Drug Delivery 2022, VOL. 29, NO. 1, 2357–2374 https://doi.org/10.1080/10717544.2022.2096712 RESEARCH ARTICLE Mimicking chronic glaucoma over 6 months with a single intracameral injection of dexamethasone/fibronectin-loaded PLGA microspheres Alba Aragón-Navasa,b,c , María J. Rodrigod,e , David Garcia-Herranza,b,c , Teresa Martinezd, Manuel Subiasd , Silvia Mendezd , Jesús Rubertef,g,h , Judit Pampalonaf,g,h , Irene Bravo-Osunaa,b,c,e , Julian Garcia-Feijooe,i, Luis E. Pablod,e , Elena Garcia-Martind,e and Rocío Herrero-Vanrella,b,c,e a Complutense University, Innovation, Therapy and Pharmaceutical Development in Ophthalmology (InnOftal) Research Group, UCM 920415, Department of Pharmaceutics and Food Technology, Faculty of Pharmacy, Complutense University of Madrid Spain, Health Research Institute of the San Carlos Clinical Hospital (IdISSC), Madrid, Spain; bDepartment of Pharmaceutics and Food Technology, Faculty of Pharmacy, Complutense University of Madrid, Madrid, Spain; cResearch Institute of the San Carlos Clinical Hospital (IdISSC), Grupo de Investigación Innovación Farmacéutica en Oftalmología, Madrid, Spain; dInstituto de Investigación Sanitaria de Aragón, Hospital Universitario Miguel Servet, Universidad de Zaragoza, Zaragoza, Spain; eNational Ocular Pathology Network (OFTARED) Carlos III Health Institute, Madrid, Spain; fCenter of Animal Biotechnology and Gene Therapy (CBATEG), Universitat Autònoma de Barcelona, Bellaterra, Spain; gCIBER de Diabetes y Enfermedades Metabólicas Asociadas (CIBERDEM), Madrid, Spain; hDepartment of Animal Health and Anatomy, School of Veterinary Medicine, Universitat Autònoma de Barcelona, Bellaterra, Spain; iDepartment of Ophthalmology, San Carlos Clinical Hospital, Health Research Institute of the San Carlos Clinical Hospital (IdISSC), Madrid, Spain ABSTRACT ARTICLE HISTORY To create a chronic glaucoma animal model by a single intracameral injection of biodegradable Received 20 April 2022 poly lactic-co-glycolic acid (PLGA) microspheres (Ms) co-loaded with dexamethasone and fibronectin Revised 8 June 2022 (MsDexaFibro). MsDexaFibro were prepared by a water-in-oil-in-water emulsion method including Accepted 14 June 2022 dexamethasone in the organic phase and fibronectin in the inner aqueous phase. To create the KEYWORDS chronic glaucoma model, an interventionist and longitudinal animal study was performed using Glaucoma animal model; PLGA forty-five Long Evans rats (4-week-old). Rats received a single intracameral injection of MsDexafibro microspheres; multiloaded; suspension (10%w/v) in the right eye. Ophthalmological parameters such as clinical signs, intraocular intraocular pressure; pressure (IOP), neuro-retinal functionality by electroretinography (ERG), retinal structural analysis neurodegeneration by optical coherence tomography (OCT), and histology were evaluated up to six months. According to the results obtained, the model proposed was able to induce IOP increasing in both eyes over the study, higher in the injected eyes up to 6 weeks (p < 0.05), while preserving the ocular surface. OCT quantified progressive neuro-retinal degeneration (mainly in the retinal nerve fiber layer) in both eyes but higher in the injected eye. Ganglion cell functionality decreased in injected eyes, thus smaller amplitudes in PhNR were detected by ERG. In conclusion, a new chronic glaucoma animal model was created by a single injection of MsDexaFibro very similar to open-angle glaucoma occurring in humans. This model would impact in different fields such as ophthalmology, allowing long period of study of this pathology; pharmacology, evaluating the neuroprotective activity of active compounds; and pharmaceutical technology, allowing the correct evaluation of the efficacy of long-term sustained ocular drug delivery systems. Introduction pressure (IOP) is the major modifiable risk factor strongly associated with the onset and progression of the disease Glaucoma is a multifactorial neurodegenerative disorder, lead- (Gaasterland et al., 2000). For this reason, the usual treatment ing cause of blindness with an estimated 60 million people worldwide currently suffering and it will be around 118 mil- of glaucoma consists in the frequent instillation of hypoten- lions in 2040 (Quigley & Broman, 2006; Tham et al., 2014). sive drugs (Conlon et al., 2017); however, it is important to It is produced by the gradual death of retinal ganglion cells remark that the return to normal intraocular pressure values (RGC) which leads to irreversible vision loss, although recent does not always stop the retinal degeneration (Pascale et al., studies have also suggested the involvement of other retinal 2012) and that, in addition, in many other cases these same cells as photoreceptors or activated glia (Asaoka et al., 2017; failures in the retina and optic nerve occur in normotensive Ramirez et al., 2017; Ashimatey et al., 2018). Intraocular patients (Anderson, 2003; Lestak et al., 2018). Sadly, there is CONTACT Rocío Herrero-Vanrell rociohv@ucm.es #Alba Aragón-Navas and María J. Rodrigo are equal contribution (first author). Supplemental data for this article is available online at https://doi.org/10.1080/10717544.2022.2096712 © 2022 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. 2358 A. ARAGÓN-NAVAS ET AL. not yet any optimal treatment for this and any other retinal the anterior chamber (Urcola et al., 2006), through the use neurodegenerative diseases. of corticoids which reduce aqueous humor outflow when Degenerative diseases of the retina begin with primary inhibiting cellular phagocytosis thus avoiding cleaning the damage, for example of the axons of retinal ganglion cells waste channels in the TM (Clark & Wordinger, 2009; Patel in the case of glaucoma. When the damaged cell dies, it et al., 2017), or by fibronectin expression in a genetically releases neurotoxic compounds in the vicinity that damage modified mouse model altering the extracellular matrix of the adjacent neurons that also end up dying in what is the TM (Roberts et al., 2020). All the previous need repeated known as “secondary neurodegeneration” (Ritch, 2000). interventions on the animal, damage aggressively or even Several decades ago, some researchers postulated the idea modify their nature. of “neuroprotection” as a complementary treatment in degen- Thus far, no research group has achieved the creation of erative diseases (Weinreb & Levin, 1999). At the level of the an animal model which simulates open-angle glaucoma com- retina neuroprotection can be defined as: “Prevention or bining the last three different strategies. With this idea in slowing down of the loss of functional integrity of the cells mind, we recently proposed an initial animal model of glau- of the retina, their axons and their axonal connections to coma by using intracameral injections of non-loaded PLGA maintain and stabilize the vision of patients as much as microspheres. The weekly injection of blank microspheres possible” (Ritch, 2000). Pardue et al. describe this neurode- produced a sustained and prolonged ocular hypertension generation of the retina in 3 stages: an adaptive first step (OHT) over six months due to a progressive physical blockage where there is only oxidative stress and neuronal dysfunction, of the trabecular meshwork. This maintained OHT produced a second stage, the “Early pathology”, where retinal structures a very progressive degeneration of the retina (Garcia-Herranz begin to damage but still reversibly, and a third stage of et al., 2021; Rodrigo et al., 2021). In a second study, dexa- "Late pathology" where blindness appears and where the methasone loaded PLGA microspheres also generated OHT damage is already irreversible. According to those authors, and neuroretinal degeneration by only two times injections “starting neuroprotective treatments at the first signs of the (Rodrigo et al., 2021). After those initial approaches, in the retinal disease would provide the most benefit in preserving present work, we propose a new model of chronic glaucoma vision” (Pardue & Allen, 2018). produced by intracameral injection of dexamethasone/ In recent decades, different neuroprotective pharmacolog- fibronectin co-loaded PLGA microspheres. This study aims to ical strategies have been explored (Thanos & Emerich, 2005; establish whether lower quantities of sustained and simul- Masuda et al., 2017; Avotri et al., 2019; Fernández-Albarral taneous co-release of those two active compounds from et al., 2019; Stankowska et al., 2019; Naik et al., 2020). PLGA microspheres in the anterior segment of the eye, com- However, one of the main problems to evaluate the efficacy bined with the mechanical blockage produced by PLGA of neuroprotective treatments resides in the absence of a microspheres themselves, would promote a progressive OHT chronic and slow degeneration glaucoma animal model that and subsequent retinal chronic degradation, only using a simulates the conditions found in the human glaucomatous single injection. The follow up of animals in terms of retinal eyes (Pang & Clark, 2020). In general, it is very difficult to structure and function is performed for six months. obtain optimal animal models of retinal pathologies. For example, in the case of glaucoma, there are the so-called “acute models”, in which a fast retinal degeneration occurs, Materials either due to genetic failures (Schlamp et al., 2006), to the Dexamethasone (DX) was supplied by Sigma-Aldrich (St. Louis direct intravitreal injection of toxic compounds (Kitaoka et al., Mo., USA) (purity >98%). Poly(D,L-lactide-co-glycolide) (PLGA) 2006), or to the direct damage of the optic nerve for example 50:50 (Inherent viscosity: 0.16–0.24 dL/g) was purchased from by crushing it (Sharma et al., 2014). Despite their utility for Evonik Industries (Essen, Germany). Polyvinyl alcohol 67,000 g/ neuroprotective studies, these acute models: (1) Do not mol (PVA) was obtained from Merck KGaA (Darmstadt, reproduce the reality of retinal chronic pathologies in human Germany) and methylene chloride from PanReac AppliChem beings; (2) Do not allow testing the long term efficacy of (Barcelona, Spain). Fibronectin (FN), fibronectin ELISA and new therapies because the retina is almost directly in the reactants (Reagent Diluent DY995, Wash Buffer WA126, Stop last step of the diseases from the very beginning of the Solution DY994 and Substrate Reagent Pack DY999) were study; and (3) It is difficult to assess the potential of modified purchased from R&D Systems (Minneapolis, MN, USA). release systems so in a few days or weeks the retina is com- pletely ruined (Cuenca et al., 2014; Smedowski et al., 2014; Pang & Clark, 2020). There are also the so-called "chronic Manufacture of dexamethasone/fibronectin-loaded models”, which present a more progressive retinal degrada- PLGA microspheres tion based on the increase in IOP secondary to a limitation of aqueous humor outflow. In rodent models, the reduction The water-in-oil-in-water (W/O/W) double emulsion solvent in aqueous humor drainage is typically obtained, either followed by extraction-evaporation technique was employed through cauterization, ligature and/or sclerosis of episcleral for dexamethasone/fibronectin-loaded PLGA microspheres veins (Morrison et al., 1997; Morgan & Tribble, 2015; Dey elaboration. In brief, micronized DX (40 mg) was added to a et al., 2018) either by mechanical blockage of the trabecular polymeric organic solution (400 mg of PLGA dissolved in 2 mL meshwork (TM) with non-biodegradable particles injected in of methylene chloride). The suspension was homogenized Drug Delivery 2359 by ultrasonication in an ice-water bath (Ultrasons; J.P. Selecta, surface of dexamethasone/fibronectin-loaded PLGA micro- Barcelona, Spain) for 5 minutes and sonication (Sonicator XL; spheres after gold sputter-coating. Heat Systems, Inc., Farmingdale, NY, USA) for 1 additional Transmission electron microscopy (TEM, Jeol 1010, Tokyo, minute in an ice-water bath. The inner aqueous phase com- Japan) was used for the assessment of the internal structure. posed of 20 µL of FN aqueous solution (containing 42.8 µg Ms were embedded in a synthetic resin medium (Spurr Low of fibronectin) was added to this organic phase and emul- Viscosity Embedding Kit) and cut in slides (50–70 nm) by sified by sonication (Sonicator XL; Heat Systems, Inc., Reichert Ultracut S ultramicrotome (Leica Microsystems Inc, Farmingdale, NY, USA) for 30 seconds at 4 °C to create the Wetzlar, Germany). initial W1/O emulsion. Afterwards, 5 mL of PVA solution 1% (w/v) were added to the mentioned W1/O emulsion and the Mean particle size and particle size distribution mixture was emulsified (Polytron®RECO, Kinematica, GmbHT Dual light scattering (Microtrac® S3500 Series Particle Size PT3000, Lucerna, Switzerland) at 7,000 rpm for 1 minute to Analyzer, Montgomeryville, PA, USA) was the technique form the final W1/O/W2 emulsion that was finally added to employed to analyze particle size and particle size distribu- 100 mL of PVA solution 0.1% (w/v) to get droplets hardening tion. The volume mean diameters (± standard deviation) by organic solvent evaporation under stirring at room tem- obtained from 3 measurements were calculated and expressed perature for 3 hours. After that, Ms were washed with MilliQ® as mean particle size. water to remove surface PVA. The desired granulometric size fraction (20–10 µm) was collected by sieving, freeze-dried (Freezing: −60 °C/15 min, drying: −60 °C/12h/0.1 mBar) and Dexamethasone quantification by HPLC/MS storage at −30 °C in dry conditions (Figure 1). Dexamethasone was quantified using a HPLC/MS system con- sisting of a liquid chromatography instrument (Waters 1525 binary HPLC pump and Waters 2707 autosampler) connected Dexamethasone/fibronectin-loaded PLGA microspheres to a MS detector (Waters 3100 single quadrupole mass spec- characterization trometer). A guard column (4 µm, 3.9 mm x 20 mm) and a Production yield percentage (PY %) Nova-Pak C18 column (4 µm, ID 2.1 mm x 150 mm) were used The production yield percentage of the chosen granulometric in-line at 45 °C in the HPLC instrument. The MS detector was fraction was calculated according to the following Eq. (1): connected to the system via Empower 2 (Waters, Milford, USA). For DX detection, the positive ion mode was chosen Weight of Ms W1  for the ESI source. Selected ion recording (SIR) DX mass (m/z) PY %  x 100 (1) 393.40 was obtained under mass spectrometer source con- weight of polymer W2   weight of DX W3   ditions of 3.5 kV electrospray voltage. The conditions for neb- weight of FN W4  ulization were 150 L/h flow rate, 120 °C source temperature, 3 V extractor voltage and for desolvation were 500 L/h flow rate, 350 °C desolvation temperature. Nitrogen gas (> 99.999%) Morphological evaluation was employed in nebulization and desolvation. For encap- Scanning electron microscopy (SEM, Jeol, JSM-6335F, Tokyo, sulation efficiency and release from Ms, an isocratic method Japan) was employed for the observation of the external was developed which was composed of 50% ammonium Figure 1. Elaboration scheme of dexamethasone-fibronectin co-loaded PLGA microspheres. Created with BioRender.com. 2360 A. ARAGÓN-NAVAS ET AL. acetate 15 mM/1 mL formic acid in MilliQ® water and 50% The study was carried out in 45 (40% males, 60% females) acetonitrile (flow rate: 0.3 mL/min). Long Evans rats with 4-week-old of age and weights ranged from 50 to 100 grams at the beginning of the study. Animals were housed in standard cages with water and food ad libi- Fibronectin quantification by enzyme-linked tum, in 12-hours dark-light cycled rooms with temperature immunosorbent assay (ELISA) (22 °C) and relative humidity (55%) controlled conditions. IOP Fibronectin quantification was performed by ELISA technique measurements were recorded under gas anesthesia with a using a Fibronectin ELISA Kit (DuoSet® Human Fibronectin mixture of 3% sevoflurane gas and 1.5% oxygen and the DY1918-05, R&D Systems, Minneapolis, MN, USA). same gas anesthesia with subcutaneous analgesia (dilution 1/10 of buprenorphine (0.05 mg/kg)) was used for performing OHT injections. General anesthesia by intraperitoneal injec- Dexamethasone encapsulation efficiency from tions (60 mg/Kg of Ketamine + 0.25 mg/Kg of dexamethasone/fibronectin-loaded PLGA microspheres Dexmedetomidine) was used for electroretinogram (ERG) and 1 mg of dexamethasone/fibronectin-loaded PLGA Ms was optical coherence tomography (OCT) and mydriatic eye drops dissolved in 2.5 mL of methylene chloride. Afterwards, 6 mL with tropicamide 10 mg/ml and phenylephrine 100 mg/ml, of methanol (MeOH) was added to cause the polymer pre- (Alcon Cusí® SA, Barcelona, Spain) to fully dilate. Surgical and cipitation, remaining the drug dissolved in the methylene sedative procedures were realized under temperature control chloride:methanol mixture. This organic solution was iso- with warm pads, antisepsia conditions, topical anesthesia lated from the solid polymer by centrifugation (5,000 rpm, with tetracaine 1 mg/ml + oxibuprocaine 4 mg/ml (Anestesico 5 minutes, 20 °C). The supernatant was collected, filtered doble Colircusi®, Alcon Cusí® SA, Barcelona, Spain) and anti- (0.22 µm) and analyzed by the HPLC/MS method previously biotic with ofloxacin 3 mg/ml (Exocin Colircusi®, Alcon Cusí® described. SA, Barcelona, Spain) eye drops and lubrication on eyes with hypromellose 2% (Methocel® OmniVision, Germany) or eryth- Dexamethasone and fibronectin in vitro release studies romycin 5 mg/g (Oftalmolosa Cusí® eritromicina, Alcon Cusí® from dexamethasone/fibronectin-loaded PLGA SA, Barcelona, Spain) and after procedures animals were let microspheres recovering in enriched 2.5% oxygen atmosphere. 2.5 mg of dexamethasone/fibronectin-loaded PLGA Ms were incubated in 1 mL of release media composed by a Injection procedure for ocular hypertension induction phosphate-buffered saline (PBS, pH 7.4) solution including Forty-five rats received 2 microliters PLGA MsDexaFibro sus- sodium azide (0.02% (w/v) (n = 4). The so-prepared samples pension (10% w/v) into the anterior chamber of the right were located in a water shaker bath (100 rpm, 37 °C, Memmert eye (RE) using a 10-μl-Hamilton® syringe and glass micropi- Shaking Bath, Memmert, Schwabach, Germany). At pre-set pette by a corneal superotemporal puncture at baseline. To times samples were gently centrifugated (5,000 rpm for 5 min, avoid reflux it is recommended placing the rat on its left 20 °C). The supernatant was extracted, filtered (0.22 μm) and side and covering the eye puncture with a surgical sponge. used for dexamethasone quantification by HPLC/MS employ- ing the method previously mentioned. The formed pellet was resuspended in the same volume of fresh release media to continue the study until the following sampling time-point. Clinical signs and intraocular pressure measurements In parallel, the same amount of dexamethasone/ Ophthalmological clinical signs such as ocular hyperemia, fibronectin-loaded PLGA Ms was suspended in cornea alteration, infection or intraocular inflammation were phosphate-buffered saline (1 ml, PBS, pH 7.4) including weekly evaluated; as well as IOP measured with the rebound sodium azide (0.02% (w/v) and bovine serum albumin (BSA) tonometer Tonolab® (Tonolab; Tiolat Oy Helsinki, Finland). (1% (w/v)). The suspension was placed in a water shaker bath The value of the IOP was the average of three consecutive (100 rpm, 37 °C, Memmert Shaking Bath, Memmert Schwabach, measurements, which resulted from the average of 6 Germany) (n = 2). Similarly, at pre-set times, the so-prepared rebounds. IOP was always recorded in the mornings (9 samples were centrifuged (5,000 rpm, 5 minutes, 20 °C), the to 13 am). supernatants removed for FN quantification by ELISA and the remaining microspheres refilled with fresh PBS/azide/ In vivo neuro-retinal examination BSA media. Electroretinography. Functionality of neuro-retinal structures was studied using electroretinography (ERG) (Roland consult® Animal’s welfare and anesthesia RETIanimal ERG, Germany) by full-field scotopic ERG and Photopic Negative Response (PhNR) protocols at baseline, The work with animals was carried out in the experimental 12 and 24 weeks. To perform the scotopic ERG test animals surgery service department of the Biomedical Research were dark-adapted for 12 hours and pupils fully dilated. Active Center of Aragon (CIBA). The experiment was previously electrodes were placed on the cornea, references at both approved by the Ethics Committee for Animal Research sides under the skin and the ground electrode nearby its (PI34/17) and carried out in strict accordance with the ARVO tail. Electrode impedance was accepted with a difference Statement for the Use of Animals in Ophthalmic and Vision <2kΩ between electrodes. Both eyes were simultaneously Research. tested by a Ganzfeld Q450 SC sphere with white LEDs Drug Delivery 2361 flashes for stimuli and seven steps with increasing intensity was performed. Finally, sections were stained with DAB of luminance and intervals were performed. Scotopic test (Sigma-Aldrich) for 3 minutes and counterstained with examined rod response: step 1: −40 dB, 0.0003 cds/m 2, Harrys Hematoxylin (Sigma-Aldrich) for 20 minutes at room 0.2 Hz [20 recordings averaged]; step 2: −30 dB, 0.003 cds/ temperature. Ganglion cells were counted in radial sec- m2, 0.125 Hz [18 recordings averaged]; step 3: −20 dB, 0.03 tions of the retina, along 2 mm of a linear region of the cds/m2, 8.929 Hz [14 recordings averaged]; step 4: −20 dB, 0.03 ganglion cell layer, at four areas, two in both sides of the cds/m2, 0.111 Hz [15 recordings averaged]; step 5: −10 dB, 0.3 optic nerve head. Images were analyzed by an operator cds/m2, 0.077 Hz [15 recordings averaged]; mixed rod–cone response: step 6: 0 dB, 3.0 cds/m2, 0.067 Hz [12 recordings blinded to treatment groups. Statistical analysis of the averaged]; and oscillatory potentials: step 7: 0 dB, 3.0 cds/m2, number of ganglion cells was conducted in R (v. 3.6.0) 29.412 Hz [10 recordings averaged]). The PhNR protocol was using a paired t-test. The results are shown as mean ± SEM. performed after light adaptation to blue background (470 nm, Values of p < 0.05 were considered statistically significant. 25 cds/m2), and a red LED flash (625 nm, −10 dB, 0.30 cds/ Procedural immunohistochemistry controls were carried m2, 1.199 Hz [20 recordings averaged]) was used as stimuli. out by omitting the primary antibody in a sequential Latency (in milliseconds) and amplitude (in microvolts) were tissue section. studied in a, b and PhNR waves. To analyze the anterior segment of the eye and the loca- tion of the microspheres, sections were deparaffinized and Optical coherence tomography. Neuro-retinal structures stained with Hematoxylin/Eosin or only with Harris were studied with the Spectralis OCT device (Heidelberg® Hematoxylin and then observed by differential interference Engineering, Germany) and with a contact lens adapted contrast (Nomarski) microscopy. In addition, paraffin sections on the rat cornea to get higher quality acquisitions. It were stained with the fluorescent dye BODIPY (Invitrogen, was performed at baseline and at 2, 4, 6, 8, 12, 18 and Carlsbad, CA, USA) at 1:50 dilution and nuclear counter- 24 weeks after the OHT injection. Protocols such as full Retina thickness posterior pole (R), Retina Nerve Fiber stained using Hoechst (Sigma-Aldrich) at 1:100 were assessed. Layer (RNFL) and Ganglion Cells Layer (GCL) with automatic For fluorescence image acquisition, a laser scanning confocal segmentation were used. These protocols analyzed an area microscope (TCS SP5; Leica Microsystems GmbH, Heidelberg, of 3 mm around the center of the optic disk by 61 b-scans Germany) was used. Image analyses were performed with and subsequent follow-up examinations were acquired at this Fiji (Schindelin et al., 2012). same location using the eye tracking software and follow-up application. The Retina and GCL were analyzed by mimicking the 9 ETDRS areas which included a central (C) 1 mm circle Statistical analysis centered in the optic disk, (though no fovea exists in rats) Data were recorded in an Excel database and statistical anal- and inner (inferior -II-, superior -IS-, nasal -IN-, temporal -IT-) ysis was performed using SPSS software version 20.0 (SPSS and outer (inferior -OI-, superior -OS, nasal -ON-, temporal -OT-) rings measuring 2 and 3 mm in diameter, respectively, Inc., Chicago, IL). The Kolmogorov-Smirnov test was used to as well as total volume (TV). The RNFL protocol provides assess sample distribution. Given the parametric distribution measurements of the 6 peripapillary sectors (inferotemporal of the data, Student’s t-test was used to evaluate the differ- -IT-, inferonasal -IN-, superotemporal -ST-, superonasal -SN-, ences between eyes, and a paired Student’s t-test was used nasal -N- and temporal -T-). Full retinal thickness posterior to compare the changes recorded in each eye over the study pole (R) comprises from the inner limiting membrane to the period. All values were expressed as means ± standard devi- retinal pigment epithelium; RNFL from the inner limiting ations. Values of p < 0.05 (expressed as *) were considered to membrane to the GCL boundaries; and GCL from RNFL to indicate statistical significance. The Bonferroni correction for the inner nuclear layer boundaries. multiple comparisons was also calculated to avoid a high Biased examinations were rejected or manually corrected false-positive rate. The level of significance for each variable if the algorithm had obviously failed. was established according to Bonferroni calculations (expressed as #). Histology Animals were euthanized under humanity conditions with Results an intracardiac injection of sodium thiopental (25 mg/ml) Production yield under general anesthesia and eyes were immediately enu- cleated. Paraffin-embedded eyes were sectioned (5 µm) Ms showed a production yield of 55.14% for the selected along the eye axis, deparaffinized and rehydrated. After 20–10 µm fraction. The particle size distribution resulted uni- several washes in phosphate-buffered saline (PBS), sec- modal with a mean particle size value of 14.81 ± 0.30 µm. tions were incubated overnight at 4 °C with mouse anti-Brn3a (Santa Cruz Biotechnology, Heidelberg, Encapsulation efficiency Germany) at 1:50 dilution. After washing the sections in PBS, slides were incubated for 2 hours at room tempera- The dexamethasone encapsulation efficiency measurements ture with biotinylated horse anti-mouse at 1:50 dilution lead to 79.13 ± 2.64% of the initial drug included during the (Vector Laboratories, Burlingame, CA, USA). Then, incuba- preparation procedure (71.94 ± 2.40 µg DX/mg Ms). tion with ABC-HRP (Thermo Fisher Scientific, Waltham Unfortunately, the fibronectin lability made impossible the Massachusetts, USA) at 1:50 dilution at room temperature real quantification of the loaded protein. 2362 A. ARAGÓN-NAVAS ET AL. Morphological evaluation time. In addition, to offer more complete information, values of the accumulated percentage of transfer are also provided SEM images evidenced the presence of spherical and throughout the graph. regular-sized Ms with surficial porous and slightly rough sur- Fibronectin underwent an initial rapid release of 35.90 ng/ faces. Moreover, the presence of internal porous was con- mg Ms in the first 10 days. After that, a sequence of rapid firmed by TEM images, typically due to the double emulsion and slow-release profiles was detected: from day 10 to day method employed for Ms elaboration (Figure 2). 28, the profile shows a release of 0.211 ng FN/mg Ms/day, followed by a slower release of 0.017 ng FN/mg Ms/day from In vitro release studies day 28 to 74. Next, from day 74 to 108, a second faster release of 0.074 ng FN/mg Ms/day was found and finally from Both dexamethasone and fibronectin in vitro release profiles day 108 to the end of the study the release rate was 0.024 ng showed a multiphasic shape typically observed for PLGA FN/mg Ms/day. microspheres. Dexamethasone also showed a high initial release of Figures 3 and 4 show the cumulative amount of both 53.42 µg/mg Ms in the first 10 days and, as mentioned before, fibronectin and dexamethasone per mg of Ms released versus a subsequent slow but continuous release of different releases Figure 2. Dexamethasone-fibronectin co-loaded PLGA microspheres morphology. Scanning (left and center) and transmission (right) electron microscopy images. Figure 3. Cumulative in vitro release of fibronectin from dexamethasone-fibronectin co-loaded PLGA microspheres. Figure 4. Cumulative in vitro release of dexamethasone from dexamethasone-fibronectin co-loaded PLGA microspheres. Drug Delivery 2363 rates, characteristic of PLGA microspheres. From day 10 to medium corticosteroid response (between 6 and 15 mmHg), 28, the in vitro release presented a slow phase with a rate (Figure 5c) with constant lineal tendency on corticosteroid of 7.22 ng DX/mg Ms/day, followed by a faster release of response in the injected eyes. Very few animals were high 22 ng DX/mg Ms/day from day 28 to 38 and the final release corticosteroid responders (>15 mmHg increase). rate was slower, 1.54 ng DX/mg DX/day from day 38 to day 77. No dexamethasone was released from this time point. In vivo neuroretinal examination Electroretinography Ophthalmological signs and intraocular pressure The induced RE showed a tendency to longer latency signals in bipolar cells (b-wave) and smaller amplitudes in photore- Forty of 43 animals used, did not show infection, intraocular ceptors (a-wave) and bipolar (b-wave) cells compared to LE inflammation, cataract formation or retinal detachment and and over the study, but did not reach statistical differences the ocular surface was well preserved. The MsDexaFibro with scotopic ERG conditions. However, the light-adapted showed a tendency to localize at the superior iridocorneal PhNR protocol that specifically measures the retinal ganglion angle, due to their lower density compared to aqueous cells (RGC) activity detected smaller amplitudes (p < 0.05) in humor and this disposition allowed a clear visual axis. One the injected RE, which it involves lower number of functional animal developed corneal leucoma and other an iridocorneal RGC (Figure 6). synechia that did not preclude proper testing and follow-up; however, another third rat developed a focus of vitreoretinitis so this animal was discarded and its results were not used. Optical coherence tomography This model caused a progressive IOP increase and ocular All the three R, RNFL and GCL protocols showed a progressive hypertensive (OHT) eyes (>20 mmHg) over the study. The decrease in neuroretinal thickness over the study, with very injected right eye (RE) showed higher measurements than few OCT sectors showed statistical significance (p < 0.05) the non-injected left eye (LE) up to 6 weeks (p < 0.05), but between RE and LE, a tendency to lower thickness measure- then both eyes similarly increased up to 24 weeks. Both eyes ments was detected in the injected RE. Full retinal thickness reached OHT at week 11 and fluctuations were observed (R) experienced the highest decrease followed by RNFL and over the study (Figure 5a). The highest percentage (88.9%) GCL and it occurred in both eyes. And an increasing fluctu- of OHT eyes was reached at 20 weeks (Figure 5b). Most rats ation was detected at week 12 especially in injected RE experienced an IOP increase (in average) according to a (Figure 7a). However, when the percentage of thickness loss Figure 5. Intraocular pressure in the dexamethasone-fibronectin co-loaded microsphere (MsDexaFibro) model. a: Intraocular pressure curve over 6 months. b: Percentage of ocular hypertensive eyes (>20 mmHg). c: Percentage of corticosteroid response and tendency in right eyes over the study. d: Averaged percentage of corticosteroid response in right eyes. Low: <6 mmHg increase; medium: 6–15 mmHg increase; high: >15 mmHg increase. Abbreviations: IOP: intraocular pressure; RE: right eye; LE: left eye; w: week; OHT: ocular hypertension; %: percentage *: p < 0.05; #: p < 0.02 (Bonferroni Correction for multiple comparisons). 2364 A. ARAGÓN-NAVAS ET AL. Figure 6. Neuroretina functionality, measured by dark- and light-adapted electroretinography (ERG), in the dexamethasone-fibronectin co-loaded microspheres (MsDexaFibro) model over 6 months of follow-up. MsDexafibro: microspheres loaded with dexamethasone and fibronectin; RE: right eye; LE: left eye; w: week; DA: dark-adapted; LA: light-adapted; μV: microvolts; ms: milliseconds. over time was analyzed, the RNFL parameter showed the and at all stages. Retina showed the highest loss rate followed highest percentage loss at every time point explored and on by RNFL and finally GCL and the biggest loss rate occurred average. It was followed by the GCL and finally the full retinal at early times. Moreover, both eyes lost similar quantity of thickness. Injected RE showed a lower percentage loss of microns in GCL thickness per every mmHg increased (Figure 8). thickness in RNFL and GCL than the non-injected LE (Figure 7b). Pathological findings The neuroretinal percentage loss by OCT sectors from RFNL, GCL and Retina was quantified and the loss tendency The count of positive Brn3a cells along with four areas of was analyzed. The inferior sector in RNFL showed the most the retina (Figure 9A), showed that the mean number of intense and frequent loss. In GCL the inner sectors showed ganglion cells per linear mm of the retina was not signifi- bigger percentage loss at any time explored in both eyes cantly different in non-intervened left eyes in comparison to and the R experienced bigger loss in outer sectors with the right eyes injected with dexamethasone-fibronectin micro- STIN averaged loss trend (see Supplementary Figures spheres (p = 0.92) (Figure 9B). However, fewer ganglion cells S1–S3). per linear mm of the retina were counted (10.12 ± 4.99) than The average loss rate expressed in microns per mmHg and in our previous model (12.16 ± 3.37) at 6 months (Rodrigo day extracted from all sectors was also quantified in both eyes et al., 2021). In order to ascertain if the injected microspheres Drug Delivery 2365 Figure 7. A Neuro-retinal thickness in microns by OCT throughout 6-month follow-up. B. Average thickness percentage loss by OCT in microspheres co-loaded with dexamethasone and fibronectin (MsDexaFibro) model up to 6-month follow-up. MsDexaFibro: microspheres co-loaded with dexamethasone and fibronectin; RE: right eye; LE: left eye; OCT: optical coherence tomography; RNFL: Retinal Nerve Fiber Layer; GCL: Ganglion Cell Layer complex; average thickness in microns (μm); w: week; %: percentage. were located at the iridocorneal angle, they were analyzed the fluorescent BODIPY, that we previously have demon- using Hematoxylin/Eosin, Nomarski and fluorescent micros- strated is specific for PLGA (Garcia-Herranz et al., 2021), fur- copy. The microspheres are transparent, and they are not ther confirmed that injected microspheres are hampered the stained by Hematoxylin or Eosin, but its presence can be still iridocorneal angle and the trabecula meshwork (Figure 10C). appreciated in the sections (Figure 10A). The characteristic Fibronectin plays a major role in the adhesion of many light refraction produced by the microspheres was used to cell types, in fact, fibronectin is considered as the extracel- readily identify them in paraffin sections using Normarski lular glue (Zollinger & Smith, 2017). Fibronectin has a RGD microscopy (Figure 10B). The staining of microspheres with loop that is a promiscuous site for many cellular integrins, 2366 A. ARAGÓN-NAVAS ET AL. Figure 8. Neuro-retinal loss rate measured by optical coherence tomography (OCT) in the co-loaded with dexamethasone and fibronectin microspheres (MsDexaFibro) model. MsDexaFibro: microspheres co-loaded with dexamethasone and fibronectin; RE: right eye; LE: left eye; w: week; RNFL: Retinal Nerve Fiber Layer; GCL: Ganglion Cell Layer complex. such as α5β1, α3β1, α8β1, and αvβ3 (Pankov & Yamada, the iridocorneal angle and the first release of the encapsu- 2002). This could explain how, unlike our other models lated substances. Thereafter, IOP slowly increased and was where microspheres without fibronectin were injected maintained, coinciding with lower load release. It was prob- (Garcia-Herranz et al., 2021; Rodrigo et al., 2021), focal adhe- ably due to extracellular matrix production by the constant sions were observed between the iris and the cornea (Figure presence of dexamethasone and the deposition of the protein 11). These synechiae are established through the fibronectin. fibronectin-containing microspheres and that would glue Microencapsulation and subsequent simultaneous release the iris and cornea (Figure 11). Obviously, these synechiae of more than one compound in modified release systems will modify the flow of the aqueous humor and therefore represent a technological challenge that has been developed its reabsorption at the level of the trabecular meshwork. in recent years as a very interesting strategy, for example, to be able to diminish the amount of polymer injected (Arranz-Romera et al., 2019; 2019). In order to encapsulate both compounds in the same formulation, it was decided Discussion to include each of them in the emulsion phase in which The model presented in this work combines some of the they were more soluble, so the protein was incorporated strategies previously used in other rodent models of glau- into the internal aqueous phase while dexamethasone was coma, such as physical blockage of the trabecular meshwork, incorporated into the organic phase. In the recent past, there the administration of corticosteroids or the increased depo- have been some efforts to encapsulate two or more sub- sition of the protein fibronectin in the extracellular matrix stances in particles for different therapies. Although studies of the trabecular meshwork (Roberts et al., 2020; Rodrigo of multiloaded nanoparticles are more frequent (Jamil et al., et al., 2021). It was decided to co-release dexamethasone 2019; Sokol et al., 2019; Madani et al., 2020), multiloaded and fibronectin as the former stimulates extracellular matrix microencapsulation is also being explored (Dutt & Khuller, production and inhibits debris clearance (Dismuke et al., 2001; Román et al., 2014; Shi et al., 2014; Baek et al., 2017). 2016; Zeng et al., 2020) and on the other hand, fibronectin The main challenge of incorporating several substances is deposition has been associated with ocular aging and glau- to accomplish a controlled release from those microsystems. coma progression by playing an important role in trabecular In this sense, our research group has successfully achieved meshwork occlusion in human glaucoma (Babizhayev & the microencapsulation and sustained release of more than Brodskaya, 1989; Kasetti et al., 2017). In the model that we one protein, the combination of proteins and low molecular present, a combination of dexamethasone with fibronectin weight compounds or even a tri-delivery of low molecular in PLGA microspheres is performed. This model progressively weight molecules using the single emulsion evaporation increased IOP over 24 weeks, reaching ocular hypertension method (García-Caballero et al., 2018; Arranz-Romera et al., at 11 weeks. There was a slight initial increase in IOP after 2019; 2021). In the present work, we have achieved the induction, coinciding with the physical blockade of Ms at release of both substances in a controlled and slow delivery Drug Delivery 2367 Figure 9. Ganglion cell analysis in glaucomatous eyes. A. Ganglion cells were counted in four areas (squares) of a radial section of the retina passing through the optic nerve. B. Two representative images of the retina marked with anti-Brn3a corresponding to a left non-intervened eye (LE) and a right eye (RE) injected with dexamethasone-fibronectin microspheres from the same animal. The mean number of ganglion cells per linear mm of retina was not different between non-intervened and injected eyes. GCL: Ganglion cell layer; INL: Inner nuclear layer; ONL: Outer nuclear layer. Scale bars: (A) 1 mm, (B) 30 µm. for several months. Regarding double emulsions, works of strategy has been successfully used in the present work; co-encapsulation of two substances in the different phases however, the potential benefit of the presence of the hydro- of the emulsion are not very numerous. Qiao et al reported phobic compound on the microencapsulation of the hydro- the co-microencapsulation of two antibiotics of different philic one has not been yet explored. polarity by double emulsion, including each compound in Obtaining models of neurodegenerative diseases of the retina the more soluble phase. They observed a slow release of is a challenge in the field of ophthalmology (Guidoboni et al., both substances and suggested that the encapsulation of 2020). In the case of glaucoma, there are several animal models, the hydrophilic compound may be enhanced with the addi- among them, rodent animal models that are more convenient. tion of the hydrophobic one (Qiao et al., 2019). The same Different mechanisms to create rat glaucoma models have been 2368 A. ARAGÓN-NAVAS ET AL. Figure 10. Detection of microspheres in injected eyes. A. Paraffin sections stained with Hematoxylin/Eosin exhibited the presence of the microspheres inte- grated into the iridocorneal angle (arrowheads), a finding never observed in the control ones. B. The characteristic light diffraction produced by microspheres using Nomarski microscopy (arrowheads) confirmed their presence in the iridocorneal angle of injected eyes. C. The Fluorescent Stain BODIPY (green) also showed the presence of microspheres hampering the iridocorneal angle and the trabecular meshwork. Nuclei were counterstained with Hoechst (blue). LE: left eye (non-intervened); RE: Right eye (dexamethasone-fibronectin microsphere injected). CB: Ciliary body; I: Iris; C: cornea; TM: Transmission mode. Scale bars: 100 µm. Drug Delivery 2369 Figure 11. Iridocorneal synechias were observed in injected eyes (arrows). A. General view of a synechia between iris and the cornea. B. Using Nomarski microscopy, the diffracted light from the microspheres (arrowheads) was observed between the cornea and iris at the level of the focal contact between them. C. Staining with BODIPY (green) confirmed the presence of microspheres gluing the cornea and iris. Nuclei were counterstained with Hoechst (blue). C: cornea; I: Iris; L: lens; TM: Transmission mode. Scale bars: 100 µm. studied, as can be seen in Table 1, however, none of them prolonged study times. This fact increases the variability of achieves a slow and progressive degeneration of the retina as results among different research groups, as well as the disad- occurs when the disease develops in humans (Pang & Clark, vantages derived from interventionism on the animal (Biswas 2020). In addition, neither of these rat models get an IOP & Wan, 2019). The present model, being single injection and increase for a long period of time. IOP-elevation induced glau- technically simpler than other inducer models (Morrison et al., coma models (Dey et al., 2018) using non-biodegradable and 2015; Dey et al., 2018), can be used by different research groups biodegradable microspheres (Morgan & Tribble, 2015; Rodrigo allowing for a more reliable comparison of treatments and ther- et al., 2021; 2021) require periodic injections into the anterior apies. It also improves animal welfare with reduced handling chamber of the eye to maintain elevated pressures over stress, reduces ocular complications such as infections or PLGA 2370 A. ARAGÓN-NAVAS ET AL. Table 1. Pressure dependent glaucoma rat models. Average Average time magnitude of Authors group and publication Average time length of IOP IOP elevation Feature of histology for retina date Mechanism of model creation Sample size Control eye for IOP to rise increase from baseline at the study end Urcola et al. (2006) Intracameral injection microbead N=6 Contralateral eye and 5 weeks 25 weeks 15 mmHg IOP 23% RGC death Control group increase Urcola et al. (2006) Intracameral injection microbead N=4 Contralateral eye and 6 weeks 24 weeks 20 mmHg IOP 27% RGC death with viscoelastic material Control group increase Samsel et al. (2011) Intracameral injection of magnetic N = 61 Contralateral eye After injection 12.8–27 days. 6 mmHg IOP 36% RGC cell loss microbead increase Moreno et al. (2005) Intracameral administration of N = 45 Contralateral eye Not specified 10 weeks 8–10 mmHg IOP Significant RGC loss and viscous agents increase damage in ON axons. Decrease in scotopic ERG activity Morrison et al (1997) and Jia Sclerosis of the outflow pathway by N = 20 Contralateral eye 10 days 7–36 days 7–28 mmHg IOP 10–100% ON axon loss et al (2000) episcleral injection of hypertonic increase saline Ueda et al (1998), WoldeMussie Sclerosis of the outflow pathway by N = 10 Control group 1 week 5 weeks 6–24 mmHg IOP 50–70% ON axon loss et al., 2001) laser photocoagulation of increase Levkovith-Verbin et al. (2002) outflow pathway Shareef et al. (1995) Laquis et al. Cautery of extraocular veins N = 18 Contralateral eye After injection 6 weeks– 13–47 mmHg 4% RGC loss per week (1998) 2.5 months IOP increase Sun et al. (2011) Transient/intermittent IOP elevation N = 31 Contralateral eye After 7 hours 25 mmHg IOP 52% RGC loss at 28 days after by corneal limbus compression compression increase the insult Morrison et al (2016) Transient/intermittent IOP elevation N = 122 Naïve group After the insult 10 days 48 mmHg IOP lesions in 83% of ON by controlled elevation of IOP increase (CEI) Shepard et al (2010) Transduction of the TM with N=7 Contralateral eye 5 days 12 days 10–15 mmHg Not reported any potential glaucoma related genes by IOP increase RGC and ON axon loss. TGFβ2 Table extracted and adapted from Pang & Clark (2020). Abbreviations: ERG = electroretinogram. IOP = intraocular pressure. ON = optic nerve. RGC = retinal ganglion cells. TM = trabecular meshwork. Drug Delivery 2371 toxicity (Dossarps et al., 2015; Park, 2017; Zhao et al., 2017); as present features of steroid-induced glaucoma and primary well as is more cost-efficient and generates less environmental open-angle glaucoma (Rozsíval et al., 1981; Babizhayev & impact (Haines et al., 2017). Brodskaya, 1989; Kasimov & Aghaeva, 2017). This model try Our model showed that the most affected parameter by to mimics then the pathophysiology of human glaucoma, OCT was RNFL followed by inner sectors of GCL and finally with predictable IOP increase and progressive neuroretinal full retina, which may reflect retrograde degeneration (Lawlor loss, by avoiding rapid increases in OHT and severe neuroret- et al., 2018). Furthermore, when considering the GCL-RNFL inal loss (Agarwal & Agarwal, 2017). All this fulfills the char- component, the greatest standardized loss per mmHg was acteristics of an ideal model and allows for long-term studies suffered by RNFL, again showing that the hypertensive noxa (Biswas & Wan, 2019). exerts early and greater damage to axonal structure (Howell The animal model presented in this work is a clear exam- et al., 2007; Guo et al., 2010). At earlier times, the Inferior ple of how pharmaceutical technology can help other disci- and Nasal sectors of RNFL seem to be the most vulnerable plines to achieve optimized results by joining forces. While to damage, and at later times they become ISNT (INFERIOR the usual employ of controlled drug release systems is > SUPERIOR > NASAL > TEMPORAL glaucoma rule) due to the focused on the treatment of pathologies, we present in this possible influence of dynamic pathogenic changes. An exam- paper another important utility of sustained-release systems, ple is the important fluctuation at 12 weeks of the study in this case not to correct damage to the body but to cause observed in this model of glaucoma as well as observed in it. The use of modified-release systems to obtain animal mod- our previous one and in healthy animals linked to neuro- els is very poorly developed yet but, as observed in this plasticity and glia involvement (Rodrigo et al., 2020; 2021; work, it presents enormous potential. Regarding this idea 2021). In the GCL, the sectors most affected on average were and consistent with a lower total dose of dexamethasone the inner sectors TSN (TEMPORAL > SUPERIOR > NASAL), which than that used in our previous model, mostly mild levels of coincide with those with the highest density of retinal gan- corticoresponse were found. But, a similar % of rats (approx. glion cells (RGC) (Salinas-Navarro et al., 2009; Guo et al., 90%) developed OHT, and even fewer ganglion cells per 2010). Glaucoma mainly affects the RGCs, and their decreased linear mm of retina (10.12 ± 4.99 vs 12.16 ± 3.37) were counted functionality has been detected in photopic conditions with at 6 months (Rodrigo et al., 2021). This supports the import- the PhNR or pattern ERG tests, and scotopic conditions with ant role of fibronectin in the development of OHT. pSTR (Porciatti, 2015; Wilsey et al., 2017). Furthermore, outer Neuroprotective therapies aims to protect neurons from retinal cells changes have also been detected in scotopic secondary degeneration, already described in the introduc- conditions with reversible high IOP (approx. 35 mmHg) (Choh tion, which is the true trigger for the amplification of neu- et al., 2016), and in patients with late glaucoma (Graham ronal degeneration in chronic retinopathies (Cheung et al., et al., 1995). In the present model, where the damage 2008). Therefore, in order to correctly evaluate neuroprotec- occurred gradually, RGC dysfunction with a statistically sig- tive therapies, it is necessary to start from experimental ani- nificant decrease in PhNR was detected early (12 weeks); and mals with healthy retinas and that these retinas slowly coinciding with mild hypertensive levels (20–22 mmHg) a degenerate, as occurs in human retinal chronic pathologies limitation of outer retinal function was also found at (Morrison et al., 2011). The rat model that is presented in early-intermediate stages. this work would allow this type of study since, by achieving A bilateral neurodegeneration after induction of unilateral a very progressive ocular hypertension and sustained over OHT has been already observed in several animal models time, the degeneration of the retina is also so. In addition, (Sapienza et al., 2016; de Hoz et al., 2018; Rodrigo et al., one of the main problems of the in vivo evaluation of intra- 2021; 2021). This can be explained because the glial activa- ocular drug delivery systems is that it is very difficult to tion, characteristic of glaucomatous eyes, spreads through demonstrate the potential of the system when applied on the visual pathway, so the degenerative factors could, in retinas as degenerated as those that appear in most animal turn, activate the glia of the contralateral eye (Howell et al., models of glaucoma (Nadal-Nicolás et al., 2016; Rodrigo et al., 2007), producing retrograde damage and also creating an 2021). In this sense, our animal model, by offering a pro- imbalance in pressure at the optic nerve head (Jóhannesson longed degeneration of the retina over at least 6 months, et al., 2018). In addition, in other works a gradual increase would be a very useful candidate to evaluate this type of in IOP in the non-induced contralateral eye has been also modified release system. In the present model, only an initial detected (Roubeix et al., 2015; Sapienza et al., 2016; Rodrigo injection is needed to produce the sustained damage and, et al., 2021; 2021), as in our case. In fact, the use of the furthermore, this injection takes place in the anterior cham- contralateral eye as a control is not recommended in glau- ber of the eye, without any other structure of the eye being coma animal models. affected. Thus, the intravitreal space remains intact to test The present study shows a rat model of glaucoma with any modified release system for long periods. ocular hypertension that is easy to induce; with a high like- Concerning the limitations of this work and future studies, lihood of being reproducible, since only one induction and (I) Due to the lability of fibronectin, encapsulation could not nature-modifying intervention is performed; biologically plau- be quantified and therefore the actual amount of fibronectin sible with minimal side effects, as the injected material is injected is unknown. However, in vitro release was in the order biocompatible as it degrades to CO2 and H2O in the Krebs of nanograms, much lower than in vivo results of fibronectin cycle (Martins et al., 2018) and its trabecular meshwork mod- quantifications in the aqueous humor of glaucoma patients ifying compounds (dexamethasone and fibronectin) are (Kim et al., 1992), which we did not perform in our study in 2372 A. ARAGÓN-NAVAS ET AL. order not to alter the model by adding other puncture and Anderson DR. (2003). Collaborative normal tension glaucoma study. Curr maintaining eye homeostasis as much as possible; (II) We have Opin Ophthalmol 14:86–90. Arranz-Romera A, Davis BM, Bravo-Osuna I, et al. (2019). Simultaneous not analyzed the differences by sex, even when its involve- co-delivery of neuroprotective drugs from multi-loaded PLGA micro- ment is linked to immune-mediated pathologies (Desai & spheres for the treatment of glaucoma. J Control Release 297:26–38. Brinton, 2019; Rodrigo et al., 2021) and glaucoma; (III) It was https://doi.org/10.1016/j.jconrel.2019.01.012. neither evaluated the potential immune role of the protein Arranz-Romera A, Esteban-Pérez S, Garcia-Herranz D, et al. (2019). fibronectin nor the anti-inflammatory effect of dexamethasone Combination therapy and co-delivery strategies to optimize treatment on it (Hernandez et al., 2020; Mzyk et al., 2022) and that it of posterior segment neurodegenerative diseases. Drug Discov Today would be interesting to study in future works; (IV) Finally, 24:1644–53. Arranz-Romera A, Hernandez M, Checa-Casalengua P, et al. (2021). A safe glaucoma is an aged-link pathology, and the rats were in their GDNF and GDNF/BDNF controlled delivery system improves migration childhood at the start of the study, so results could deviate in human retinal pigment epithelial cells and survival in retinal gan- from translational human interpretation. glion cells: potential usefulness in degenerative retinal pathologies. In conclusion, a new chronic glaucoma animal model was Pharmaceuticals 14:50. created by a single injection of PLGA co-loaded MsDexaFibro Asaoka R, Murata H, Yanagisawa M, et al. (2017). The association between very similar to open-angle glaucoma. This chronic rat model photoreceptor layer thickness measured by optical coherence tomog- would have an important impact in ophthalmology, so it raphy and visual sensitivity in glaucomatous eyes. PLoS One 12:e0184064. allows long periods of study of this pathology, in pharma- Ashimatey BS, King BJ, Swanson WH. (2018). Retinal putative glial alter- cology because it could be used to evaluate the neuropro- ations: implication for glaucoma care. Ophthalmic Physiol Opt 38:56– tective activity of active compounds and also in pharmaceutical 65. https://doi.org/10.1111/opo.12425. technology, so it could allow the correct evaluation of the Avotri S, Eatman D, Russell-Randall K. (2019). Effects of resveratrol on efficacy of sustained intraocular drug delivery systems. inflammatory biomarkers in glaucomatous human trabecular mesh- work cells. Nutrients 11:984. Babizhayev MA, Brodskaya MW. (1989). Fibronectin detection in drainage outflow system of human eyes in ageing and progression of Disclosure statement open-angle glaucoma. Mech Ageing Dev 47:145–57. Babizhayev MA, Brodskaya MW. (1989). Fibronectin detection in drainage No potential conflict of interest was reported by the authors. outflow system of human eyes in ageing and progression of open-angle glaucoma. Mech Ageing Dev 47:145–57. Baek JS, Choo CC, Tan NS, Loo SCJ. (2017). Sustained-releasing hollow Funding microparticles with dual-anticancer drugs elicit greater shrinkage of tumor spheroids. Oncotarget 8:80841–52. This work was supported by Rio Hortega Research Grant M17/00213, Biswas S, Wan KH. (2019). Review of rodent hypertensive glaucoma PI17/01726, PI17/01946 (Instituto de Salud Carlos III), Grants models. Acta Ophthalmol 97:e331–340. MAT2017-83858-C2-2 and MAT2017-83858-C2-1 funded by MCIN/ Cheung W, Guo LI, Cordeiro MF. (2008). Neuroprotection in glaucoma: AEI/10.13039/501100011033 and by ‘ERDF A way of making Europe’, drug-based approaches. Optom Vis Sci 85:406–16. Grants PID2020-113281RB-C21 and PID2020-113281RB-C22 funded by Choh V, Gurdita A, Tan B, et al. (2016). Short-term moderately elevated MCIN/AEI/10.13039/501100011033. A.A.N thanks for the grant intraocular pressure is associated with elevated scotopic electroretino- PRE2018-083951 funded by MCIN/AEI/10.13039/501100011033 and by gram responses. Invest Ophthalmol Vis Sci 57:2140–51. ‘ESF Investing in your future’. D.G.H. acknowledges to UCM-Santander Clark AF, Wordinger RJ. (2009). The role of steroids in outflow resistance. fellowship (CT17/17-CT17-18). The sponsor or funding organization had Exp Eye Res 88:752–9. no role in the design or conduct of this research. This work was funded Conlon R, Saheb H, Ahmed IIK. (2017). Glaucoma treatment trends: a by Spanish Ministry of Science and Innovation. review. Can J Ophthalmol 52:114–24. https://doi.org/10.1016/j. jcjo.2016.07.013. Cuenca N, Fernández-Sánchez L, Campello L, et al. (2014). Cellular re- ORCID sponses following retinal injuries and therapeutic approaches for neurodegenerative diseases. Prog Retin Eye Res 43:17–75. Alba Aragón-Navas http://orcid.org/0000-0002-7466-8914 de Hoz R, Ramírez AI, González-Martín R, et al. (2018). Bilateral early María J. Rodrigo http://orcid.org/0000-0002-4009-3075 activation of retinal microglial cells in a mouse model of unilateral David Garcia-Herranz http://orcid.org/0000-0003-2092-5927 laser-induced experimental ocular hypertension. Exp Eye Res 171:12– Manuel Subias http://orcid.org/0000-0002-2344-016X 29. Silvia Mendez http://orcid.org/0000-0001-6072-1545 Desai MK, Brinton RD. (2019). Autoimmune disease in women: endocrine Jesús Ruberte http://orcid.org/0000-0003-1540-1432 transition and risk across the lifespan. Front Endocrinol (Lausanne) Judit Pampalona http://orcid.org/0000-0001-9454-0734 10:265. Irene Bravo-Osuna http://orcid.org/0000-0003-3133-7872 Dey A, Manthey AL, Chiu K, Do CW. (2018). Methods to induce chronic Luis E. Pablo http://orcid.org/0000-0003-2389-8282 ocular hypertension: reliable rodent models as a platform for cell Elena Garcia-Martin http://orcid.org/0000-0001-6258-2489 transplantation and other therapies. Cell Transplant 27:213–29. Rocío Herrero-Vanrell http://orcid.org/0000-0001-9764-7975 Dismuke WM, Klingeborn M, Stamer WD. (2016). Mechanism of fibronec- tin binding to human trabecular meshwork exosomes and its mod- ulation by dexamethasone. PLoS One 11:e0165326. References Dossarps D, Bron AM, Koehrer P, et al. (2015). Endophthalmitis after intravitreal injections: incidence, presentation, management, and vi- Agarwal R, Agarwal P. (2017). Rodent models of glaucoma and their sual outcome. Am J Ophthalmol 160:17–25. applicability for drug discovery. Expert Opin Drug Discov 12:261–70. Dutt MI, Khuller GK. (2001). Chemotherapy of mycobacterium tubercu- https://doi.org/10.1080/17460441.2017.1281244. losis infections in mice with a combination of isoniazid and rifampi- Drug Delivery 2373 cin entrapped in Poly(DL-lactide-co-glycolide) microparticles. J Levkovith-Verbin H, Quigley HA, Martin KRG, Valenta D, Baumrind LA. Antimicrob Chemother 47:829–35. (2002). Translimbal laser photocoagulation to the trabecular meshwork Fernández-Albarral JA, Ramírez AI, De Hoz R, et al. (2019). Neuroprotective as a model of glaucoma in rats. Invest Ophthalmol Vis Sci 43:402–10. and anti-inflammatory effects of a hydrophilic saffron extract in a https://iovs.arvojournals.org/article.aspx?articleid=2162526. model of glaucoma. IJMS 20:4110. Madani F, Esnaashari SS, Bergonzi MC, et al. (2020). Paclitaxel/metho- Gaasterland DE, Ederer F, Beck A, et al. (2000). The Advanced Glaucoma trexate co-loaded PLGA nanoparticles in glioblastoma treatment: for- Intervention Study (AGIS): 7. The relationship between control of mulation development and in vitro antitumor activity evaluation. Life intraocular pressure and visual field deterioration. Am J Ophthalmol Sci 256:117943. 130:429–40. https://doi.org/10.1016/S0002-9394(00)00538-9. Martins C, Sousa F, Araújo F, Sarmento B. (2018). Functionalizing PLGA García-Caballero C, Lieppman B, Arranz-Romera A, et al. (2018). and PLGA derivatives for drug delivery and tissue regeneration ap- Photoreceptor preservation induced by intravitreal controlled delivery plications. Adv Healthcare Mater 7:1701035. of gdnf and gdnf/melatonin in rhodopsin knockout mice. Mol Vis Masuda T, Shimazawa M, Hara H. (2017). Retinal diseases associated with 24:733–45. oxidative stress and the effects of a free radical scavenger (edaravone). Garcia-Herranz D, Rodrigo MJ, Subias M, et al. (2021). Novel use of plga Oxid Med Cell Longev 2017:9208489. microspheres to create an animal model of glaucoma with progressive Moreno MC, Aldana Marcos HJ, Croxatto JO, et al. (2005). A new exper- neuroretinal degeneration. Pharmaceutics 13:237. imental model of glaucoma in rats through intracameral injections Graham SL, Goldberg I, Buckland L, Hollows FC. (1995). Flash and pattern of hyaluronic acid. Exp Eye Res 81:71–80. electroretinogram changes with optic atrophy and glaucoma. Exp Eye Morgan JE, Tribble JR. (2015). Microbead models in glaucoma. Exp Eye Res 60:697–706. Res 141:9–14. https://doi.org/10.1016/j.exer.2015.06.020. Guidoboni G, Sacco R, Szopos M, et al. (2020). Neurodegenerative dis- Morrison JC, Cepurna WO, Johnson EC. (2015). Modeling glaucoma in orders of the eye and of the brain: a perspective on their rats by sclerosing aqueous outflow pathways to elevate intraocular fluid-dynamical connections and the potential of mechanism-driven pressure. Exp Eye Res 141:23–32. modeling. Front Neurosci 14:566428. Morrison JC, Cepurna WO, Tehrani S, et al. (2016). A period of controlled Guo L, Normando EM, Nizari S, et al. (2010). Tracking longitudinal retinal elevation of IOP (CEI) produces the specific gene expression respons- changes in experimental ocular hypertension using the cSLO and es and focal injury pattern of experimental rat glaucoma. Invest spectral domain-OCT. Invest Ophthalmol Vis Sci 51:6504–13. Ophthalmol Vis Sci 57:6700–11. Haines A, Harris F, Kasuga F, Machalaba C. (2017). Future Earth - Linking Morrison JC, Cepurna Ying Guo WO, Johnson EC. (2011). Pathophysiology research on health and environmental sustainability. BMJ 357:j2358. of human glaucomatous optic nerve damage: insights from rodent Hernandez H, Roberts AL, McDowell CM. (2020). Nuclear factor-kappa models of glaucoma. Exp Eye Res 93:156–64. beta signaling is required for transforming growth factor Beta-2 in- Morrison JC, Moore CG, Deppmeier LMH, et al. (1997). A rat model of duced ocular hypertension. Exp Eye Res 191:107920. chronic pressure-induced optic nerve damage. Exp Eye Res 64:85–96. Howell GR, Libby RT, Jakobs TC, et al. (2007). Axons of retinal ganglion Mzyk P, Zalog EG, McDowell CM. (2022). A20 attenuates the fibrotic cells are insulted in the optic nerve early in DBA/2J glaucoma. J Cell response in the trabecular meshwork. IJMS 23:1928. Biol 179:1523–37. https://doi.org/10.1083/jcb.200706181. Nadal-Nicolás FM, Rodriguez-Villagra E, Bravo-Osuna I, et al. (2016). Jamil A, Aamir Mirza M, Anwer MK, et al. (2019). Z. Iqbal, Co-delivery Ketorolac administration attenuates retinal ganglion cell death after of gemcitabine and simvastatin through PLGA polymeric nanoparticles axonal injury. Investig Ophthalmol Vis Sci 57:1183–92. https://doi. for the treatment of pancreatic cancer: in-vitro characterization, cel- org/10.1167/iovs.15-18213. lular uptake, and pharmacokinetic studies. Drug Dev Ind Pharm Naik S, Pandey A, Lewis SA, et al. (2020). Neuroprotection: a versatile 45:745–53. https://doi.org/10.1080/03639045.2019.1569040. approach to combat glaucoma. Eur J Pharmacol 881:173208. Jóhannesson G, Eklund A, Lindén C. (2018). Intracranial and intraocular Pang IH, Clark AF. (2020). Inducible rodent models of glaucoma. Prog pressure at the lamina cribrosa: gradient effects. Curr Neurol Neurosci Retin Eye Res 75:100799. Rep 18:25. Pankov R, Yamada KM. (2002). Fibronectin at a glance. J Cell Sci Kasetti RB, Maddineni P, Millar JC, et al. (2017). Increased synthesis and 115:3861–3. deposition of extracellular matrix proteins leads to endoplasmic re- Pardue MT, Allen RS. (2018). Neuroprotective strategies for retinal disease. ticulum stress in the trabecular meshwork. Sci Rep 7:14951. https:// Prog Retin Eye Res 65:50–76. doi.org/10.1038/s41598-017-14938-0. Park K. (2017). Tolerance levels of PLGA microspheres in the eyes. J Kasimov EM, Aghaeva FA. (2017). Cortisol levels in plasma and aqueous Control Release 266:365. humour of patients with steroid induced and other glaucomas. Vestn Pascale A, Drago F, Govoni S. (2012). Protecting the retinal neurons from Oftalmol 133:39–45. glaucoma: lowering ocular pressure is not enough. Pharmacol Res Kim KS, Lee BH, Kim IS. (1992). The measurement of fibronectin concen- 66:19–32. https://doi.org/10.1016/j.phrs.2012.03.002. trations in human aqueous humor. Korean J Ophthalmol 6:1–5. Patel GC, Phan TN, Maddineni P, et al. (2017). Dexamethasone-induced Kitaoka Y, Kitaoka Y, Kwong JMK, et al. (2006). TNF-α-induced optic nerve ocular hypertension in mice: effects of myocilin and route of admin- degeneration and nuclear factor-κB p65. Investig Ophthalmol Vis Sci istration. Am J Pathol 187:713–23. https://doi.org/10.1016/j.aj- 47:1448–57. path.2016.12.003. L. Jia, Cepurna WO, Johnson EC, Morrison JC, Patterns of intraocular Porciatti V. (2015). Electrophysiological assessment of retinal ganglion pressure elevation after aqueous humor outflow obstruction in rats. cell function. Exp Eye Res 141:164–70. (2014). Invest Ophthalmol Vis Sci. 41 (2000) 1380–1385. https://iovs.arvojour- Qiao ZW, Yuan Z, Zhang W, et al. (2019). Preparation, in vitro release nals.org/article.aspx?articleid=2123102#89531022 (accessed May 26, and antibacterial activity evaluation of rifampicin and 2022). moxifloxacin-loaded poly(D,L-lactide-co-glycolide) microspheres. Artif Laquis S, Chaudhary P, Sharma SC. (1998). The patterns of retinal gan- Cells Nanomed Biotechnol 47:790–8. glion cell death in hypertensive eyes. Brain Res 784:100–4. Quigley H, Broman AT. (2006). The number of people with glaucoma Lawlor M, Danesh-Meyer H, Levin LA, et al. (2018). Glaucoma and the worldwide in 2010 and 2020. Br J Ophthalmol 90:262–7. https://doi. brain: trans-synaptic degeneration, structural change, and implications org/10.1136/bjo.2005.081224. for neuroprotection. Surv Ophthalmol 63:296–306. Ramirez AI, de Hoz R, Salobrar-Garcia E, et al. (2017). The role of mi- Lestak J, Jiraskova N, Zakova M, et al. (2018). Normotensive glaucoma. croglia in retinal neurodegeneration: Alzheimer’s disease, Parkinson, Biomed Pap 164. https://doi.org/10.5507/bp.2018.039. and glaucoma. Front Aging Neurosci 9:214. 2374 A. ARAGÓN-NAVAS ET AL. Ritch R. (2000). Neuroprotection: is it already applicable to glaucoma traocular pressure and reduces outflow facility in rodent Eyes. Invest therapy? Curr Opin Ophthalmol 11:78–84. Ophthalmol Vis Sci 51:2067–76. Roberts AL, Mavlyutov TA, Perlmutter TE, et al. (2020). Fibronectin extra Shi X, Li C, Gao S, et al. (2014). Combination of doxorubicin-based che- domain A (FN-EDA) elevates intraocular pressure through Toll-like motherapy and polyethylenimine/p53 gene therapy for the treatment receptor 4 signaling. Sci Rep 10:9815. https://doi.org/10.1038/ of lung cancer using porous PLGA microparticles. Colloids Surfaces s41598-020-66756-6. B Biointer faces 122:498–504. https://doi.org/10.1016/j.col- Rodrigo MJ, Garcia-Herranz D, Aragón-Navas A, et al. (2021). Long-term surfb.2014.07.020. corticosteroid-induced chronic glaucoma model produced by intra- Smedowski A, Pietrucha-Dutczak M, Kaarniranta K, Lewin-Kowalik J. cameral injection of dexamethasone-loaded PLGA microspheres. Drug (2014). A rat experimental model of glaucoma incorporating Deliv 28:2427–46. rapid-onset elevation of intraocular pressure. Sci Rep 4:5910. https:// Rodrigo MJ, Garcia-Herranz D, Subias M, et al. (2021). Chronic glaucoma doi.org/10.1038/srep05910. using biodegradable microspheres to induce intraocular pressure Sokol MB, Nikolskaya ED, Yabbarov NG, et al. (2019). Development of elevation. six-month follow-up. Biomedicines 9:682. novel PLGA nanoparticles with co-encapsulation of docetaxel and Rodrigo MJ, Martinez-Rincon T, Subias M, et al. (2020). Effect of age and abiraterone acetate for a highly efficient delivery into tumor cells. J sex on neurodevelopment and neurodegeneration in the healthy eye: Biomed Mater Res 107:1150–8. longitudinal functional and structural study in the long–evans rat. Stankowska DL, Nam MH, Nahomi RB, et al. (2019). Systemically admin- Exp Eye Res 200:108208. istered peptain-1 inhibits retinal ganglion cell death in animal mod- Rodrigo MJ, Martinez-Rincon T, Subias M, et al. (2021). Influence of sex els: implications for neuroprotection in glaucoma. Cell Death Discov on neuroretinal degeneration: six-month follow-up in rats with chron- 5:112. ic glaucoma. Invest Ophthalmol Vis Sci 62:9. Sun H, Wang Y, Pang IH, et al. (2011). Protective effect of a JNK inhibi- Román BS, Gómez S, Irache JM, Espuelas S. (2014). Co-encapsulated CpG tor against retinal ganglion cell loss induced by acute moderate oligodeoxynucleotides and ovalbumin in PLGA microparticles; an in ocular hypertension. Mol Vis 17:864. /pmc/articles/PMC3081797/ (ac- vitro and in vivo study. J Pharm Pharm Sci 17:541–53. cessed May 26, 2022). Roubeix C, Godefroy D, Mias C, et al. (2015). Intraocular pressure reduc- Tham YC, Li X, Wong TY, et al. (2014). Global prevalence of glaucoma tion and neuroprotection conferred by bone marrow-derived mesen- and projections of glaucoma burden through 2040: a systematic re- chymal stem cells in an animal model of glaucoma. Stem Cell Res view and meta-analysis. Ophthalmology 121:2081–90. Ther 6:177. https://doi.org/10.1186/S13287-015-0168-0. Thanos C, Emerich D. (2005). Delivery of neurotrophic factors and ther- Rozsíval P, Hampl R, Obenberger J, et al. (1981). Aqueous humour and apeutic proteins for retinal diseases. Expert Opin Biol Ther 5:1443–52. plasma cortisol levels in glaucoma and cataract patients. Curr Eye Ueda J, Sawaguchi S, Hanyu T, et al. (1998). Experimental glaucoma Res 1:391–6. model in the rat induced by laser trabecular photocoagulation after Salinas-Navarro M, Mayor-Torroglosa S, Jiménez-López M, et al. (2009). an intracameral injection of India. Ink Jpn J Ophthalmol 42:337–44. A computerized analysis of the entire retinal ganglion cell population Urcola JH, Hernández M, Vecino E. (2006). Three experimental glaucoma and its spatial distribution in adult rats. Vision Res 49:115–26. models in rats: comparison of the effects of intraocular pressure el- Samsel PA, Kisiswa L, Erichsen JT, et al. (2011). A novel method for the evation on retinal ganglion cell size and death. Exp Eye Res 83:429– induction of experimental glaucoma using magnetic microspheres. 37. Invest Ophthalmol Vis Sci 52:1671–5. WoldeMussie WLAE; Ruiz G; Wijono M, Neuroprotection of retinal gan- Sapienza A, Raveu AL, Reboussin E, et al. (2016). Bilateral neuroinflam- glion cells by brimonidine in rats with laser-induced chronic ocular matory processes in visual pathways induced by unilateral ocular hypertension. Invest Ophthalmol Vis Sci 42 (2001) 2849–2855. https:// hypertension in the rat. J Neuroinflammation 13:44. https://doi. iovs.arvojournals.org/article.aspx?articleid=2123016. org/10.1186/s12974-016-0509-7. Weinreb RN, Levin LA. (1999). Is neuroprotection a viable therapy for Schindelin J, Arganda-Carreras I, Frise E, et al. (2012). Fiji: an open-source glaucoma? Arch Ophthalmol 117:1540–4. platform for biological-image analysis. Nat Methods 9:676–82. Wilsey L, Gowrisankaran S, Cull G, et al. (2017). Comparing three differ- Schlamp CL, Li Y, Dietz JA, et al. (2006). Progressive ganglion cell loss ent modes of electroretinography in experimental glaucoma: diag- and optic nerve degeneration in DBA/2J mice is variable and asym- nostic performance and correlation to structure. Doc Ophthalmol metric. BMC Neurosci 7:66. 134:111–28. Shareef SR, Garcia-Valenzuela E, Salierno A, et al. (1995). Chronic ocular Zeng W, Wang W, Wu S, et al. (2020). Mitochondria and autophagy hypertension following episcleral venous occlusion in rats. Exp Eye dysfunction in glucocorticoid-induced ocular hypertension/glaucoma Res 61:379–82. mice model. Curr Eye Res 45:190–8. Sharma TP, McDowell CM, Liu Y, et al. (2014). Optic nerve crush induces Zhao M, Rodríguez-Villagra E, Kowalczuk L, et al. (2017). Tolerance of spatial and temporal gene expression patterns in retina and optic high and low amounts of PLGA microspheres loaded with mineralo- nerve of BALB/cJ mice. Mol Neurodegener 9:14. https://doi. corticoid receptor antagonist in retinal target site. J Control Release org/10.1186/1750-1326-9-14.[PMC][24767545] 266:187–97. Shepard AR, Cameron Millar J, Pang IH, et al. (2010). Adenoviral gene Zollinger AJ, Smith ML. (2017). Fibronectin, the extracellular glue. Matrix transfer of active human transforming growth factor-β2 elevates in- Biol 60–61:27–37.