Inhibition of TLR3-Mediated Proinflammatory Effects by Alkylphosphocholines in Human Retinal Pigment Epithelial Cells

Abstract

Purpose.: To elucidate the role of Toll-like receptor 3 (TLR3) in the pathogenesis of age-related macular degeneration (AMD) and to investigate the effect of alkylphosphocholines (APCs) on the TLR3-mediated expression of cytokines and growth factors in human retinal pigment epithelial (RPE) cells.

Methods.: Confluent cultures of human RPE cells (ARPE-19) were stimulated with poly (I:C) RNA as a well-established ligand for TLR3. Cytokine profiles were determined by RT-PCR on the activation of TLR3. RPE cells were transfected with siRNA specific for TLR3 and RIG-1 to determine the receptors involved. The effect of preincubation of RPE cells with APCs on the expression level of target genes was assessed.

Results.: Poly (I:C) RNA stimulation led to a dose-dependent increase in the expression of TLR3 and RIG-I. A significant increase in expression levels of IL-6, TNF-α, IL-8, MCP-1, ICAM-1, and BFGF was observed after poly (I:C) RNA stimulation (P < 0.05). This effect was time and dose dependent. No effect on PEDG or VEGF expression was seen. Transfection of RPE cells with siRNA specific for TLR3 reduced poly (I:C) RNA-induced mRNA expression of the genes (P < 0.05). Preincubation of RPE cells with APCs significantly reduced the poly (I:C) RNA-induced expression of the target genes (P < 0.05).

Conclusions.: The authors demonstrate that the expression of proinflammatory cytokines and chemokines in RPE cells depends on the activation of TLR3. The induction of downstream gene expression is blocked by siRNA specific for TLR3 and alkylphosphocholines. Therefore, TLR3 should be considered a novel target in AMD therapy.

Innate immunity important for discriminating self- and non–self-antigens consists of the viral receptors Toll-like receptors (TLRs) and the alternative complement pathway. Both systems act as mediators between innate and adaptive immunity and can be activated as response to foreign antigens and in autoimmunity. In particular, TLR3 recognizes the dsRNA of viral origin and polyriboinosinic/polyribocytidylic acid (poly (I:C) RNA), a synthetic analog of viral dsRNA. ¹ TLR3 is expressed on retinal pigment epithelial (RPE) cells, ² which play a central role in local immune defense by acting as antigen-presenting cells and which express a variety of cytokines, chemokines, and growth factors. There is growing evidence of cross-talk between TLRs and complement pathways in that a regulatory role for complement in TLR-induced cytokine responses has been shown. ³ Adding on the potential impact of the expression of viral receptors, RPE cells represent targets for different infectious agents, including viruses. Age-related macular degeneration (AMD) exemplifies the clinical relevance of these findings. Two distinct forms of AMD are known: the “dry” atrophic form, characterized by RPE loss and formation of extracellular deposits called drusen in the central region of the retina, ⁴ and the “wet” neovascular form, characterized by new vessel formation from the choroid ⁵ mediated by an increased level of vascular endothelial growth factor (VEGF) in the aqueous. ⁶ The pathogenesis of AMD is supposed to be defined by chronic inflammatory processes. However, mechanisms initiating and perpetuating inflammation remain to be elucidated. Drusen characteristic for early AMD contain cellular debris, structural proteins of RPE cells, and complement components such as complement factor H (CFH). ⁷ Furthermore, genetic variants of both TLR3 and the key complement regulatory protein CFH have recently been found to confer protection against AMD. ^8,9 Thus, TLR3 and the complement system are considered to be relevant in the pathogenesis of AMD regardless of whether these components of innate immunity are activated by self-antigens or foreign antigens. This study was intended to analyze the expression regulation of TLR3 and dependent cytokines and chemokines in RPE cells. In addition, the role of another viral receptor recognizing viral dsRNA, helicase retinoic acid-inducible gene I (RIG-I), was examined. ¹⁰ The selection of target genes was based on previous evidence for their relevance in acute and chronic inflammatory processes and in the pathogenesis of AMD. Expression of CFH on RPE cells is known to be reduced by the proinflammatory cytokines IL-6 and TNF-α. ¹¹ Furthermore, a systemic increase in the levels of the inflammatory markers CRP and IL-6 is independently associated with the progression of AMD. ¹² Fibronectin fragments, which are associated with the development of degenerative diseases of ocular tissues, upregulate IL-6 and MCP-1 in RPE cells. ¹³ In addition, intraocular concentrations of MCP-1 and ICAM-1 have been found to be elevated in AMD. ¹⁴ For IL-8, a promoter polymorphism has been associated with an increased risk for AMD. ¹⁵ Stimulation of RPE cells with amyloid β (Aβ), another constituent of drusen, upregulates proinflammatory mediators such as IL-8. ¹⁶  

Concomitantly, in this work, a novel therapeutic option in blocking inflammatory responses by alkylphosphocholines (APCs) was identified. APCs are synthetic phospholipid derivatives that act as membrane-targeted drugs influencing signal transduction in proliferative ophthalmic, ^{17 –20} neoplastic, ²¹ and protozoal diseases. ²² They have been demonstrated to inhibit protein kinase C (PKC), mitogen-activated kinase (MAPK/p38), and protein kinase B (PI3K/AKT); lipid raft formation and inhibition of phosphatidylcholine synthesis are discussed as potential cell membrane targets. 

Materials and Methods

Cell Culture

ARPE-19 cells, a human RPE cell line, were purchased from the American Type Culture Collection (Manassas, VA) and were grown in a 1:1 mixture of DMEM and Ham's F12 medium (DMEM/Ham's F12; Biochrome, Cambridge, UK), supplemented with 10% FCS, penicillin (100 U/mL), and streptomycin sulfate (100 μg/mL). The ARPE-19 cell line is well established in AMD research and arose spontaneously from primary RPE cells obtained from a human donor. Experiments were performed in accordance with the Declaration of Helsinki. The cells have a highly epithelial morphology and form the characteristic hexagonal monolayer. Cells were grown at 37°C in a humidified 5% CO₂ atmosphere and were split twice a week when 90% confluence was reached. Cells were obtained at passage 20 and used at passages 22 to 29. Before experimental procedures, the ARPE-19 cells were kept under serum-free conditions for 24 hours. 

Alkylphosphocholines

An APC composed of a carbonyl chain with 22 C atoms bound to phosphocholine was synthesized and kindly donated by Hansjörg Eibl (Max-Planck-Institute for Biophysical Chemistry, Göttingen, Germany). The substance was of analytical grade, as determined by high-performance liquid chromatography. A stock solution of APC in phosphate-buffered saline (PBS; pH 7.2) was prepared under sterile conditions at a concentration of 1 mM and was stored at 4°C. Further dilutions were obtained in PBS, and PBS only served as a control for all subsequent experiments. 

Quantitative Reverse Transcriptase-Polymerase Chain Reaction Analysis

Quantitative RT-PCR analysis was conducted as described. ²³ For quantitative RT-PCR, 2 μg isolated total RNA underwent random primed reverse transcription using a modified Moloney murine leukemia virus reverse transcriptase (Superscript; Life Technologies, Darmstadt, Germany). In parallel, 2-μg aliquots were processed without reverse transcription to control for contaminating genomic DNA. Real-time RT-PCR was performed on a sequence detection system (TaqMan ABI 7700; PE Applied Biosystems, Darmstadt, Germany). GAPDH was used as a reference gene. All water controls were negative for target and housekeeper. Sequences with the following GenBank accession numbers served for the design of the predeveloped Taq Man assay reagents or primers and probe, purchased from Applied Biosystems: NM003265/U88879 Seq AOD: TAGCAGTCATCCAACAGAATCATGAG (human TLR3); NM014314 Seq AOD: GACCATGCAGGTTATTCTGGACTTT (human RIG-I); NM000600 (human IL-6), Hs00174128_m1, NM_000594, NM_000594, T: ATGTTGTAGCAAACCCTCAAGCTGA (human TNF-α), Z11686 (human IL-8); NM002982.3/M24545.1 Seq AOD: TCAGCCAGATGCAATCAATGCCCCA (human MCP-1), NM_000201 (human ICAM-1), AGCCAGGTAACGGTTAGCACACACTCCTT (human BFGF), NM_002615.4 (human PEDG), NM_03376 (human VEGF), and M33197 (human GAPDH). 

Knockdown of Gene Expression with Short-Interfering RNA

Predesigned short-interfering RNA (siRNA) specific for TLR3 and RIG-I were purchased from Ambion (Tokyo, Japan). Transfection of siRNA into the cells was performed using transfected agent (siPORT-NeoFX; Ambion), as previously described. ²⁴ Scrambled siRNA was used as the nonspecific negative control of siRNA (Ambion). Cells were pretreated with siRNA for 24 hours and then washed once using cell culture medium to remove remaining transfection agent. Poly (I:C) RNA was mixed into the cell culture medium as indicated. 

Cell Death Assay

For analyses of the effect of APC and poly (I:C) RNA on the apoptosis of RPE cells, cells were incubated with culture medium with or without APC (2 μM, 4 μM) for 24 hours and subsequently incubated with poly (I:C) RNA (5 μg/mL) alone or in combination with APC (2 μM, 4 μM) for 12 hours. The amount of dead cells was determined (Cell Death Detection ELISA kit; Roche, Basel, Switzerland) in accordance with the protocol provided by the company. 

Statistical Analysis

Values are provided as mean ± SD. Statistical analysis was performed by unpaired t-test if applicable or by ANOVA. Significant differences are indicated for P < 0.05 and P < 0.01, respectively. 

Results

Effect of poly (I:C) RNA on the Expression of the Viral Receptors TLR3 and RIG-I on Human RPE Cells

First we tested the expression of the viral receptors TLR3 and RIG-I on human RPE cells under basal and proinflammatory conditions. To simulate a proinflammatory milieu in vitro, the cytokines IFN-γ, TNF-α, and IL-1β were chosen. RPE cells were incubated without and with the cytokines IFN-γ (20 ng/mL), TNF-α (25 ng/mL), and IL-1β (10 ng/mL) alone or in combination for 24 hours. Expression of TLR3 and RIG-I was analyzed by real-time RT-PCR. RPE cells showed a basal expression of TLR3 that was not significantly influenced by any of the cytokines alone. The cytokine combination led to a strong increase in TLR3 expression (Fig. 1A). Basal expression of RIG-I was significantly increased by IFN-γ, TNF-α, and the combination of the cytokines, with the latter having the most pronounced effect (Fig. 1B). Given that RPE cells exhibit a basal expression of the viral receptors TLR3 and RIG-I and poly (I:C) RNA can act as a ligand for these receptors, we tested whether the expression of TLR3 and RIG-I could be influenced by poly (I:C) RNA. RPE cells were stimulated with the synthetic analog of viral RNA, poly (I:C) RNA (10 μg/mL), for different time intervals (3, 6, 9, 12, 24 hours) RT-PCR showed a basal expression of the viral receptors TLR3 and RIG-I that did not change significantly over incubation time. TLR3 expression was increased on stimulation of RPE cells with poly (I:C) RNA for 6 hours and a maximal expression level reached after 24 hours of poly (I:C) RNA stimulation (Fig. 1C). RPE cells showed a comparable increase in RIG-I expression after 6, 9, 12, and 24 hours of poly (I:C) RNA stimulation (Fig. 1D). Next, RPE cells were stimulated without and with different concentrations of poly (I:C) RNA (0.5, 5, 10 μg/mL) for 12 hours, and the expression of TLR3 and RIG-I was analyzed by RT-PCR. Poly (I:C) RNA stimulation led to a dose-dependent increase in the expression of TLR3 and RIG-I, with a maximum under 10 μg/mL poly (I:C) RNA stimulation (Figs. 1E, 1F). 

Figure 1.

 

Effect of poly (I:C) RNA on the expression of the viral receptors TLR3 and RIG-I in human RPE cells. RPE cells were cultivated under basal conditions (basal) or stimulated with the proinflammatory cytokines IFN-γ (20 ng/mL), TNF-α (25 ng/mL), or IL-1β (10 ng/mL) alone or in combination (comb) for 24 hours. Expression of TLR3 and RIG-I was analyzed by RT-PCR. (A) Basal expression of TLR3 was not significantly influenced by IFN-γ, TNF-α, or IL-1β alone but was strongly increased by the combination of these cytokines. (B) Basal expression of RIG-I was significantly increased by IFN-γ and TNF-α stimulation as well as the combination of the cytokines, which led to the strongest increase in RIG-I expression. RPE cells were stimulated without (basal) and with poly (I:C) RNA (10 μg/mL) for different time intervals (3, 6, 9, 12, 24 hours), and the expression of TLR3 and RIG-I was analyzed by RT-PCR. RPE cells showed a basal expression of TLR3 that was increased by poly (I:C) RNA stimulation after 6, 9, 12, and 24 hours, with a maximum at 24 hours of poly (I:C) RNA stimulation (C). RPE cells showed a comparable increase in RIG-I expression after 6, 9, 12, and 24 hours of poly (I:C) RNA stimulation, and no increase in RIG-I expression was observed after 3 hours of poly (I:C) RNA stimulation (D). Next RPE cells were stimulated without (basal) and with different concentrations of poly (I:C) RNA (0.5, 5, 10, μg/mL) for 12 hours, and the expression of TLR3 and RIG-I was analyzed by RT-PCR. Poly (I:C) RNA stimulation led to a dose-dependent increase in the expression of TLR3 (E) and RIG-I (F), with a maximum under 10 μg/mL poly (I:C) RNA stimulation. Results are mean ± SD of three incubations for each condition. rRNA served as the reference gene. Comparable results were obtained in two series of independent experiments. *P < 0.05, **P < 0.01.

Effect of poly (I:C) RNA on Cytokine and Chemokine mRNA Levels on Human RPE Cells

Next we tested the effect of poly (I:C) RNA stimulation on mRNA levels of selected cytokines and chemokines. RPE cells were stimulated with poly (I:C) RNA (10 μg/mL) for different time intervals (3, 9, 12, 24 hours) and with different concentrations of poly (I:C) RNA (0.5, 5, 10 μg/mL) for 12 hours, and the expression of IL-6, TNF-α, IL-8, and MCP-1 was analyzed by RT-PCR. Poly (I:C) RNA stimulation led to a time-dependent increase in mRNA expression of the genes, reaching a maximum after 12 hours; expression levels of IL-6 and MCP-1 were still significantly elevated after 24 hours of poly (I:C) stimulation. Stimulation with different concentrations of poly (I:C) RNA (0.5, 5, 10 μg/mL) for 12 hours led to a dose-dependent increase in IL-6, TNF-α, IL-8, and MCP-1 expression with a maximum under 10 μg/mL poly (I:C) RNA stimulation (Fig. 2). 

Figure 2.

 

Effect of poly (I:C) RNA on cytokine and chemokine mRNA levels in human RPE cells. RPE cells were stimulated without (basal) and with poly (I:C) RNA (10 μg/mL) for different time intervals (3, 9, 12, 24 hours), and the expression of IL-6 (A), TNF-α (C), IL-8 (E), and MCP-1 (G) was analyzed by RT-PCR. Poly (I:C) RNA stimulation led to a time-dependent increase in mRNA expression of these genes. After 24 hours of poly (I:C) RNA stimulation, gene expression decreased again but was still significantly elevated for IL-6 and MCP-1. Stimulation with different concentrations (0.5, 5, 10 μg/mL) of poly (I:C) RNA for 12 hours led to a dose-dependent increase in IL-6 (B), TNF-α (D), IL-8 (F), and MCP-1 (H) expression, with a maximum under 10 μg/mL poly (I:C) RNA stimulation. Results are mean ± SD of three incubations for each condition. rRNA served as the reference gene. Comparable results were obtained in two series of independent experiments. *P < 0.05, **P < 0.01.

Effect of poly (I:C) RNA on mRNA Levels of ICAM-1, BFGF, PEDG, and VEGF on Human RPE Cells

Furthermore, the expression of mediators important for the adhesion of inflammatory cells, fibrosis, and vascular growth were analyzed. RPE cells were stimulated with poly (I:C) RNA (10 μg/mL) for different time intervals (3, 9, 12, 24 hours) and with different concentrations of poly (I:C) RNA (0.5, 5, 10 μg/mL) for 12 hours, and the expression of ICAM-1, BFGF, PEDG, and VEGF was analyzed by RT-PCR. ICAM-1 expression was increased by poly (I:C) RNA after 3, 9, 12, and 24 hours of stimulation time; the maximal increase was observed after 12 hours of stimulation. BFGF expression was increased by poly (I:C) RNA after 9, 12 and 24 hours, with a maximum after 12 hours. Poly (I:C) RNA had no effect on the basal expression of PEDG or VEGF. Stimulation with different concentrations of poly (I:C) RNA (0.5, 5, 10 μg/mL) for 12 hours led to a dose-dependent increase in ICAM-1 and BFGF expression, with a maximum under 10 μg/mL poly (I:C) RNA stimulation. Because we did not observe an increase in PEDG and VEGF on stimulation with poly (I:C) RNA, we did not test expression levels of these genes with different concentrations of poly (I:C) RNA (Fig. 3). 

Figure 3.

 

Effect of poly (I:C) RNA on mRNA levels of ICAM-1, BFGF, PEDG, and VEGF in human RPE cells. RPE cells were stimulated without (basal) and with poly (I:C) RNA (10 μg/mL) for different time intervals (3, 9, 12, 24 hours), and expression of ICAM-1 (A), BFGF (C), PEDG (E), and VEGF (F) was analyzed by RT-PCR. ICAM-1 expression was increased by poly (I:C) RNA after 3, 9, 12, and 24 hours of stimulation time. The maximum increase was observed after 12 hours of stimulation (A). BFGF expression was increased by poly (I:C) RNA after 9, 12, and 24 hours, with a maximum after 12 hours of stimulation time (C). Poly (I:C) RNA had no effect on the basal expression of PEDG (E) and VEGF (F). Stimulation with different concentrations (0.5, 5, 10 μg/mL) of poly (I:C) RNA for 12 hours led to a dose-dependent increase in ICAM-1 (B) and BFGF (D) expression, with a maximum under 10 μg/mL poly (I:C) RNA stimulation. Results are mean ± SD of three incubations for each condition. rRNA served as the reference gene. Comparable results were obtained in two series of independent experiments. *P < 0.05, **P < 0.01.

Effect of Transfection with siRNA Specific for TLR3 and RIG-I on Poly (I:C) RNA Induced Changes in the Expression of Cytokines, Chemokines, Adhesion Molecules, and Factors Important for Fibrosis

To identify the viral receptor responsible for poly (I:C)-induced changes in expression of the target genes, RPE was transfected with siRNA specific for TLR3, RIG-I, and scrambled RNA as negative control for 24 hours and was stimulated with poly (I:C) RNA (10 μg/mL) for 12 hours. Expression of IL-6, TNF-α, IL-8, and MCP-1 as well as with ICAM-1 and BFGF was analyzed by RT-PCR. The significant increase in the expression of all these genes induced by poly (I:C) RNA stimulation was reduced after the transfection of RPE cells with siRNA specific for TLR3. siRNA specific for RIG-I and negative controls containing unspecific RNA had no effect on poly (I:C) RNA-induced gene expression (Fig. 4). 

Figure 4.

 

Effect of transfection with siRNA specific for TLR3 and RIG-I on poly (I:C) RNA-induced changes in the expression of cytokines, chemokines, adhesion molecules, and factors important for fibrosis. RPE cells were transfected with siRNA specific for TLR3 and RIG-I and for scrambled RNA as a negative control for 24 hours, as described in Materials and Methods. RPE cells were stimulated with poly (I:C) RNA (10 μg/mL) for 12 hours, and the expression of IL-6 (A), TNF-α (B), IL-8 (C), MCP-1 (D), ICAM-1 (E), and BFGF (F) was analyzed by RT-PCR. Poly (I:C) RNA stimulation led to a significant increase in the expression of these genes. Transfection of RPE cells with siRNA specific for TLR3 reduced poly (I:C) RNA-induced mRNA expression of these genes. siRNA specific for RIG-I and negative controls containing unspecific RNA had no effect on poly (I:C) RNA-induced gene expression. Results are mean ± SD of three incubations for each condition. rRNA served as the reference gene. Comparable results were obtained in two series of independent experiments. *P < 0.05, **P < 0.01.

Effect of APC on Poly (I:C) RNA Induced Changes in the Expression of Viral Receptors, Cytokines and Chemokines, Adhesion Molecules, and Factors Important for Fibrosis

RPE cells were pretreated without and with APC in different concentrations (2 μM, 4 μM) for 24 hours, washed with PBS, and stimulated with different concentrations of poly (I:C) RNA (0.5, 5, 10 μg/mL) alone or in combination with APCs in different concentrations (2 μM, 4 μM) for 12 hours. The structural formula of APCs is shown in Figure 5. Expression of the viral receptors TLR3 and RIG-I, the selected cytokines and chemokines IL-6, TNF-α, IL-8, and MCP-1, and of ICAM-1 and BFGF was analyzed by RT-PCR. RPE cells showed a basal expression of these genes that was not influenced by APCs in the two concentrations tested. Poly (I:C) RNA stimulation led to a dose-dependent increase in the expression of all these genes that was significantly reduced by APCs. The effect of APCs was comparable for the two concentrations tested. APC had no toxic effect on cell viability tested by a cell death assay, as described in Materials and Methods (Fig. 6). 

Figure 5.

 

Chemical structure of the APC oleyl-phosphocholine (C18:1-PC).

Figure 6.

 

Effect of APC on poly (I:C) RNA-induced changes in the expression of viral receptors, cytokines and chemokines, adhesion molecules, and factors important for fibrosis. RPE cells were pretreated without and with APCs in different concentrations (2 μM, 4 μM) for 24 hours, washed with PBS, and stimulated with different concentrations of poly (I:C) RNA (0.5, 5, 10 μg/mL), alone or in combination with APC (2 μM, 4 μM) for 12 hours. Poly (I:C) RNA stimulation led to a dose-dependent increase in the expression of TLR3 (A) and RIG-I (B), IL-6 (C), TNF-α (D), IL-8 (E), MCP-1 (F), ICAM-1 (G), and BFGF (H). RNA expression was analyzed by RT-PCR. RPE cells showed a basal expression of these genes that was not influenced by APC. APCs significantly reduced poly (I:C) RNA-induced increase in expression of the genes mentioned. APCs had no toxic effect on cell viability tested by cell death assay, as described in Materials and Methods (I). Results are mean ± SD of three incubations for each condition. rRNA served as the reference gene. Comparable results were obtained in two series of independent experiments. *P < 0.05, **P < 0.01.

Discussion

Based on the emerging consensus of local inflammation conferring a substantial risk for the development of AMD, we investigated TLR3 as a potential target in this degenerative disease process. Because genotype analyses point to a particular relevance of components of innate immunity including TLR3, we aimed at defining TLR3-dependent mediators in RPE considered to be crucial modulators of retinal immune processes. 

By showing an increase in TLR3 expression levels on stimulation with proinflammatory cytokines and dose- and time-dependent upregulation of TLR3 by the ligand poly (I:C) RNA, we substantiated for the first time the functionality of this viral receptor on human RPE cells. Corresponding results are found for the viral receptor RIG-I. Mediators, adhesion molecules, and growth factors known to be TLR3 dependent and relevant in inflammation and fibrosis were analyzed. The cytokines IL-6, IL-8, and TNF-α, the chemokine MCP-1, the adhesion molecule ICAM-1, and the growth factor BFGF were shown to exhibit time- and dose-dependent responses to the activation of viral receptors with poly (I:C). Expression levels of PEDF and VEGF were unaffected. Observation of an early induction of proinflammatory mediators peaking at 12 hours corresponded with the results of in vivo data from the TLR3 knockout mouse with increased apoptosis and loss of RPE cells as early as 48 hours after the injection of poly (I:C) in the vitreous. ⁸ Knockdown experiments with siRNA specific for the viral receptors have provided evidence for these effects being mediated predominantly by TLR3. 

Dysregulation of the complement system is considered to be pivotal in the pathogenesis of AMD as the best-characterized clinical example of a progressively degenerative disorder. Factors other than drusen are found to contain components of the complement system such as CFH (complement factor H), which binds to polyanionic structures and inactivates complement C3b, thereby controlling local inflammatory processes. This is estimated to be of general relevance for local inflammation because both IL-6 and TNF-α have been shown to reduce CFH expression in RPE cells. ¹¹ Furthermore, local accumulation of protein degradation products, such as Aβ and fibronectin fragments, are both known to induce IL-8. In RPE cells, IL-6 and MCP-1 can perpetuate inflammation in advanced degenerative disease. ^13,16 It is thus reasonable to assume that the proinflammatory cytokines IL-6, IL-8, and TNF-α and the chemokine MCP-1 play a role in the development and progression of chronic inflammation. This might be a simple amplification of local inflammatory responses or their interference with complement regulation. The adjuvant effect of BFGF in local glial proliferation and matrix production leading to focal subretinal fibrosis is evident. Given that the regulation of all these mediators is dependent on and thereby converges to TLR3 in RPE cells, we infer it to play a key role in pathogenesis of local inflammatory processes, ultimately leading to retinal degeneration. We can only speculate about the relative role of nuclear acid fragments originating from cell debris and viral RNA originating from RNA viruses or generated in the course of DNA virus replication ²⁵ as a trigger for the imbalance of the local immune system favoring inflammation by the activation of TLR3. 

APCs may be of therapeutic interest for AMD because they broadly inhibit TLR3-mediated upregulation of proinflammatory and profibrotic mediators in RPE. However, autoimmunity and viral infections represent only some of the many causes considered risk factors for AMD development. Additional studies are warranted to further define the mode of action possibly involving lipid raft organization and receptor cycling between intracellular compartments and the cell membrane. In addition, interference of APCs with the complex structure of TLR3, ²⁶ necessitating a correct positioning of the N- and C-terminal ends, is conceivable. 

Footnotes

 Supported by Freunde und Förderer der Augenklinik der Ludwig-Maximilians-Universität München (KE); Else Kröner-Fresenius-Stiftung, Fritz-Bender-Stiftung, Deutsche Vereinigung zur Bekämpfung von Viruskrankheiten (MW).

Footnotes

 Disclosure: M. Wörnle, None; M. Merkle, None; A. Wolf, None; A. Ribeiro, None; S. Himmelein, None; M. Kernt, None; A. Kampik, None; K.H. Eibl-Lindner, None

The authors thank Katja Obholzer for expert technical assistance.