Neutralizing Tumor Necrosis Factor-α Activity Suppresses Activation of Infiltrating Macrophages in Experimental Autoimmune Uveoretinitis

Abstract

purpose. During experimental autoimmune uveoretinitis (EAU), infiltrating macrophages become activated to express nitric oxide synthase (NOS)-2 and generate nitric oxide (NO). The current study was designed to determine whether neutralizing TNF activity with a soluble fusion protein of TNFp55 receptor (sTNFr-IgG) inhibits macrophage activation, thereby contributing to reduced tissue damage observed with such treatment.

methods. EAU was induced in Lewis rats by active immunization with soluble retinal extract (RE) and pertussis toxin (intraperitoneally), and animals were treated on days 6 and 8 after immunization with either sTNFr-IgG or human (hu)IgG. Disease course and severity were noted clinically, and eyes were enucleated for histologic scoring, including TUNEL immunofluorescence, at various stages of disease. Infiltrating retinal macrophages were isolated through a density gradient and subsequently phenotyped by flow cytometry, analyzed for ability to produce nitrite, either spontaneously or after cytokine stimulation, and assayed by PCR for cytokine gene expression.

results. Neutralizing TNF activity suppressed tissue damage without impeding myeloid cell infiltrate. Moreover, with sTNFr-IgG treatment, infiltrating macrophages demonstrated reduced nitrite production at the height of disease, and the level of apoptosis within the retina of both ED1⁺ cells and resident cells was reduced. PCR analysis demonstrated a significant increase in TGFβ signal and absent or low TNF signal throughout the disease course after treatment with sTNFr-IgG.

conclusions. sTNFr-IgG successfully suppresses retinal damage and impairs macrophage activation but not trafficking during EAU. sTNFr-IgG–mediated suppression of NO production results in reduced levels of apoptosis of inflammatory cells and reduction in photoreceptor damage.

Tumor necrosis factor (TNF)-α is a pleiotropic, pivotal proinflammatory cytokine and is implicated in the orchestration of many autoimmune and inflammatory diseases, and therefore neutralizing its activity has been successful in the treatment of rheumatoid arthritis ¹ and, more recently, uveitis. ² TNF-α activity can be inhibited in vivo and in vitro by specific monoclonal antibodies (mAbs) or by administration of soluble TNF-α receptor (sTNFr-IgG), as demonstrated by successful inhibition of tissue damage in EAU and EAE. ^{3 4 5 6} Experimentally neutralizing TNF-α activity with mAb prevents the upregulation of adhesion molecules, such as vascular cell adhesion molecule (VCAM)-1 on vascular endothelium, ⁷ and thereby reduces the inflammatory infiltrate. In addition, inhibition of TNF-α directly suppresses tissue damage ⁸ and cell death. ⁹ In contrast to other reports using anti-TNF-α mAb, we and others have shown with sTNFr-IgG that, despite the suppression of clinical disease and reduction in tissue damage during EAU, there is no reduction in T-cell infiltrate. ^{3 4 5 6} sTNFr-IgG therapy suppresses T helper (Th)-1 cell responses in EAU, ¹⁰ and, in these experiments, it was noted that macrophage phenotypes, as represented by MHC class II expression within the infiltrate, was reduced, inferring that treatment may also downregulate macrophage activity. 

EAU induced by soluble retinal antigens closely resembles human disease, ¹¹ and is mediated by CD4⁺ T cells and activated macrophages. ^{12 13 14} Macrophages are major effectors of tissue damage. They are found in the outer retina during retinal inflammation and have been identified as a major source of nitric oxide synthase (NOS)-2. ^{15 16 17} Recent data have shown that inhibition of NOS2 significantly inhibits tissue damage, ^{16 18} which can be attributed to a dramatic reduction in photoreceptor apoptosis, despite retention of T cells during inhibition of NOS2. It appears therefore that formation of peroxynitrites and reactive oxygen species contributes to photoreceptor death but is also necessary for activation-induced cell death and elimination of effector T cells. Both TNF-α and IFN-γ activate macrophages and elicit strong NO responses. ¹⁹ Cytokines act hierarchically to elicit macrophage activation and a maximum NO response, whereas TNF-α alone is unable generate significant NO response. In the retina, the resident myeloid cell population (microglia, MG) does not constitutively express NOS2 and thus does not generate NO spontaneously. ¹⁷ During EAU, infiltrating myeloid cells under the influence of pronounced cytokine release from the Th1-cell infiltrate ^{20 21 22 23} express NOS2 and nitrotyrosine and constitutively release NO. ¹⁷ Given our observations of TNF neutralization, ^{4 10} we wanted to determine whether macrophage infiltrate function was simultaneously downregulated, suppressing generation of NO, and accounting for the limited tissue damage observed during sTNFr-IgG therapy. 

Methods

Induction and Clinical Assessment of EAU and sTNFr-IgG Treatment

Female Lewis rats used throughout the study were treated humanely according to the UK Animal License Act and to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. EAU was induced in adult female Lewis rats (6–8 weeks) by a 0.1-mL immunizing intradermal injection into the footpad of 5 to 7 mg/mL retinal extract (RE; vol/vol) in Freund’s complete adjuvant, along with an intraperitoneal (IP) injection of 1 μg pertussis toxin (Centre for Applied Microbiology and Research [CAMR], Salisbury, UK). Soluble bovine RE was prepared by hypotonic lysis of freshly dissected bovine retinas in the dark, as described previously. ²⁴ Ocular examination was performed daily and EAU graded ²⁴ by slit lamp biomicroscopy, from day 6 after immunization. The extent of anterior chamber and retroiridal inflammation was assigned a clinical score on a scale of 0 to 4. ²⁴ Clinical observation allowed the day of maximum disease to be recorded in control groups, thus allowing comparison with suppression of clinical disease in treated groups at the same time points, as previously observed. ^{4 10} sTNFr-IgG was kindly provided by and produced at the Therapeutic Antibody Centre, University of Oxford (Oxford, UK) as a chimeric molecule containing the soluble domain of the human p55 TNFr coupled to the CH₂ and CH₃ constant domains of human IgG₁. In repeated experiments, animals (three to four per group) were treated with IP injection on days 6 and 8 after immunization with 2 and 1 mg sTNFr-IgG, respectively, or with equivalent amounts of human immunoglobulin G (huIgG) as a control. 

Antibodies and Cytokines

Rat rIFN-γ was obtained from Bradsure Biologicals, Ltd. (Loughborough, UK) and human rTNF-α and rTGF-β₁ were obtained from Sigma Chemical Co. (Poole, UK). Rat recombinant (r)IL-4 was produced in house, as described previously, ²⁵ using a CHO cell line generously donated by Neil Barclay (MRC Cellular Immunology Unit, Oxford, UK). Unless otherwise stated, mouse mAbs specific for rat cell surface markers were obtained from Serotec (Oxford, UK), including mouse anti-rat I-A (OX6):RPE and mouse anti-rat CD4 (w3/25). Purified anti-rat CD86 (B7-2), biotinylated mouse IgG1κ (negative isotype control), biotinylated mouse anti-rat mononuclear phagocyte mAb (IC7), anti-rat CD11b/c (OX42), and streptavidin-alkaline phosphatase (AP) complex (SA-APC) were obtained from BD PharMingen (San Diego, CA). Anti-mouse IgG (whole molecule)-FITC conjugate was obtained from Sigma Chemical Co. 

Isolation and Preparation of Macrophage Cultures

Retinal Myeloid-Cell Populations.

Animals were killed by CO₂ asphyxiation from days 7 to 21 after immunization. The retinas of each animal were dissected microscopically in cold 1% BSA/PBS and then mechanically disrupted through a sieve (50-μm mesh) to obtain a single-cell suspension. Retinal cells from each animal were kept separate throughout isolation procedures. After further washes in 1% BSA/PBS, cells were subsequently isolated using a graduated density gradient (Percoll; Pharmacia, Uppsala, Sweden), as previously described, ^{26 27} that enriches leukocytic populations within a mixed population of cells containing resident retinal cells, such as ganglion cells and neuronal elements. Cells were collected from the 1.072-and 1.088-g/mL layers (enriched retinal cell population) and washed, and viability was determined by trypan blue exclusion. At this point, 1 to 2 × 10⁶ cells were separated for flow cytometric analysis (cells from each animal were assessed). For culture, the purity of CD11b/c⁺ macrophage cells was confirmed by single-color flow cytometric analysis, and macrophage-enriched populations were plated at 2 × 10⁵ CD11b/c⁺ macrophages per milliliter in 96-well tissue culture plates and rested for 1 hour at 37°C, after which RPMI complete medium was then replaced, removing nonadherent cells, and macrophages were incubated for a further 24 hours, with or without cytokines. 

Cytokine Stimulation of Macrophage Cultures.

Assay culture conditions were optimized using splenic macrophages (derived from immunized animals at time retinal cells were isolated), which were cultured and stimulated as previously described. ^{17 19} Cytokines used were IFN-γ (20 U/mL), TNF-α (5 ng/mL), IL-4 (5 μL), and sTNFr-IgG (13 μg/mL), alone or administered sequentially in combination, with the administration of each cytokine separated by a 4-hour period. Macrophage function was assessed 24 hours after addition of the first cytokine. IFN-γ, followed 4 hours later by TNF-α (IFNγ/TNFα), was used as a positive stimulation of NO production, confirming previous in vitro studies. ¹⁹ Where stated, cytokine additions were given relative to IFN-γ/TNF-α (positive control) stimulation. Cytokines were not removed before macrophage function was assessed. In retinal specimens, the number of cytokine stimulation assays that could be performed was restricted by the number of macrophages isolated. 

Flow Cytometric Phenotypic Analysis of Retinal Macrophages

Immunophenotyping of infiltrating leukocytes was performed using mouse mAb specific for the rat cell surface markers to be listed later. mAbs used were either unconjugated or conjugated to biotin, phycoerythrin (PE), or FITC for three-color immunofluorescence. Unconjugated mAbs were detected with rat anti-mouse FITC, and biotinylated mAbs were detected with streptavidin-APC (SA-APC). Staining was performed, with fluorescence-activated cell sorting (FACS) buffer (PBS, BSA, and 10 mM NaN₃) used for washes. All reagents, buffers, and incubations were performed and kept at 4°C. Negative isotype control and single positive control experiments were performed to allow accurate breakthrough compensation. Primary unconjugated mAbs included CD4, CD11b/c, and CD86. Cells were incubated sequentially with primary mAb, anti-mouse IgG (whole molecule)-FITC conjugate in the presence of 10% NRS, blocked with 10% normal mouse/normal rat serum (NMS/NRS), biotinylated second antibody (CD45, CD11b/c, or IC7) with PE-conjugated third antibody (mouse anti-I-A [mouse/rat]), and finally SA-APC. Acquisition of data was performed by flow cytometer (FACS Calibur; BD Biosciences; Plymouth, UK) and the accompanying software (CellQuest; BD Biosciences). A total of 10,000 events were recorded and gates and instrument settings were set according to forward- and side-scatter characteristics. Fluorescence analysis was performed after further back-gating to exclude dead cells, most neutrophils, and background staining. 

Quantification of NO Synthesis

NO generation was estimated after 24 hours in culture by assaying culture supernatants for the stable reaction product of NO (nitrite) against a sodium nitrite standard on the same plate. A 200-μL total volume of each cell-free supernatant (diluted 1:4 with dH₂O) was incubated with 50 μL of Griess reagent (0.5% sulfanilamide, 0.05% N-(1-naphtyl) ethylenediamine dihydrochloride in 2.5% phosphoric acid) in 96-well flat-bottomed plates for 10 minutes at room temperature. The optical densities were measured at 540 and 690 nm to account for background. 

Immunohistochemistry

Eyes removed from normal or immunized animals at the various stages of EAU were embedded in optimal cutting temperature (OCT) compound, snap frozen, and stored at −30°C, or processed for routine paraffin embedding for histologic scoring and immunofluorescence, for the detection and quantification of apoptosis. Histologic scoring was performed, as previously described. ²⁴ In brief, three sections per slide were examined, and scoring was based on the extent of infiltration and the extent of structural damage—principally, photoreceptor and nuclear cell loss and disruption of retinal architecture. Apoptosis was detected in sections with a kit (TACS 2 TdT-Fluor In Situ Apoptosis Detection; Trevigen, Inc., Gaithersburg, MD; AMS Biotechnology, Oxon, UK), using deionized water throughout and according to the manufacturer’s instructions. Briefly, slides were first dewaxed with xylene (BDH Laboratory Supplies, Poole, UK) and reducing concentrations of ethanol, before rehydration in PBS for 15 minutes, incubation with proteinase K for 15 minutes, and subsequently 1× TdT labeling buffer. Slides were finally incubated in streptavidin-fluorescein at 1:50 dilution for 30 minutes. After two washes in PBS, slides were either mounted in antifade medium (Vectashield; Vector Laboratories, Burlingame, CA) or further incubated with ED1 mAb (Serotec), detected with streptavidin-Texas red (1:50; Amersham Life Science, Amersham, UK). The number of TUNEL-positive cells within the retina and subretinal space was counted throughout all the sections under a ×20 objective with the appropriate excitation filters of a confocal microscope (BH2-RFC; Olympus, Tokyo, Japan), according to the manufacturer’s instructions. Cells in three sections per eye were counted. 

Cytokine Expression

RNA was extracted from whole retina (one to two per sample; RNA-Bee (Biogenesis, Poole, UK), an improved version of the single-step method of RNA isolation. All components used had been treated with diethyl pyrocarbonate (DEPC). The final concentration of RNA was determined by spectrophotometric analysis. Viable cDNA was then generated for each sample. The expression of multiple cytokines in retina, including IL-6, TNF, IL-1β, TGFβ, and granulocyte-macrophage–colony-stimulating factor (GM-CSF) was determined by the use of commercial quantitative PCR detection kits (Biosource, Camarillo, CA), according to the manufacturer’s instructions. All components were provided in the kits, with the exception of the Taq DNA polymerase (Promega, Southampton, UK). The primers used were as follows: GAPDH, 532 bp; IL-6, 432 bp; TNF-α, 352 bp; IL-1β, 295 bp; TGF-β, 250 bp; and GM-CSF, 210 bp. A total reaction volume of 25 μL per sample was used, which included 7.25 μL of sample or control cDNA. The PCR products from the multicytokine detection kit described earlier were visualized on a 2% agarose gel by ethidium bromide staining. To determine fragment size, a DNA ladder (Roche Diagnostics Ltd., Lewes, UK) was run on the same gel. A positive control was also run on each gel, to allow complete analysis of the samples. Gels were visualized and further analyzed with a commercial system (Syngene; Synoptics, Ltd., Cambridge, UK), whereby the fluorescence intensity of the ethidium bromide stained PCR products was measured under ultraviolet light, with the area of fluorescence intensity being equivalent to the cDNA levels present in the sample. A ratio of each cytokine-to-GAPDH intensity was obtained for the raw volume of each sample, from which results were analyzed. 

Statistical Analysis

Differences between groups of NO production were analyzed by nonparametric Kruskal-Wallis analysis of variance and the Dunnett posttest. Histologic scores and level of apoptosis from counts of TUNEL positivity was assessed by nonparametric Mann-Whitney analysis, in which P < 0.05 was considered significant. 

Results

Effect on Retinal Damage of Neutralizing TNF with sTNFr-IgG

These data confirm previous experimental data that treatment of rats or mice with sTNFr-IgG after immunization suppresses EAU. ^{4 5 10} We now show that during the course of EAU there was no selective inhibition of myeloid cell infiltration into the retina after treatment with sTNFr-IgG (Fig. 1 , Table 1 ), although macrophages were delayed in entering the retina (maximum day 11 in huIgG controls and day 13 in TNFr-IgG–treated animals). Further flow cytometric phenotype assessment for expression of major histocompatibility complex (MHC) class II, IC7, CD4, and CD86 was not significantly different between control and treated animals at any of the time points tested (data not shown). Despite myeloid cell infiltration in sTNFr-IgG–treated animals, histologic scores representing structural damage of the retina and loss of photoreceptor outer segments were significantly lower, both at the height and the resolution of disease, compared with control animals (3.17 ± 0.17 huIgG control and 0.5 ± 0.29 sTNFr-IgG–treated, day 11 after immunization; P = 0.03), again confirming previous reports of suppression. In subsequent experiments, when the dose of immunizing antigen (RE) was increased to 7 mg/mL, significant protection was noted only at the height of disease, but at resolution, there was no significant difference in the extent of structural photoreceptor damage observed (data not shown), inferring that large doses of antigen could overcome the therapeutic effect of sTNFr-IgG. 

Effect of sTNFr-IgG on Generation of NO by Infiltrating Myeloid Cells

The ability to isolate retinal macrophage populations from normal retina and during EAU permits further detailed assessment of macrophage activation in vitro. Previously, we have shown that resident macrophages (mainly MG ²⁸ ) behave as though conditioned by TGFβ and remain unresponsive to further stimulation, generating little NO. ¹⁷ Infiltrating macrophages within retina generate NO only during peak disease, at which time they are still unresponsive to further cytokine stimulation. ¹⁷ Consequently, we wanted to determine whether sTNFr-IgG inhibits NO production, the suppression of which may account for the reduced retinal damage observed with treatment, particularly at the phase of maximum disease. Confirming previous data, splenic macrophages failed to constitutively generate NO but were stimulated to do so with IFN/TNF (Fig. 2A) . Treatment of splenic macrophages in vitro with sTNFr-IgG did not inhibit NO production after stimulation with IFN/TNF, whether administered at the same time or 4 hours before stimulation with IFN/TNF. By contrast, resident retinal macrophages and MG from normal eyes did not generate NO spontaneously and remained unresponsive to cytokine stimulation in vitro (Fig. 2B) . Macrophages isolated from inflamed EAU retinas have the capacity to generate NO, and Figure 3A shows that when macrophages, (including inflammatory macrophages) were isolated from retinas at peak disease during EAU, animals treated with sTNFr-IgG showed significantly suppressed generation of NO and, as with control animals, did not respond to further cytokine stimulation (P = 0.03). By day 15 after immunization, there was comparable NO production in both the huIgG control and sTNFr-IgG–treated animals (P = 0.06), and again macrophages from both groups of animals remained unresponsive to subsequent cytokine stimulation (Fig. 3B) . PCR assessment of cytokine production from retinas revealed that sTNFr-IgG treatment profoundly suppressed TNF signal. The effect was maximum on day 11 after immunization in the huIgG control (Fig. 4) . Control animals showed maximum signals for IL-6, TNF, IL-1β, and GM-CSF when clinically and histologically maximum disease was observed (see day 11 after immunization; Fig. 1 ). By contrast, when TNF activity was neutralized, there was an increase in TGFβ signal. Only by day 15 after immunization was there any notable increase in IL-1β signal in treated animals, whereas TNF and IL-6 remained comparatively low throughout, never achieving the signal intensity observed in control animals at day 11 after immunization. 

Effect of sTNFr-IgG on Apoptosis

During EAU, TNF and NO induce apoptosis of both resident retinal cells and infiltrating leukocytes, predominantly T cells and macrophages ^{10 18} during the severest phases of disease. TUNEL staining of retinal sections to assess whether apoptosis is also reduced with therapy was therefore performed, given that sTNFr-IgG suppressed NO production and increased TGFβ, which culminated in protection against photoreceptor damage. Figure 5 shows that during peak phases of EAU (days 10 and 11), huIgG-treated control animals displayed a greater extent of apoptosis than sTNFr-IgG–treated animals (Fig. 5A ; P = 0.05, Mann-Whitney). Throughout the time course of EAU, the overall rate of apoptosis (combining all days examined; day 8 to 13 after immunization) was higher in control than sTNFr-IgG–treated animals, but the difference did not achieve statistical significance (Fig. 5B ; P = 0.08, Mann–Whitney). Figure 5C is representative of cells undergoing apoptosis, showing that both macrophages (ED1⁺ cells) and resident retinal cell populations were identified. 

Discussion

Macrophages are central to the tissue destruction that follows antigen-specific CD4⁺ Th1-cell–mediated inflammation, such as occurs in EAU. ^{12 13} Macrophages develop a coordinated set of properties depending on the microenvironment and cytokine activation directing a multiplicity of responses involved in inflammation and tissue repair. Classic macrophage activation after stimulation with IFNγ and TNF refers to the ability of a macrophage to express NOS2 and generate nitrite, peroxynitrites, and superoxides, which in turn induce lipid peroxidation of cell membranes and cell death. In the rat, naïve bone marrow–derived macrophages respond to cytokines in a hierarchical fashion, requiring, for example, both IFNγ and TNF for optimal priming of NO production, whereas TNF alone is largely insufficient. ^{17 19} Recently, in EAU, we have shown that infiltrating macrophages, conditioned by the microenvironment, are activated to generate NO during peak disease. ¹⁷ Furthermore, inhibition of NOS2 suppresses EAU ^{16 18} and prevents peroxynitrite formation by macrophages, thereby protecting photoreceptor cells from apoptosis. The role of NO and peroxynitrite in ocular inflammation and tissue damage has been confirmed by others. These present data further our understanding of macrophage conditioning during EAU and the mechanisms by which anti-TNF therapies are active during suppression of target organ damage. To add further to our previous observations, we have shown that after sTNFr-IgG therapy, entry of macrophages into the retina was not impaired, but generation of NO was, concomitant with suppression of photoreceptor damage. 

Given that macrophage activation, inducing NOS2 expression and NO generation, is driven by IFNγ/TNF stimulation, it is perhaps not surprising to confirm that neutralizing TNF activity results in diminished macrophage-derived NO production within the retina. One notion is that sTNFr-IgG binds to membrane TNF on macrophages inhibiting TNF production and subsequent macrophage activation. However, as we have showed in the current study, macrophages generated NO after stimulation with IFNγ alone, and thus administering sTNFr-IgG to macrophages in vitro does not block generation of NO. ^{29 30} IFNγ/TNF-induced NO production. How does sTNFr-IgG therapy suppress macrophage activity in vivo? We have shown that during sTNFr-IgG therapy, there is pronounced inhibition of retinal T-cell–derived IFNγ and T-cell activation, despite maintained proliferation ¹⁰ and retention of T cells within retina as a result of reduced apoptosis. During EAU, there is pronounced production of Th1 cytokine, ^{20 21 22 23} which are available to activate infiltrating macrophages at the height of disease. It is possible in light of our current observations that sTNFr-IgG binds and inhibits both soluble TNF (solTNF) and membrane TNF (memTNF) on T cells, thus preventing T-cell activation and IFNγ production and subsequent IFNγ/TNF–mediated activation of infiltrating macrophages during EAU. TNF-mediated activities are numerous, and recent views dictate a fundamental role in the control of leukocyte movement by virtue of a role in chemokine expression, ³¹ in addition to other recognized roles including control of T-cell activation and survival. ^{32 33 34}  

TNF is first produced as a transmembrane molecule that is cleaved by metalloproteinases to produce solTNF, ³⁵ and both memTNF and solTNF interact with both TNF receptors. ³⁶ Generally, however, solTNF is regarded as the ligand for p55 or TNFr1. ³⁷ TNFr1 contains death domains that mediate apoptosis on engagement, ³⁸ but also under other circumstances may mediate cell survival. ³³ Recent evidence strongly infers that memTNF supports the generation of lymphoid structures, whereas solTNF is required for generation of the full phenotype of inflammatory lesions, as shown in experimental models of autoimmunity within the central nervous system (CNS). ³⁹ That sTNFr-IgG treatment in these and previous experiments delayed but did not impair cell movement into the retina, yet suppressed the full phenotype of tissue damage, implies in the main that the principal action is against solTNF. If that is the case, then the main action opposes T-cell production of TNF, which in turn reduces the activation of macrophages that are infiltrating the retina. TGFβ cytokine expression was present in normal retina (Fig. 4) , and significantly increased in treated animals compared with the control at peak disease. The increase in IFNγ production ¹⁰ and decrease in TGFβ after treatment contributes to alternative macrophage programming and suppression of NO. ¹⁷ In addition, although we could did not demonstrate differences in expression of MHC class II and costimulatory expression with treatment, further evidence is needed to assess whether higher TGFβ and lower IFNγ levels result in downregulation of antigen-presenting cell (APC) function contributing to regulation of disease. 

Suppression of NO production, accounting for reduction in target organ damage is consistently observed with sTNFr-IgG therapy and is supported by other related findings in which NOS2 inhibition, induced with nonspecific inhibitors of NOS results in significant suppression of clinical disease and structural damage. ¹⁸ In these experiments, nitrotyrosine formation was seen, indicating that large quantities of both NO and superoxide were generated and reacted to form peroxynitrite, which is highly toxic, particularly to neuronal cells. Inducible NOS2 and NO have both regulatory and effector functions in autoimmunity, providing both a pathogenic and regulatory role during disease evolution and remission. ^{40 41} Tissue-specific expression of NOS is essential for the regulation of immune responses in the periphery, depending on the amount of NO produced. During EAU, until resolution there is retention of ED1⁺ macrophages, ^{17 18} which is also observed whether inhibited with sTNFr-IgG or NOS2 inhibitors, implying that any macrophage necrosis at the height of disease does not significantly affect progression and outcome of disease. This notion is further supported by our observation that macrophage apoptosis was reduced at peak disease in these current experiments. Simultaneously, apoptosis of resident neuronal cells and photoreceptors was also reduced after sTNFr-IgG therapy and with NOS2 inhibitors. Furthermore, NO appears to modulate mitochondria-mediated apoptosis that leads to photoreceptor death and the Fas or death receptor pathway that results in T-cell apoptosis. ³⁴ Similarly, although not identified directly in these experiments, sTNFr-IgG may target solTNF, also preventing TNFr1-induced apoptosis and accounting for the retention of T cells that we observed previously. The role of TNF in retinal inflammation remains undefined. The recent differentiation between functions of memTNF and solTNF, ³⁹ taken with the current data, helps to explain why sTNFr-IgG delays but does not inhibit leukocyte trafficking, as has been seen in other models ^{3 6 42} or in TNF^−/− animals, ^{43 44} and directs where more exquisitely specific targeting of TNF responses could be used therapeutically in the future. Although clinical reports indicate a possible benefit in the treatment of uveitis and retinal vasculitis related to, for example, Behçet’s disease, there remains concerns that in some patients inflammation may be perpetuated by such therapy. These findings, taken in the context of this present data, may of course relate to the retention of inflammatory cells rather than ongoing tissue damage. However, secondary effects as a result of persistence of inflammatory cells, such as edema, may still result in an absence of perceived clinical success. Neutralizing TNF activity therefore is more likely to be beneficial in combination with other immunosuppression until we refine the ability to target TNF responses vis-à-vis memTNF versus solTNF and their respective receptors. 

In summary, in the current studies sTNFr-IgG inhibited TNF activity, resulting in the incomplete formation of an inflammatory response, probably through inhibition of solTNF. The result was downregulation of macrophage-derived, NO-mediated cell death and presumably solTNF-mediated cell death. 

 Supported by the Iris Fund for the Prevention of Blindness, United Kingdom.

 Submitted for publication November 13, 2002; revised December 20, 2002; accepted January 22, 2003.

 Disclosure: M. Robertson, None; J. Liversidge, None; J.V. Forrester, None; A.D. Dick, 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.

 

sTNFr-IgG therapy suppressed target organ destruction without impairing retinal myeloid cell infiltrate. By flow cytometry, the CD11b⁺ cell population identified infiltrating macrophages after appropriate gates were applied on a scatter dot plot to exclude granulocytes. (A) Representative experiment showing that during the course of EAU, equivalent percentages of CD11b⁺ macrophages infiltrated the retina, despite sTNFr-IgG therapy (see Table 1 for absolute numbers). Therapy caused a consistent slight delay in infiltration (mean ± SEM; n = 3 to 4 each group). (B) Representative experiment displaying histologic scoring of structural changes demonstrating that, despite equivalent macrophage infiltration, there was marked reduction in structural damage in the retina at the height of disease (mean ± SEM; n = 3 to 4; P < 0.03). Animals were immunized IP with 5 μg/mL of RE-CFA (vol/vol) and 1 μg pertussis toxin.

Table 1.

 

Numbers of Infiltrating Cells in Control versus sTNFr-IgG–Treated Animals

Days after Immunization	Total Enriched Retinal Cell Number (×105)	Leukocyte Number (CD45+; ×105)	Macrophage Number (CD11b/c+; ×105)
8
Control	18.75 ± 5.73	5.84 ± 2.42	3.53 ± 1.38
sTNFr	4.00 ± 0.29	0.34 ± 0.04	0.36 ± 0.04
10
Control	28.17 ± 3.11	27.33 ± 3.05	19.79 ± 2.84
sTNFr	11.6 ± 0.52	7.03 ± 1.40	4.98 ± 1.30
11
Control	43.30 ± 5.36	42.10 ± 5.38	22.07 ± 1.85
sTNFr	30.3 ± 5.92	27.71 ± 5.45	19.28 ± 4.18
13
Control	29.00 ± 0.23	25.46 ± 0.86	13.23 ± 0.31
sTNFr	58.87 ± 0.71	46.50 ± 3.00	23.66 ± 3.71
15
Control	9.8 ± 0.76	8.03 ± 0.59	3.00 ± 0.23
sTNFr	28.07 ± 6.00	25.08 ± 5.80	11.42 ± 4.29
21
Control	9.88 ± 1.56	7.80 ± 1.29	3.42 ± 0.60
sTNFr	11.50 ± 1.04	7.31 ± 1.50	1.84 ± 0.52

Figure 2.

 

Comparison of nitrite production by splenic and retinal macrophages after cytokine stimulation. (A) Splenic macrophages (similar to bone marrow–derived macrophages) were activated when stimulated with IFNγ or IFNγ/TNF together. sTNFr-IgG did not protect macrophages against activation with IFNγ/TNF, whether administered before (sTNFr+4h) or at the same time (sTNFr+0h) as a cytokine cocktail (n = 6). (B) Macrophages (principally MG) from normal retina were constitutively low nitrite producers (note differences in the magnitude of response) and were not stimulated by further cytokine addition (n = 3). All samples represented as mean ± SEM.

Figure 3.

 

sTNFr-IgG treatment suppressed generation of nitrite by infiltrating macrophages at the height of disease. (A) At peak disease (day 11) there was a significant suppression of nitrite production in treated animals (P = 0.03). (B) By day 15 after immunization (entering resolution phase of disease) nitrite production had decreased in control animals and there was no difference compared with treated animals (n = 3). Data are the mean ± SEM.

Figure 4.

 

PCR analysis of retinal cells confirms suppression of TNF signal after treatment, with a concomitant early increase in TGFβ. (A) Index ratio of cytokine signal intensity compared with GAPDH demonstrates suppression of TNF signal after treatment throughout the course of EAU. Early in disease there was an increase in TGFβ signal in treated animals and marked IL-1β signal in the early-resolution phase of disease (day 15). (B) Representative of repeated experiments showing quantitative PCR detection gel displaying: A, GAPDH (532 bp); B, IL-6 (453 bp); C, TNF (352 bp); D, IL-1β (295 bp); E, TGFβ (250 bp); and F, GM-CSF (210 bp). Lanes 1 to 3: individual day-10 control animals; lanes 4 to 5: individual day-10 treated animals; lanes 6 to 8: day-11 control animals, and lane 12: a single example of day-11 animals. Lanes G and H: positive control lane and DNA ladder, respectively.

Figure 5.

 

sTNFr-IgG treatment reduces apoptosis within retina in both resident cell and infiltrating ED1⁺ macrophage populations. (A) In control animals, apoptosis (TUNEL⁺ cells) increased dramatically during peak disease and was observed in both ED1⁺ cells and resident cell populations (day 11; P = 0.05). Apoptosis throughout the course of EAU was not statistically increased (B). (C) Representative sections to display (left) dual positive ED1⁺TUNEL⁺ (yellow) cells and TUNEL⁺ (Green-FITC) ED1⁻ cells. Right: comparative sections from treated animals in which TUNEL positivity was absent, despite extensive ED1⁺cell infiltration early in the disease course. Magnification, ×200.