Pigment Epithelial Growth Factor Suppresses Inflammation by Modulating Macrophage Activation

Abstract

purpose. To study the contribution of murine retinal pigment epithelial (RPE) cells to the innate immune-privilege status of the subretinal space as determined by the ability of pigment epithelial–derived factor (PEDF) and somatostatin (SOM), produced by RPE, to regulate macrophage-mediated inflammation.

methods. Serum-free medium was added to RPE eyecups (a healthy monolayer of RPE resting on choroid and sclera) and the supernatants were removed after 24 hours (RPE SN). The RPE SN was assayed for the presence of PEDF and SOM and for its ability to regulate interleukin (IL)-12, IL-10, and nitric oxide (NO) production by resting and activated macrophages. A group of mice received intradermal injection of lipopolysaccharide (LPS) and PEDF in one ear and LPS alone in the other ear. Ear thickness was measured before- and 24 hours after ear injections.

results. Soluble factors present in the RPE SN inhibited IL-12 production and substantially increased IL-10 while having minimal effects on NO production by activated macrophages. The message for PEDF, SOM, and IL-10 was detected in RPE cells, and the protein for these factors was found in the RPE SN. The stimulation of IL-10 and suppression of IL-12 production by RPE-SN–treated macrophages was neutralized by anti-PEDF antibodies. Neutralization of SOM in the RPE SN, suppressed NO production by activated macrophages. Intradermal injection of PEDF substantially inhibited LPS-induced inflammatory response.

conclusions. PEDF inhibits LPS-driven macrophage activation in vitro and in vivo. By producing PEDF, the RPE contributes to innate immune privilege of the eye.

Immune privilege is an evolutionary adaptation of the eye to protect itself from the ravages of excessive inflammation. In the eye, immune privilege suppresses the induction of immunogenic inflammation protecting ocular cells that are incapable of regeneration from the damaging effect of inflammation. 

Adaptive immune privilege is well documented in all ocular compartments ^{1 2} and is characterized by prolonged survival of allogeneic grafts, inhibition of the proinflammatory Th1 pathway and downregulation of complement-fixing antibodies, while at the same time, specifically produced cytotoxic T cells and noncomplement-fixing antibodies protect the eye from pathogenic damage. ¹ Existence of immune privilege to innate cells (innate immune privilege) has also been demonstrated recently in the anterior segment of the eye: Apte and Niederkorn ³ showed that even though corneal endothelial cells are vulnerable to lysis by NK cells, aqueous humor inhibited their NK-mediated cytotoxicity; Chen et al. ⁴ demonstrated that components of aqueous humor strongly inhibit CD95L-dependent activation of neutrophils; both TGF-β₂ and α-MSH constituents of aqueous humor inhibit neutrophil-mediated killing of corneal endothelial cells ⁵ ; and suppression of NO production by macrophages has been demonstrated by calcitonin gene-related peptide in the aqueous humor ⁶ ; and α-melanocyte stimulating hormone α-MSH in aqueous humor has been found to inhibit LPS-stimulated Toll-like receptor-4 signaling in macrophages. ⁷  

Although suppression of adaptive immunity in the subretinal space has been demonstrated previously, we now address whether the subretinal space suppresses innate immunity by investigating the effects of RPE eyecup supernatants (RPE SN) on the inflammatory activity on macrophages. The RPE are a monolayer of pigmented neuroepithelial cells that line the outermost boundary of the retina. The RPE monolayer is strategically placed, not only to act as the outer blood–retinal barrier, but also as an immunologic barrier by expressing cell surface molecules and secreting soluble mediators that influence the immune system. ⁸  

To evaluate the immunomodulatory properties of subretinal space, we used RPE eyecup cultures in which the RPE monolayer remains intact, resting on the choroid and the sclera. ⁹ We used this technique to demonstrate that the RPE produces TGF-β and thrombospondin to suppress Th1 cell activation and to promote systemic tolerance to antigens placed into the ocular microenvironment. ¹⁰  

Besides TGF-β and thrombospondin and their suppressive activity on adaptive immunity, the RPE may also produce other immunosuppressive factors that affect innate immune activity. Two of these potential factors are pigment epithelium-derived factor (PEDF) and somatostatin (SOM). Recently, PEDF, secreted by the RPE into the interphotoreceptor matrix (subretinal space), has been shown to inhibit proliferation of innate immune cell such as macrophages. ¹¹ PEDF is a 50-kDa protein member of serine protease inhibitor family and is found in the RPE, ciliary body, and parts of the cornea and retina by immunofluorescence and Western blot analysis. ¹² PEDF is a potent antiangiogenic factor and can inhibit the growth of blood vessels in the eye. ^{13 14} Both choroidal neovascularization and diabetic vitreoretinopathy have been associated with low levels of PEDF in the eye. ^{15 16} Intravitreal injection of PEDF has also shown in a diabetic rat model to cause a decrease vascular permeability that correlated with reduction in levels of VEGF and its receptor, together with reduction in proinflammatory cytokines such as MCP-1, TNF-α, and ICAM-1. ¹⁷  

As well as being major mediators of inflammation, macrophages have been shown to be both anti- ¹⁸ and proangiogenic. ^{19 20} Thus, it is important to know the influence of PEDF on behavior of these innate immune cells. SOM, a neuropeptide present in aqueous humor, was reported to mediate the induction of regulatory T cells via production of α-MSH. ²¹ Transcripts for SOM and its receptors have been reported to be present within the human retina and RPE. ²²  

We examined whether RPE can release both PEDF and SOM and whether these factors contribute to the suppression of innate immunity in the subretinal space. 

Methods

Mice

Adult male C57BL/6 mice aged 6 to 8 weeks were obtained from Taconic (Germantown, NY). All experimental procedures conformed to the ARVO Statement for the Use of Animals for Ophthalmic and Vision Research, and were approved by the Schepens Institutional Animal Care and Use Committee. 

Preparation of RPE Eyecups

RPE eyecups were prepared as described previously. ⁹ Briefly, enucleated eyes of C57BL/6 mice were placed in Ca²⁺-Mg²⁺–free Hanks’ balanced sale solution (HBSS) on ice for 30 minutes. Subsequently, the anterior segment of the eye including the cornea, iris, ciliary body, and lens were excised with microscissors. The remaining tissue was placed in 0.01 U/mL of chondroitinase ABC (Sigma-Aldrich) for 30 minutes at 37°C, then placed on ice and washed three times in HBSS. The neural retina was gently lifted off the RPE layer with microsurgical forceps. Posterior eyecups consisting of sclera, choroid, and a healthy monolayer of RPE were placed in individual wells of a microculture plate (S plate; Nalge Nunc International, Naperville, IL) for 24 hours, diluted 1:10 and used for further experiments. 

Cell Cultures

A mouse monocytic leukemia cell line, RAW 264.7, was obtained from ATCC (Manassas, VA) and grown in complete DMEM. In all experiments, 5 × 10⁵ cells were incubated in each well of a 96-well plate (Fisher Scientific, Pittsburgh, PA) in serum-free medium, with or without the addition of lipopolysaccharide (LPS; Sigma-Aldrich), RPE SN, recombinant PEDF, SOM, or antibodies against PEDF and SOM. The cultures were incubated for 18 hours at 37°C. Subsequently, the culture supernatants were assayed for IL-10, IL-12, or nitric oxide. 

IL-10 and IL-12 ELISA

The concentrations of IL-10 and IL-12 were assayed by using specific sandwich ELISAs (quantikine murine IL-10 and mouse IL-12p70; R&D Systems, Minneapolis, MN). In brief, samples and standard recombinant IL-10 or IL-12p70 were added to the wells of precoated 96-well plates and incubated for 2 hours at room temperature. The plates were washed and IL-10 or IL-12p70 antibody conjugates were added, incubated for 2 hours, and washed. Substrate solution was added to each well and incubated for 30 minutes at room temperature. The reaction was stopped by the addition of 1.0 N H₂SO₄. A plate reader (MicroQuant; Bio-Tek Instruments, Winooski, VT) was used to read the optical density of the color change at a wavelength of 450 nm. The concentration of cytokine in the sample was calculated from a standard based on the optical density of the curve made from the optical density versus the corresponding standard cytokine concentration run with the samples. 

SOM ELISA

The concentration of SOM in the SN of the RPE eyecups was measured using a competitive ELISA method. ²¹ A 96-well flat-bottomed plate (Corning-Costar, Corning, NY) was coated with 1: 500 dilution of anti-Rb IgG (Sigma-Aldrich) overnight. The plate was then blocked and incubated with a 1:400 dilution of anti-SOM antibody. RPE SN were mixed with 2 ng/mL of biotinylated SOM and added to the wells. To prepare the standard curve, known quantities of SOM protein (0.003–10 ng/mL) were mixed with the biotinylated-SOM. The buffer to block, wash the wells, and dilute the antibodies was a 1% BSA (Sigma-Aldrich) solution in 0.1 M PBS (PBS-BSA). The plate was incubated for 2 hours at room temperature and washed. Diluted (1:1000) streptavidin-β galactosidase (Invitrogen-Gibco, Gaithersburg, MD) was added to the wells, and the plate was incubated for 30 minutes at room temperature. After washing, the substrate chlorophenyl-red-β-d-galactoside (Invitrogen-Gibco BRL) was added, and the optical density of the color change was read 1 hour later by a plate reader (MicroQuant; Bio-Tek Instruments). An equation fitted to the polynomial regression of known SOM concentrations was used to calculate the concentration of SOM in the SN of the RPE eyecup from the sample optical density. The sensitivity of the competitive ELISA was 3 pg/mL. 

RNA Isolation

Total RNA was isolated from RPE cells harvested from eyecups and from the confluent second passage of cultured monodispersed RPE (RNA STAT-60 kit; Tel-Test, Inc., Friendswood, TX) according to the manufacturer’s instructions provided. This kit uses a single-step method by acid guanidinium thiocyanate-phenol-chloroform extraction. 

Reverse Transcription–Polymerase Chain Reaction

cDNA was synthesized by reverse transcribing RNA with random hexamers and AMV reverse transcriptase (Promega, Madison, WI). For PCR amplification of SOM F-ppSOM, TGG CTT TGG GCG GTG TCA, and R-ppSOM, CAG CCA GCT TTG CGT TCC (265 bp). Primers for GAPDH were, F-GAPH, GGTGAAGGTCGGTGTGAACGGA; R-GAPDH, TGTTAGTGGGGTCTCGCTCCTG (245 bp). PCR reactions were performed in a 50-μL amplification mixture containing 1× polymerase buffer, 2.5 mM MgCl₂, 0.2 mM each dNTP, 1 μM of forward and reverse primers, 1.25 U Taq polymerase (Perkin Elmer, Wellesley, MA). The PCR thermal profile was performed in a thermal cycler (GeneAmp PCR System 2400; Perkin Elmer): 1 cycle for 5 minutes at 94°C and 5 minutes at 60°C; 40 cycles for 2 minutes at 72°C, 1 minute at 94°C, and 1 minute at 58°C; and 1 cycle, 10 minutes at 72°C, hold at 4°C. PCR products were then separated by 1.5% agarose gel electrophoresis. 

Immunoblot

Recombinant PEDF protein (BioProducts Maryland, Inc., Middletown, MD) was used as a positive control. RPE SN (final dilution 1:20) and control samples (2.5 ng) were subjected to SDS-PAGE in 4% to 12% Bis-Tris gradient gels (Invitrogen Inc., Carlsbad, CA) followed by electrophoretic transfer of separated proteins to nitrocellulose membranes (Pierce, Rockford, IL). Immunoblot analysis was performed with anti-PEDF antibody (BioProducts Maryland, Inc.). Antibodies bound to proteins on the membrane were detected with horseradish peroxidase–conjugated secondary antibody (Santa Cruz Biotechnology, Santa Cruz, CA) and chemiluminescent substrate (ECL detection reagents; GE Healthcare, Piscataway, NJ). Membranes were then exposed to light film (Biomax Eastman Kodak, Rochester, NY) to detect the chemiluminescent signal. 

Nitric Oxide Assay

Accumulation of nitrite in the culture supernatant of LPS-activated macrophages were assayed as an indicator of nitric oxide production. Macrophages, in phenol-free DMEM supplemented with 10% fetal bovine serum (Hyclone, Logan, UT), were placed in the wells of flat-bottomed 96-well plate at the concentration of 5 × 10⁵ cells per well for 2 hours at 37°C. The culture media were replaced with 100 μL of phenol-red–free DMEM with 0.5% of FBS. Experimental wells contained 50 μL of 1 μg/mL of LPS and RPE SN. Some experimental wells also received 1 μg/mL of anti-PEDF or anti-SOM antibodies. Recombinant PEDF (250 ng/mL) or SOM (2 μg/mL) were added to some cultures, with or without LPS. The cultures were incubated for 18 hours at 37°C. The supernatants were assayed for NO by mixing 100 μL of supernatants with 100 μL Griess reagent (1% sulfanilamide-0.1% naphthylethylene diamine dihydrochloride in 2% H₃PO₄) in a 96-well plate. After 15 minutes of incubation at room temperature, the plate was read at 550 nm using the plate reader (MicroQuant; Bio-Tek Instruments). The concentration of nitrite in the culture supernatant was determined from a standard curve of known sodium nitrite concentrations (0.003–100 μM). 

Induction and Assessment of Endotoxin-Induced Inflammation

Groups of C57BL/6 mice (n = 5) received intradermal injections of LPS (1 μg/5 μL) together with 5 μL of recombinant PEDF (250 or 125 ng/mL) intradermally into the right ear pinnae. As a control, the mice received 1 μg of LPS in 10 μL in the left ear. Both ear pinnae were measured with an engineer’s micrometer (Mitutoyo, Tokyo, Japan) immediately before and 24 hours after the ear injection. The measurements were performed in triplicate. The suppression of inflammation was measured as the change in ear swelling (24-hour minus 0-hour measurement in the PEDF- and LPS-injected ear) relative to positive control animals (24-hour measurement minus 0-hour measurement in the LPS-injected ear). A two-tailed Student’s t-test was used, with significance assumed at P ≤ 0.05. 

Results

Effect of RPE SN on IL-10, IL-12, and NO Production by Macrophages

Since RPE SN inhibits proinflammatory T-cell activation, ¹⁰ we wished to determine whether soluble factors produced by RPE could also suppress inflammatory activity of macrophages. We assayed for interleukin (IL)-10 as an anti-inflammatory cytokine and, interleukin (IL)-12 as a proinflammatory cytokine to assess the influence of RPE SN on macrophage activation. In addition, we assayed for nitric oxide (NO) synthesis, because NO produced by infiltrating macrophages was shown to play a central role in experimental autoimmune uveitis (EAU). ²³  

RAW cell macrophages were either incubated alone, with LPS, or with RPE SN and LPS. Some cultures contained only the RPE SN. RPE SN did not contain any IL-12, but resting macrophages produced a background level of IL-12, which rose significantly in the presence of LPS (Fig. 1A) . Addition of RPE SN significantly inhibited IL-12 production by both resting and LPS-activated macrophages. 

To assess the influence of RPE SN on IL-10 production, macrophages were either incubated alone or with RPE SN. Some wells contained only the RPE SN. There were detectable levels of IL-10 in the RPE SN. Although untreated resting macrophages did not produce IL-10, treatment with RPE SN significantly enhanced IL-10 production. This increase in IL-10 level correlated with increased message for IL-10 in RPE-SN–treated macrophages. A twofold increase in IL-10 expression was detectable by RT-PCR in SN-treated macrophages compared with untreated macrophages (data not shown). 

The effect of RPE SN on NO production was determined by coculturing RAW cell macrophages with LPS and RPE SN. Control cultures contained either macrophages alone or macrophages treated with LPS. As demonstrated in Figure 1C , untreated macrophages do not produce NO, but production of NO was significantly increased after treatment with LPS. The RPE SN contained a significant but small amount of NO, and macrophages treated with RPE SN, in presence of LPS, had a significant but modestly higher level of NO than with LPS alone (Fig. 1C) . These data demonstrate that soluble factors produced by RPE affect macrophage innate activity by inducing IL-10 and significantly inhibiting IL-12 production, while having a slight effect on NO generation. 

Role of PEDF in IL-10 and IL-12 Production by Macrophages

It is known that PEDF is produced and secreted by RPE. ²⁴ To detect the presence of PEDF in our RPE SN, immunoblot assay for PEDF was performed. The results in Figure 2show that there is a detectable amount of PEDF in the RPE SN in concentrations of 500 to 250 ng/mL. 

To see whether PEDF mediates any of the effects of RPE SN on the RAW cell macrophages, the cells were either cultured with RPE SN alone or with RPE SN with anti-PEDF antibody. To determine whether recombinant PEDF behaves similarly to the RPE SN, some macrophages were treated with PEDF (250 ng/mL) alone. The macrophages produced background levels of IL-10 (Fig. 3A) . There was a tripling of IL-10 when the macrophages were treated with RPE SN. A similar increase in IL-10 was seen when the macrophages were treated with recombinant PEDF (Fig. 3A) . Addition of anti-PEDF antibody significantly reversed RPE SN–induced IL-10 production. 

To assess the role of PEDF in suppressing IL-12 production by macrophages, cultures of RAW cell macrophages were treated with the RPE SN as before, with or without the addition of anti-PEDF antibody. The RPE SN treatment inhibited LPS-stimulated IL-12 production (Fig. 3B) . Neutralizing PEDF in the RPE SN significantly upregulated IL-12 production by the macrophages. Treating LPS-stimulated macrophages with PEDF at a concentration of 250 ng/mL, suppressed IL-12 production (Fig. 3B) . This result demonstrates that PEDF content of the SN is responsible for inhibitory effect of RPE SN on IL-12 secretion by macrophages. 

Because IL-10 is known to have an inhibitory effect on IL-12 production by macrophages, ²⁵ we assayed for the possibility that PEDF suppression of IL-12 production is via PEDF-induced IL-10 in the macrophages. The anti-IL-10 antibody was added to cultures of LPS-stimulated macrophages treated with RPE SN. As before, addition of either RPE SN or PEDF to LPS-stimulated macrophage cultures suppressed IL-12 production (Fig. 4) . The neutralization of IL-10 with anti-IL-10 antibody significantly increased IL-12 production by the RPE-SN–treated macrophages (Fig. 4) . Similarly, neutralization of IL-10 in the cultures of LPS-stimulated macrophages treated with recombinant PEDF blocked PEDF inhibition of IL-12 production. The data demonstrate that PEDF indirectly suppresses IL-12 secretion by stimulating an immunosuppressive autocrine pathway of IL-10 production by macrophages. 

Effect of PEDF and SOM on NO Production by Macrophages

To see the role of PEDF in NO production, anti-PEDF antibody was added to RPE-SN–treated RAW cell macrophage cultures. Neutralization of PEDF did not alter LPS-stimulated generation of nitric oxide by RPE-SN–treated macrophages; however, treatment of LPS-stimulated macrophages with recombinant PEDF inhibited NO generation (Fig. 5) . That recombinant PEDF inhibited NO production, but the RPE SN that contained the PEDF did not inhibit NO production by LPS-activated macrophages suggests that there is another factor in the RPE SN that prevents the suppression of NO production. 

Human RPE cells have been found to produce SOM and express some of its receptors. In addition, SOM regulated NO production in the cultured human RPE cells. ²⁶ We looked for the presence of SOM in our RPE SN and SOM production by the murine RPE. Using RT-PCR, we demonstrate that RPE cells both in monodispersed culture and harvested from eyecups expressed mRNA for SOM (Fig. 6A) . SOM protein, detectable by competitive ELISA, was present in a concentration of 10.5 ± 5.2 ng/mL in the RPE SN (data not shown). To determine the effect of SOM on NO production by RPE-SN–treated macrophages, anti-SOM antibody was added to macrophage cultures. The RPE SN suppressed NO production significantly, with the addition of the anti-SOM antibody (Fig. 6B) . The macrophages treated with SOM alone did not induce NO production and when stimulated with LPS, the SOM-treated macrophages produced NO at levels that were no different from those in untreated LPS-stimulated macrophages (Fig. 6B) . To test whether PEDF and SOM have an antagonistic relationship in regulation of NO production, RAW cell macrophages were cultured with LPS, with LPS plus recombinant PEDF at various concentrations, or with recombinant SOM (2 μg/mL). The results depicted in Figure 6Cdemonstrate that PEDF inhibits NO production by LPS-treated RAW cell and that addition of SOM reverses the inhibitory actions of PEDF. We conclude that PEDF and SOM act antagonistically in regulation of NO production by macrophages. 

Effect of PEDF on Macrophages In Vivo

To test whether inhibition of macrophage function by PEDF also takes place in vivo, we used an endotoxin-induced inflammation model. LPS activates macrophages, which in turn induce acute inflammation with polymorphonuclear leukocyte (PMNL) infiltration. ²⁷ Groups of C57BL/6 mice received 5 μL of recombinant PEDF, at different concentrations, together with 5 μL of LPS (1 μg/5 μL) administered intradermally into the right ear pinnae. As a control, the mice received 1 μg of LPS in 10 μL in the left ear. Ear thickness was measured before and 24 hours after ear injection. As demonstrated in Figure 7 , there was a significant reduction of swelling in the right ear, which received PEDF and LPS compared with the left ear which received only LPS. Moreover, the reduction in the degree of swelling was dependent on the concentration of PEDF. The result points to potent anti-inflammatory activity of PEDF in vivo. 

Discussion

In an earlier study, we showed that the suppression of adaptive immunity by RPE is mediated by RPE production of TGF-β and thrombospondin. ¹⁰ The results we present herein demonstrate, for the first time, that RPE suppression of innate immunity is, at least in part, through their production of PEDF. The PEDF, produced by the RPE, induced IL-10 production by the macrophages to further suppress IL-12 production in an autocrine manner, an action of PEDF never before documented. 

The finding that RPE can stimulate IL-10 production by macrophages is important, because the immunoregulatory activity of IL-10 is very similar to the ocular factors that mediate immune privilege in the eye. IL-10 inhibits the central proinflammatory transcription factor NFκB in macrophages and CD4⁺ T cells. In contrast, IL-10 promotes activation and migration of CD8⁺ cytotoxic T cells and may mediate the induction of regulatory T cells. ^{28 29 30 31} In dendritic cells, IL-10 inhibits expression of class II major histocompatibility complex (MHC) molecules ³² and the CD28 costimulatory pathway of T-cell activation, ³³ while promoting production of TGF-β, ³⁴ a well-known anti-inflammatory cytokine. Therefore, the induction of IL-10 further promotes the mechanisms of immune privilege to suppress the induction of inflammation by innate and adaptive immunity. 

Our results are the first to demonstrate SOM production by murine RPE and that the role of SOM in suppressing innate immunity is different from its role in aqueous humor suppression of Th1 cells. Although SOM alone does not induce NO production by macrophages and does not suppress NO production by LPS-stimulated macrophages, its presence prevented RPE SN suppression of NO production by LPS-stimulated macrophages. We demonstrated that PEDF, another factor in RPE SN, inhibits NO production by LPS-stimulated macrophages, and only when SOM was neutralized in the RPE SN was there suppression of NO production. It appears that PEDF and SOM are antagonistic to each other in regulating LPS-stimulated NO production. 

Although our results suggest that NO generation by macrophages cannot be suppressed in the subretinal space by the RPE, there is a possibility that NO functions in the subretinal space as an immunosuppressive factor. There is a delicate balance between NO proinflammatory and immunosuppressive affects. How this balance is regulated in the eye is unknown; however, there is evidence that NO may play a beneficial role in immune privilege of the subretinal space. RPE cells from rats cocultured with activated lymphocytes produce high amounts of NO that are cytotoxic to the lymphocytes. ²³ In addition, Zech et al. ³⁵ demonstrated that NO elevates transepithelial electrical resistance (TER), thus promoting the integrity of the tight junctions between the RPE cells and strengthening the ocular outer blood–retinal barrier. These reports indicate that NO has a beneficial role in maintaining the immune privilege of the subretinal space. 

The PEDF induction of IL-10 by macrophages may well be an example of the evolutionary adaptation of the eye to mediate immune privilege with factors that have functions that are necessary for ocular physiology. The most well-understood action of PEDF is its antiangiogenic activity. PEDF inhibits endothelial cell migration in response to a diverse group of proangiogenic factors ³⁶ and augments endothelial cell expression of FasL, leading to selective apoptosis of endothelial cells involved in angiogenesis. ³⁷ Also, PEDF downregulates nuclear factor of activated T cells (NFAT), an essential transcription factor for vascular epithelial growth factor (VEGF) induction of angiogenesis. ³⁸ In addition, PEDF could render some of its antiangiogenic activity through induction of IL-10 in macrophages. It has been reported that IL-10 is antiangiogenic by its ability to decrease VEGF, TNF-a, or MMP-9 synthesis in a rabbit corneal angiogenesis model, and in mice hindlimb ischemia-induced angiogenesis. ^{39 40 41} A recent study has also linked the antiangiogenic function of PEDF with its ability to decrease proinflammatory factors such as TNF-α, ICAM-1, and MCP-1. ¹⁷  

To test the effect of PEDF on macrophages in vivo, we injected LPS into the left ear pinnae of a group of mice, whereas the other ear received LPS and PEDF. LPS induces macrophage activation, which in turn produce PMNL infiltration. ²⁷ We have demonstrated that PEDF substantially inhibited the inflammation induced by LPS in vivo. In a previous report, we noted that RPE-produced thrombospondin-1 is a potent immunosuppressive factor in the subretinal space by activation of TGF-β and inhibition of Th1 activation. ⁹ In the present study, by production of PEDF, RPE profoundly inhibited proinflammatory activity of macrophages. This action of PEDF was also demonstrated in an in vivo model of LPS-induced inflammation. Our results further contribute to defining immune privilege within the ocular microenvironment to be the suppression of inflammation mediated by both innate and adaptive immunity. 

³  Deceased March 15, 2004.

 Supported in part by National Eye Institute Grant EY05678 and Grant 04037006 from the Department of Defense.

 Submitted for publication September 26, 2005; revised February 24 and April 17, 2006; accepted June 23, 2006.

 Disclosure: P. Zamiri, None; S. Masli, None; J.W. Streilein, None; A.W. Taylor, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

Effect of SN from C57BL/6 RPE eyecups on IL-10, IL-12, and NO production by macrophages (Mφ). Twenty-four hours after incubation of RPE eyecups with SFM, SN was collected. (A) Macrophages were incubated either alone or with RPE SN. Some wells contained RPE SN only. The cultures were incubated for 18 hours, and the culture supernatants of the wells were assayed for IL-10 (A). Macrophages were incubated either alone, with LPS or with RPE SN with or without LPS. Some wells contained only RPE SN. The cultures were incubated for 18 hours and the supernatants of the wells were assayed for IL-12 (B) or for nitric oxide (C). The results are the mean ± SEM of triplicate wells from a representative experiment. *Significantly different results (P < 0.05).

Figure 2.

 

Detection of PEDF in RPE SN. Diluted RPE SN (1:200) was analyzed by immunoblot with anti-PEDF antibody. Recombinant PEDF (2.5 ng) was used as a positive control. A chemiluminescence assay was used to detect horseradish peroxidase (HRP)-labeled secondary antibody bound to anti-PEDF antibody on the membrane. PEDF was detected at 45 kDa (arrow).

Figure 3.

 

Effect of PEDF on IL-10 and IL-12 production by macrophages. Macrophages (Mφ) were incubated alone, with RPE SN with or without the addition of anti-PEDF antibody, or with recombinant PEDF. The cultures were incubated for 18 hours, and the culture supernatants of the wells were assayed for IL-10 (A). Similarly, macrophages were incubated with either RPE SN or recombinant PEDF (250 ng/mL) in the presence or absence of LPS. Some cultures in addition received anti-PEDF antibody. The cultures were incubated for 18 hours, and their supernatant was assayed for IL-12 (B). The results are the mean ± SEM of triplicate wells from a representative experiment. *Significantly different results (P < 0.01).

Figure 4.

 

Regulation of IL-12 production by IL-10. Macrophages (Mφ) were incubated alone, with LPS, with RPE SN, or with recombinant PEDF. In some experiments, anti-PEDF or anti-IL-10 antibodies were added to the cultures. The wells were incubated for 18 hours, and supernatants of cultures were assayed for IL-12 by ELISA. The results are the mean ± SEM of triplicate wells from a representative experiment. *Significantly different results (P < 0.05).

Figure 5.

 

Role of PEDF in NO production by macrophages. Macrophages (Mφ) were incubated alone or with LPS, RPE SN, or recombinant PEDF. Some wells in addition to macrophages and RPE SN received anti-PEDF antibody. The wells were incubated for 18 hours, and culture supernatants were assayed for NO production with Griess reagent. The results are the mean ± SEM of triplicate wells from a representative experiment. *Significantly different results (P < 0.05).

Figure 6.

 

The presence of SOM in RPE SN and its role in NO production. Total RNA was isolated from RPE grown in cultures or from the RPE eyecups and message for somatostatin (ppSomatostatin) was detected by RT-PCR. The PCR products were resolved by electrophoresis on 1.5% agarose gel followed by ethidium bromide staining (A). Macrophages were incubated alone or with LPS, RPE SN, or with recombinant SOM. Anti-PEDF antibody was added to some cultures. The cultures were incubated for 18 hours, and supernatants of the cultures were assayed for NO production with Griess reagent (B). Macrophages were incubated alone or with LPS, RPE SN, or recombinant SOM (2 μg/mL) or recombinant PEDF. The cultures were incubated for 18 hours, and supernatants of the cultures were assayed for NO production with Griess reagent (C). The results are the mean ± SEM of triplicate wells from a representative experiment. *Significantly different (P < 0.05). **Significantly different (P < 0.01).

Figure 7.

 

Capacity of PEDF to suppress inflammation in vivo. All C57BL/6 mice received intradermal injection of LPS into the left ear pinnae. A group of mice were injected with LPS and PEDF (250 ng/mL) into the right ear pinnae, whereas the other group received LPS and PEDF (125 ng/mL). Ear thickness was measured before and 24 hours after injection. The data depicted are expressed as the differences in ear thickness measurement between the two ears. *,**Significant difference between ear thickness in the two ears (P < 0.05, P < 0.01, respectively).