Pharm. 57 (1989) 163–168. doi:10.1016/0378-5173(89)90305-0.
[13]
K. Abduljalil, M. Diestelhorst, O. Doroshyenko, A. Lux, A. Steinfeld, S. Dinslage, R. Süverkrüp, U. Fuhr, Modelling ocular pharmacokinetics of fluorescein administered as lyophilisate or conventional eye drops, Eur. J. Clin. Pharmacol. 64 (2008) 521–529. doi:10.1007/s00228-0070457-3.
[14]
H. Refai, R. Tag, Development and characterization of sponge-like acyclovir ocular minitablets,
Drug Deliv. 18 (2011) 38–45. doi:10.3109/10717544.2010.509364. [15]
R.M. Moosa, Y.E. Choonara, L.C. Du Toit, P. Kumar, L.K. Tomar, C. Tyagi, T.R. Carmichael, V. Pillay, In vitro pharmaceutical characterization and statistical optimization of a novel topically applied instantly-soluble solid eye drop matrix, Pharm. Dev. Technol. 20 (2015) 854–862. doi:10.3109/10837450.2014.930488.
[16]
O. Saher, D.M. Ghorab, N.M. Mursi, Levofloxacin hemihydrate ocular semi-sponges for topical treatment of bacterial conjunctivitis: Formulation and in-vitro/in-vivo characterization, J. Drug Deliv. Sci. Technol. 31 (2016) 22–34. doi:10.1016/j.jddst.2015.11.004.
[17]
G. Di Prima, M. Licciardi, F. Carfì Pavia, A.I. Lo Monte, G. Cavallaro, G. Giammona, Microfibrillar polymeric ocular inserts for triamcinolone acetonide delivery, Int. J. Pharm. 567 (2019) 118459. https://www.sciencedirect.com/science/article/pii/S0378517319304934?via%3Dihub (accessed October 3, 2019).
[18]
Gagandeep, T. Garg, B. Malik, G. Rath, A.K. Goyal, Development and characterization of nanofiber patch for the treatment of glaucoma, Eur. J. Pharm. Sci. 53 (2014) 10–16. doi:10.1016/J.EJPS.2013.11.016.
[19]
J. Sun, Z. Zhou, A novel ocular delivery of brinzolamide based on gellan gum: In vitro and in vivo evaluation, Drug Des. Devel. Ther. 12 (2018) 383–389. doi:10.2147/DDDT.S153405.
[20]
A. Fernández-Ferreiro, M. González Barcia, M. Gil-Martínez, A. Vieites-Prado, I. Lema, B. Argibay, J. Blanco Méndez, M.J. Lamas, F.J. Otero-Espinar, In vitro and in vivo ocular safety and eye surface permanence determination by direct and Magnetic Resonance Imaging of ion-sensitive hydrogels based on gellan gum and kappa-carrageenan, Eur. J. Pharm. Biopharm. 94 (2015) 342–351. doi:10.1016/j.ejpb.2015.06.003.
[21]
I.D. Rupenthal, C.R. Green, R.G. Alany, Comparison of ion-activated in situ gelling systems for ocular drug delivery. Part 1: Physicochemical characterisation and in vitro release, Int. J. Pharm. 411 (2011) 69–77. doi:10.1016/j.ijpharm.2011.03.042.
[22]
I.D. Rupenthal, C.R. Green, R.G. Alany, Comparison of ion-activated in situ gelling systems for ocular drug delivery. Part 2: Precorneal retention and in vivo pharmacodynamic study, Int. J. Pharm. 411 (2011) 78–85. doi:10.1016/j.ijpharm.2011.03.043.
[23]
P.-E. Jansson, B. Lindberg, P.A. Sandford, Structural studies of gellan gum, an extracellular polysaccharide elaborated by Pseudomonas elodea, Carbohydr. Res. 124 (1983) 135–139. doi:10.1016/0008-6215(83)88361-X.
[24]
E.R. Morris, K. Nishinari, M. Rinaudo, Gelation of gellan – A review, Food Hydrocoll. 28 (2012) 373–411. doi:10.1016/J.FOODHYD.2012.01.004.
[25]
H. Grasdalen, O. Smidsrød, Gelation of gellan gum, Carbohydr. Polym. 7 (1987) 371–393. doi:10.1016/0144-8617(87)90004-X.
[26]
H. Moritaka, H. Fukuba, K. Kumeno, N. Nakahama, K. Nishinari, Effect of monovalent and divalent cations on the rheological properties of gellan gels, Food Hydrocoll. 4 (1991) 495– 507. doi:10.1016/S0268-005X(09)80200-2.
[27]
M. Rostami, M. Ghorbani, M. Aman mohammadi, M. Delavar, M. Tabibiazar, S. Ramezani, Development of resveratrol loaded chitosan-gellan nanofiber as a novel gastrointestinal delivery system, Int. J. Biol. Macromol. 135 (2019) 698–705. doi:10.1016/j.ijbiomac.2019.05.187.
[28]
B.J. Catley, A. Ramsay, C. Servis, Observations on the structure of the fungal extracellular polysaccharide, pullulan, Carbohydr. Res. 153 (1986) 79–86. doi:10.1016/S00086215(00)90197-6.
[29]
H. Bender, J. Lehmann, K. Wallenfels, Pullulan, ein extracelluläres Glucan von Pullularia pullulans, Biochim. Biophys. Acta. 36 (1959) 309–316. doi:10.1016/0006-3002(59)90172-6.
[30]
T. Kimoto, T. Shibuya, S.S.-F. and C. Toxicology, undefined 1997, Safety studies of a novel starch, pullulan: chronic toxicity in rats and bacterial mutagenicity, Elsevier. (n.d.).
[31]
A.C. Stijnman, I. Bodnar, R. Hans Tromp, Electrospinning of food-grade polysaccharides, Food Hydrocoll. 25 (2011) 1393–1398. doi:10.1016/J.FOODHYD.2011.01.005.
[32]
M.R. Karim, H.W. Lee, R. Kim, B.C. Ji, J.W. Cho, T.W. Son, W. Oh, J.H. Yeum, Preparation and characterization of electrospun pullulan/montmorillonite nanofiber mats in aqueous solution, Carbohydr. Polym. 78 (2009) 336–342. doi:10.1016/J.CARBPOL.2009.04.024.
[33]
J. Li, S. Tian, Q. Tao, Y. Zhao, R. Gui, F. Yang, L. Zang, Y. Chen, Q. Ping, D. Hou, Montmorillonite/chitosan nanoparticles as a novel controlled-release topical ophthalmic delivery system for the treatment of glaucoma., Int. J. Nanomedicine. 13 (2018) 3975–3987. doi:10.2147/IJN.S162306.
[34]
M. Paulsson, H. Hägerström, K. Edsman, Rheological studies of the gelation of deacetylated gellan gum (Gelrite(®)) in physiological conditions, Eur. J. Pharm. Sci. 9 (1999) 99–105. doi:10.1016/S0928-0987(99)00051-2.
[35]
A. Torre-Muruzabal, L. Daelemans, G. Van Assche, K. De Clerck, H. Rahier, Creation of a nanovascular network by electrospun sacrificial nanofibers for self-healing applications and its effect on the flexural properties of the bulk material, Polym. Test. 54 (2016) 78–83. doi:10.1016/J.POLYMERTESTING.2016.06.026.
[36]
J.M. Deitzel, J. Kleinmeyer, D. Harris, N.C. Beck Tan, The effect of processing variables on the morphology of electrospun nanofibers and textiles, Polymer (Guildf). 42 (2001) 261–272. doi:10.1016/S0032-3861(00)00250-0.
[37]
E.R. Morris, A.N. Cutler, S.B. Ross-Murphy, D.A. Rees, J. Price, Concentration and shear rate dependence of viscosity in random coil polysaccharide solutions, Carbohydr. Polym. 1 (1981) 5–21. doi:10.1016/0144-8617(81)90011-4.
[38]
P. Vashisth, P.A. Pruthi, R.P. Singh, V. Pruthi, Process optimization for fabrication of gellan based electrospun nanofibers, Carbohydr. Polym. 109 (2014) 16–21. doi:10.1016/J.CARBPOL.2014.03.003.
[39]
S.L. Shenoy, W.D. Bates, H.L. Frisch, G.E. Wnek, Role of chain entanglements on fiber formation during electrospinning of polymer solutions: Good solvent, non-specific polymerpolymer interaction limit, Polymer (Guildf). 46 (2005) 3372–3384. doi:10.1016/j.polymer.2005.03.011.
[40]
T. Han, A.L. Yarin, D.H. Reneker, Viscoelastic electrospun jets: Initial stresses and elongational rheometry, Polymer (Guildf). 49 (2008) 1651–1658. doi:10.1016/J.POLYMER.2008.01.035.
[41]
J. Carlfors, K. Edsman, R. Petersson, K. Jörnving, Rheological evaluation of Gelrite® in situ gels for ophthalmic use, Eur. J. Pharm. Sci. 6 (1998) 113–119. doi:10.1016/S0928-0987(97)000742.
[42]
C. Faber, E. Scherfig, J.U. Prause, K.E. Sørensen, Corneal thickness in pigs measured by
ultrasound pachymetry in vivo, Scand. J. Lab. Anim. Sci. (2008). [43]
J. HEICHEL, F. WILHELM, K.S. KUNERT, T. HAMMER, Topographic Findings of the Porcine Cornea, Med. Hypothesis, Discov. Innov. Ophthalmol. 5 (2016) 125.
[44]
P.M. Kiely, G. Smith, L.G. Carney, Meridional variations of corneal shape, Am. J. Optom. Physiol. Opt. (1984). doi:10.1097/00006324-198410000-00001.
[45]
H. Hiratani, A. Fujiwara, Y. Tamiya, Y. Mizutani, C. Alvarez-Lorenzo, Ocular release of timolol from molecularly imprinted soft contact lenses, Biomaterials. 26 (2005) 1293–1298. doi:10.1016/J.BIOMATERIALS.2004.04.030.