Article Text

Download PDFPDF
Toll-like receptor 3 activation drives the inflammatory response in oxygen-induced retinopathy in rats
  1. Min Cai1,2,
  2. Xuedong Zhang1,2,
  3. Yingyuan Li1,2,
  4. Haiyan Xu1,2
  1. 1Department of Ophthalmology, The First Affiliated Hospital of Chongqing Medical University, Chongqing, People's Republic of China
  2. 2Chongqing Key Laboratory of Ophthalmology, Chongqing Eye Institute, Chongqing, People's Republic of China
  1. Correspondence to Dr Xuedong Zhang, The First Affiliated Hospital of Chongqing Medical University, No. 1 Youyi Road, You Yi Road, Yu Zhong District, Chongqing 400016, People's Republic of China; zxued{at}sina.com

Abstract

Aims Ischaemia is one of the most important causes of blindness. The purpose of this study was to investigate the potential role and mechanisms by which toll-like receptor 3 (TLR3) influences the progression of the inflammatory response in a rat model of oxygen-induced retinopathy (OIR).

Methods OIR rat models were successfully established and received single intravitreal injections of polyinosine-polycytidylic acid (poly (I:C)) and anti-TLR3 antibody, respectively, on postnatal day 17 (P17). Pathological retinal neovascularisation was evaluated by haematoxylin and eosin staining and immunohistochemistry with Isolectin B4 FITC (fluorescein isothyocyanate). Retinal expressions of TLR3 and nuclear factor kappa B (NF-κB) were measured using real-time PCR, immunohistochemistry and western blot. Furthermore, interleukin-6 (IL-6) and tumour necrosis factor α (TNFα) expression levels were assessed with real-time PCR and ELISA.

Results Both gene and protein expression levels of TLR3 and NF-κB were significantly elevated in the retinas of OIR rats compared to the controls. Increased IL-6 and TNFα expression levels were also observed in the retinas of OIR rats. Furthermore, TLR3 signalling pathway components, including NF-κB and IL-6/TNFα, were markedly upregulated upon stimulation with poly(I:C). In addition, the pre-treatment of TLR3 neutralising antibody in OIR models significantly decreased TLR3 and NF-κB expressions, as well as related inflammatory factors IL-6/TNFα expression.

Conclusions Our results suggest that upregulation of the TLR3 signalling pathway is involved in the pro-inflammatory response in OIR rat retinas.

  • Inflammation
  • Retina
  • Experimental &#8211 animal models

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Introduction

Ischaemic retinopathy is a leading cause of blindness worldwide. This condition leads to microvascular degeneration and hypoxia-driven angiogenesis in the retina,1 which is a common pathway in the development of retinopathy of prematurity (ROP), proliferative diabetic retinopathy (PDR), and age-related macular degeneration (AMD). Oxygen-induced retinopathy (OIR), a well-established model for these diseases, is always induced to investigate the pathological changes within the retinal tissue.2 Large numbers of observational and randomised studies have consistently shown that chronic low grade inflammation plays an important role in the development of various ischaemic retinopathies, including ROP, PDR, and AMD,2–4 by releasing various related cytokines and chemokines. However, the exact mechanisms involved in these retinal diseases that initiate and regulate the inflammatory process remain to be explored. Further investigations are needed to elucidate the underlying pathologic mechanisms to identify newer targets for the treatment of these retinopathies.

The toll-like receptors (TLRs) are a group of proteins that exist either on the cell surface or within distinct intracellular compartments. They belong to a family of innate immune system receptors that recognise distinct pathogen-associated molecular patterns (PAMPs). In particular, TLR3 recognises viral-associated double-stranded (ds) RNA5 and endogenous ligands, resulting in the activation of nuclear factor kappa B (NF-κB) and other transcription factors, which in turn stimulate the transcription of inflammatory cytokines and apoptotic genes.6 Over the past decade, increasing evidence has suggested that TLR3 is involved in many inflammatory diseases; TLR3 activation plays a crucial role in rheumatoid arthritis via inducing the secretion of pro-inflammatory cytokines and chemokines.7 Moreover, Wang et al8 demonstrated that TLR3 activation initiates pro-inflammatory signalling pathways leading to airway inflammation. In addition, recent studies have explored the role of TLR3 in retinal innate immunity. Kumar et al9 suggested that the stimulation of retinal pigment epithelial (RPE) cells with dsRNA induces the release of various cytokines, chemokines and adhesion molecules in a TLR3-dependent manner. Similarly, cells cultivated from human epiretinal membranes (ERMs) obtained from eyes with PDR release pro-inflammatory cytokines after polyinosine-polycytidylic acid (poly(I:C)) stimulation.10 These observations led us to explore further the role of TLR3 in the inflammatory responses associated with some ischaemic retinopathies.

In this study, we established a rat model of OIR to explore the potential role of TLR3 and its downstream signalling components in mediating hypoxia-induced ischaemic retinopathy.

Methods

OIR rat model

To induce retinopathy, newborn Sprague Dawley rat pups, together with their nursing mothers, were placed in an airtight polypropylene container equipped with inlet and outlet ports in which hyperoxic conditions (80%±1.3% O2) were alternated with hypoxic conditions (21%±1.5% O2) for 1 day each cyclically from postnatal day 1 (P1) to P14. The pups were then returned to normal atmospheric conditions for 4 days. The control group was maintained in room air. OIR rat pups were euthanased at different time points. These studies adhered to the Association for Research in Vision and Ophthalmology (ARVO) Statement for the Use of Animals in Ophthalmic and Vision Research. The Institutional Animal Ethics Committee of Chongqing Medical University approved all of the animal procedures.

Retinal structure by haematoxylin and eosin staining

Rat pups were killed at different point times (ie, on P14, P16, P18, and P20). Their eyes were enucleated and fixed for 24 h in 4% paraformaldehyde (PFA). The eyes were embedded in paraffin, and 5 μm thick sagittal cross-sections were cut from each eye. Samples were stained with haematoxylin and eosin (H&E) to allow for the identification of preretinal neovascular tuft nuclei. Final images were taken with an upright microscope (Olympus CX31, Tokyo, Japan) under 400× magnification using a digital camera (Sony DSC-H3, Tokyo, Japan).

Retinal vasculature staining

After the rat pups were killed, both eyes were harvested and fixed for 60 min in 4% PFA. Retinas were dissected and permeabilised with 20% goat serum and 0.5% Triton X-100 in phosphate-buffered saline (PBS) for 3 h at room temperature, followed by incubation overnight with 1:200 fluorescein-labelled Isolectin B4 (Vector Labs, Burlingame, California, USA) at 4°C. After washing with PBS, the retinas were flat-mounted onto microscopic slides and photographed using a Nikon Eclipse E600 fluorescent microscope with a 4× objective and a Qimaging Retiga EXi camera. The size of the avascular area and the neovascularisation were measured using Image-Pro Plus software and Adobe Photoshop (Adobe Systems, Inc).

Immunohistochemistry

Deparaffinised retinal sections were subjected to antigen retrieval. Retinas were washed four times for 5 min each with PBS and then treated with goat serum for 30 min at 37°C. Sections were incubated with rabbit anti-TLR3 polyclonal antibody (1:500, Abcam, Cambridge, UK) at 4°C overnight. After washing with PBS, the sections were incubated with goat anti-rabbit secondary antibody conjugated with horseradish peroxidase (HRP)-streptavidin (Beyotime, Shanghai, China) and were then developed with diaminobenzidine (DAB) and counterstained with haematoxylin. Isotype IgG was used as the negative control antibody. The final images were taken with an Olympus BX60 microscope (Olympus Optical Co Ltd, Tokyo, Japan) under 400×magnification.

Administration of TLR ligand and TLR3 neutralising antibody to OIR and normal rats

OIR rat pups and normal rat pups at P17 were used for poly(I:C) and anti-TLR3 neutralising antibody administration experiments. The rat pups were injected intraperitoneally with pentobarbital sodium (35 mg/kg). After mydriasis, 1.5 μg (dilution in 1.5 μL saline) of poly(I:C) (Sigma-Aldrich, St Louis, Missouri, USA) or equal concentration of rabbit anti-TLR3 neutralising antibody (Abcam, Cambridge, UK) was injected into the vitreous cavity of the right eye of each rat pup. The same volume of rabbit non-immune IgG was injected intravitreally into the left eye to serve as a control. The intravitreal injections were delivered 1 mm posterior to the limbus with a 33-gauge needle. Rat pups were killed 24 h after treatment.

Quantitative real-time reverse transcriptase PCR

Total mRNA was extracted from the retinas using Trizol Reagent (Life Technologies, New York, New York, USA) according to the manufacturer's protocol. Appropriate concentrations of RNA obtained from each sample were converted to cDNA using reverse transcriptase (RT) (Takara Biotechnology, Dalian, China) according to the manufacturer's instructions. After reverse transcription, equal quantities of cDNA were subjected to real-time PCR using the SYBR Premix Ex Taq (Takara Biotechnology, Dalian, China) and an ABI 7500 real-time instrument (Applied Biosystems, ABI, Foster City, California, USA) under the following conditions: 1 cycle of initial denaturation at 95°C for 30 s, 40 cycles at 95°C for 3 s followed by 60°C for 30 s, 1 cycle at 95°C for 15 s followed by 60°C for 60 s and 95°C for 15 s. Oligonucleotides for the RT-PCR were synthesised by Life Technologies Corporation (Shanghai); the primers used are shown in table 1. Relative expression levels were calculated using the formula 2−ΔΔCt method.

Table 1

Details of the PCR conditions used

Western blot

Retinas were dissolved with radioimmunoprecipitation assay (RIPA) lysis buffer and mixed with 5X loading buffer (Beyotime, Shanghai, China). Proteins were loaded and separated on an 8–10% polyacrylamide gel and transferred to polyvinylidene fluoride membranes. Membranes were blocked with 5% skim milk at 37°C for 2 h and were then incubated with primary antibodies to TLR3 (Abcam, Cambridge, UK) at a 1:700 dilution, and NF-κB p65 (Santa Cruz Biotechnology, California, USA) at a 1:500 dilution at 4°C overnight. After washing with 0.1% Tween 20 diluted with Tris-buffered saline (TBS) buffer, the membranes were incubated for 60 min with secondary antibodies (Sigma, St Louis, Missouri, USA) at a 1:5000 dilution at 37°C. An anti-β-actin antibody (Abcam, Cambridge, UK) was used as a loading control. Immunoreactivity was visualised using an enhanced chemiluminescence solution and quantified by densitometry analysis (quality one software).

ELISA

Proteins from rat retinas were extracted as mentioned above and the level of interleukin-6 (IL-6) and tumour necrosis factor α (TNFα) was measured by their respective ELISA kit (R&D Systems, Minneapolis, Minnesota, USA) according to the manufacturer's instructions.

Statistical analysis

Statistical analyses of the data were performed using SPSS Software (SPSS V.17.0) (SPSS, Inc, Chicago, Illinois, USA). All data were presented as mean±SD unless otherwise specified. The statistical analysis of differences between the experimental groups was performed using analysis of variance followed by a least significant difference test. A value of p<0.05 was regarded as statistically significant.

Results

Quantification of pathologic neovascular tuft formation

The results of H&E staining showed obvious retinal lesions in the ganglion cell layer (GCL), such as retinal neovascularisation (NV) in both the OIR group and the OIR-poly(I:C) group (figure 1A). The number of preretinal neovascular tufts was significantly higher in the OIR-poly(I:C) group compared with the OIR group (figure 1B, p<0.05). The maximum number of preretinal neovascular tufts was observed at P18. Staining with Isolectin B4 FITC (fluorescein isothiocyanate) also showed the maximum NV quantified at P18, while the maximum vaso-obliteration was quantified at P14 (figure 2A–F). Furthermore, we found that poly(I:C) remarkably increased NV and anti-TLR3 antibody attenuated retinal NV in OIR models compared with controls (figure 2G–L).

Figure 1

H&E stained retinal sections of OIR, OIR-poly(I:C) models and the amount of pathologic neovascular tuft formation. Rat pups were exposed to hyperoxic conditions and were then returned to room air to induce OIR. Equal poly(I:C) (1.5 μg) was injected into one eye of each rat one day before. Eyes from the models were collected at different time points (P14, P16, P18 and P20) and then embedded for H&E staining. (A) Retinal structures were different in OIR, OIR-poly(I:C) and normal retinas at P14, P16, P18 and P20. Original magnification for all panels, 400×. (B) The number of preretinal neovascular tuft nuclei penetrating the internal limiting membrane in OIR-poly(I:C) models was significantly higher than that in OIR controls, with maximum production at P18 compared with controls. Data are shown as mean±SEM. N=12. *p<0.05; **p<0.01. H&E, haematoxylin and eosin; GCL, ganglion cell layer; INL, inner nuclear layer; IPL, inner plexiform layer; OIR, oxygen-induced retinopathy; ONL, outer nuclear layer; P, postnatal day; poly(I:C), polyinosine-polycytidylic acid.

Figure 2

Pathological retinal neovascularisation in OIR rat models. (A–F) Whole mounted retinas from OIR rat pups stained with Isolectin B4 FITC showed the size of the central avascular area, with maximum at P14 compared with controls. The maximum neovascularisation was observed at P18. (G–L) Retinal flat mounts of rat pups injected with poly(I:C) showed significantly more neovascular tufts than controls. OIR models injected with anti-toll-like receptor 3 (TLR3) antibody displayed decreased revascularisation compared to controls. There was no difference in the size of central avascular area among these groups. The white arrows indicate NV staining. Data are shown as mean±SEM. N=18. *p<0.05; **p<0.01. Scale bar: 1000 μm. FITC, fluorescein isothyocyanate; NV, neovascularisation; OIR, oxygen-induced retinopathy; P, postnatal day; poly(I:C), polyinosine-polycytidylic acid.

Significantly upregulated TLR3 and NF-κB expression in OIR rat retinas

We found significant increases in TLR3 mRNA expression in OIR retinas at P14, P16, P18, and P20 (figure 3) and a maximum induction at P18 compared with the normal controls. As shown in figure 4B, C, quantification of western blots revealed a significant increase in the TLR3 protein level in the OIR group. To verify our findings further, immunohistochemistry was used to visualise the localisation of TLR3 expression. TLR3 was mainly expressed in the GCL and the inner nuclear layer (figure 5). Because NF-κB has been implicated as the most common downstream effector in the TLR3 signalling pathway, we further evaluated its protein expression. NF-κB expression was distinctly upregulated in the OIR group after the exposure to hypoxia compared with the control group (figure 4B, D).

Figure 3

Toll-like receptor 3 (TLR3) mRNA expression in the OIR model. Sprague Dawley rat pups were exposed to hyperoxic conditions and then returned to room air to induce OIR. Retinas from the OIR model collected at different time points (P14, P16, P18 and P20) were used for RT-PCR. TLR3 mRNA was significantly increased in the OIR group, with maximum induction at P18. Values are expressed as mean±SD, n=4 experiments. *p<0.05 compared with controls; **p<0.01 compared with controls. OIR, oxygen-induced retinopathy; P, postnatal day; RT, reverse transcriptase.

Figure 4

Toll-like receptor 3 (TLR3) and nuclear factor kappa B (NF-κB) p65 protein expression with poly(I:C) or anti-TLR3 antibody intravitreal injection. TLR3 and NF-κB p65 expression were analysed by RT-PCR and western blot at P18. (A) RT-PCR revealed significant upregulation of TLR3 mRNA expression in the retinas of the OIR- and control-poly(I:C) groups compared with the control group, while reduction of TLR3 mRNA expression was induced by injection of anti-TLR3 antibody. (B–D) Western blots showing significant activation of TLR3 and NF-κB p65 protein expression in the retinas of both the OIR group and the poly(I:C)-stimulated group compared to the control group. However, TLR3 and NF-κB p65 protein expression in the OIR-poly(I:C) group were inhibited by injection of anti-TLR3 antibody. Values are expressed as mean±SD, n=4 experiments. *p<0.05; **p<0.01. β-actin was used as an internal control. OIR, oxygen-induced retinopathy; P, postnatal day; poly(I:C), polyinosine-polycytidylic acid; RT, reverse transcriptase.

Figure 5

Locations of toll-like receptor 3 (TLR3) expression in rat retinas. Immunohistochemical staining showing the locations of TLR3 expression in the retinas of OIR and normal (N) rat pups. The brown granules marked with arrows indicate specific staining (×400); an isotype-matched IgG was used as a negative control. Scale bar: 50 μm. OIR, oxygen-induced retinopathy.

Increased IL-6 and TNFα production in the OIR model in rat retinas

To determine the inflammatory response in OIR rat retinas, we evaluated the gene and protein expression levels of IL-6 and TNFα by quantitative real-time RT-PCR and ELISA, respectively. The gene and protein expressions of IL-6 and TNFα were significantly upregulated in the OIR model compared with normal controls (figure 6, p<0.01 and p<0.05, respectively).

Figure 6

Expression of related inflammatory factors. IL-6 and TNFα expression levels were analysed by RT-PCR and ELISA. (A) IL-6 mRNA expression levels were distinctly elevated in retinas of the OIR- and control-poly(I:C) groups compared with the control groups. However, hypoxia induced IL-6 mRNA expression was reduced by injection of anti-toll-like receptor 3 (TLR3) antibody. (B) TNFα mRNA expression levels were significantly upregulated in the retinas of the OIR- and control-poly(I:C) groups compared to the normal groups. Hypoxia induced TNFα mRNA expression was decreased by injection of anti-TLR3 antibody. (C) IL-6 protein expression levels were distinctly elevated in retinas of the OIR- and OIR-poly(I:C) groups compared with the control groups. (D) TNFα production was significantly upregulated in the retinas of the OIR- and OIR-poly(I:C) groups compared to the normal groups. Data are shown as mean±SD, n=3–5 experiments. *p<0.05; **p<0.01. β-actin was used as an internal control. IL-6, interleukin-6; OIR, oxygen-induced retinopathy; poly(I:C), polyinosine-polycytidylic acid; RT, reverse transcriptase; TNFα, tumour necrosis factor α.

Poly(I:C) increases while anti-TLR3 antibody reduces TLR3 and NF-κB p65 expression in rat retinas

As shown in figure 4A, C, poly(I:C) significantly upregulates TLR3 mRNA and protein expression levels compared to the controls, whereas there were no statistically significant increases in the mRNA or protein levels between the OIR-saline group and the normal-poly(I:C) group. The western blot results revealed enhanced NF-κB p65 expression in the poly(I:C)-stimulated group compared with the controls (figure 4B, D). In addition, poly(I:C), added together with hypoxia, resulted in even higher expressions of TLR3 and NF-κB. However, treatment of anti-TLR3 antibody to rat retinas with OIR significantly reduced these alternations (figure 4).

Poly(I:C) increases while anti-TLR3 antibody alleviates IL-6 and TNFα production in rat retinas

Figure 6 shows that poly(I:C) significantly upregulated the mRNA and protein expressions of IL-6 and TNFα in rat retinas compared with the controls. However, anti-TLR3 antibody reduced IL-6 and TNFα production compared with OIR-poly(I:C) group.

Discussion

In this study, we have shown that hypoxia-induced elevated TLR3 expression might induce NF-κB and inflammatory cytokine expression in the OIR model in rat retinas. We examined the role of TLR3 in the retinas of Sprague Dawley rat pups with OIR and observed that TLR3 expression was markedly increased in the retinas of rats with OIR compared to the controls, which is consistent with the development of pathological retinopathy. These findings indicate that components of the TLR3 signalling pathway play an important role during hypoxia-induced retinopathy.

Recent studies have revealed the correlation between hypoxic stress and TLR signalling mediated innate immunity. Increased expression of TLR3 in OIR rat models may be due to two reasons. First, it is well recognised that hypoxic stress plays a key role in the development of many chronic inflammatory ocular diseases and induces cell death and tissue necrosis.11 Exposure of rat pups to hyperoxia causes retinal dysfunction and retinal cell necrosis, followed by severe vasoconstriction and vaso-obliteration.12 TLR3 in the retina recognises the RNAs released from necrotic cells, and further promotes the inflammatory reaction, thus aggravating the retinopathy process. Secondly, it has become clear that hypoxia-inducible factor-1α (HIF-1α), a key transcriptional factor in hypoxia, plays an important role in the regulation of immunity and inflammation. Under hypoxic conditions, disruption of HIF-1α in Müller cells attenuates the increases of retinal vascular leakage and inflammation.13 Fanlei Hu et al demonstrated that hypoxia and HIF-1α potentiate TLR signalling-induced inflammation in rheumatoid arthritis.14 TLR signalling-induced innate immunity is linked to the hypoxic response through NF-κB in macrophages.15 This could explain the effects of hypoxia on TLR3 signalling-induced increased expressions of NF-κB p65, along with downstream pro-inflammatory cytokines, such as IL-6 and TNFα in the OIR rat model. Nevertheless, the exact mechanism—for example, how hypoxia activates NF-κB downstream signalling components—remains to be further explored.

Increasing evidence has demonstrated that TLR3 could play a crucial role in the underlying immune mechanisms of various inflammatory diseases. Recent studies have shown that TLR3 recognises endogenous RNAs released from damaged cells and plays an important role in pro-inflammatory responses in skin injury16 and acute ischaemic injury.17 TLR3 is also expressed in ocular tissues, and its activation causes inflammation and leads to cell death in both human RPE cells and primary mouse RPE cells. It has also been reported that TLR3 is expressed in the choroidal neovascular membranes from patients with wet AMD.18 Consistent with our results, we also noted increased expression of TLR3 in retinas in the OIR rat model, a well-accepted model for ischaemic retinopathy. Furthermore, our findings showed that hypoxia-induced TLR3 protein expression was mainly in the GCL and the inner nuclear layer of retinas, which chiefly consist of neurones, astrocytes and Müller cells. It is consistent with the recent study by Pan et al19 that TLR3 signalling in astrocytes can be activated by cerebral ischaemic preconditioning. We also found that the inhibition of TLR3 reduced retinal angiogenesis in OIR models, while there was no significant difference in vascular regression in the retinas injected with anti-TLR3 antibody.

In addition, we observed an anticipated phenomenon—namely, the activation of TLR3 by poly(I:C) promotes NF-κB p65 expression and the secretion of inflammatory cytokines in the retinas of both normal and OIR rats. Moreover, there was no statistically significant difference in the level of TLR3 between the OIR-saline group and the normal-poly(I:C) group. It is proposed that both hypoxia and poly(I:C)-induced activation of the TLR3 signalling pathway may play a similar role in the progression of retinopathy in vivo.

The cytokines IL-6 and TNFα are regarded as the functional pro-inflammatory markers for the activation of the TLR3 signalling pathway. TNFα and IL-6 are significantly upregulated in the retina, which correlates with vascular leakage in both animal OIR models20 and patients with diabetic retinopathy.21 In agreement with previous reports, we found that TNFα and IL-6 production was significantly upregulated in OIR rats and controls stimulated with poly(I:C). It is noteworthy that intravitreally injecting TLR3 neutralising antibody could not totally inhibit the production of IL-6 and TNFα; therefore, there is likely to be another pathway involved in the hypoxia-induced inflammatory response.

The main limitation of our current study is that we only investigated the role of TLR3 in vivo. A series of assays in vitro is also essential, and we plan to conduct such studies in the near future.

In conclusion, we have shown that TLR3 is a key player in the development of the retinal inflammatory response based on the result that both hypoxia and poly(I:C) upregulate pro-inflammatory cytokines through a TLR3/NF-κB-dependent pathway in vivo. Our findings suggest that the TLR3 signalling cascade may play an important role in the pathogenesis of ischaemic retinopathy.

Acknowledgments

The authors would like to express their gratitude to Mr Peizeng Yang and Mr Bo Lei for helpful advice and technical assistance.

References

Footnotes

  • Contributors M-C: in charge of the animal experiments, as well as writing the manuscript. X-DZ: designed the experiments and revised the manuscript. Y-YL: performed the animal experiments. H-Y X: specimen collection and statistical analysis.

  • Funding This work was supported by the fund projects of National Natural Science Foundation of China (81371843), the Key Research Project of the Health Bureau of Chongqing (2011-1-029), and the special fund of Chongqing Key Laboratory of Ophthalmology (CSTC).

  • Competing interests None.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Linked Articles

  • At a glance
    Keith Barton James Chodosh Jost Jonas