Bioinspired Fibrillary Hydrogel with Controlled Swelling Behavior

Subscriber access provided by University of Rhode Island | University Libraries

Article

Bioinspired Fibrillary Hydrogel with Controlled Swelling Behavior: Applicability as an Artificial Vitreous Sruthi Santhanam, Ying-Bo Shui, Jessica J. Struckhoff, Bedia Begum Karakocak, Paul D. Hamilton, George J. Harocopos, and Nathan Ravi ACS Appl. Bio Mater., Just Accepted Manuscript • DOI: 10.1021/acsabm.8b00376 • Publication Date (Web): 29 Nov 2018 Downloaded from http://pubs.acs.org on December 11, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Bio Materials

Bioinspired Fibrillary Hydrogel with Controlled Swelling Behavior: Applicability as an Artificial Vitreous

Sruthi Santhanam a, Ying-Bo Shui b, Jessica Struckhoff b, Bedia Begum Karakocak a, Paul D. Hamilton b, George J. Harocopos b,c, Nathan Ravi a,b,d* a

Energy, Environmental and Chemical Engineering, Washington University in St. Louis, St.

Louis, MO, USA b

Department of Ophthalmology and Visual Sciences, Washington University School of

Medicine, St. Louis, MO, USA c

Department of Pathology and Immunology, Washington University School of Medicine, St.

Louis, MO, USA d

Department of Veterans Affairs, St. Louis Medical Center, St. Louis, MO, USA

*Correspondence to: Nathan Ravi, Department of Ophthalmology & Visual Sciences, Washington University School of Medicine; 517 South Euclid Ave, Campus Box 8096, St. Louis 63110, USA. Tel.: +1 314-747-4458; fax: +1 314 -747-5073; e-mail: [email protected]

1 ACS Paragon Plus Environment

ACS Applied Bio Materials 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 35

Abstract The vitreous humor of the eye is mainly composed of fibrillary collagen and semi-flexible hyaluronic acid (HA). To mimic this macromolecular composition of the vitreous, we previously developed an injectable two-component hydrogel composed of a fibrillary gellan and a semi-flexible polyelectrolyte, poly[methacrylamide-co-(methacrylic acid)], both endowed with thiol cross-linkers. We optimized the hydrogel formulations for optical, physical, mechanical, and transport properties approximating those of the vitreous. Here, we studied 11 hydrogel formulations with varying concentrations of each component, and as expected, we found that they all swelled in physiological solution. The two formulations that most closely matched the vitreous properties were investigated further. Judged against nonsurgical control and silicone oil, a clinically-accepted vitreous replacement, both hydrogel formulations were biocompatible in rabbits for 30 days. Both hydrogels maintained optical clarity, physiological intra-ocular pressure, and intact retinal layers that displayed normal electroretinography. The swelling behavior of the gel led us to postulate that the native vitreous may also exhibit controlled swelling, where ionic HA’s swelling capacity is restricted by fibrillary collagen. In conclusion, the two hydrogels merit further in-vivo evaluation as an artificial vitreous for an extended duration and additionally in mini-pigs for their similarity to human eyes in size. Keywords: hydrogel, vitreous humor, biocompatible, biomimetic, swelling.


Page 3 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


1.

Introduction In the past four decades, vitrectomy, the removal of vitreous humor of the eye, has

become an important treatment option for various ocular diseases, such as diabetic vitreous hemorrhage, retinal tear and detachment, uveitis, and ocular trauma involving the vitreous.1 Vitreous replacements are an essential component of the vitrectomy surgery. Silicone oil,2 heavy silicone oil,3 liquid perfluorocarbons,4 sulfur hexafluoride and other gases,5 and saline buffers6 have been clinically successful as temporary vitreous replacements in treating uncomplicated retinal detachments. Silicone oils are predominantly used for complex or refractory retinal detachments and are clinically accepted as a long-term vitreous replacement. The substitutes perform well; however, they do exhibit adverse effects such as elevation in intra-ocular pressure, retinal toxicity, hydrophobicity, emulsification of oils, and post-operative cataracts.7-10 There is a continuing search for substitutes with improved tolerance and bio-performance for long-term use. Ideal vitreous substitutes would be easily implantable materials that mimic the structure and functions of the native vitreous and be biocompatible with the surrounding tissues. The native vitreous is a hydrogel primarily composed of water ( > 99.9 % vol.) and two biopolymers, collagen – a fibrillary biopolymer that forms a rigid scaffold, and hyaluronic acid (HA) – a relatively flexible polyelectrolyte that stabilizes the collagen network.11 Essential functions of the vitreous include maintaining optical transparency12, acting as a viscoelastic damper that protects surrounding ocular tissues from immediate mechanical impact,13, 14 supporting the growth of the eye,15 and mediating nutrient and oxygen transport.16-18 Baino19, Kleinberg et al.20, and Su et al.21 have extensively reviewed the use of hydrogels, cross-linked water-soluble polymers, as promising vitreous substitutes. Recently, Barth et al.22 investigated the use of a cross-linked sodium hyaluronic acid (HealaflowTM) as a vitreous substitute. The hydrogel was optically clear and biocompatible,



Page 4 of 35

but slightly raised the intra-ocular pressure and had a short-residence time in vivo. Hayashi et al.23 synthesized a hydrogel with polymeric clusters of thiol and maleimide reactive tetraarmed poly(ethylene glycol). Formulations of this hydrogel with low (7 g·L-1) polymer concentrations were observed to gel in situ within 10 minutes, to be biocompatible, and to produce low swelling pressure for retinal tamponade. Though these biocompatible hydrogels achieved varying degrees of vitreous biomimicry, they lacked a fibrillary component, and thus provided only an incomplete understanding of the role of macromolecular structure and functions of the native vitreous. A hydrogel that includes both a fibrillary and a semi-flexible part would be a truer biomimetic, which may be advantageous in three ways: First, they would meet the essential requirements of an ideal vitreous substitute. Second, understanding the interplay of these two components would provide more significant insights into the physiology and pathophysiology of the native vitreous. Third, it would explain the rigidflexible architecture observed in other soft tissues and thus enable us to propose a unifying design principle used by nature. Previously, we examined the macromolecular structure of the native vitreous, and hypothesized that a two-component hydrogel consisting of rigid-rod-like gellan (an analogue of collagen) and a polyelectrolyte poly[methacrylamide-co-(methacrylic acid)] (an analogue of HA), both endowed with thiol groups for in vivo reversible gelation, would be worthy of testing as a biomimetic vitreous substitute.24 Gellan, in the polymeric mixture, instantaneously transitions from a random coil to a helical structure upon cooling to physiological temperature, and physically traps the semi-flexible copolymer network. Within minutes, the thiols in the physically-crosslinked gel oxidize to form mixed disulfide covalent bonds and establish a permanent hydrogel of two dissimilar networks. The phase-transition, mechanical, and osmotic swelling properties of this two-component polymeric system were a function of the degree of thiolation, and the concentration of each component in the mixture.


Page 5 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


As we observed previously,24 for a constant degree of thiolation, the concentration of thiolated gellan governs the sol-gel phase transition properties and the stiffness of the hydrogels, while the concentration of the copolymer influences the swelling capacity. Thus, we evaluated the properties of 11 hydrogel formulations by varying the concentration of each component to match the native vitreous. Two formulations of our hydrogel had optical clarity, transparency, density, and viscoelastic properties that approximated the features of the native vitreous. They expanded in physiological fluids due to Donnan osmotic pressure. Furthermore, the hydrogels were biocompatible in vitro with ocular cells and non-degradable in contact with enzymatic solutions, and consequently, we chose them for in vivo evaluation. In this work, we expanded our investigations on the osmotic swelling behavior of the original 11 formulations of the two-component hydrogels. By varying the concentration of each component, we ascertained the swelling behavior in a poly(vinyl pyrrolidone) solution of 3 kPa osmotic pressure. For a selected formulation, we compared the degree of swelling (DoS) of the two-component hydrogel to its individual components in physiological solutions. By understanding the osmotic swelling behavior of the two-component hydrogel, we sought to gain insights into the swelling characteristics of the native vitreous humor. Furthermore, we evaluated the applicability of the biomaterial as an artificial vitreous in a one-month study in Dutch-Belted rabbits. We removed and replaced the vitreous, in the right eye of the rabbits, with two optimized formulations of our hydrogel. Also, we compared the results of these formulations to silicone oil, and to the non-surgical left eye, which served as controls. 2.

Materials and Methods

2.1.

Materials

Gentamicin, amphotericin, and all reagents for hydrogel synthesis were purchased from Sigma Aldrich Co. (St. Louis, MO) and used as received unless otherwise mentioned.



Page 6 of 35

Methacrylic acid (MAA) was purchased from Sigma-Aldrich and distilled before use. Dulbecco’s Modified Eagles’ Medium/Nutrient Mixture F-12 Ham (DMEM/F-12) was purchased from Thermo Fisher Scientific (Great Island, NY). Dutch-Belted male rabbits were purchased from Myrtle’s Rabbitry/Covance Research Products Inc. (Denver, PA), and silicone oil 5000 (SO) was from Bausch and Lomb Incorporated (Rochester, NY). The drugs and supplies for animal surgery were purchased from the Division of Comparative Medicine Pharmacy, Washington University School of Medicine (St. Louis, MO). 2.2.

Synthesis of Hydrogels

The first component, thiolated gellan (11% mol of thiol), was synthesized by reacting deacylated gellan with cystamine dihydrochloride, using carbodiimide chemistry as described in detail in our previous study.24 The second component, poly[methacrylamide-co(methacrylic acid)-co-N’,N’-bis(methylacryloyl-cystamine)] (CoP), was synthesized by free radical polymerization of 78% (mol) MAM, 20% (mol) MAA, and 2% (mol) BMAC, as described in detail in our previous work.25 The two hydrogel formulations, namely (1) 0.9 g·L-1 thiolated gellan (G) and 12 g·L-1 copolymer thiolated poly[methacrylamide-co-(methacrylic acid)] (CoP), represented as 0.9G_12CoP, and (2) 1.5 g·L-1 G and 10 g·L-1 CoP, represented as 1.5G_10CoP were tested for for in vivo biocompatibility (Supporting Information S-1). The thiolated gellan and the CoP were dissolved separately at 2X the concentration of the final formulations, in nitrogen (N2) bubbled water and 2X PBS respectively. The pH of the solutions was adjusted to 7.5. The solutions were prepared under sterile conditions, N2 bubbled and either used immediately for osmotic swelling studies or stored at 4 °C for 8 hours before rabbit surgery. Equal volumes of both the solutions were separately warmed at 48 °C for 15 minutes and then mixed to obtain the desired composition.


Page 7 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


For the rabbit surgery, 2 mL of the composite polymer solution at 48 °C was drawn up with an 18-gauge needle, and approximately 0.5-0.6 mL was injected through an infusion line into a partially emptied rabbit vitreous cavity. The temperature of the polymer solution decreased to ~ 42 °C as it traveled through the infusion line; however, it remained above the sol-gel transition temperature (~ 40 °C) as a liquid. The polymer solution was injected slowly, within two minutes. 2.3.

Osmotic Swelling Behavior of Hydrogels

The osmotic swelling pressure of polymeric hydrogels was determined by the macroscopic swelling/de-swelling technique.26-28 This technique is based on the principle that the swelling pressure of a hydrogel becomes equal to the known osmotic pressure exerted by the surrounding macromolecular polymer solution at equilibrium. We performed two sets of experiments to determine the swelling/de-swelling of hydrogels. In the first experiment, 17 hydrogels of 11 different formulations, designed previously24 using response surface methodology (RSM),29, 30 were investigated by immersing them in poly(vinyl pyrrolidone) solution (dissolved in 1X PBS) (PVP, 29 kDa),31 at 3 kPa osmotic pressure. In the second experiment, the two optimized hydrogel formulations, 0.9G_12CoP and 1.5G_10CoP (n=5 each), were swelled/de-swelled in poly(ethylene glycol) solutions (dissolved in 1X PBS) (PEG, 20kDa),32 at different osmotic pressures. The polymer volume fraction of the gel at equilibrium with the surrounding solution was measured and normalized to the initial polymer volume fraction, to obtain the volume ratio of polymer injected into the cavity to that of the vitreous removed. In both the experiments, 3 mL of each polymeric solution at 48 °C was injected into the dialysis cassettes with semi-permeable membranes having a molecular weight cut off (MWCO) of 10 kDa, gelled, and swelled in 1X PBS at 37 °C until it reached equilibrium. At equilibrium, there was no more change in the weight of the hydrogel. Each swollen gel was 7 ACS Paragon Plus Environment


Page 8 of 35

then immersed in a macromolecular solution of known osmotic pressure at 25 °C until it reached equilibrium. The dialysis membrane selectively prevented the entry of external polymeric macromolecules into the hydrogel. The degree of swelling/de-swelling of the hydrogel was normalized to the DoS of the initially formed gel, as given in Equation (1):

(

)

(

)

(

)

[Equation (1)]

where Ws is the final swollen weight of the hydrogel in equilibrium with the surrounding solution, Wd is the dry weight of the hydrogel, and Wi is the initial weight of the fully formed hydrogel. Based on Eq. (1), hydrogels that neither swell nor de-swell have a value equivalent to 1. Swollen gels have a value > 1, in contrast to de-swollen gels, which have a value < 1. 2.4.

In vitro Bio-degradation of Hydrogel

We tested the degradation and stability of the two-component hydrogels in the presence of primary porcine retinal pigment epithelial (ppRPE) cells for 30 days. The ppRPE cells, extracted as described in our previous study, were grown on the apical side of microporous collagen-coated Millicell-CM insert membranes (0.4 µM).24 Two mL of the polymer solution was pipetted into each cell of a six-well tissue culture tray, incubated at 37 °C for three days to fully gel, then allowed to swell in DMEM medium (DMEM/F-12 with 10% fetal calf serum, 1X gentamicin, and 0.1X amphotericin) for seven days. The insert membranes containing the ppRPE cells were then placed over the hydrogel and incubated in DMEM medium for 30 days. Every 3-4 days, media from the apical side of the cultured cells on the membrane was transferred to the basolateral side containing the gel, to allow any enzymes on the apical side of the epithelial cells to come into contact with the hydrogel. The goal was to accurately represent the enzymes that the hydrogel may come in contact with inside the eye. The ppRPE cells alone were the controls. The cells were fixed using 4%


Page 9 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


formaldehyde, stained for nucleus and actin fibers using (4', 6-diamidino-2-phenylindole dihydrochloride, DAPI) and phalloidin, and imaged using confocal microscopy.33, 34 2.5.

Animal Preparation and Study Protocol

All studies were performed in accordance with the Association for Research in Vision and Ophthalmology (ARVO) resolution on the use of animals in vision and ophthalmic research. Thirty-two Dutch-Belted male rabbits, 5-6 months of age, were used in this experiment. Eleven rabbits (n=11) were investigated for each formulation of hydrogel, 0.9G_12CoP and 1.5G_10COP, and 10 rabbits with silicone oil 5000 (n=10) (SO) served as the vitreous substitute control. Vitrectomy followed by replacement was performed on the right eye of each animal, while the left eye was the non-surgical eye – referred to as a non-surgical control in this paper. We performed detailed ophthalmic examinations by slit lamp and measured intra-ocular pressure (IOP) before surgery and on days 1, 4, 7, and 30 postoperation. Before surgery and 30 days after, electroretinogram (ERG) tests were performed on a subset of study rabbits (six were injected with SO, six with 0.9G_12CoP, and eight with 1.5G_10CoP hydrogel). Ocular coherence tomography (OCT) and fundus examinations were performed on a subset of rabbits (n=4 for each experimental group) 30 days after surgery. All rabbits were euthanized after the final clinical examinations, and the eyeballs were excised for histologic evaluations. 2.6.

Vitrectomy and Injection of Vitreous Substitute

Dr. Ying-Bo Shui, an experienced vitreoretinal surgeon, performed all surgical procedures. The surgeries were performed under general anesthesia using isoflurane and ketamine. Rabbits were prepared and draped in a standard surgical manner, their eyes were dilated, and a topical antibiotic was instilled. While anesthetized, the rabbits underwent a complete ophthalmic examination by the retinal surgeon.



Page 10 of 35

After establishing baseline ocular conditions, a partial two-port pars plana vitrectomy was performed on the right eye of the rabbit. Approximately 0.5 to 0.6 mL of vitreous from the temporal side of the vitreous cavity was removed using a 23-gauge vitrector (Pro Care Plus Vitrectomy System, Vision Care Devices, Redding, CA), followed by injection of air, which is subsequently replaced with approximately the same volume of the vitreous substitute as the removed vitreous (Figure S1). Due to the relatively large volume of the lens in the rabbit eye, only a partial vitrectomy could be performed, to avoid touching the lens and forming a cataract. The rabbit vitreous is known for its tight adherence to the retina, rendering complete hyaloid removal unlikely. 2.7.

Examinations

2.7.1. Slit Lamp Examinations Slit-lamp examinations were performed before surgery and at 1, 4, 7, and 30 days postoperation. 2.7.2. Intraocular Pressure Measurements Rabbit eyes were anesthetized with topical 0.5% proparacaine hydrochloride (Alcaine, Alcon-Couvreur, Belgium). IOP was measured using a tonometer (Tono-Pen XL, Nicosia, Cyprus) on the days mentioned in the study protocol. 2.7.3. Funduscopic and Ocular Coherence Tomography Rabbits were anesthetized by intramuscular administration of 15-20 mg·kg-1 ketamine with 25 mg·kg-1 xylazine and maintained with IV ketamine (5-10 mg·kg-1) as needed during the procedure. The rabbit fundus was visualized and imaged with an endoscope camera, and OCT was performed on both eyes using Leica’s Envisu R2210 Spectral domain ophthalmic imaging system (Buffalo Grove, Illinois) to obtain the cross-sectional images (B Scan) of the retina. The thickness of the retina was determined by measuring the length of the retinal layer between the inner limiting membrane (ILM) and the retinal pigment epithelial cell layer


Page 11 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(RPE) at 10 different points using ImageJ (Image processing and analysis in Java, version 1.52g, NIH). 2.7.4. Electroretinogram Analysis At least 7 days prior to surgery and 30 days post-surgery, electroretinogram examinations were performed on the study rabbits (n = 6 each for SO and 0.9G_12CoP hydrogel, and n =8 for 1.5G_10CoP hydrogel). Briefly, rabbits were sedated by intramuscular administration of 15–20 mg·kg-1 ketamine with 2–5 mg·kg-1 xylazine and were monitored for constant physiologically appropriate body temperature and vital signs. A human ERG-Jet corneal contact lens electrode (Universo SA, La Chaux-de-Fonds, Switzerland) was placed on each eye with 2.5% hypromellose (“Gonak,” Akorn Inc., Lake Forest, IL). A reference needle electrode (The Electrode Store, Inc., Buckley, WA) was placed under the skin between the ears at the vertex of the skull, and a ground electrode was placed subcutaneously on the left flank. The rabbit was positioned with its head inside the Ganzfeld dome of an LKC E3000 Electro-diagnostic system (LKC Technologies, Gaithersburg, MD), and ERG testing was performed in five steps on both eyes. The rabbit was dark-adapted (DA) for 20 minutes before scotopic flash ERG was performed. The average response from six flash stimuli (0.0101 cd·s·m-2) was obtained and will be referred to as DA 0.01 ERG in the paper. After two additional minutes of dark adaptation, ERG was measured from an average of five 2.53 cd·s·m-2 flash stimuli presented at intervals of 30.3 ms (DA 3 ERG), and finally, standard dark-adapted oscillatory potentials (DA 3 OP) were averaged from responses of three 2.53 cd·s·m-2 flash stimuli. Next, the rabbit was adapted to a background light intensity of 2.30 log cd·m-2 for 10 minutes and then tested using photopic flash ERG (n=5, 2.53 cd·s·m-2 intensity, referred to as LA 3 ERG). The photopic flicker ERG consisted of an average response from 10 flickering stimuli of the same intensity at 30 Hz, at the same background light level. If needed, Yobine® was



Page 12 of 35

administered after ERG to assist in the recovery of the rabbits. In the electroretinography, the a-wave amplitude was measured from the baseline to the trough of the a-wave, and the bwave amplitude was measured from the bottom of the a-wave to the peak of the b-wave. The analyzed data were used to compare the treated right eye retinal responses post-surgery with their respective pre-surgical activity. 2.7.5. Histology Analysis After 30 days, the rabbits were euthanized, and the eyeballs were excised, and then fixed in 4% formalin, pH 7.3, in 0.1 M Sørensen’s phosphate buffer for 1 hour. The specimens were then washed with 1X PBS twice, sectioned at 12 µm on a cryostat, and every tenth slide was stained with hematoxylin and eosin (H&E) according to standard procedures. An experienced pathologist from Seventh Wave Laboratories LLC (St. Louis, MO) conducted the histopathological evaluations. Tissues were scored from 0 (null) to 4 (severe) for five histopathological criteria: (1) Needle track or suture line through the ora ciliaris, (2) Mononuclear cell inflammation in the vitreous, (3) Focal rupture of the posterior lens capsule, (4) Displacement of the nuclei rods and cones into the inner or outer segment of the photoreceptor layer, and (5) Hemorrhage. Furthermore, representative specimens from each of the study cohorts were additionally reviewed by our university’s ophthalmic pathologist (GJH). Bright-field images were captured with a charge-coupled camera attached to a microscope (Olympus BX51, Olympus America, Inc., Melville, NY). The artifacts in the images are due to a delay in the fixation process. 2.8

Statistical Analysis

We statistically analyzed the results of the osmotic swelling behavior of hydrogel and models that fit the response surface using analysis of variance (ANOVA) with 95% confidence limits using Design-Expert®. All other experimental results were reported as the mean ± SE (standard error), where error bars denote the SE. The standard error was the standard


Page 13 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


deviation divided by the square root of the number of repeats for each experimental group. The experimental groups for the rabbit studies were the hydrogel of formulations 0.9G_12CoP and 1.5G_10CoP, while SO and the non-surgical left eye served as controls. The F-test was performed to analyze the variance between the experimental groups and the control, followed by a two-tailed t-test (assuming equal or unequal variances, based on the result from the F-test). The p-value < 0.05 was considered to be statistically significant. 3.

Results

3.1.

Osmotic Swelling Behavior of Hydrogels

3.1.1 Osmotic Swelling of Hydrogels Designed Using Response Surface Methodology (RSM) Thiolated gellan hydrogels exhibited negligible swelling in 1X PBS, with a normalized DoS nearly equivalent to 1, in contrast to the thiolated poly[methacrylamide-co-(methacrylic acid)], which were highly swellable (normalized DoS 2 (Figure 1A). On the other hand, the two-component hydrogels swelled in 1

S had a normalized DoS (normalized DoS 1.

lower than that of the copolymer, CoP, and higher than that of thiolated gellan, G (Figure 1A). This intermediate degree of swelling demonstrates that in a two-component hydrogel, the copolymer’s swelling capacity is restricted by the minimally swellable thiolated gellan fibrils, resulting in a swelling behavior that we refer to as “controlled swelling behavior” in this paper. In our previous work 24, we showed that all the 11 hydrogel formulations design using RSM swelled in the presence of 1X PBS. In other words, the normalized DoS of the hydrogels were > 1. Furthermore, the normalized DoS varied linearly with the concentrations of G and the CoP, thereby fitting a linear surface model using ANOVA analysis (Refer to Supporting Information S-2, Figure S2A). The overall contribution of [CoP] was significant (p < 0.01) for hydrogel swelling in this defined concentration range. 13 ACS Paragon Plus Environment


Page 14 of 35

On the other hand, we found that all the 11 hydrogel formulations de-swelled with normalized DoS < 1 when immersed in PVP solutions at 3 kPa osmotic pressure (Figure S2B). From the ANOVA analysis in the design of experiments, the concentrations of both the thiolated poly[methacrylamide-co-(methacrylic acid)] (p < 0.01) and the fibrillary thiolated gellan (p < 0.05) were significant in controlling the degree of de-swelling of the two-component hydrogels. Furthermore, the experimental data fits a linear surface model, whose empirical equation, predicted using the least square fit estimation, is given by Eq. (2). Normalized DoS = 0.22 + 0.10 × [G] + 0.04 × [CoP]

([CoP] >> [G])

[Equation (2)]

where 0.22 is the intercept for the defined concentration ranges of [G] and [CoP]. We conclude that both components offer compressive resistance against an external load.

Figure 1. The normalized degree of swelling (N.DoS) of hydrogels. (A) N. DoS of the twocomponent hydrogel, 1.5G_15CoP, compared to that of each of its component, thiolated gellan, and CoP, in 1X PBS. (B) Swelling pressure produced by two optimized hydrogel formulations (n = 5 each) for varying ratios of injected polymer solution volume vs. volume of the native vitreous removed from the eye. 14 ACS Paragon Plus Environment

Page 15 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


3.1.2 Osmotic Swelling Pressure of the Two Optimal Hydrogel Formulations In our system, the osmotic swelling pressure, produced by the hydrogel, is the driving mechanism for retinal tamponade. Furthermore, the swelling pressure produced by the hydrogel was observed to vary by tailoring the volumetric ratio of the polymer solution injected into the vitreous cavity to the volume of native vitreous removed. For a constant hydrogel composition, the volumetric ratio is directly related to the final polymer concentration of the hydrogel, which decreases upon swelling in contact with a physiological solution. At a volumetric ratio equivalent to 1, the two hydrogels, 0.9G_12CoP, and 1.5G_10CoP, produced swelling pressures of 11.4 and 12.9 mmHg, respectively. The swelling pressures produced by both the hydrogels are comparable to that of the overall eye pressure (11 – 21 mmHg35), and hence, the swelling pressure is within the acceptable limits for their use as a vitreous substitute. 3.2.

In vitro Degradation of the Hydrogels

Both the hydrogels, 0.9G_12CoP and 1.5G_10CoP, were stable and did not degrade for 30 days in contact with the insert membrane supporting ppRPE cells on its apical side. The ppRPE cells stayed alive and maintained a monolayer of intact cobblestone morphology with central nuclei in co-culture with the hydrogels (without physical contact). The results were comparable to that of the control, which was cells alone (Figure 2). Furthermore, the actin density, which is linked to the overall health of the cells36, was not compromised in the presence of the hydrogels. Thus, there was no apparent toxic effect of the hydrogels on the ppRPE cells.



Page 16 of 35

Figure 2. Degradation study of the hydrogel. Confocal images of the ppRPE cells. The image shows the monolayer of cells stained for nuclei (blue; DAPI) and cell-matrix – actin fibers (red; phalloidin) in (i) Control medium (without hydrogel co-culture) and in co-culture (without physical contact) with (ii) 1.5G_10CoP hydrogel and (iii) 0.9G_12CoP hydrogel. No toxic effect was observed. 3.3.

Rabbit Studies

The in-situ forming hydrogels, 0.9G_12CoP and 1.5G_10CoP, were easily incorporated into standard vitrectomy procedures via injection. No significant inflammation in the anterior segment of the eye was observed in slit-lamp examinations performed on days 1, 7, and 30 after surgery. The conjunctiva was red on day 1 post-operation, but this redness was significantly reduced on day 4, and the conjunctiva was clear and colorless by post-operative day 7. The lenses were clear, and the Y-suture was visible in rabbits treated with the hydrogels. However, six silicone-oil-treated (n=10), three 0.9G_12CoP-treated (n=11), and six 1.5G_10CoP-treated (n=11) rabbits had a slight posterior polar cataract at 30 days postoperation. Specifically, in three of the six 1.5G_10CoP hydrogel-treated rabbits, the cataract was highly localized at the surgical entrance of vitrectomy. Therefore, the posterior polar cataract was more likely due to a local trauma that occurred during the surgery. A global cataract would be expected with vitreous substitute toxicity. The retina and the optic nerve appeared normal. With one exception, funduscopic examinations on a postoperative day 30


Page 17 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


revealed no vitreous opacity, vitreous hemorrhage, membrane formation, or chorioretinal lesions (Figure 3A). In one rabbit treated with 1.5G_10CoP hydrogel, vitreous hemorrhage occurred during surgery and was still present on day 30.



Page 18 of 35

Figure 3. (A) Fundoscopic and (B) Optical coherence tomography (OCT) images of rabbit vitreous cavities and retinas. (1) Non-surgical control, (2) Silicone oil vitreous replacement, (3) 0.9G_12CoP replacement, and (4) 1.5G_10CoP hydrogel replacement. Data represent the mean ± SE. There was no vitreous opacity, membrane formation, chorioretinal lesions, or atrophy in retinal layers. Optical coherence tomography was performed on four rabbits in each group and showed no retinal detachment 30 days post-operation. Column B in Figure 3 shows the OCT retinal right-eye cross-section images of rabbits treated with 0.9G_12CoP, 1.5G_10CoP hydrogels, and SO, as well as an image of non-surgical control (left eye) at 30 days postoperation.

ased on a student’s two-tailed t-test, there was no significant difference in the

retinal thickness of the hydrogel treated eye compared to that of the untreated-control eye and silicone oil control. There was no atrophy in the retinal thickness of either the hydrogeltreated rabbits or the silicone oil-treated rabbits.

Figure 4. Pre-surgical and post-surgical IOP measurements in all experimental groups. Data represent the mean ± SE. The significant differences in IOPs of the experimental groups compared to the controls are represented with *. At 30 days post-operation, no significant difference was observed in IOPs of the hydrogel-treated rabbits (n=10 for 0.9G_12CoP and 18 ACS Paragon Plus Environment

Page 19 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


n=8 for 1.5G_10CoP gel) compared to the controls (n=10 for silicone oil and n=24 for nonsurgical control). Pre-surgical and post-surgical IOP measurements were comparable for all experimental groups (Figure 4). On a postoperative day 1, 0.9G_12CoP hydrogel showed a decrease in IOP, but this difference recovered by day 4. The literature value for an average rabbit IOP is 15.44 ± 2.16 mmHg,35 with a range of 11–21 mmHg during the light phase of the diurnal cycle. All IOP measurements fell within the published rabbit IOP ranges,35 except the immediate post-operative measurements in all experimental groups including the SO control group. Although the immediate post-operative experimental values were slightly higher than the reported values, the values were within acceptable range for that particular time period, and the spike in IOP usually settles within 24 hours,37 which is consistent with our observations. A two-tailed t-test at different time points showed that there was a significant difference in the IOPs of the experimental group compared to the non-surgical control on day 1, 4, and 7 (Figure 4). Although the difference is significant, the values of the experimental groups were not greater than the controls and were within the acceptable range of IOP values.35 Furthermore, on day 30, no significant difference (p > 0.10) in the final IOPs (day 30 post-operation) of the experimental groups compared to the silicone oil and the non-surgical control.



Page 20 of 35

Figure 5. Electroretinogram (ERG) response (n=5) of a rabbit retina elicited by light stimuli of 2.53 cd·s·m-2 intensity under dark-adapted conditions. The light stimuli elicit both rods and cones responses. The responses were obtained before and 30 days post-operation, where rabbit vitreous was partially replaced with a 0.9G_12CoP hydrogel. Data represent the mean amplitudes (n=5) ± SE. Electroretinogram analysis revealed a normal retinal function in the rabbits treated with our hydrogels. The rabbits showed a characteristic ERG wave pattern in response to light stimuli, with an initial corneal-negative deflection (a-wave, due to hyperpolarization of the photoreceptor layer) followed by a positive amplitude peak (b-wave, due to depolarization of the Müller cells) 7 days prior and 30 days post-operation (Figure 5).


Page 21 of 35

Table 1. Amplitudes and implicit times of characteristic waves in response to light stimuli in dark-adapted and light-adapted conditions. DA 3 ERG

DA 0.01 ERG

LA 3 ERG

a-Wave

a-Wave

b-Wave

b-Wave

b-Wave

b-Wave

b-Wave

b-Wave

Amplitude

Implicit

Amplitude

Implicit

Amplitude

Implicit

Amplitude

Implicit

(µV)

time (ms)

(µV)

time (ms)

(µV)

time (ms)

(µV)

time (ms)

171.04

11.75

346.31

76.33

225.01

48.04

161.29

56.50

±3.75

±0.39

±21.63

±7.31

±30.55

±2.75

±8.10

±21.47

179.66

11.08

345.88

73.42

217.64

59.86

157.10

29.58

±10.38

±0.58

±33.44

±7.49

±29.22

±5.75

±8.68

±0.51

160.93

11.08

359.06

70.33

244.78

61.85

143.81

28.50

±8.00

±0.42

±28.48

±7.51

±27.79

±6.12

±15.06

±0.82

164.55

10.08

412.35

60.50

287.97

87.73

181.65

28.25

±9.32

±0.30

±38.04

±8.45

±24.43

±7.14

±11.81

±0.65

145.44

11.19

268.59

59.25

150.40

64.68

129.33

50.19

±11.17

±10.38

±30.54

±9.12

±19.58

±4.01

±14.99

±19.90

131.21

13.50

295.72

69.19

156.46

50.40

128.12

30.44

±4.35

±1.11

±19.82

±10.88

±31.97

±15.67

±18.44

±1.05

0.9G_12CoP

Silicone oil

Pre

1.5G_10CoP

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Post

Pre

Post

Pre

Post

Table 1 summarizes the amplitudes and implicit times (time-to-peak from the onset of flash stimuli) of the characteristic waves in response to light stimuli under scotopic (darkadapted, DA), and photopic (light-adapted, LA) conditions. The data summarized are raw data without any processing such as short time Fourier transformation, continuous wavelet transformation, or discrete wavelet transformation, and hence are likely to have a considerable standard error. The amplitudes for SO treated-rabbits, including the deviations, were comparable to those found by Mackiewicz et al.3 To understand how the amplitude and implicit time of the experimental group compared to that of the controls, we calculated the ratio of the post-operative to the pre-operative ERG parameters (Figure 6). 21 ACS Paragon Plus Environment


Page 22 of 35

Figure 6. The characteristic ratios of post- to pre-operative a-wave and b-wave amplitudes (A, C, D), and corresponding implicit times (B) of the vitreous-substitute-treated (right, R) eye and the control (left, L) eye, when stimulated with light stimuli of 2.526 cd·s·m-2 intensity (A, B), and 0.01 cd·s·m-2 intensity (C) in dark-adapted conditions and photopic flash in light-adapted conditions (D). (A) compares the amplitude between the substitutes, and (B) compares the implicit times between the substitutes. The photoreceptor activity was normal. The ERG responses (amplitudes and implicit times) of the experimental groups (hydrogels) stimulated with a standard mesopic light flash were not significantly different (p > 0.10) from the SO control (Figure 6A, 6B), except for the b-wave amplitude of 0.9G_12CoP (p = 0.04) hydrogel. The b-wave amplitude of the 0.9G_12CoP treated eye is higher than the SO control, and is statistically comparable to the left-eye control, suggesting normal photoreceptor activity. Furthermore, the amplitudes and implicit times of the treated 22 ACS Paragon Plus Environment

Page 23 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(right eye) are comparable to those of the non-surgical control (left eye), except for the awave amplitude (p = 0.04) and the corresponding minimum implicit time (minT, p = 0.03) of the 0.9G_12CoP hydrogel. The increase in the a-wave amplitude and the corresponding decrease in the implicit time (minT) of the 0.9G_12CoP gel are both statistically significant in comparison to the non-surgical control, indicating that the photoreceptors have better functional activity than the control. Therefore, we concluded that rabbits treated with hydrogels exhibited normal rod and cone activity. The exclusive activity of rods stimulated by 0.010 cd·s·m-2 scotopic light flashes in the dark-adapted rabbits was comparable to those of the controls (both SO and non-surgical eye) (Figure 6C). Furthermore, the activity of cones was confirmed to be normal through the photopic flashes and flicker response in lightadapted conditions. The b-wave amplitude ratio was nearly 1 in response to photopic flashes (Figure 6D). Also, there was no absolute difference in flicker responses. Statistically, there was no significant difference between the ERG responses of the experimental groups when stimulated with different light flashes compared to the silicone oil and the non-surgical controls, indicating that the hydrogel-substitutes were biocompatible in the rabbit eyes and exhibited similar retinal activity to the controls.



Page 24 of 35

Figure 7. Bright-field microscopic image (H&E staining, Scale bar = 50µm) of rabbit retinas. (A) Non-surgical control. (B) Silicone oil, (C) 0.9G_12CoP hydrogel, and (D) 1.5G_10CoP hydrogel. B, C, and D are retinas of rabbit eyes with partial vitreous replacements. All retinal layers were intact, without any retinal atrophy. Histopathological examinations revealed that the retinal layers were intact in the rabbits treated with the hydrogels. Figure 7 shows the retinal structure of rabbit eyes treated with 0.9G_12 CoP and 1.5G_10CoP hydrogel, and with SO, at 30 days post-operation. The nuclei of the cells are stained dark-blue, and the cytoplasm is stained red. The integrity of the retinal layers was preserved, and no significant pathological inflammation or deformation, such as retinal edema, epiretinal membranes, atrophic changes of the retinal layers, hemorrhage, or mononuclear cell inflammation was observed in the experimental group compared to the SO group (Figure 8). When qualitatively scored, the pathological changes in the posterior eye of the experimental groups were found to be less than 1 (0 = no change and 4 = severe change). A minimal number of mononuclear inflammatory cells in the vitreous were common. The presence of mononuclear inflammatory cells is perhaps due to focal trauma to the posterior lens capsule, and to the presence of a suture line near the ora ciliaris that may have introduced the foreign material. Both of these sources of inflammation are likely to occur during the surgical procedure, due to the large size of the normal rabbit lens. The H&E-stained sections displayed intact retinal morphology, without any degenerative changes in the retina.


Page 25 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 8. Histopathology scores for rabbit eye with silicone oil (n=9), 0.9G_12CoP (n=8), and 1.5G_10CoP (n=8) hydrogel as vitreous replacements. Scores from 0 to 4 indicate severity, from null to severe. Data represent the mean ± SE. No significant difference was observed between the hydrogels and the controls, except the mononuclear cell inflammation, which was common in the vitreous, but minimal (score 1 . 4.

Discussion

Current substitutes are not based on the physiological properties of the vitreous, and consequently, provide no insight into developing better and more physiologically similar prostheses. An ideal substitute would mimic the properties of the native vitreous and be permanent. Inspired by the macromolecular design and composition of the native vitreous, we previously developed an injectable two-component hydrogel composed of fibrillary gellan (analogous to fibrillary collagen) and polyelectrolyte poly[methacrylamide-co-(methacrylic acid)] (similar to HA), both endowed with thiol groups.24, 38 Eleven statistically designed formulations of this hydrogel, with water contents ranging from 98.6 to 99.5% (vol), gelled instantaneously at low total concentrations of the two polymer components ranging from 5.5 g·L-1 to 17.5 g·L-1. The gels were optically transparent viscoelastic solids, with optical, physical, mechanical and transport characteristics similar to the native vitreous.24,38

Finally, 25

ACS Paragon Plus Environment


Page 26 of 35

three hydrogel formulations, 1.5G_5CoP, 0.9G_12CoP, and 1.5G_10CoP, whose properties most closely matched those of the native vitreous were selected for further testing and found to be biocompatible in vitro with various ocular cell lines.24 In this study, we found the two-component hydrogel exhibited controlled swelling in 1X PBS, where the higher swelling capacity of the copolymer, due to its higher charge density, was restricted by the minimally swellable fibrillary gellan network (Figure 1A). The behavior resulted in a swollen network, whose degree of swelling was initially regulated by varying the concentration of each component (Figure S2A). In addition to free-swelling of the hydrogels in PBS, we determined the hydrogel’s swelling behavior under isotropic compression by immersing them in PVP solutions at 3 kPa osmotic pressures. We assumed that vitreous replacement might be subjected to a maximum compression of 3 kPa.39 In the presence of an external compressive load, the hydrogels de-swelled when their internal swelling pressure was less than the external compressive load. Moreover, the components of the hydrogel exerted a combined compressive resistance against the load, which potentially mimics the physiological functions of the native vitreous, such as dampening ocular motion and protecting the eye against injury. In addition to osmotic swelling behavior, we investigated the biodegradability of the hydrogel. The two-component hydrogels were nondegradable in vitro in contact with the insert membrane supporting ppRPE cells on its apical side. This finding was in agreement with our previous degradation studies of the hydrogel in contact with enzymatic solutions.24 The two hydrogels, 0.9G_12CoP, and 1.5G_10CoP, that closely matched the properties of the native vitreous, were biocompatible in rabbit eyes for 30 days. Rabbits were inexpensive, easy to handle, and very sensitive to toxicity,40 which is ideal for our purpose. Further, NIH recommends the use of the rabbit for the ocular toxicity test. Although rabbits serve as the best model, the high lens to vitreous-chamber volume ratio of rabbits (1.27),


Page 27 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


compared to that of humans (0.25), results in low accessibility and the inadvertent trauma to the lens during standard vitrectomy. Therefore, a partial two-port vitrectomy was performed as an alternative to the standard vitrectomy to avoid surgical complications. In the partial vitrectomy, about 50% of the vitreous was removed from the vitreous base to the core (Figure S1) when compared to the standard procedure. With an exception, we found that the 22 hydrogel-treated rabbits had normal irides and conjunctiva, without corneal opacity and elevation in IOP. Furthermore, the hierarchy of retinal layers was intact, without any disturbance in the functional activity of the photoreceptors. In a most striking observation, iatrogenic retinal tears in two hydrogel-treated rabbits were found to be healed and had not resulted in detachment. Iatrogenic retinal tears in rabbits, without the perfusion of blood inside the vitreous, develop into larger retinal tears or detachments within 3-4 weeks in rabbits.41 In our study, direct visual examination by our vitreoretinal surgeon, OCT, and ERG confirmed that the layers of the retina were intact, with the normal activity of the retinal photoreceptors at 4 weeks post-surgery. We believe the osmotic swelling pressure exerted by the swollen hydrogel network in the vitreous cavity played an important role in retinal reattachment. For confirmation, additional in vivo studies in retinal detachment models are recommended. The role of the vitreous in generating osmotic swelling pressure remains poorly understood. Experimental techniques to predict the swelling pressure lack due to the fragile biopolymer network of the native vitreous. In light of our findings on the behavior of biomimetic vitreous substitutes, we feel it is reasonable to predict the swelling behavior of the native vitreous. We propose that the native vitreous exhibits controlled Donnan swelling behavior, similar to the biomimetic substitute. The native vitreous, due to its hydrophilic composition and ionic HA, swells in the presence of physiological fluids.11, 42 The swelling



Page 28 of 35

force of HA is counter-balanced by the rigid elastic collagen fibers, resulting in a swollen vitreous gel. According to the Flory-Rehner theory,43 the pressure generated by the swelling HA (HA) is the summation of three main components: (1) Osmotic pressure due to the biopolymer-solvent mixing in the vitreous gel (os), (2) Elastic pressure due to deformation of network chains into an elongated state during swelling (el), and (3) ionic pressure induced by non-uniform distribution of mobile ions between the vitreous gel and the external solution (ion) (Eq. (3)).

HA = os, HA + el, HA + ion, HA

[Equation (3)]

The ionic pressure generated due to the Gibbs-Donnan effect is positive and favors swelling. The osmotic pressure for a polyelectrolyte gel-like HA is also positive and depends on the polymer volume fraction and the polymer-solvent interaction parameter. In contrast, the elastic pressure is negative in magnitude and restricts swelling. Due to the dominating positive force that favors swelling, HA has a high swelling capacity. This is congruous with our observations on the ionic thiolated poly[methacrylamide-co-(methacrylic acid)] where its degree of swelling was high (normalized DoS ≈ 2; Figure 1A . On the other hand, collagen forms a minimally-swellable network in physiological fluids.

Lloyd et al. showed that fresh collagen fibrils from ox-hides swell minimally,

retaining their constant dimensions, in solutions (HCl/ NaOH) with pH ranging between 5.5 to 9.0.44 Therefore, in physiological solutions of neutral pH, the net swelling pressure produced by the collagen fibrils are most likely a minimal positive value; a negative swelling pressure would be indicated by shrinkage in the volume of collagen fibrils. This observation is in agreement with our findings on gellan, where the degrees of swelling of thiolated gellan fibrils at neutral pH were minimal (normalized DoS ≈ 1, Figure 1A and not negative.


Page 29 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


The native vitreous comprises of the swelling HA and the rigid minimally-swellable collagen fibrils. The net swelling pressure produced by the vitreous is perhaps given by,

vitreous = HA COL

[Equation (4)]

swell, vitreous = os, COL + os, HA + el, COL+ el, HA + ion, COL + ion, HA

[Equation (5)]

Here, the swelling force is due to the net (os + ion), which is counter-balanced by the restraining/compressive force primarily contributed by the el of collagen fibrils, resulting in a swollen vitreous gel. Furthermore, the swelling pressure (swell) produced by the vitreous gel may play a central role in (1) Stimulating the growth and development of the eye and (2) Securing the retina in position. The vitreous swelling hypothesis is also consistent with observations of other investigators. Halfter et al.45 disrupted the vitreous body and the inner limiting membrane (ILM) of chick embryos with collagenase and observed an enlargement (predominantly axial) in the eye size. They also speculated that the cortical vitreous body and ILM of the retina provided the mechanical strength to withstand the pressure emanating from the core or the central vitreous that is relatively rich in hyaluronic acid, during scleral development.

The pressure within the vitreous cavity may originate from the Donnan

swelling pressure (vitreous), as given in Eq. (4). When collagen is disrupted, the restraining elastic pressure of collagen (el,COL) approaches zero, and swell, vitreous increases due to the high ionic, HA. The ionic HA in the vitreous forms a highly swellable network that increases the weight of the vitreous body and the size of the eye-orbit. Excess or inadequate swelling pressure of the vitreous in infants may alter the development of the eye, resulting in myopia or hyperopia, respectively. However, as the eye reaches adult size, we believe that the vitreous continues to exert a non-zero osmotic Donnan swelling pressure, which is capable of holding the retina against the posterior wall of the eye but may not affect the growth of the eye.

This hypothesis is congruous with the suggestion of Nickerson et al.42 that the



Page 30 of 35

hydrostatic force exerted by the swelling HA approaches equilibrium against the tension from the collagen fibrils. Furthermore, the interplay between a rigid component (collagen) and a semi-flexible component (HA) in vitreous swelling seems universal in biological hydrogels, for example, articular cartilage,46,

47

the nucleus pulposus,26 and the corneal stroma.48

Therefore, by designing biomimetic prosthesis of the native vitreous, we can potentially maximize functional recovery in individuals who suffer from a vitreous-retinal-involving traumatic eye injury and also fundamentally predict the structure-property relationship of the natural vitreous. 5.

Conclusion

Inspired by the native vitreous, we developed a two-component hydrogel that mimics its macromolecular composition, optical transparency, refractive index, density, rheology, elastic diffusion coefficient, and the frictional coefficient. Our non-degradable hydrogels are designed for controlled swelling in contact with a physiological fluid and exert Donnan osmotic swelling pressure that has the potential to re-attach the retina as well as prevent retinal detachment. The hydrogel was well tolerated by the rabbit’s eye and was non-toxic to the retina for at least 30 days. Furthermore, it overcomes the limitations of silicone oil and has comparable in vivo biocompatibility, hence proving to be superior to the clinically accepted current substitute. Due to the biomimetic composition and properties of the native vitreous, the hydrogel provides additional insights into the physiology of the vitreous. Two hydrogel formulations are recommended for further evaluation in a mini-pig model eye for a longer duration. 6.

Supporting Information

Supporting Information is available for the schematic representation of the anatomy of a rabbit eye where the vitrectomy area is labeled. The selection of hydrogel formulations for


Page 31 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


the biocompatibility studies and the osmotic swelling findings for the hydrogels designed using response surface methodology is provided. 7.

Acknowledgments

This research was supported by an NIH grant, EY021620, and a Department of Veterans Affairs rehab merit review grant, RX000657-01 to NR. Also, Research to Prevent Blindness, Inc. and NIH Core Grant P30 EY02687 provided support. We are thankful to Prof. James Ballard and Dr. Rithwick Rajagopal for their feedback and constructive comments to improve the manuscript. Veteran Affairs Office of Research and Development IO1BX007080 grant to Kelle H. Moley supported confocal imaging. 8.

Competing Interests

The authors declare no competing financial interests.



9.

Page 32 of 35

References

Mahdavi, P.; Tornambe, P. E., Pars Plana Vitrectomy. In Operative Dictations in 1. Ophthalmology, Rosenberg, E. D.; Nattis, A. S.; Nattis, R. J., Eds. Springer International Publishing: Cham, 2017; pp 297-299. 2. Giordano, G. G.; Refojo, M. F., Silicone Oils as Vitreous Substitutes. Prog. Polym. Sci. 1998, 23, (3), 509-532. 3. Mackiewicz, J.; M hling, .; Hiebl, .; Meinert, H.; Maaijwee, K.; Kociok, N.; L ke, .; agorski, .; Kirchhof, .; Joussen, A. M., In ivo Retinal Tolerance of arious Heavy Silicone Oils. Invest. Ophthalmol. Visual Sci. 2007, 48, (4), 1873-1883. 4. Chang, S.; Zimmerman, N. J.; Iwamoto, T.; Ortiz, R.; Faris, D., Experimental Vitreous Replacement With Perfluorotributylamine. Am. J. Ophthalmol. 1987, 103, (1), 2937. 5. Norton, E. W., Intraocular Gas in the Management of Selected Retinal Detachments. Trans. - Am. Acad. Ophthalmol. Otolaryngol. 1973, 77, (2), OP85-98. 6. Grafton, E. G.; Guyton, J. S., The Value of Injecting Saline into the Vitreous as an Adjunct to Diathermy Operations for Retinal Detachment. Am. J. Ophthalmol. 1948, 31, (3), 299-303. 7. Nguyen, Q. H.; Lloyd, M. A.; Heuer, D. K.; Baerveldt, G.; Minckler, D. S.; Lean, J. S.; Liggett, P. E., Incidence and Management of Glaucoma after Intravitreal Silicone Oil Injection for Complicated Retinal Detachments. Ophthalmology 1992, 99, (10), 1520-1526. 8. Federman, J. L.; Schubert, H. D., Complications Associated with the Use of Silicone Oil in 150 Eyes after Retina-vitreous Surgery. Ophthalmology 1988, 95, (7), 870-876. Leaver, P. K.; Grey, R. H.; Garner, A., Silicone oil Injection in the Treatment of 9. Massive Preretinal Retraction. II. Late Complications in 93 Eyes. Br. J. Ophthalmol. 1979, 63, (5), 361-367. 10. Riedel, K. G.; Gabel, V.-P.; Neubauer, L.; Kampik, A.; Lund, O.-E., Intravitreal Silicone Oil Injection: Complications and Treatment of 415 Consecutive Patients. Graefe's Arch. Clin. Exp. Ophthalmol. 1990, 228, (1), 19-23. 11. Bishop, P. N., Structural Macromolecules and Supramolecular Organisation of the Vitreous Gel. Prog. Retinal Eye Res. 2000, 19, (3), 323-344. 12. Swann, D. A., Chemistry and Biology of the Vitreous Body. Int. Rev. Exp. Pathol. 1979, 22, 1-64. 13. Lee, B.; Litt, M.; Buchsbaum, G., Rheology of the Vitreous Body. Part I: Viscoelasticity of Human Vitreous. Biorheology 1992, 29, (5-6), 521-533. 14. Weber, H.; Landwehr, G.; Kilp, H.; Neubauer, H., The Mechanical Properties of the Vitreous of Pig and Human Donor Eyes. Ophthalmic Res. 1982, 14, (5), 335-343. 15. Sebag, J., Age-related Changes in Human Vitreous Structure. Graefe's Arch. Clin. Exp. Ophthalmol. 1987, 225, (2), 89-93. 16. Shui, Y.-B.; Holekamp, N. M.; Kramer, B. C.; Crowley, J. R.; Wilkins, M. A.; Chu, F.; Malone, P. E.; Mangers, S. J.; Hou, J. H.; Siegfried, C. J., The Gel State of the Vitreous and Ascorbate-dependent Oxygen Consumption: Relationship to the Etiology of Nuclear Cataracts. Arch. Ophthalmol. 2009, 127, (4), 475-482. 17. Holekamp, N. M., The Vitreous Gel: More than Meets the Eye. Am. J. Ophthalmol. 2010, 149, (1), 32-36. 18. Reddy, V. N., Dynamics of Transport Systems in the Eye. Friedenwald Lecture. Invest. Ophthalmol. Visual Sci. 1979, 18, (10), 1000-1018. 19. Baino, F., Towards an Ideal Biomaterial for Vitreous Replacement: Historical Overview and Future Trends. Acta Biomaterialia 2011, 7, (3), 921-935.


Page 33 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


20. Kleinberg, T. T.; Tzekov, R. T.; Stein, L.; Ravi, N.; Kaushal, S., Vitreous Substitutes: A Comprehensive Review. Surv. Ophthalmol. 2011, 56, (4), 300-323. 21. Su, X.; Tan, M. J.; Li, Z.; Wong, M.; Rajamani, L.; Lingam, G.; Loh, X. J., Recent Progress in Using Biomaterials as Vitreous Substitutes. Biomacromolecules 2015, 16, (10), 3093-3102. 22. Barth, H.; Crafoord, S.; Andréasson, S.; Ghosh, F., A Cross-linked Hyaluronic Acid Hydrogel (Healaflow®) as a Novel Vitreous Substitute. Graefe's Arch. Clin. Exp. Ophthalmol. 2016, 254, (4), 697-703. 23. Hayashi, K.; Okamoto, F.; Hoshi, S.; Katashima, T.; Zujur, D. C.; Li, X.; Shibayama, M.; Gilbert, E. P.; Chung, U.-i.; Ohba, S.; Oshika, T.; Sakai, T., Fast-forming Hydrogel with Ultralow Polymeric Content as an Artificial Vitreous Body. Nature Biomedical Engineering 2017, 1, 0044. 24. Santhanam, S.; Liang, J.; Struckhoff, J.; Hamilton, P. D.; Ravi, N., Biomimetic Hydrogel with Tunable Mechanical Properties for Vitreous Substitutes. Acta Biomaterialia 2016, 43, 327-337. 25. Liang, J.; Struckhoff, J.; Du, H.; Hamilton, P.; Ravi, N., Synthesis and Characterization of In Situ Forming Anionic Hydrogel as Vitreous Substitute. J. Biomed. Mater. Res., Part B 2016, 105, (5), 977-988. 26. Urban, J. P.; McMullin, J. F., Swelling Pressure of the Intervertebral Disc: Influence of Proteoglycan and Collagen Contents. Biorheology 1985, 22, (2), 145-157. 27. Shi, Y.; Xiong, D.; Liu, Y.; Wang, N.; Zhao, X., Swelling, Mechanical and Friction Properties of PVA/PVP Hydrogels after Swelling in Osmotic Pressure Solution. Mater. Sci. Eng., C 2016, 65, 172-180. Horkay, F.; Magda, J.; Alcoutlabi, M.; Atzet, S.; Zarembinski, T., Structural, 28. Mechanical and Osmotic Properties of Injectable Hyaluronan-based Composite Hydrogels. Polymer (Guildf) 2010, 51, (19), 4424-4430. 29. Karakocak, B. B.; Liang, J.; Kavadiya, S.; Berezin, M. Y.; Biswas, P.; Ravi, N., Optimizing the Synthesis of Red-emissive Nitrogen-doped Carbon Dots for Use in Bioimaging. ACS Appl. Nano Mater. 2018, 1, (7), 3682-3692. 30. Santhanam, S.; Liang, J.; Baid, R.; Ravi, N., Investigating Thiol-modification on Hyaluronan via Carbodiimide Chemistry using Response Surface Methodology. J. Biomed. Mater. Res., Part A 2015, 103, (7), 2300-2308. 31. Vink, H., Precision Measurements of Osmotic Pressure in Concentrated Polymer Solutions. Eur. Polym. J. 1971, 7, (10), 1411-1419. 32. Chahine, N. O.; Chen, F. H.; Hung, C. T.; Ateshian, G. A., Direct Measurement of Osmotic Pressure of Glycosaminoglycan Solutions by Membrane Osmometry at Room Temperature. Biophys. J. 2005, 89, (3), 1543-1550. 33. Xia, L. Y.; Zhang, X. D.; Cao, M.; Chen, Z.; Wu, F. G., Enhanced Fluorescence Emission and Singlet Oxygen Generation of Photosensitizers Embedded in Injectable Hydrogels for Imaging Guided Photodynamic Cancer Therapy. Biomacromolecules 2017, 18, (10), 3073-3081. 34. Karakocak, B. B.; Liang, J.; Biswas, P.; Ravi, N., Hyaluronate Coating Enhances the Delivery and Biocompatibility of Gold Nanoparticles. Carbohydr. Polym. 2018, 186, 243251. 35. Pereira, F. Q.; Bercht, B. S.; Soares, M. G.; da Mota, M. G. B.; Pigatto, J. A. T., Comparison of a Rebound and an Applanation Tonometer for Measuring Intraocular Pressure in Normal Rabbits. Vet. Ophthalmol. 2011, 14, (5), 321-326. 36. Tezel, G.; Seigel, G. M.; Wax, M. B., Density-dependent Resistance to Apoptosis in Retinal Cells. Curr. Eye Res. 1999, 19, (5), 377-388.



Page 34 of 35

37. Desai, U. R.; Alhalel, A. A.; Schiffman, R. M.; Campen, T. J.; Sundar, G.; Muhich, A., Intraocular Pressure Elevation after Simple Pars Plana Vitrectomy. Ophthalmology 1997, 104, (5), 781-786. 38. Morozova, S.; Hamilton, P.; Ravi, N.; Muthukumar, M., Development of a Vitreous Substitute: Incorporating Charges and Fibrous Structures in Synthetic Hydrogel Materials. Macromolecules 2016, 49, (12), 4619-4626. 39. Karimi, A.; Razaghi, R.; Navidbakhsh, M.; Sera, T.; Kudo, S., Quantifying the Injury of the Human Eye Components due to Tennis Ball Impact using a Computational Fluid– Structure Interaction Model. Sports Engineering 2016, 19, (2), 105-115. 40. Gad, S. C., The Rabbit. In Animal Models in Toxicology, CRC Press: Boca Raton, 2016; p 1152. 41. Cleary, P. E.; Ryan, S. J., Experimental Posterior Penetrating Eye Injury in the Rabbit. I. Method of Production and Natural History. Br. J. Ophthalmol. 1979, 63, (5), 306. 42. Nickerson, C. S.; Park, J.; Kornfield, J. A.; Karageozian, H., Rheological Properties of the Vitreous and the Role of Hyaluronic Acid. Journal of Biomechanics 2008, 41, (9), 1840-1846. 43. Flory, P. J.; Rehner, J., Statistical Mechanics of ross‐Linked olymer Networks II. Swelling. J. Chem. Phys. 1943, 11, (11), 521-526. 44. Lloyd, D. J.; Marriott, R. H.; Pleass, W. B., The Swelling of Protein Fibres. Part I. The Swelling of Collagen. Trans. Faraday Soc. 1933, 29, (140), 554-563. 45. Halfter, W.; Winzen, U.; Bishop, P. N.; Eller, A., Regulation of Eye Size by the Retinal Basement Membrane and Vitreous Body. Invest. Ophthalmol. Visual Sci. 2006, 47, (8), 3586-3594. Maroudas, A., Balance between Swelling Pressure and Collagen Tension in Normal 46. and Degenerate Cartilage. Nature 1976, 260, (5554), 808-809. 47. Horkay, F.; Basser, P. J., Osmotic Properties of Cartilage. In Biophysics and Biochemistry of Cartilage by NMR and MRI, 2016; 44-61. 48. Hart, R. W.; Farrell, R. A., Structural Theory of the Swelling Pressure of Corneal Stroma in Saline. Bull. Math. Biophys. 1971, 33, (2), 165-186.


Page 35 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Graphical Abstract


Bioinspired Fibrillary Hydrogel with Controlled Swelling Behavior

Recommend Documents