Journal of Clinical Medicine Article New Insights into the Mechanisms of Action of Topical Administration of GLP-1 in an Experimental Model of Diabetic Retinopathy Joel Sampedro 1,2,† , Patricia Bogdanov 1,2,† , Hugo Ramos 1 , Cristina Solà-Adell 1,2 , Mireia Turch 1,2 , Marta Valeri 3 , Olga Simó-Servat 1,2 , Carmen Lagunas 4 , Rafael Simó 1,2,5, * and Cristina Hernández 1,2,5, * 1 Diabetes and Metabolism Research Unit, Vall d’Hebron Research Institute, 08035 Barcelona, Spain; joel.sampedro@vhir.org (J.S.); patricia.bogdanov@vhir.org (P.B.); hugo.ramos@vhir.org (H.R.); cristina.sola@vhir.org (C.S.-A.); mireia.turch@vhir.org (M.T.); olga.simo@vhir.org (O.S.-S.) 2 Centro de Investigación Biomédica en Red de Diabetes y Enfermedades Metabólicas Asociadas (CIBERDEM), Instituto de Salud Carlos III (ICSIII), 28029 Madrid, Spain 3 Unit of High Technology, Vall d’Hebron Research Institute, 08035 Barcelona, Spain; marta.valeri@vhir.org 4 Ferrer Advanced Biotherapeutics, Ferrer, Diagonal 549, 08029 Barcelona, Spain; clagunas@ferrer.com 5 Department of Medicine, Universitat Autònoma de Barcelona, 08193 Barcelona, Spain * Correspondence: rafael.simo@vhir.org (R.S.); cristina.hernandez@vhir.org (C.H.); Tel.: +34-934-894-172 (C.H.) † These authors contributed equally to this study.  Received: 28 January 2019; Accepted: 7 March 2019; Published: 11 March 2019  Abstract: The main goals of this work were to assess whether the topical administration of glucagon-like peptide-1 (GLP-1) could revert the impairment of the neurovascular unit induced by long-term diabetes (24 weeks) in diabetic mice and to look into the underlying mechanisms. For that reason, db/db mice were treated with eye drops of GLP-1 or vehicle for 3 weeks. Moreover, db/+ mice were used as control. Studies performed in vivo included electroretinogramand the assessment of vascular leakage by using Evans Blue. NF-κB, GFAP and Ki67 proteins were analyzed by immunofluorescence (IF). Additionally, caspase 9, AMPK, IKBα, NF-κB, AKT, GSK3, β-catenin, Bcl-xl, and VEGF were analyzed by WB. Finally, VEGF, IL-1β, IL-6, TNF-α, IL-18, and NLRP3 were studied by reverse transcription-polymerase chain reaction (RT-PCR) and immunofluorescence. We found that topical administration of GLP-1 reverted reactive gliosis and albumin extravasation, and protected against apoptosis and retinal dysfunction. Regarding the involved mechanisms, GLP-1 exerted an anti-inflammatory action by decreasing NF-κB, inflammosome, and pro-inflammatory factors. In addition, it also decreased VEGF expression. Furthermore, GLP-1 promoted cell survival by increasing the anti-apoptotic protein Bcl-xl and the signaling pathway Akt/GSK3b/β-catenin. Finally, Ki67 results revealed that GLP-1 treatment could induce neurogenesis. In conclusion, the topical administration of GLP-1 reverts the impairment of the neurovascular unit by modulating essential pathways involved in the development of diabetic retinopathy (DR). These beneficial effects on the neurovascular unit could pave the way for clinical trials addressed to confirm the effectiveness of GLP-1 in early stages of DR. Keywords: diabetic retinopathy; glucagon-like peptide-1; Retinal neurodegeneration; experimental diabetes 1. Introduction Diabetic retinopathy (DR), the most common complication of diabetes, is the primary cause of blindness among the working-age population of western countries [1]. J. Clin. Med. 2019, 8, 339; doi:10.3390/jcm8030339 www.mdpi.com/journal/jcm J. Clin. Med. 2019, 8, 339 2 of 20 The current treatments recommended for arresting the progression of DR (i.e., laser photocoagulation, intravitreal injections of antiangiogenic factors such as anti-vascular endothelial growth factor (VEGF) or corticosteroids, vitreoretinal surgery) are specifically addressed to very advanced stages of DR. In addition, all these treatments are quite expensive, present relevant adverse effects, and a vitreoretinal specialist is required [2]. Consequently, new treatments for treating early stages of the disease are urgently needed. In recent years, a myriad of experimental studies have indicated that neurodegeneration is an early event in the diabetic retina and several mechanisms linking neurodegeneration and microvascular disease has been proposed [3–5]. In addition, the pivotal role of neurovascular unit (NVU) impairment in the pathogenesis of DR has been underlined [6]. As a result of all this evidence, the American Diabetes Association has recently defined DR as a highly specific neurovascular disease of the retina rather than a merely microvascular disease [7]. The clinical impact of this new concept will be important not only in terms of screening based on measuring retinal neurodegeneration but also in the development of drug discovery and new strategies for the early treatment of the disease. Numerous experimental treatments based on neuroprotection have been reported in the last decade. However, only a few have been able to have an impact on microvascular disease; among them, GLP-1 receptor agonists (GLP-1RAs). In this regard, we previously reported that the topical administration of GLP-1 or GLP-1RAs using eye-drops prevented retinal neurodegeneration and early vascular leakage in db/db mice 12 weeks old [8]. In this experiment, the treatment was initiated when db/db mice were 10 weeks old, a stage in which only minimal if any diabetes-induced abnormalities in the retina could be detected [9]. Therefore, the beneficial effect of GLP-1 or GLP-1RAs was considered mainly preventative. It should be noted that topical treatment was unable to reduce blood glucose levels and, therefore, the observed effects of GLP-1 or GLP-1RAs were attributed to a local direct effect unrelated to an improvement of metabolic control. In the present study we wanted to assess in the same model whether the topical administration of GLP-1 reverts the impairment of the NVU when retinal neurodegeneration and microvascular impairment are already well-established. In addition, the underlying mechanisms have been examined. 2. Experimental Section 2.1. Experimental Design A total of 30 diabetic male db/db (BKS.Cg-Dock7m +/+ Leprdb/J) mice and 15 non-diabetic mice (db/+; control group) were purchased (Charles River Laboratories, Calco, Italy) for the study. Animals had free access to ad libitum food (ENVIGO Global Diet Complete Feed for Rodents, Mucedola, Milan, Italy) and filtered water. Mice were maintained under tight environmental conditions of temperature (20 ◦ C) and humidity (60%). Moreover, they had cycles of 12 h/12 h light/darkness. In order to minimize variability, animals were randomly housed (block randomization) in groups of four mice per cage. Each cage contained absorbent bedding and nesting material (BioFresh Performance Bedding 1/800 Pelleted Cellulose, Absorption Corp, Ferndale, WA, USA). 2.2. Interventional Study When the mice were aged 21 weeks, GLP-1 eye-drops (n = 15) or vehicle eye-drops (n = 15) were randomly administered directly onto the superior corneal surface of each eye using a syringe. One drop (5 µL) of GLP-1 (2 mg/mL), or vehicle (5 µL phosphate-buffered saline (PBS), pH 7.4) was administered twice daily for three weeks in each eye. On the last day (24 weeks of age), one drop of either GLP-1 or vehicle was administered to the eyes 1 h before euthanasia. The evaluation of the results was performed by investigators unaware of treatment received by the mice. This study was approved by the Animal Care and Use Committee of VHIR (Vall d’Hebron Research Institute). All the experiments were performed in accordance with the tenets of the European Community (86/609/CEE) and the Association for Research in Vision and Ophthalmology (ARVO). J. Clin. Med. 2019, 8, 339 3 of 20 2.3. Electroretinogram Full-field electroretinogram recordings were measured using the Ganzfeld ERG platform (Phoenix Research Laboratories, Pleasanton, CA, USA). Animals were dark adapted for at least 8 h prior to ERG recording and then anesthetized with isoflurane. Tropicamide (1%) was applied to each eye prior to the test. A cutaneous ground electrode was placed near the base of the tail, a needle electrode was placed cutaneously on the head between the two eyes and a cornel electrode was placed near each eye. Carboxymethylcellulose (1%) drops were applied to the interior surface of the contact lens electrodes prior to their placement on the eyes. The ERG parameters were measured as defined by the International Society for Clinical Electrophysiology of Vision [10]. 2.4. Tissue Processing The mice were killed by cervical dislocation. For mRNA and protein assessments the retinas were separated immediately after enucleation, frozen in liquid nitrogen, and stored at −80 ◦ C. Retinas used for immunohistochemical analysis were obtained from mice after transcardial perfusion with p-formaldehyde 4%. In these cases, intraperitoneal injection of anaesthesia (1 mL ketamine and 0.3 mL xylazine) was previously administered. 2.5. RNA Isolation and Quantitative Reverse Transcription Polymerase Chain Reaction (RT-PCR) Assay Total RNA from mice was extracted using Trizol® reagent (Invitrogen, Madrid, Spain) according to the manufacturer’s protocol. Then, RNA samples were treated with DNase (Qiagen, Madrid, Spain) to remove genomic contamination and purified on a RNeasy MinElute column (Qiagen, Madrid, Spain). RNA quantity was measured on a Nanodrop spectrophotometer, and integrity was determined on an Agilent 2100 Bioanalyzer. The single strand cDNA was synthesized as described in Prime Script® RT Master Mix kits. Real-time reverse transcription polymerase chain reaction (RT-PCR) was performed using SYBR Green PCR Master Mix (Applied Biosystems, Warrington, UK) using the 7.900 HT Sequence Detection System in 384-well optical plates with specific primers displayed in Table 1. For each sample, RT-PCRs were performed in triplicate and relative quantities were calculated using ABI SDS 2.0 RQ software, analyzed by the 2−∆∆Ct method, and given as ratio compared to β-actin as the endogenous control. Table 1. Primers used for RT-PCR. Gen Symbol Sequence Primer IL-1 Forward (50 -30 ) 50 -GCAACTGTTCCTGAACTCAACT-30 IL-1 Reverse (50 -30 ) 50 -ATCTTTTGGGGTCCGTCAACT-30 IL-6 Forward (50 -30 ) 50 -TAGTCCTTCCTACCCCAATTTCC-30 IL-6 Reverse (50 -30 ) 50 -TTGGTCCTTAGCCACTCCTTC-30 TNF Forward (50 -30 ) 50 -CCCTCACACTCAGATCATCTTCT-30 TNF Reverse (50 -30 ) 50 -GCTACGACGTGGGCTACAG-30 VEGFa Forward (50 -30 ) 50 -GAGTACCCCGACGAGATAGA-30 VEGFa Reverse (50 -30 ) 50 -GGCTTTGGTGAGGTTTGAT-30 NLRP-3 Forward (50 -30 ) 50 -ATTACCCGCCCGAGAAAGG-30 NLRP-3 Reverse (50 -30 ) 50 -TCGCAGCAAAGATCCACACAG-30 IL-18 Forward (50 -30 ) 50 -AGCAGTCCCAACTAAGCAGTA-30 IL-18 Reverse (50 -30 ) 50 -CAGCCAGTAGAGGATGCTGA-30 Actin Forward (50 -30 ) 50 -CTAAGGCCAACCGTGAAAG-30 Actin Reverse (50 -30 ) 50 -ACCAGAGGCATACAGGGACA-30 2.6. Western Blotting A lysis buffer was prepared (RIPA buffer: phenylmethanesulfonylfluoride (PMSF), 1 mM; Na3VO4, 2 mM; NaF, 100 mM) with a 1× protease inhibitor cocktail (Sigma, St Louis, MO, USA). Then, proteins were extracted from the neuroretinas in 80 µL of the aforementioned buffer. A total of 25 µg protein was resolved by 10% SDS-PAGE and transferred to a polyvinylidene difluoride (PVDF) membrane (Bio-Rad Laboratories, Madrid, Spain). The primary antibodies (Table 2) were incubated overnight at 4 ◦ C. J. Clin. Med. 2019, 8, 339 4 of 20 The following day, the suitable horseradish peroxidase (HRP) secondary antibody anti-rabbit or anti-mouse (Dako Agilent, Santa Clara, CA, USA) was incubated for 1 h at room temperature. Anti-cyclophilin A (1:10,000; BML-SA296; Enzo, NY, USA) was used to normalize protein levels. Densitometric analysis of the western blot bands was performed with Image J software. Table 2. Primary antibodies and specific dilutions used in Western Blot analysis. Antibody Description p-AMPK 1:1000; ab80039; Abcam, Cambridge, UK AMPK 1:1000; ab133448; Abcam, Cambridge, UK p-IKBα 1:1000; ab133462; Abcam, Cambridge, UK IKBα 1:5000; ab32518; Abcam, Cambridge, UK NF-κB (p65) 1:1000; sc-8008; Santa Cruz, Dallas, Texas, USA p-AKT 1:1000; #2965; Cell Signaling, Leiden, The Netherlands AKT 1:5000; #9272; Cell Signaling, Leiden, The Netherlands p-GSK3β 1:1000; ab75745; Abcam, Cambridge, UK β-catenin 1:1000; GTX132611; Genetex, California, USA Bcl-xL 1:1000; 610211; BD Biosciences, Sparks, MD, USA VEGF 1:5000; ab46154; Abcam, Cambridge, UK 2.7. Immunohistochemical Analysis Firstly, paraffined sections were deparaffinized in xylene and rehydrated in ethanol. Sections were fixed in acid methanol (−20 ◦ C) for 1 minute and washed with 0.01M 4 phosphate buffered saline (PBS) at pH 7.4. Then, sections were incubated in blocking solution (3% bovine serum albumin (BSA), Tween 0.05% PBS) for 1 h at room temperature and afterwards, they were incubated overnight at 4 ◦ C with specific primary antibodies (Table 3). Table 3. Targets, dilution, and sources of applied primary antibodies in immunofluorescence. Antibody Description GFAP rabbit monoclonal; 1:500; ab7260 (Abcam, Cambridge, UK) Ki67 rabbit polyclonal; 1:500; ab15580 (Abcam, Cambridge, UK) NFκB mouse monoclonal; 1:100; sc-8008 (Santa Cruz, Dallas, Texas, USA) IL-1β rabbit polyclonal; 1:100; ab9722 (Abcam, Cambridge, UK) IL-6 rabbit polyclonal; 1:100; ab7737 (Abcam, Cambridge, UK) TNF-α mouse monoclonal; 1:100; ab8348 (Abcam, Cambridge, UK) Collagen IV rabbit polyclonal; 1:200; ab6586 (Abcam, Cambridge, UK) NLRP3 rabbit polyclonal; 1:200; ab214185 (Abcam, Cambridge, UK) Bcl-xL rabbit polyclonal; 1:100; 610211 (BD Biosciences, Sparks, MD, USA) VEGFA rabbit monoclonal; 1:100; ab52917 (Abcam, Cambridge, UK) GFAP: Glial fibrillary acidic protein; NFkB: Nuclear factor kappa B; IL: Interleukin; TNF-α: Tumor necrosis factor α; NLRP3: NLR family pyrin domain containing 3; Bcl-xL: B-cell lymphoma-extra large; VEGFA: vascular endothelial growth factor A. The following day, after washing, sections were incubated with a fluorescent ALEXA 488 or 594 as a secondary antibody (anti-rabbit or anti-mouse) (Life Technologies S.A, Madrid, Spain) in blocking solution for 1 h and washed again. Nuclei were counterstained using Hoechst 33342 (bisbenzimide), an organic compound that emits blue fluorescence when bound to DNA (Thermo Fisher Scientific, Eugene, OR, USA). Finally, samples were mounted in Mounting Medium Fluorescence (Prolong, Invitrogen, Thermo Fisher Scientific, Eugene, OR, USA) with a coverslip. Images were acquired with a confocal laser scanning microscope (FV1000; Olympus, Hamburg, Germany) at a resolution of 1024 × 1024 pixels. Five fields (three corresponding to the central and two to the peripheral retina) from each section were analyzed using ImageJ software (National Institutes of Health, Bethesda, MD, USA). In order to evaluate glial reactivity, the degree of glial activation was analyzed using a scoring system based on the extent of GFAP (glial fibrillary acidic protein) staining previously described [11]. This scoring system (Table 4) has been used previously by our group [8,9,12]. The fluorescence intensity of the images was quantified by ImageJ. J. Clin. Med. 2019, 8, 339 5 of 20 Table 4. Scoring system to evaluate glial activation. Glial Fibrillary Acidic Description Protein (GFAP) Score 1 Müller cell endfeet region/GCL only 2 Müller cell endfeet region/GCL plus a few proximal processes 3 Müller cell endfeet plus many processes, but not extending to ONL 4 Müller cell endfeet plus processes throughout with some in the ONL 5 Müller cell endfeet plus lots of dark processes from GCL to the outer margin of ONL For apoptosis assessment, the TUNEL (Terminal Transferase dUTP Nick-End Labeling) DeadEnd Fluorometric System kit (PROMEGA, Madison, WI, USA) was carried out using 488 nm as excitation wavelength and the range for detecting positive cells in the confocal laser scanning microscope was 515–565 nm (green). Finally, the number of green positive cells and total cells (in blue) was counted. In order to evaluate cell proliferation, Ki67 positive cells were counted. In addition, co-labeling immunofluorescence combining Ki67 and NeuN (as a neuron specific marker) and CD-31 (as a marker for endothelium of the blood vessels) was performed. To assess the nuclear translocation of NFκB, Z-series stacks of retinal section were captured at 1024 × 1024 pixels and orthogonal projections of z-stack were evaluated using laser confocal microscopy (FluoView ASW 1.4, Olympus, Hamburg, Germany). 2.8. Retinal Vascular Permeability The permeability of retinal vasculature was examined ex vivo by assessing albumin leakage from the blood vessels into the retina using the Evans Blue albumin method. For this purpose, three animals per group were intravenously injected in the tail with a solution of Evans Blue (E2129 SIGMA, Sant Louis, MO, USA) (5 mg/mL dissolved in PBS pH 7.4). After injection, the animals turned blue, confirming dye uptake and distribution. After 2h, the mice were euthanized by cervical dislocation and the eyes were enucleated. The retinas of each animal (n = 3 per group) were isolated, weighed and rapidly protected from light. Flat-mounted slides were obtained, and cover slipped with a drop of the mounting medium Prolong Gold antifade reagent (Invitrogen, Thermo Fisher Scientific, Eugene, OR, USA). Digital images from different random fields of all retinas were acquired using a confocal laser scanning microscope (FV1000; Olympus, Hamburg, Germany) at ×60 using the 561-nm laser line, and each image was recorded with identical beam intensity at a size of 1024 pixels × 1024 pixels. For quantitative analysis of the albumin-bound Evans Blue, the number of extravasations per field of ×60 was counted. We used these images to assess the number of capillary branches and sprouts following a previously reported method [13]. 2.9. Statistical Analysis Comparisons of continuous variables were performed using ANOVA and the Bonferroni Post Hoc test with SPSS software (SPSS/Windows version 18; SPSS, Chicago, IL, USA). The statistical significance level was set at p < 0.05 (*). 3. Results 3.1. Topical Administration of Glucagon-Like Peptide-1 (GLP-1) Ameliorates Electroretinogram Abnormalities Induced by Diabetes The amplitude of the b-wave, which is predominantly produced by Müller and bipolar cells, was significantly lower in diabetic mice in comparison with non-diabetic mice (Figure 1A,B). Treatment with GLP-1 was able to ameliorate these functional abnormalities induced by diabetes. The a-wave, which is derived from photoreceptor function, was also significantly lower in diabetic mice treated with vehicle than in non-diabetic mice at all intensities (Figure 1A,C). The amplitude of a-wave in the diabetic mice treated with GLP-1 was significantly higher than in diabetic mice treated with vehicle J. Clin. Med. 2019, 8, 339 6 of 20 at 10 cd·s·m−2 flash intensity. In addition, GLP-1 treatment ameliorated the amplitude of oscillatory J.potentials (OPs), Clin. Med. 2019, which 8, x FOR reflect PEER REVIEWthe neuronal synaptic activity of amacrine cells and other neurons of 7 of 21 the inner retina (Figure 1D,E). Flash intensity 10 cd·s·m-2 Flash intensity 40 cd·s·m-2 Flash intensity 80 cd·s·m-2 A 200 200 200 150 150 150 Amplitude (µV) Amplitude (µV) Amplitude (µV) 100 100 100 50 50 50 0 0 0 142 189 236 283 1 48 95 133 177 221 265 309 1 45 89 124 165 206 247 288 1 42 83 -50 -50 -50 -100 -100 -100 Time (ms) -150 Time (ms) Time (ms) -150 -150 200 B C a-Wave amplitud (µV) 350 * b-Wave amplitud (µV) 300 ** * 150 * ** * * 250 * * 200 100 * 150 100 50 50 0 0 10 40 80 10 40 80 Flash intensity (cd·s·m-2 ) Flash intensity (cd·s·m-2 ) 40 E Flash intensity 10 cd·s·m-2 D Flash intensity 10 cd·s·m-2 * Amplitude ∑OPs 60 30 Oscillatory potentials OP 2 OP 3 30 OP 4 20 (µV) OP 1 0 (μV) 10 50 80 110 -30 0 Time (ms) db/db vehicle db/db GLP-1 db/+ control -60 40 Flash intensity 40 cd·s·m-2 Amplitude ∑ OPs Flash intensity 40 cd·s·m-2 30 * 60 OP 2 * (µV) Oscillatory potentials OP 3 30 20 OP 4 OP 1 10 (µV) 0 50 80 110 -30 0 db/db vehicle db/db GLP-1 db/+ control Time (ms) -60 30 Flash intensity 80 cd*sec/m2 Flash intensity 80 cd·s·m-2 Amplitude ∑ OPs * Oscillatory potentials 60 * OP 2 OP 3 20 (µV) 30 OP 4 (µV) OP 1 10 0 50 80 110 -30 0 db/db vehicle db/db GLP-1 db/+ control -60 Time (ms) Figure 1. Electroretinogram. (A) Electroretinogram traces in response to low, medium and high Figure 1. Electroretinogram. − (A) Electroretinogram traces in response to low, medium and high stimulus intensities (10 cd·s·m 2 , 40 cd·s·m−2 and 80 cd·s·m−2 ) in a representative non-diabetic mouse stimulus intensities (10 cd·s·m−2, 40 cd·s·m−2 and 80 cd·s·m−2) in a representative non-diabetic mouse (green), a db/db mouse treated with vehicle (red) and a db/db mouse treated with glucagon-like (green), a db/db mouse treated with vehicle (red) and a db/db mouse treated with glucagon-like peptide-1 (GLP-1) eye-drops (blue); (B) Quantitative analyses of b-wave amplitude in db/db mice peptide-1 (GLP-1) eye-drops (blue); (B) Quantitative analyses of b-wave amplitude in db/db mice treated with vehicle (black bars), db/db mice treated with GLP-1 eye drops (grey bars) and non-diabetic treated with vehicle (black bars), db/db mice treated with GLP-1 eye drops (grey bars) and non- mice (white bars). (n = 4 mice per group) * p < 0.05 in comparison with db/db vehicle; ** p < 0.01 in diabetic mice (white bars). (n = 4 mice per group) * p < 0.05 in comparison with db/db vehicle; ** p < comparison with db/db vehicle. (C) Quantitative analyses of a-wave amplitude in db/db mice treated 0.01 in comparison with db/db vehicle. (C) Quantitative analyses of a-wave amplitude in db/db mice with vehicle (black bars), db/db mice treated with GLP-1 eye drops (grey bars) and non-diabetic mice treated with vehicle (black bars), db/db mice treated with GLP-1 eye drops (grey bars) and non- (white bars). (n = 4 mice per group) * p < 0.05 in comparison with db/db vehicle; (D) Oscillatory diabetic mice (white bars). (n = 4 mice per group) * p < 0.05 in comparison with db/db vehicle; (D) potential traces in response to low, medium and high stimulus intensities (10 cd·s·m−2 , 40 cd·s·m−2 Oscillatory potential traces in response to low, medium and high stimulus intensities (10 cd·s·m−2, 40 and 80 cd·s·m−2 ) in a representative non-diabetic mouse (green), a db/db mouse treated with vehicle cd·s·m−2 and 80 cd·s·m−2) in a representative non-diabetic mouse (green), a db/db mouse treated with (red) and a db/db mouse treated with GLP-1 eye-drops (blue); (E) Quantitative analyses of oscillatory vehicle (red) and a db/db mouse treated with GLP-1 eye-drops (blue); (E) Quantitative analyses of potentials at 10 cd·s·m−2 , 40 cd·s·m−2 and 80 cd·s·m−2 in db/db mice treated with vehicle (black bars), oscillatory potentials at 10 cd·s·m−2, 40 cd·s·m−2 and 80 cd·s·m−2 in db/db mice treated with vehicle db/db mice treated with GLP-1 eye drops (grey bars) and non-diabetic mice (white bars). (n = 4 mice (black bars), db/db mice treated with GLP-1 eye drops (grey bars) and non-diabetic mice (white bars). per group) * p < 0.05 in comparison with db/db vehicle. (n = 4 mice per group) * p < 0.05 in comparison with db/db vehicle. 3.2. Neurodegeneration is Inhibited by Glucagon-Like Peptide-1 (GLP-1) Eye-Drops in Diabetic Mice In non-diabetic mice, GFAP expression was confined to the retinal ganglion cell layer (GCL), and therefore, the GFAP score was <2 (Figure 2A,B). By contrast, diabetic mice treated with vehicle J. Clin. Med. 2019, 8, 339 7 of 20 3.2. Neurodegeneration Is Inhibited by Glucagon-Like Peptide-1 (GLP-1) Eye-Drops in Diabetic Mice J. Clin. Med. 2019, 8, x FOR PEER In non-diabetic mice,REVIEW 8 of 21 GFAP expression was confined to the retinal ganglion cell layer (GCL), and therefore, the GFAP score was <2 (Figure 2A,B). By contrast, diabetic mice treated with vehicle presented presentedananaberrant GFAPextent, aberrant GFAP extent,this this being being the the GFAPGFAP scorescore ≥3, which ≥3, which was significantly was significantly higher higher than than in non-diabetic mice (Figure 2A,B). Eye drops containing GLP-1 led to a significant decrease of glial of in non-diabetic mice (Figure 2A,B). Eye drops containing GLP-1 led to a significant decrease glial activation activation (GFAP (GFAP scorescore ≤2) (Figure ≤2) (Figure 2A,B). 2A,B). Glialactivation. Figure2.2.Glial Figure activation.(A) (A)Comparison Comparison of of glial glial fibrillary fibrillaryacidic acidicprotein protein(GFAP) (GFAP)immunoreactivity immunoreactivity (green) among representative samples from diabetic mice treated with vehicle or glucagon-like 20µm (green) among representative samples from diabetic mice treated with vehicle or glucagon-like peptide-1 (GLP-1) eye-drops and from a non-diabetic mouse. Scale bars, 20 µm; (B) Quantification peptide-1 (GLP-1) eye-drops and from a non-diabetic mouse. Scale bars, 20 µm; (B) Quantification of of glial activation based on the extent of GFAP staining. n = 4 mice per group (five retinal sections glial activation based on the extent of GFAP staining. n = 4 mice per group (five retinal sections per per mouse). mouse). Retinal apoptosis, analyzed by TUNEL, was significantly higher in diabetic mice treated with Retinal vehicle apoptosis, than analyzed in non-diabetic mice byinTUNEL, the innerwas significantly (INL) higher in and outer nuclear diabetic layer (ONL) mice treated (Figure with 3A,B). vehicle than in non-diabetic The treatment with GLP-1 led micetoina the inner (INL) significant and of decrease outer nuclearinlayer apoptosis (ONL) (Figure 3A,B). the aforementioned nuclearThe treatment layers (Figure 3A,B) and also in the ratio of cleaved caspase 9/total caspase 9, thus supporting layers with GLP-1 led to a significant decrease of apoptosis in the aforementioned nuclear the (Figure 3A,B) and anti-apoptotic also in effects the ratioinofTUNEL observed cleavedassay caspase 9/total (Figure caspase 9, thus supporting the anti-apoptotic 3E,F). effects observed in TUNEL assay (Figure 3E,F). J. Clin. Med. 2019, 8, 339 8 of 20 Figure 3. (A) Terminal Transferase dUTP Nick-End Labeling (TUNEL) positive immunofluorescence (green) in a representative db/db mouse treated with vehicle, a db/ db mouse treated with glucagon-like peptide-1 (GLP-1) eye-drops and a non-diabetic mouse. Nuclei were labelled with Hoechst stain (blue). Scale bars, 20 µm; (B) Quantification of TUNEL-positive cells in the neuroretina. Results are means expressed as a percentage. n = 4 mice per group. * p < 0.001 in comparison with the other groups; (C) Comparison of hematoxylin-eosin retinas among representative samples from diabetic mice treated with vehicle or GLP-1 eye-drops and from a non-diabetic mouse. Scale bars, 50 µm. n = 3; (D) Quantification of the number of total cells in ganglion cell layer (GCL), inner nuclear layer (INL) and outer nuclear layer (ONL) calculated in hematoxylin-eosin staining. n = 3 mice per groups; * p < 0.05 db/db vehicle in comparison with other groups, ** p < 0.01 db/db vehicle in comparison with other groups; (E) Western blotting bands of cleaved and total caspase 9 and (F) densitometric analyses of retinas from db/db mice treated with vehicle (black bars), db/db mice treated with GLP-1 eye drops (grey bars) and non-diabetic mice (white bars) n = 3 * p < 0.05 db/db vehicle in comparison with the other groups; GCL, ganglion cell layer; INL, inner nuclear layer; IPL, inner plexiform layer; ONL, outer nuclear layer; OPL, outer plexiform layer. J. Clin. Med. 2019, 8, 339 9 of 20 J. Clin. Med. 2019, 8, x FOR PEER REVIEW 10 of 21 3.3. 3.3. Topical Topical AdministrationofofGlucagon-Like Administration Glucagon-Like Peptide-1 Peptide-1 (GLP-1) (GLP-1) Preserves PreservesRetinal RetinalThickness Thicknessand Promotes and Promotes Neurogenesis Neurogenesis AsAsexpected, expected,the thenumbers numbersofoftotal total cells GCL, INL, cells in GCL, INL,and andONL ONLwerewerelower lower inin diabetic diabetic mice mice than than in in non-diabeticmice. non-diabetic mice.The Thenumber number ofof retinal cells cells detected detectedinindiabetic diabeticmice micetreated treatedwith with GLP-1 GLP-1waswas similartotothat similar thatthe theobserved observedininnon-diabetic non-diabetic mice mice (Figure (Figure3C,D). 3C,D).These Theseresults resultssuggest suggest that topical that topical administrationofofGLP-1 administration GLP-1could couldpromote promote neurogenesis. neurogenesis. InInthis thisregard, regard,Ki67, Ki67,a marker a marker of of proliferation, proliferation, was significantly increased in GLP-1 treated diabetic mice in comparison was significantly increased in GLP-1 treated diabetic mice in comparison with the other with the other groups groups (Figure 4A,B). (Figure 4A,B). Figure Figure4. 4.(A) (A)Comparison Comparisonof ofKi67 Ki67 (red) (red) immunofluorescence colabelingwith immunofluorescence colabeling with NeuN NeuN (neuron (neuron specific specific marker) marker) (green) in retinas among representative samples from diabetic mice treated with vehicle or or (green) in retinas among representative samples from diabetic mice treated with vehicle glucagon-like glucagon-likepeptide-1 peptide-1(GLP-1) (GLP-1) eye-drops and from eye-drops and fromaanon-diabetic non-diabeticmouse. mouse. Scale Scale bars, bars, 20 20 µm. n =n4;= 4; µm. (B)(B) Quantification × field. n = 4 mice per groups; * p Quantification of Ki67-positive cells per ×60 field. n = 4 mice per groups; * p < 0.05 db/db GLP-1 in in of Ki67-positive cells per 60 < 0.05 db/db GLP-1 comparison comparison with withother othergroups; groups;GCL, GCL,ganglion cell layer; ganglion cell layer;INL, INL,inner innernuclear nuclearlayer; layer; IPL, IPL, inner inner plexiform plexiform layer; ONL, layer; ONL,outerouternuclear nuclearlayer; layer;OPL, OPL, outer outer plexiform layer. plexiform layer. 3.4. Treatment with Glucagon-Like Peptide-1 (GLP-1) down-Regulates Inflammation Induced by Diabetes 3.4. Treatment with Glucagon-Like Peptide-1 (GLP-1) down-Regulates Inflammation Induced by Diabetes through through the Inhibition of NF-κB, Proinflammatory Cytokines and the NLRP3 Inflammasome Pathway the Inhibition of NF-κB, Proinflammatory Cytokines and the NLRP3 Inflammasome Pathway We found that topical administration of GLP-1 down-regulated the phosphorylation of IKBα We found that topical administration of GLP-1 down-regulated the phosphorylation of IKBα induced by diabetes, thus producing a down-regulation of NF-κB in the retina. The ratio induced by diabetes, thus producing a down-regulation of NF-κB in the retina. The ratio phospho- phospho-IKBα/total-IKBα was significantly lower in retinas from diabetic mice treated with GLP-1 J. Clin. Med. 2019, 8, 339 10 of 20 than in retinas from diabetic mice treated with vehicle, and similar to non-diabetic mice (Figure 5A,B). Consequently, diabetic mice treated with GLP-1 exhibited a significantly lower content of active NF-κB (p65) protein than diabetic mice treated with vehicle and similar to non-diabetic mice (Figure 5A,B). Immunofluorescence analysis also confirmed that the active NF-κB (p65) translocated to the nucleus in diabetic retina. Treatment with GLP-1 also reduced nuclear translocation of NF-κB (p65) in retinal cells (Figure 5C). In addition, we observed that activation of the AMPK signaling pathway was one of the mechanisms involved in NF-κB inhibition by GLP-1. Thus, retinas from db/db mice treated with GLP-1 eye-drops exhibited a higher ratio phospho-AMPK/ total-AMPK than retinas from db/db mice treated with vehicle (Figure 5A,B). Figure 5. Anti-inflammatory effects. (A) Western blotting bands of phospho-AMPK, total AMPK, phospho-IKBα, total IKBα, NF-κB (p65) and (B) densitometric analyses in mouse retinas in db/db mice treated with vehicle (black bars), db/db mice treated with glucagon-like peptide-1 (GLP-1) eye drops (grey bars) and non-diabetic mice (white bars) n = 4 * p < 0.05 db/db vehicle in comparison with other groups, (C) Orthogonal view of confocal images from NF-kB p65 in a representative db/db mouse treated with vehicle, a db/db mouse treated with GLP-1 eye-drops and a non-diabetic mouse by immunofluorescence. Red: NF-kB p65 antibody stained. Blue: Hoechst stained. J. Clin. Med. 2019, 8, 339 11 of 20 J. Clin. Med. 2019, 8, x FOR PEER REVIEW 12 of 21 Asaaresult As resultofofNF-κB NF-κBactivation activationinduced induced byby diabetes, diabetes, retinas retinas fromfrom diabetic diabetic micemice treated treated withwith vehicle showed a significant upregulation of several proinflammatory cytokines vehicle showed a significant upregulation of several proinflammatory cytokines (mRNA and protein) (mRNA and protein) suchasasIL-1β such IL-1βandandIL-6 IL-6(Figure (Figure6A–D). 6A–D). Topical Topical treatment treatment with with GLP-1 GLP-1 dramatically dramatically abrogated abrogated thisthis upregulation,leading upregulation, leadingtotolevels levelslike like the the non-diabetic non-diabetic control control group group (Figure (Figure 6A–D). 6A–D). TNFα TNFα waswasalsoalso increasedininretinas increased retinasfrom fromdiabetic diabeticmice miceinin comparison comparison with with control control mice, mice, andand a lower a lower abundancy abundancy waswas found foundininthe thegroup groupofofmice micetreated treatedwith withGLP-1 GLP-1 (Figure (Figure6E,F). However, 6E,F). However, in both cases in both the the cases differences differences did didnot notreach reachstatistical statisticalsignificance. significance. Figure6.6.(A) Figure (A)Immunofluorescence Immunofluorescence of of IL-1β. IL-1β. Scale Scale bars, 30 µm. µm. Nuclei Nuclei were werelabelled labelledwith withHoechst Hoechststain (blue); stain (B) Real-time (blue); quantitative (B) Real-time quantitativereverse transcription reverse polymerase transcription polymerasechainchain reaction (RT-PCR) reaction analysis (RT-PCR) of IL-1βofin analysis db/db IL-1β mice mice in db/db treated withwith treated vehicle (black vehicle bars), (black db/db bars), db/dbmicemicetreated treated with with glucagon-like glucagon- like peptide-1 peptide-1 (GLP-1) (GLP-1) eye eye drops(grey drops (greybars) bars)and andnon-diabetic non-diabetic mice mice (white (whitebars), bars),nn= =4. 4. * p*