FTY720 Attenuates Retinal Inflammation and Protects Blood–Retinal Barrier in Diabetic Rats

Abstract

Purpose: FTY720 has shown a protective effect in several diseases via inhibiting inflammation and decreasing vascular permeability. The purpose of this study was to assess the impact of FTY720 on inflammation and the blood–retinal barrier (BRB) in diabetic rats.

Methods: Male Wister rats were induced to develop diabetes by streptozotocin, and FTY720 was administered by oral gavage daily for 12 weeks. All experiments were performed at 12 weeks after model establishment. Gene expression was assessed by real-time PCR. Protein expression and/or distribution were assessed by Western blotting and/or immunohistochemistry. The BRB breakdown was determined by staining of retinal whole mounts and quantified using Evans blue.

Results: FTY720 induced lymphopenia in diabetic rats. Proinflammatory cytokines (TNF-α, IL-6, and IL-1β) and adhesion molecules (inter-cellular cell adhesion molecule-1 and vascular cell adhesion molecule-1) were increased in retinas of diabetic rats. FTY720 significantly inhibited the up-regulation of these inflammatory factors. FTY720 also suppressed nuclear factor-κB activation seen in retinas of diabetic rats. Additionally, FTY720 prevented BRB breakdown and reduction of tight junction proteins (ZO-1, Occludin, and Claudin-5) in the retinas of diabetic rats. Down-regulation of S1P1 and S1P3 was also reversed by FTY720 in retinas of diabetic rats.

Conclusions: FTY720 provides protection against diabetic retinopathy (DR), which may involve its anti-inflammatory and barrier-enhancing effects. The S1PR modulation may serve as a novel approach to treat patients with DR.

Diabetic retinopathy (DR) is a major microvascular complication of diabetes mellitus (DM) as a leading cause of blindness in the working-age population worldwide.¹ Microvascular abnormalities are the major early features of DR that consist of progressive thickening of basement membrane, dysfunction, and loss of endothelial cells, pericytes, and vascular smooth muscle.² As such, these abnormalities can result in increased vascular permeability and capillary nonperfusion. Blood–retinal barrier (BRB) breakdown is a hallmark of DR, leading to vascular leakage and development of retinal edema. Diabetic macular edema is a serious complication of microvascular leakage and occlusion,^3,4 which can cause blindness in patients with DR without proper treatment.⁵ 

Inflammatory response plays an important role in early development of DR.^6–9 Both an increase of inflammatory factors and leukocyte adhesion contribute to BRB breakdown. In diabetic patients and experimental diabetic animals, many proinflammatory factors are up-regulated in retinas and vitreous, including TNF-α, IL-6, and IL-1β. Such factors are associated with vascular hyperpermeability. Furthermore, it has been recognized that hyperglycemia damages vasculature via inducing endothelial activation and proinflammatory phenotype of endothelial cells,^10,11 which is characterized by up-regulation of cell surface adhesion molecules such as inter-cellular cell adhesion molecule-1 and vascular cell adhesion molecule-1 (ICAM-1 and VCAM-1). Such molecules facilitate the adhesion and transmigration of leukocyte across endothelium, leading to retinal vascular inflammation, capillary occlusion,^12–14 and endothelial cell death.^15,16 These events finally lead to BRB breakdown,^15,17 associated with the initiation and progression of diabetic microangiopathy. Therefore, an ideal anti-inflammatory treatment for DR should simultaneously target multiple cellular and soluble components. 

FTY720 (Fingolimod) acts as a modulator for four of five known sphingosine 1-phosphate (S1P) receptor subtypes (S1P1, S1P3, S1P4, and S1P5).¹⁸ Numerous preclinical studies have reported the therapeutic efficacy and anti-inflammatory effects of FTY720 in neuroinflammatory diseases, such as experimental autoimmune encephalomyelitis¹⁹ and experimental autoimmune optic neuritis.²⁰ The anti-inflammatory effect of FTY720 is to prevent the egress of lymphocytes from lymph nodes and reduce their homing to target tissues.^21,22 In clinical trials, FTY720 has been shown to be effective in multiple sclerosis²³ and renal transplantation,²⁴ representing a new class of immunomodulatory agents. 

Other than immune cells, S1PRs are also expressed in neural lineages (astrocyte, microglial, oligodendrocyte, neuron)^25,26 and endothelial cells. Growing evidence has shown that FTY720 can penetrate the blood–brain barrier (BBB) and exert regenerative^26–28 and anti-inflammatory effects, which may contribute to the protective effect of FTY720. Of note, S1PRs expressed in endothelial cells participate in maintenance of vascular integrity. Reportedly, FTY720 has a barrier-enhancing effect to prevent vascular leakage via S1PR signaling in acute lung injury^29,30 and cardiac microvascular dysfunction of diabetes.^31,32 

Given the immunomodulatory and barrier-enhancing effect of FTY720, we reasoned that FTY720 may provide protection in DR In this study, we sought to assess the impact of FTY720 on inflammation and BRB damage in streptozotocin (STZ)-induced diabetic rats. 

Materials and Methods

Animals and Induction of Diabetes

Male Wistar rats (12–14 weeks old, 250–300 g) were purchased from the Academy of Military Medical Sciences (Beijing, China). All procedures were performed in accordance with the ARVO statement for the use of animals in ophthalmic and vision research, and the protocol was approved by the Animal Care and Use Committee of Tianjin Medical University. Rats were given a single intravenous injection via tail vein of STZ (50 mg/kg in 5 mM, pH 4.5, citrate buffer; Sigma-Aldrich Corp., St. Louis, MO, USA) to induce diabetes. Establishment of diabetes was confirmed 48 hours after STZ injection by measuring blood glucose levels using a glucometer (Precision PC; Medic, Cambridge, UK). Only rats that developed sustained hyperglycemia exceeding 16.7 mM were included in the study. Body weight and blood glucose of all the rats were measured every 4 weeks. All the rats were euthanized at 12 weeks after model establishment. 

Administration of FTY720

FTY720 (Sigma-Aldrich Corp.) was administered at a dose of 0.3 mg/kg in distilled water by oral gavage daily for 12 weeks after model establishment. This dosing regimen has been reported to effectively inhibit retinal inflammation and restore vascular integrity.³³ To determine the effect of FTY720 on DR (DM+ FTY720 group, n = 40), age-matched normal rats were used as normal controls (control group, n = 40) and diabetic rats were used as model controls (DM group, n = 40). 

Western Blot Analysis

Retinas were homogenized and lysed in Radio Immunoprecipitation Assay buffer with phenylmethanesulfonyl fluoride, and the protein concentration was measured using a standard BCA assay (Pierce, Rockford, IL, USA). Samples were separated by SDS-PAGE and transferred to polyvinylidene fluoride membranes. After blocking with 5% skim milk, the membranes were incubated with specific primary antibodies against TNF-α (1:1000; Gene Tex, Irvine, CA, USA), IL-6 (1:1000; Abcam, Cambridge, MA, USA), IL-1β (1:1000; Gene Tex), nuclear factor (NF)-κB (p65) (1:500; Gene Tex), ICAM-1 (1:1000; Proteintech, Chicago, IL, USA), VCAM-1 (1:1000; Abcam), ZO-1 (1:200; Invitrogen, Camarillo, CA, USA), Occludin (1:250; Abcam), and Claudin-5 (1:500; Invitrogen) overnight at 4°C and subsequently with appropriate secondary antibody for 1 hour at room temperature. Finally, the labeled proteins were detected by the ChemiDoc MP System (BIORAD; Hercules, CA, USA). Band densitometry was analyzed by ImageJ software (http://imagej.nih.gov/ij/; provided in the public domain by the National Institutes of Health, Bethesda, MD, USA), and the gray values of protein bands were calculated by normalization to β-actin (1:1000; Santa Cruz Biotechnology, Dallas, TX, USA). The protein levels presented here are the relative protein expression (/control). 

Quantitative Real-Time PCR

Total RNA was extracted from retinas of each group with Trizol regent (Invitrogen), and cDNA was synthesized using a reverse transcription system kit (TransGen Biotech, Beijing, China) according to the manufacturer's directions. Gene-specific primers used for ICAM-1, VCAM-1, NF-κB (p65), S1P1, S1P3, and β-actin were designed and developed by Sangon Biotech (Shanghai, China). The primers used were GCCTGGGGTTGGAGACTAAC (sense) and CTCGCTCTGGGAACGAATACA (antisense) for ICAM-1, GGAAATGCCACCCTCACCTT (sense) and AACAGTAAATGGTTTCTCTTGAACA (antisense) for VCAM-1, CTGGCCATGGACGATCTGTT (sense) and TGATCTTGATGGTGGGGTGC (antisense) for NF-κB (p65), TTCTGCGGGAAGGAAGTATG (sense) and TGCTGCCGTTGTGTAGTTTC (antisense) for S1P1, GCCACCCGCCAGTCTTG (sense) and GCCAGCTTCCCCACGTAAT (antisense) for S1P3, and CCCATCTTAGAGGGTTACGC (sense) and TTTAATGTCACGCACGATTTC (antisense) for β-actin. The PCR was performed with equal amounts of cDNA in the CFX Connect Real-Time system using the TransStart TOP Green qPCR Super Mix (TransGen Biotech). The PCR conditions were 94°C for 30 minutes, followed by 40 cycles at 95°C for 20 seconds; 57°C for 20 seconds; and 72°C for 20 seconds. We performed three independent experiments on five rats of each group. The relative gene expression was normalized to β-actin and analyzed using the △△CT method. 

Immunohistochemistry

Rats were euthanized, and eyes were enucleated and embedded in paraffin after being fixed in fresh 4% paraformaldehyde with 10% glacial acetic acid. Five-micrometer serial sections were prepared for immunochemical staining. After deparaffinization in xylol, dehydration in graded alcohols, and washing with PBS, the specimens were incubated with 3% hydrogen peroxide to block nonspecific endogenous peroxidase activity. Antigen retrieval was achieved by being subjected to boiling citrated buffer (pH 6.0) for 30 minutes. Subsequently, the sections were incubated with 5% bovine serum albumin for 1 hour at room temperature to block the nonspecific binding. Then the sections were incubated with primary antibodies against ICAM-1 (1:100; Proteintech), VCAM-1 (1:200; Abcam), ZO-1 (1:100; Invitrogen), Occludin (1:100; Abcam), and Claudin-5 (1:100; Invitrogen) overnight at 4°C. After washing with PBS, the sections were incubated with biotinylated secondary antibody (1:200; Vector Laboratories, Burlingame, CA, USA) for 2 hours at 37°C, followed by incubated with the solution of horseradish peroxidease streptavidin (Vector Laboratories) for 1 hour. A diaminobenzidine substrate kit (ZSGB-BIO, Beijing, China) was used for visualization of immunoreactivity and then counterstaining with hematoxylin for cell nucleus. 

Determination of BRB Breakdown Using Evans Blue

Evans blue (EB) leakage was performed as previously described,³³ with minor modification. Briefly, EB (20 mg/mL in saline; Sigma-Aldrich Corp.) was injected through the tail vein at a dose of 45 mg/kg. Rats were euthanized 2 hours later, and the eyes were enucleated and immersed in fresh 4% paraformaldehyde with 10% glacial acetic acid for 2 hours. Then the retinal whole mounts were prepared, and EB leakage was examined by confocal fluorescence microscope (Olympus, Japan). 

The BRB breakdown was quantified according to the procedure described by Xu et al.,³⁴ with minor modifications. Briefly, after the dye was injected through the tail vein and circulated for 2 hours, rats were perfused through the left ventricle. Immediately after perfusion, both eyes were enucleated, and retinas were separated and dried. Each retina was incubated in 120 μL formamide (Sigma-Aldrich Corp.) for 18 hours at 37°C to extract EB, and the supernatant was used to quantify the concentration of extracted dye by spectrophotometer in triplicate at 620 and 740 nm. The concentration of extracted dye was calculated from a standard curve of EB in formamide. Finally, BRB breakdown was calculated as the following equation: 120 μL × ng/μL EB/mg retinal dry weight, with results being expressed as nanograms per milligram. 

Statistical Analysis

All data are presented as mean ± SD. Differences among groups were compared using 1-way ANOVA, followed by the least significant differences test. P < 0.05 was considered statistically significant. 

Results

Effects of FTY720 on Body Weight and Blood Glucose Level

Effects of FTY720 on body weight and blood glucose in rats were studied once every 4 weeks. Body weight of the diabetic group and FTY720-treated diabetic group was significantly lower than that of normal controls (P < 0.05). There was less weight gain in the diabetic group compared with FTY720-treated diabetic rats (P < 0.05; Table). Blood glucose levels were significantly increased in the diabetic group and FTY720-treated diabetic group compared with normal controls throughout the study (P < 0.05). Additionally, there was no significant difference in blood glucose levels between the diabetic group and FTY720-treated diabetic group (P > 0.05; Table), demonstrating that FTY720 has no effect on blood glucose levels. 

Table

 

Changes in Body Weight and Blood Glucose for 4, 8, and 12 Weeks (Mean ± SD, n = 40)

 

FTY720 Induces Lymphopenia

To further verify whether FTY720 could induce lymphopenia in diabetic rats, we analyzed the whole blood samples at 12 weeks by the veterinary hematology system (HEMAVET), and the results (K/μL: 1000/μL) showed that the count of circulating lymphocytes in FTY720-treated diabetic rats was significantly reduced compared with diabetic rats (2.6 ± 0.47 vs. 4.09 ± 0.95; P < 0.05). 

FTY720 Inhibits the Increase of Inflammatory Cytokines in Retinas of Diabetic Rats

The expression of inflammatory cytokines, including TNF-α, IL-6, and IL-1β, was significantly increased in diabetic rats compared with normal controls (P < 0.05). FTY720 treatment significantly reduced their expression compared with diabetic rats (P < 0.05; Fig. 1). 

Figure 1

 

FTY720 inhibits the expression of inflammatory cytokines in retinas of STZ-induced diabetic rats. Tumor necrosis factor-α, IL-6, and IL-1β expression was determined by Western blot analysis (n = 5 per group). Results showed that FTY720 significantly inhibited expression of TNF-α (1.63 ± 0.48 vs. 2.78 ± 0.76), IL-6 (1.87 ± 0.44 vs. 2.97 ± 0.96), and IL-1β (2.85 ± 1.04 vs. 2.97 ± 0.96) in retinas compared with the diabetic group. *P < 0.05 versus control group; ^#P < 0.05 versus DM group.

FTY720 Attenuates the Up-Regulation of Adhesion Molecules in Retinas of Diabetic Rats

The effect of FTY720 on the expression of adhesion molecules was analyzed at both protein and mRNA levels as shown in Figure 2. Immunohistochemical analysis showed that ICAM-1– and VCAM-1–positive staining was much higher in diabetic rats than normal controls and FTY720-treated diabetic rats, and they were mainly in the ganglion cell layer (GCL) and nerve fiber layer (NFL) (Fig. 2A). Western blot analysis showed that expression of ICAM-1 and VCAM-1 was significantly increased in diabetic rats compared with normal controls (P < 0.05), whereas their expression was significantly decreased in FTY720-treated diabetic rats compared with diabetic rats (P < 0.05; Fig. 2B). Consistent with protein expression, the mRNA expression of ICAM-1 and VCAM-1 was also significantly increased in diabetic rats compared with normal controls and this was reversed by FTY720 treatment (P < 0.05; Fig. 2C). 

Figure 2

 

FTY720 attenuates adhesion molecules expression in retinas of STZ-induced diabetic rats. (A) Immunochemistry staining of ICAM-1 and VCAM-1: ICAM-1 and VCAM-1 were abundantly expressed in retinas of diabetic rats, mainly in GCL and NFL. (B) Western blot analysis of ICAM-1 and VCAM-1 (n = 6 per group). FTY720 significantly inhibited expression of ICAM-1 (1.16 ± 0.15 vs. 1.56 ± 0.17) and VCAM-1 (1.34 ± 0.52 vs. 2.26 ± 0.82) compared with diabetic rats. (C) Expression of ICAM-1 and VCAM-1 mRNA determined by RT-PCR (n = 5 per group). FTY720 significantly attenuated the retinal mRNA expression of ICAM-1 (1.01 ± 0.05 vs. 2.23 ± 0.17) and VCAM-1 (1.24 ± 0.28 vs. 2.42 ± 0.33) compared with diabetic rats. *P < 0.05 versus control group; ^#P < 0.05 versus DM group.

Effect of FTY720 on NF-κB Signaling in Retinas of Diabetic Rats

The effect of FTY720 on NF-κB was determined at both protein and mRNA levels as shown in Figure 3. In the diabetic group, the expression of NF-κB at the protein level was significantly increased in diabetic rats compared with normal controls (P < 0.05). Such an increase was attenuated by FTY720 treatment (P < 0.05; Fig. 3A). Similar to the protein level, the mRNA expression of NF-κB was higher than in the control group (P < 0.05), whereas FTY720 significantly inhibited its expression (P < 0.05; Fig. 3B). 

Figure 3

 

FTY720 inhibits NF-κB activation. The expression of NF-κB (p65) in retinas was analyzed using Western blot (A) (n = 5 per group) and RT-PCR (B) (n = 5 per group). FTY720 significantly inhibited NF-κB expression at both protein (1.03 ± 0.2 vs. 1.55 ± 0.18) and mRNA (1.23 ± 0.09 vs. 1.92 ± 0.11) levels compared with diabetic rats. *P < 0.05 versus control group; ^#P < 0.05 versus DM group.

FTY720 Attenuates Retinal Microvascular Permeability

To assess the effects of FTY720 on BRB, we evaluated the microvascular permeability by EB staining. Retinal flat-mounts staining showed diffused EB leakage from retinal vasculature in diabetic rats, whereas FTY720 treatment attenuated the leakage of EB (Fig. 4A). We also quantified the mean of retinal EB leakage expressed as nanograms/milligram retina (Fig. 4B). The mean of diabetic retinas was 30.1 ± 2.3 ng/mg, which was much higher than that of normal controls (10.95 ± 1.95 ng/mg; P < 0.05). FTY720 treatment of diabetic retinas showed a mean of 17.28 ± 2.39 ng/mg, indicating up to 42.57% inhibition of blood–retinal leakage (P < 0.05). 

Figure 4

 

FTY720 attenuates retinal vascular leakage. (A) Microvascular permeability was assessed by EB staining of retinal whole mounts (n = 5 per group). The dye leaked from retinal vascular to the outside in diabetic rats, whereas almost no leakage of EB could be seen in the FTY720-treated diabetic group. (B) Quantification of vascular leakage of EB into the retina (n = 6 per group). The amount of retinal EB leakage was significantly higher in diabetic rats than normal rats, and FTY720 could inhibit the leakage up to 42.57%. *P < 0.05 versus control group; ^#P < 0.05 versus DM group.

FTY720 Prevents the Loss of Tight-Junction Proteins

To analyze the effect of FTY720 on BRB integrity, the expression of tight-junction proteins such as ZO-1, Occludin, and Claudin-5 was detected by both immunohistochemistry and Western blot. As shown in Figure 5A, ZO-1, Occludin, and Claudin-5 were expressed abundantly in normal rats and mainly in the GCL, NFL, outer plexiform layer (OPL), and inner nuclear layer (INL). However, their expression was significantly decreased in retinas of diabetic rats in comparison with normal controls (P < 0.05). This decrease was reversed by FTY720 treatment (P < 0.05; Fig. 5B). Thus, the reduced loss of tight-junction protein expression suggests the protective effect of FTY720 on BRB. 

Figure 5

 

FTY720 increases the expression of tight junction proteins. (A) Immunochemical staining of ZO-1, Occludin, and Claudin-5. In the normal control group, these tight junction proteins were expressed mainly in GCL, NFL, OPL, and INL, whereas their expression was decreased in retinas of diabetic rats. (B) Expression of ZO-1, Occludin, and Claudin-5 was determined using Western blot analysis (n ≥ 5 per group). FTY720 significantly up-regulated expression of ZO-1(0.84 ± 0.05 vs. 0.59 ± 0.1), Occludin (0.82 ± 0.02 vs. 0.63 ± 0.1), and Claudin-5 (0.77 ± 0.1 vs. 0.57 ± 0.06) in retinas of diabetic rats. *P < 0.05 versus control group; ^#P < 0.05 versus DM group.

FTY720 Reverses Down-Regulation of S1P1 and S1P3

To explore whether the effect of FTY720 on DR was related to S1PR modulation, mRNA levels of S1P1 and S1P3 were assessed (Fig. 6). Gene transcription of S1P1 and S1P3 was found in retinas of normal controls. However, their mRNA levels were significantly decreased in retinas of diabetic rats in comparison with normal controls (P < 0.05), which was significantly reversed by FTY720 (P < 0.05). 

Figure 6

 

FTY720 reverses down-regulation of S1P1 and S1P3. The mRNA levels of S1P1 and S1P3 were assessed by RT-PCR (n = 5 per group). FTY720 reversed down-regulation of S1P1 (2.26 ± 0.14 vs. 0.07 ± 0.01) and S1P3 (2.24 ± 0.22 vs. 0.17 ± 0.00) in retinas of the diabetic group. *P < 0.05 versus control group; ^#P < 0.05 versus DM group.

Discussion

We demonstrated a previously unrecognized finding that FTY720 provides protective effects in DR. FTY720 reduces production of inflammatory factors and preserves tight-junction and BRB integrity. These findings indicate that FTY720 might be a promising treatment option in clinical management of DR patients by targeting multiple processes during the pathogenesis of DR. 

T lymphocytes participate in DR. High-level infiltration of T lymphocytes has been found in vitreous and fibrovascular membranes of patients with proliferative diabetic retinopathy,^35–37 and the CD4/CD8 ratio is higher in vitreous compared with peripheral blood.^35,38 T-lymphocyte infiltration is associated with the severity of retinopathy and visual prognosis. Reportedly, FTY720 provides protection in many neurologic diseases involving inflammation, including autoimmune diseases in the central nervous system,^19,20 intracerebral hemorrhage,^39,40 and cerebral ischemia.^41,42 The beneficial effect of FTY720 may result from lymphopenia and attenuated inflammatory cell infiltration. Consistent with these previous findings, we also observed FTY720-induced lymphopenia. Our data support the possibility that FTY720-induced lymphopenia may contribute to the beneficial effect of FTY720 in DR. 

Many proinflammatory factors are up-regulated in retinas of diabetic rats and vitreous of patients with DR.^43,44 In agreement with these findings, we found an increase in TNF-α, IL-6, and IL-1β in DR rats. Tumor necrosis factor-α is a potent mediator of leukostasis in retinal vasculature⁴⁵ and mediates apoptosis of retinal endothelial cells and pericytes, both of which contribute to BRB breakdown.⁴⁶ One known mechanism of TNF-α regulating vascular permeability is through PKCζ/NF-κB signaling to down-regulate both claudin-5 and ZO-1 expression.⁴⁷ Inhibition of TNF-α can suppress NF-κB activation, leukostasis, and BRB breakdown in retinas of diabetic rats.^46,48,49 Interleukin-1β contributes to the increase of vascular permeability via NF-κB activation,⁵⁰ retinal capillary cell apoptosis,⁵¹ and leukocyte adhesion. Inhibition of IL-1β signaling prevents DR progression. The level of IL-6 in diabetic patients is also significantly correlated with severity and progression of DR.⁴³ In the present study, we showed the increase of these proinflammatory factors, together with the loss of tight-junction proteins and increased vascular permeability in DR FTY720, significantly attenuates TNF-α, IL-6, and IL-1β expression and preserves subsequent BRB breakdown. 

In agreement with our results, the levels of adhesion molecules including ICAM-1 and VCAM-1 are increased in both serum and vitreous of patients with DR.^52–54 Experimental and clinical evidence has demonstrated that adhesion of leukocyte to retinal endothelium is increased in diabetic rats (leukostasis).^13,55 Such cells contribute to early DR, including retinal capillary degeneration, pericyte loss, and increased vascular permeability.^56,57 Adhesion molecules expressed on the vascular endothelial cell surface including ICAM-1 and VACM-1 contribute to activation of granulocytes, monocytes/macrophages, and lymphocytes in the damaged site. Additionally, they recruit circulating leukocytes toward injured retina and play an important role in firm attachment and transendothelial migration of leukocytes. In experimental DR models, deletion of leukocyte adhesion molecules CD18 and ICAM-1 can significantly reduce retinal vascular lesions.¹⁶ Apart from induction of lymphopenia, a novel molecular mechanism by which FTY720 exerts anti-inflammatory effect has been reported: suppressing gene transcription of endothelial adhesion molecules and thereby preventing adhesion of immune cells to endothelial cells and subsequent extravasation through activation of S1P3.⁵⁸ Our results showed that FTY720 inhibits ICAM-1 and VCAM-1 expression while promoting gene transcription of S1P3 in diabetic rats. We therefore postulate that FTY720 may suppress leukocyte adhesion and subsequent BRB breakdown via up-regulation of S1P3, contributing to its protection in DR. 

Nuclear factor-κB is a transcription factor in control of the expression of many genes involved in the inflammatory response. Proinflammatory cytokines such as TNF-α and IL-1β have been demonstrated to induce the expression of ICAM-1 and VCAM-1 via NF-κB activation. The observed NF-κB activation in the epiretinal membrane of patients with DR plays an important role in the pathogenesis of DR.^59,60 In the present study, we observed that FTY720 treatment inhibits NF-κB activation in the retinas. We reason that the anti-inflammatory effect of FTY720 may be partially caused by the reduced activation of NF-κB. 

The S1PRs expressed on endothelial cells, especially S1P1 and S1P3,^61–63 are main mediators cooperating in maintenance of endothelial barrier integrity. FTY720, once phosphorylated by sphingosine kinase 2, can act on endothelial S1PRs to decrease vascular permeability and preserve the integrity of biological barriers. In experimental diabetic models, FTY720 has been shown to restore coronary blood flow³¹ and microvascular barrier impairment³² via S1PR modulation. BBB is regulated by S1PRs signaling,⁶⁴ and FTY720 can directly repair BBB under inflammatory conditions by up-regulation of claudin-5 and down-regulation of VCAM-1 in brain microvascular endothelial cells through modulation of S1P1.⁶⁵ However, it remains unknown whether FTY720 can protect BRB. We show that FTY720 can reduce retinal microvascular permeability and preserve BRB integrity. Further, expression of tight-junction proteins including ZO-1, Occludin, and Claudin-5 is up-regulated by FTY720, together with the decreased expression of adhesion molecules. Additionally, we found that gene transcription of S1P1 and S1P3 is suppressed in retinas of diabetic rats, which is reversed by FTY720. Therefore, down-regulation of S1P1 and S1P3 may be an important signal responsible for BRB breakdown in diabetes. The potential protective effect of FTY720 on BRB may involve maintenance of the endothelial barrier via S1P1/S1P3 modulation apart from attenuating inflammation. 

Apart from widely analyzed vasculopathy, DR is also recognized as a neurodegenerative disease of the eye,⁶⁶ which is associated with neuroinflammation.⁶⁷ Retinal microglia are the major component of the innate immune cells in the retina, and they are activated as a proinflammatory phenotype in DR. Activated microglia are considered a major source of proinflammatory and neurotoxic mediators, as well as a potential culprit contributing to early inflammation in DR.⁶⁸ FTY720 has been shown to inhibit microglia activation directly via binding S1P1, down-regulating production of proinflammatory cytokines and neurotrophic factors.⁶⁹ Thus, FTY720 can suppress neuroinflammatory responses in DR, which contributes to reduction of neuronal death. FTY720 also has been shown to protect retinal ganglion cells directly.⁷⁰ Neuronal apoptosis and neuroinflammation in retina can in turn cause BRB breakdown. Above all, FTY720 may directly inhibit neuroinflammation and neuronal apoptosis in DR, which in turn protects BRB. 

In conclusion, we showed that FTY720 provides protection in DR, and S1PR modulation may serve as a novel approach to treat patients with DR. Further studies are warranted to elaborate the operating mechanisms of FTY720 in DR. 

Acknowledgments

Supported by Natural Science Foundation of Tianjin Grants 12 JCYBJC33900 and 14JCYBJC28000 and National Natural Science Foundation of China Grants 81371038 and 91442124. 

Disclosure: L. Fan, None; H. Yan, None