IFN-γ stimulates macrophage activation and NO production, which leads to destruction of the retina in experimental autoimmune uveoretinitis. In this study, we investigate the mechanism of disease resistance in TNF p55 receptor-deficient animals. We show that although T cell priming is relatively unaffected, macrophages lacking the TNF p55 receptor fail to produce NO following IFN-γ stimulation because of a requirement for autocrine TNF-α signaling through the TNF p55 receptor. In contrast to the impaired activation of NO synthesis, MHC class II up-regulation was indistinguishable in wild-type and TNFRp55−/− mice stimulated with IFN-γ. These defects could be overcome by stimulating macrophages with LPS. Together, these results show that selected aspects of IFN-γ activation are controlled by autocrine secretion of TNF-α, but that this control is lost in the presence of signals generated by pathogen-associated molecular patterns recognizing receptors.

Studies of experimental autoimmune uveoretinitis (EAU)3 have provided abundant evidence that TNF-α plays a pivotal role in the pathogenesis of intraocular inflammation and retinal tissue destruction. EAU is an organ-specific autoimmune disease in which Th1 CD4+ T cells, directed toward retinal Ags, produce cytokines, such as IFN-γ, which activate resident and infiltrating mononuclear cells essential for maximal tissue destruction (1). When macrophages are depleted using dichloromethylene diphosphonate (Cl2MDP; clodronate), during the “effector” stage of the disease, i.e., 9–11 days following immunization with retinal Ag, this causes a delay in the onset and severity of EAU, supporting the notion that blood-borne activated macrophages are the major effectors of tissue damage during EAU (2). During peak inflammation, infiltrating macrophages adapt to environmental Th1 T cell cytokines (IFN-γ/TNF-α), and generate NO. This suggests in vivo programming of macrophages within the retina (3) and highlights the NO effector response as a mediator of tissue damage (4), in the generation of a persistent inflammatory response, and as a regulator of T cell apoptosis (5). These inflammatory macrophages induced by a Th1-dominated immune response also secrete other inflammatory mediators (e.g., IL-1, IL-6, and TNF-α) and possess cytotoxic and antimicrobial effector functions (6).

In patients with uveitis, elevated levels of TNF-α are found in the ocular fluid and serum (7), corroborating the conclusion that TNF-α plays a critical role in retinal tissue destruction. In EAU, TNF-α blockade with the p55 TNFR fusion protein (sTNFr-Ig) leads to profound suppression of retinal inflammation and structural damage and suppresses Th1 activity (reduced IFN-γ production) and cell- and macrophage-derived NO production within the retina (8, 9, 10, 11, 12). Clinically, blockade with sTNFr-Ig successfully suppresses autoimmune posterior segment intraocular inflammation with a reduction of peripheral IFN-γ+ CD4+ T cells and increasing numbers of IL-10+CD4+ T cells in responders (10, 11).

The biological activities of TNF-α are mediated by two structurally related but functionally distinct receptors, p55 and p75, which both belong to the TNFR gene family (13, 14, 15). The two receptors are coexpressed on the surface of most cell types, and both can be released by proteolysis to generate soluble molecules capable of binding TNF. Signaling through the TNF p75 receptor leads primarily to cell proliferation. Experiments using receptor-specific Abs (16), -ligands (17), and mice deficient in TNFRp55 and/or TNFRp75 (18) indicate that p55 is the primary signaling receptor on most cell types through which the majority of inflammatory responses classically attributed to TNF occur (19). Furthermore, the discovery that soluble TNF more than membrane-bound TNF is required to generate full expression of inflammatory lesions, as observed in experimental autoimmune encephalomyelitis (EAE; Ref.20), highlights the potential importance of targeting TNFRp55, since it is the principal ligand for soluble TNF (21).

In this paper, we have studied EAU in mice deficient in the TNFRp55. We show that despite relatively normal T cell priming, TNFRp55−/− mice are resistant to EAU induction. Because TNFRp55−/− macrophages have selective defects in their responses to IFN-γ, they do not generate NO without additional stimuli from innate immune receptors. This is not due to a defect in the IFN-γ receptor but rather because of a failure of autocrine TNF-α signaling which is necessary for full macrophage activation in the absence of signals from the innate immune system. Macrophage activation in a proinflammatory milieu may therefore be modified by the presence of pathogen-derived signals, and may be different in autoimmune disease where inflammation occurs in the absence of foreign immune system activators.

TNFRp55-deficient mice (TNFRp55−/−) of background strain C57BL/6 were obtained from The Jackson Laboratory. A breeding colony was established within the Animal Services unit at Bristol University (Bristol, U.K.). Animals were specific pathogen-free, isolator-reared, and maintained in accordance to Home Office Regulations for Animal Experimentation, U.K. C57BL/6 wild-type (TNFRp55+/+) mice were purchased from Harlan Olac. All animals were immunized between 6 and 8 wk of age. For each experiment, female animals were immunized by a s.c. injection of 500 μg/mouse of interphotoreceptor retinoid-binding protein (IRBP) peptide 1–20 (GPTHLFQPSLVLDMAKVLLD; Sigma Genosys), emulsified in v/v CFA (supplemented with 2.5 mg/ml Mycobacterium tuberculosis; Ref.22), with an additional i.p. injection of 1.5 μg of Bordetella pertussis toxin on day 0. In some experiments IRBP 161–180 (SGIPYIISYLHPGNTILHVD; Sigma Genosys) was used as a control peptide. Mice were sacrificed by CO2 asphyxiation on the indicated days following immunization. Eyes were enucleated for histological grading (23) by immunohistochemistry. Histological disease score is graded in two forms: cellular infiltrate and structural changes. Cellular infiltrate is scored within the ciliary body, vitreous, vessels, rod outer segments, and choroid, while structural change is scored within the rod outer segments, neuronal layers, and retinal morphology. Both scores are added together to get a final disease grade (see Table I).

Table I.

Summary of EAU disease scoring

LocationFindingScore
Cellular infiltration   
 Ciliary body Cell infiltrate <5 cells 
 Mild thickening 
 Moderate thickening 
 Gross thickening 
 Vitreous Cells <5 
 Cells 5–25 
 Cells 25–50 
 Cells 50–100 
 Cells >100 
 Vasculitis (mural or extravascular cells) <10% vessels involved 
 10–25% 
 25–50% 
 50–75% 
 >75% 
 Cells in or around wall 
 Mild perivascular cuffing 
 Moderate cuffing 
 Gross cuffing 
 Rod out segments Cell infiltrate 
 Partial loss 
 Moderate loss 
 Subtotal loss 
 Total loss 
 Choroid Cell infiltrate 
 Mild thickening 
 Moderate thickening 
 Gross thickening 
 Granulomas 1 
 Granulomas 2–5 
 Granulomas >5 
Structural/morphological changes   
 Rod outer segments Cell infiltrate 
 Partial loss 
 Moderate loss 
 Subtotal loss 
 Neuronal layers Cell infiltrate 
 Partial loss 
 Moderate loss 
 Subtotal loss 
 Total loss 
 Retinal morphology Folds <10% 
 Folds 10–50% 
 Folds >50% 
LocationFindingScore
Cellular infiltration   
 Ciliary body Cell infiltrate <5 cells 
 Mild thickening 
 Moderate thickening 
 Gross thickening 
 Vitreous Cells <5 
 Cells 5–25 
 Cells 25–50 
 Cells 50–100 
 Cells >100 
 Vasculitis (mural or extravascular cells) <10% vessels involved 
 10–25% 
 25–50% 
 50–75% 
 >75% 
 Cells in or around wall 
 Mild perivascular cuffing 
 Moderate cuffing 
 Gross cuffing 
 Rod out segments Cell infiltrate 
 Partial loss 
 Moderate loss 
 Subtotal loss 
 Total loss 
 Choroid Cell infiltrate 
 Mild thickening 
 Moderate thickening 
 Gross thickening 
 Granulomas 1 
 Granulomas 2–5 
 Granulomas >5 
Structural/morphological changes   
 Rod outer segments Cell infiltrate 
 Partial loss 
 Moderate loss 
 Subtotal loss 
 Neuronal layers Cell infiltrate 
 Partial loss 
 Moderate loss 
 Subtotal loss 
 Total loss 
 Retinal morphology Folds <10% 
 Folds 10–50% 
 Folds >50% 
TOTAL
Infiltrative Grading: Infiltrative 30 Structural Grading: Structural 12 
0.5–2 <2 
3–5 2–5 
6–12 5–8 
13–18 8–10 
19–24 10–12 
25–30   
TOTAL
Infiltrative Grading: Infiltrative 30 Structural Grading: Structural 12 
0.5–2 <2 
3–5 2–5 
6–12 5–8 
13–18 8–10 
19–24 10–12 
25–30   

Bone marrow cells were obtained by flushing the femurs of mice with DMEM. Cells were cultured as previously described (24) in hydrophobic Teflon bags (a gift from M. Munder, Institute of Immunology, Ruprecht-Karls-University, Heidelberg, Germany) in DMEM containing 10% heat-inactivated FCS, 5% normal horse serum, and the supernatant of M-CSF-secreting L929 fibroblasts at a final concentration of 15% (v/v). Cells were removed from culture bags, washed, and stained with F4/80 to determine purity by flow cytometry. Macrophages were used as described.

Assay culture conditions were optimized using BM-Mφ, which were cultured and stimulated as previously described (25). Cytokines used were IFN-γ (20 U/ml), TNF-α (20 U/ml; PeproTech), and TGF-β (10 ng/ml; R&D Systems), alone or administered sequentially in combination, with the administration of each cytokine separated by a 4-h period. Macrophage function was assessed 24 h after addition of the first cytokine LPS (1 μg/ml–0.1 pg/ml; Sigma Genosys) or CpG-ODN (10 μM/ml) (25).

NO generation was measured after 24 h by assaying culture supernatants for the stable reaction product of NO (NO2; nitrite) against a sodium nitrite standard on the same plate. A 100-μl total volume of each cell-free supernatant was incubated v/v with Greiss reagent (0.5% sulfanilamide, 0.05% N-(1-napthyl) ethylenediamine dihydrochloride in 2.5% phosphoric acid) in 96-well flat-bottom plates for 10 min at room temperature. The ODs were measured at 540 nm, with a reference filter of 630 nm.

Spleens were mechanically disrupted and filtered through a 70-μM cell strainer. RBC were lysed using 0.2 and 1.6% sodium chloride. Cells were washed in RPMI 1640 and seeded at 5 × 105 in complete media (RPMI 1640 supplemented with 10% FCS, 2 mM l-glutamine, 100 U/ml penicillin, 100 μg/ml streptomycin, 25 mM HEPES, 50 μM 2-ME, 20 μg of gentamicin, and 1 mM sodium pyruvate) with the indicated concentration of IRBP1–20 or purified protein derivative (tuberlin PPD; Veterinary Laboratories Agency) for a total of 72 h. For the last 16 h, cells were pulsed with 0.5 μCi of [3H]thymidine. Control cultures contained PBS/2% DMSO in place of peptide or 2.5 μg/ml Con A, added to positive control cultures to demonstrate optimum growth conditions and cell viability during the assay. Results were calculated as mean cpm ± 1 SEM of quadruplicate cultures. For cytokine profile analysis of responding cells, parallel 1-ml cultures were set up with 5 × 106 cells and the indicated concentration of IRBP1–20. After 72 h supernatants were harvested, debris removed by centrifugation, and the samples aliquoted and frozen at −70°C until assayed.

In some experiments, wild-type (WT) and TNFRp55−/− animals were immunized with IRBP1–20 as described. On day 11 postimmunization, lymph nodes (LNs) and spleens were removed and CD4 cells were isolated using MACS-positive separation columns (LS+), these cells were used as CD4+ T cells. Spleens and LN cells (LNCs) were taken from nonimmunized mice. CD4 cells were depleted by MACS separation and these cells were used as a source of APCs. CD4 T cells from both WT and TNFRp55−/− were cocultured with APCs from WT or TNFRp55−/− mice, with a range of concentrations of IRBP1–20 or IRBP161–180 (control peptide; SGIPYIISYLHGNTILHVD).

IL-2, IFN-γ, IL-10, IL-12p40, and TNF-α production were assayed by capture ELISA. Briefly, capture Ab (in 0.2 M phosphate buffer) was applied overnight at 4°C (Nunc Immuno plate; Fisher Scientific) then nonspecific binding sites were blocked using 1% PBSA for 1 h at 37°C. The supernatant was added for 1 h at 37°C, the plate washed with 0.5% PBS/Tween 20, and the appropriate detection Ab added for 1 h at room temperature. Extra-avidin peroxidase (Sigma Genosys) was added for 30 min at room temperature followed by chromogen substrate (citrate buffer, with 3, 3′, 5, 5′-tetramethylbenzidine and H2O2), and the reaction stopped using 2 N H2SO4. The mAb pairs for capture and biotinylated detection were obtained from BD Pharmingen. Recombinant cytokines were used as standards, with curves generated from 1/2 dilutions from 0.196–10 ng/ml IL-2, 19.5–1000 pg/ml for IL-10, 1.95–1000 pg/ml for IL-12p40, 0.0048–5 ng/ml for IFN-γ, and 0.097–25 ng/ml for TNF-α.

Eyes removed from animals at various stages of EAU were embedded in optimal cutting temperature (RA Lamb) compound, snap-frozen, and stored at −80°C. Serial 10-μm cryostat sections of the eyes were taken onto poly(l-lysine)-coated slides. Before staining, tissue sections were fixed in acetone and sections were single stained for CD45. Following primary mAb, the appropriate biotinylated secondary Ab was then applied, followed by an avidin-biotin-HRP complex and Vector 3,3′-diaminobenzidine. Stained slides were counterstained with hematoxylin, dehydrated, and mounted in Histomount.

Staining was performed as previously described (8). Briefly, 1 × 105 −5 × 105 cells were stained in combination with F4/80-FITC (CI:A3–1; Serotec), CD40-PE (3/23; BD Pharmingen), I-Ab-biotin (KH74; BD Pharmingen) and streptavidin-conjugated allophycocyanin (BD Pharmingen), for 30 min on ice in PBS containing 1% FCS and 0.1% sodium azide (FACS buffer). Cells were preincubated with purified anti-CD16/CD32 (2.4G2, rat igG2b) to block FcR-mediated binding of staining mAb. Unconjugated mAb were stained by two steps of 30 min on ice with washing (FACS buffer) in between. 7-aminoactinomycin D (BD Pharmingen) was used to assess cell viability.

Statistical analysis was performed by two-tailed unpaired t tests (GraphPad Instant software) among the groups and p values ≤0.05 were considered significant, unless otherwise stated. Results are expressed as mean ± SEM. Disease incidence was compared using Fisher’s exact test (StatsDirect).

It has already been shown that mice lacking the TNFp55 receptor are resistant to EAE (26, 27) and NOD mice with this genotype do not develop diabetes (28). In rats, sTNFr-Ig suppresses EAU and so initially we wished to determine whether TNFRp55−/− mice were resistant to EAU induction.

TNFRp55−/− and WT mice were immunized with IRBP1–20, as previously described (22). Mice were sacrificed on days 11, 14, 18, and 21 postimmunization, eyes were enucleated, and stained with anti-CD45 for histological disease assessment (23). The disease incidence in the TNFRp55−/− mice (4 of 9) was significantly reduced compared with WT mice (11 of 11), implying resistance rather than suppression or a change in disease kinetics (Fig. 1). EAU in WT (C57BL/6) mice typically peaks between days 18 and 21 postimmunization, and in these experiments, disease expression in WT mice was greatest at day 18 with a mean disease score of 11.3 ± 0.8, while on this day TNFRp55−/− mice had a statistically significant reduced disease score of 3.5 ± 0.5 (p = 0.003) (Fig. 1). Furthermore, although WT mice had a lower disease score by day 21 (6.2 ± 0.8), TNFRp55−/− mice continued to have a statistically significant reduced disease score (3.5 ± 0.5; p = 0.03) compared with this group.

FIGURE 1.

A, TNFRp55−/− mice have reduced EAU histological scores. EAU was induced in TNFRp55−/− or WT animals, and eyes were taken at days 18 and 21 postimmunization. Retinal inflammation and structural damage was determined by histological score in animals that developed disease as described in Materials and Methods. TNFRp55−/− mice showed a statistically significant reduction in disease score (p = 0.003) at day 18, which remained until at least day 21 (p = 0.031) (n = 9). Disease incidence was lower in TNFRp55−/− when compared with WT mice (p = 0.008; n = 11; Fisher’s exact test on pooled data). B and C, Pathology of disease. Micrographs B and C are cryosections of retina from mice 21 days postimmunization, stained for CD45. They show severe pathology with CD45-positive cell infiltrate within the vitreous (arrow, 1), retinal vasculitis with further leukocyte infiltration (arrow, 2) of the inner retina (B; WT). In TNFRp55−/− mice significant protection is seen, inflammation is restricted to fewer cells in the inner retina (arrow, 1), fewer retinal folds and reduced retinal detachment (C; TNFRp55−/−).

FIGURE 1.

A, TNFRp55−/− mice have reduced EAU histological scores. EAU was induced in TNFRp55−/− or WT animals, and eyes were taken at days 18 and 21 postimmunization. Retinal inflammation and structural damage was determined by histological score in animals that developed disease as described in Materials and Methods. TNFRp55−/− mice showed a statistically significant reduction in disease score (p = 0.003) at day 18, which remained until at least day 21 (p = 0.031) (n = 9). Disease incidence was lower in TNFRp55−/− when compared with WT mice (p = 0.008; n = 11; Fisher’s exact test on pooled data). B and C, Pathology of disease. Micrographs B and C are cryosections of retina from mice 21 days postimmunization, stained for CD45. They show severe pathology with CD45-positive cell infiltrate within the vitreous (arrow, 1), retinal vasculitis with further leukocyte infiltration (arrow, 2) of the inner retina (B; WT). In TNFRp55−/− mice significant protection is seen, inflammation is restricted to fewer cells in the inner retina (arrow, 1), fewer retinal folds and reduced retinal detachment (C; TNFRp55−/−).

Close modal

Splenocytes from immunized TNFRp55−/− and WT mice were assessed during the course of EAU for proliferation and cytokine production. There was no difference in proliferation and IL-2 production on day 10 (Fig. 2). Interestingly, in some experiments, IFN-γ production on day 12 was greater in the TNFRp55−/− mice. By day 21, TNFRp55−/− splenocytes demonstrate reduced peptide-specific proliferation and cytokine production (Fig. 2), although these differences are not statistically significant. We also assessed responses to PPD and found the same pattern of equivalent responses on day 12 but lower responses by the TNFRp55−/− splenocytes on day 21 (data not shown). IL-4 and IL-10 production were beneath the sensitivity of their respective assays.

FIGURE 2.

During disease, splenocyte proliferation and cytokine production is reduced in TNFRp55−/− mice. Splenocyte proliferation assays were performed as described in the presence of titrated concentrations of IRBP1–20. At 21 days postimmunization, TNFRp55−/− mice (▴) had markedly reduced proliferation and subsequently reduced IL-2 and IFN-γ production when compared with WT (▪).

FIGURE 2.

During disease, splenocyte proliferation and cytokine production is reduced in TNFRp55−/− mice. Splenocyte proliferation assays were performed as described in the presence of titrated concentrations of IRBP1–20. At 21 days postimmunization, TNFRp55−/− mice (▴) had markedly reduced proliferation and subsequently reduced IL-2 and IFN-γ production when compared with WT (▪).

Close modal

To assess priming of TNFRp55−/− CD4+ T cells further, TNFRp55−/− and WT mice were immunized with IRBP1–20, 10 days later, CD4+ T cells were isolated from spleen and LNs and APCs were derived from CD4-depleted naive spleens. CD4+ T cells (WT or TNFRp55−/−) were cocultured with APCs (WT or TNFRp55−/−) and IRBP1–20 for 72 h. Proliferation was assessed by [3H]thymidine uptake, and supernatants assessed for cytokine production by ELISA. TNFRp55−/− CD4 T cells activated with APCs from naive mice showed reduced peptide-specific proliferation and IL-2 production, when compared with WT CD4+ T cells, regardless of whether the APCs were WT or TNFRp55−/− in origin. However, there was no reduction in IFN-γ production and in these experiments TNFRp55−/− CD4+ T cells produced equivalent or greater levels of TNF-α than WT T cells (Fig. 3). Therefore, disease reduction cannot be solely due to a lack of T cell-derived proinflammatory cytokines.

FIGURE 3.

WT and TNFRp55−/− T cells produce equivalent amounts of proinflammatory cytokine. Ten days postimmunization, T cells from WT and TNFRp55−/− were isolated from spleens and lymph nodes, with nonimmunized T cell-depleted spleens and LNCs were used as APCs. T cells from both WT and TNFRp55−/− mice were cocultured with APCs from either WT or TNFRp55−/− immunized mice. Cultures containing WT CD4 T cells had higher proliferation when compared with cultures containing TNFRp55−/− T cells (A). Similarly, cultures containing WT CD4 T cells had higher IL-2 production (B), while culture containing TNFRp55−/− CD4 T cells had higher TNF-α production (C). All cultures had very similar IFN-γ production (D). Representative of two independent experiments.

FIGURE 3.

WT and TNFRp55−/− T cells produce equivalent amounts of proinflammatory cytokine. Ten days postimmunization, T cells from WT and TNFRp55−/− were isolated from spleens and lymph nodes, with nonimmunized T cell-depleted spleens and LNCs were used as APCs. T cells from both WT and TNFRp55−/− mice were cocultured with APCs from either WT or TNFRp55−/− immunized mice. Cultures containing WT CD4 T cells had higher proliferation when compared with cultures containing TNFRp55−/− T cells (A). Similarly, cultures containing WT CD4 T cells had higher IL-2 production (B), while culture containing TNFRp55−/− CD4 T cells had higher TNF-α production (C). All cultures had very similar IFN-γ production (D). Representative of two independent experiments.

Close modal

Because of the discrepancy between near normal T cell priming but significant disease reduction, and because of the importance of macrophage activation to disease pathology in EAU, we examined the response to TNFRp55−/− macrophages to inflammatory mediators. BM-Mφ were generated, seeded at 5 × 105/ml, and stimulated with medium, TGFβ, TNF-α, IFN-γ, or IFN-γ followed by TNF-α 4 h later (IFN-γ/TNF-α). After 24 h, supernatant was taken and tested for nitrite, TNF-α, IL-6, and IL-12p40 production. Following IFN-γ or IFN-γ/TNF-α stimulation, BM-Mφ from WT mice generated significant quantities of nitrite, as expected. However, TNFRp55−/− BM-Mφ did not respond (p < 0.005) (Fig. 4 A). This may indicate that in TNFRp55−/− mice, there is a defect in IFN-γ-induced nitrite production either via an effect on the IFN-γ receptor or because of indirect requirement for signaling through the TNFRp55 (29), which is absent in the knockout mice. However, when we measured levels of TNF-α induced by IFN-γ, these were low compared with those reached following activation with LPS (data not shown).

FIGURE 4.

A, TNFRp55−/− BM-Mφ are unable to produce nitrite in response to IFN-γ stimulation. BM-Mφ were set-up and stimulated with cytokines as previously described (25 ). WT BM-Mφ show the expected response to IFN-γ stimulation with a spike in nitrite production. In contrast, TNFRp55−/− BM-Mφ demonstrated a blunted nitrite response (representative data of eight independent experiments). B, Preincubation of sTNFr-Ig before IFN-γ stimulation results in WT BM-Mφ having a very similar nitrite response as TNFRp55−/− BM-Mφ, by neutralizing TNF production. □, WT response; ▪, TNFRp55−/− response. (Representative of three independent experiments).

FIGURE 4.

A, TNFRp55−/− BM-Mφ are unable to produce nitrite in response to IFN-γ stimulation. BM-Mφ were set-up and stimulated with cytokines as previously described (25 ). WT BM-Mφ show the expected response to IFN-γ stimulation with a spike in nitrite production. In contrast, TNFRp55−/− BM-Mφ demonstrated a blunted nitrite response (representative data of eight independent experiments). B, Preincubation of sTNFr-Ig before IFN-γ stimulation results in WT BM-Mφ having a very similar nitrite response as TNFRp55−/− BM-Mφ, by neutralizing TNF production. □, WT response; ▪, TNFRp55−/− response. (Representative of three independent experiments).

Close modal

To determine whether IFN-γ-induced TNF-α is required to induce nitrite production from BM-Mφ following IFN-γ stimulation, we used soluble TNF p55 receptor-Ig fusion protein (sTNFr-Ig) which binds both soluble and membrane-bound TNF (and therefore prevents any TNFR signaling). sTNFr-Ig suppresses EAU in rats (8), mice (30), and in humans (10). In these experiments, sTNFr-Ig was added to TNFRp55−/− and WT BM-Mφ 1 h before cytokine stimulation. Nitrite production by TNFRp55−/− BM-Mφ remained undetectable in the presence of sTNFr-Ig. The addition of sTNFr-Ig to WT BM-Mφ abolished NO production following IFN-γ or TNF-α/IFN-γ stimulation, resulting in a phenotype identical to TNFRp55−/− BM-Mφ responses (Fig. 4,B). This supports a model in which NO production depends on low-level autocrine production of TNF-α, and given data shown in Fig. 4 A, where exogenous TNF-α does not restore NO in TNFRp55−/− macrophages, infers TNFRp55 dependency. The lack of IFN-γ-mediated NO response was not due to differences in macrophage expression of IFN-γ receptor (CD119), which was equivalent between WT and TNFRp55−/− macrophages (data not shown).

Because IFN-γ-stimulated NO production is critically dependent on locally produced TNF-α, we investigated other molecules important in macrophage effector function. The full expression of macrophage activation includes up-regulation of MHC class II, CD40, and IL-6 production (6, 31). In TNFRp55−/− mice, IFN-γ-induced up-regulation of MHC class II is equivalent to WT (Fig. 5). However, when macrophages are incubated in the presence of IFN-γ and sTNFr-Ig, MHC class II expression in both WT and TNFRp55−/− mice falls to background levels, implying a requirement for TNF-α signaling via TNFRp75. The expression patterns of the costimulators, CD80 and CD86, are equivalent in TNFRp55−/− BM-Mφ and WT BM-Mφ activated with IFN-γ (data not shown). In TNF-α-deficient animals, CD40 up-regulation is known to be defective (32, 33), and this phenotype is corrected by the addition of TNF-α. TNFRp55−/− macrophages stimulated with IFN-γ have a partial defect in CD40 up-regulation compared with WT mice. This partial up-regulation in WT cells is exactly recapitulated in WT cells in the presence of sTNFr-Ig, showing that maximal CD40 up-regulation depends on signals via the TNFRp55 (Fig. 5).

FIGURE 5.

TNFRp75 dependence of MHC II up-regulation and TNFRp55 dependence of maximal CD40 up-regulation. Unstimulated (filled histogram) macrophages or IFN-γ-stimulated WT-Mφ (unbroken line) or TNFRp55−/−-Mφ (broken line) were activated for 24 h in the presence or absence of sTNFr-Ig. MHC II up-regulation is independent of TNFRp55 but blocked by sTNFr-Ig implying TNFRp75 dependence. Maximal CD40 up-regulation is TNFRp55 dependent.

FIGURE 5.

TNFRp75 dependence of MHC II up-regulation and TNFRp55 dependence of maximal CD40 up-regulation. Unstimulated (filled histogram) macrophages or IFN-γ-stimulated WT-Mφ (unbroken line) or TNFRp55−/−-Mφ (broken line) were activated for 24 h in the presence or absence of sTNFr-Ig. MHC II up-regulation is independent of TNFRp55 but blocked by sTNFr-Ig implying TNFRp75 dependence. Maximal CD40 up-regulation is TNFRp55 dependent.

Close modal

IFN-γ is secreted by Th1 T cells and is a critical determinant of the cytokine microenvironment, but in infection other signaling molecules also play an essential role defining the phenotype of the response. One such “stage setting” family of receptors is the family responsible for binding pathogen-associated molecular pattern (PAMP) motifs. To determine whether macrophages derived from TNFRp55−/− mice are capable of generating nitrite on activation through innate immune receptors, we stimulated them with LPS or CpG-ODN and measured their responses. Bone-marrow macrophages stimulated with CpG-ODN did not produce nitrite, unlike macrophage/monocyte cell lines (34). In contrast, bone-marrow macrophages taken from either WT or TNFRp55−/− mice, and stimulated with LPS, produced increased amounts of nitrite, from both WT and TNFRp55−/− macrophages (Fig. 6)

FIGURE 6.

TNFRp55−/− macrophages produce NO in response to LPS. LPS was titrated from 1 μg/ml–0.1 pg/ml to assess macrophage function through PAMPs (for example, LPS through TLR4; CpG (10 μM/ml) through TLR9). The concentration of LPS correlated with nitrite production. Bars indicate IFN-γ or CpG-stimulated production of NO from WT or TNFRp55−/− macrophages. Representative of four independent experiments.

FIGURE 6.

TNFRp55−/− macrophages produce NO in response to LPS. LPS was titrated from 1 μg/ml–0.1 pg/ml to assess macrophage function through PAMPs (for example, LPS through TLR4; CpG (10 μM/ml) through TLR9). The concentration of LPS correlated with nitrite production. Bars indicate IFN-γ or CpG-stimulated production of NO from WT or TNFRp55−/− macrophages. Representative of four independent experiments.

Close modal

Several cytokines have been implicated in the pathogenesis of EAU, but TNF-α plays a predominant role (8, 12), which has lead to studies examining TNF blockade in EAU (10, 11) and other autoimmune diseases such as EAE (26, 27, 35). Although it is known that TNFRp55 is involved in the pathogenesis of posterior uveitis (36), the mechanisms remain unclear.

In this study, we investigated what role TNFRp55 plays in the pathogenesis of EAU using TNFRp55−/− mice and found that this mutation suppresses full expression of disease, not principally by affecting T cell priming or IFN-γ production but through suppression of IFN-γ-induced NO production by macrophages.

TNFRp55−/− mice had reduced disease scores and lower disease incidence. Early in disease (day 10), splenocyte proliferation in response to stimulation with autoantigen or PPD in WT and TNFRp55−/− mice was equivalent but by peak day of disease (day 21), splenocyte proliferation and IFN-γ production were significantly lower (Fig. 2). These data indicate that although T cell priming was successful in TNFRp55−/− mice at day 10, by peak day of disease it was markedly reduced. This may be because disease in TNFRp55−/− mice resolves more quickly than disease in WT mice or because there is less tissue damage in TNFRp55−/− mice and therefore less amplification of the inflammatory autoimmune response. But the detection in some experiments of more IFN-γ (Fig. 2) and more TNF-α (Fig. 3) may also indicate excess negative feedback through the IFN-γ receptor when TNFRp55 is absent, lending support to the idea that TNFRp75 signals are insufficient to compensate for the lack of TNFRp55. Similar findings of increased IFN-γ have also been reported in the spinal cords of TNFRp55-deficient animals with EAE (37). When primed T cells from immunized TNFRp55−/− or WT mice are activated with naive APCs, there is no deficiency in the effector cytokines IFN-γ and TNF-α secreted by TNFRp55−/− T cells although there are slightly lower levels of T cell proliferation (Fig. 3). Because this deficiency is not seen using whole splenocyte preparations, it may reflect less efficient costimulation by unactivated APCs, compared with APCs derived from the spleen of an immunized mouse.

As retinal destruction was markedly reduced in TNFRp55−/− animals, we wanted to determine whether macrophage activation was impaired, since macrophages are the main mediator for tissue damage, because of their ability to cause peroxidation of the lipid membrane, by the production of NO. We generated BM-Mφ in Teflon bags; these F4/80-positive macrophages were nonactivated and had a resting phenotype (CD40low and MHCIIlow; data not shown). TNFRp55−/− macrophages were unable to produce nitrite in response to IFN-γ stimulation or IFN-γ followed by TNF-α stimulation. The inability of macrophages to produce NO under these circumstances helps explain the lack of retinal destruction in vivo. Following disease induction in murine EAU, CD4-positive IFN-γ and TNF-α-secreting T cells are found in the retina, this cytokine milieu classically activates infiltrating macrophages and resident microglia resulting in the production of NO (23, 36, 38). In TNFRp55−/− mice, although T cells are able to infiltrate the eye during EAU, the inability of macrophages to respond to IFN-γ leads to reduced retinal destruction.

To investigate the mechanism by which TNFRp55−/− macrophages fail to produce NO in response to IFN-γ stimulation, we used a soluble TNF-fusion protein (sTNFr-Ig), which binds free TNF-α and inhibits even low levels of TNFR-mediated signaling. This confirmed that in this system TNF-α was necessary for NO production. However, it also raised the possibility that NO production is not the only effector function which may be effected in the absence of the TNFRp55−/−-mediated autocrine signaling pathway. This proposal is consistent with data in the literature demonstrating that TNF-α is required for full up-regulation of CD40 induced by IFN-γ, which has been demonstrated in TNF-α knockout mice (39, 40, 41). To explore this further, we examined CD40 up-regulation in WT and TNFRp55−/− mice and showed that signaling through this receptor was required for full CD40 up-regulation. In contrast, IFN-γ-induced MHC II up-regulation was independent of TNFRp55−/− signaling, but was almost completely blocked in the presence of sTNFr-Ig (Fig. 5). This indicates that that there is a non-TNFRp55 TNF-α-dependent signaling pathway that induces MHC II up-regulation, which may therefore be TNFRp75.

Overall, these results are completely consistent with earlier studies of TNFRp55−/− mice, which found that TNF-α synergizes with IFN-γ in wild-type but not knockout mice (42). However, in this and other studies of TNFRp55−/− mice infected with intracellular parasites (42, 43), in the presence of pathogens, macrophages do produce NO, indicating that there is no absolute requirement for TNFRp55 for NO secretion.

Together, these results show that selected aspects of IFN-γ activation are controlled by autocrine secretion of TNF-α, as shown in Fig. 7. In this system, NO production and MHC II up-regulation are both critically dependent on autocrine secretion of TNF-α, but only NO secretion requires signals from the TNFRp55. CD40 up-regulation is enhanced by signaling via p55 but low-level up-regulation in response to IFN-γ occurs even in the absence of TNF-α. In addition the patterns of response seen following activation with IFN-γ alone can be overridden by the presence of signals generated by selected PAMPs, recognizing receptors, such as TLR4, but not others such as TLR9 (Fig. 6). This may represent discrimination in the type of effector response likely to destroy a pathogen based on cues from the different types of pathogens themselves. This raises the question of the role of these signals in promulgating autoimmune responses. As the spectrum of exogenous and endogenous ligands for these receptors continues to grow (44), whether endogenous ligands for these receptors mimic the effects of exogenous stimuli is an important outstanding question. In this case, we hypothesize that in TNFRp55−/− mice, activation sufficient for priming occurs in the presence of mycobacterial-derived products in draining LNs, but not when TNFRp55-deficient macrophages are exposed to IFN-γ in their absence in the retina.

FIGURE 7.

Selective autocrine signaling from p55 and p75 induced by IFN-γ. Ligation of the IFN-γ receptor produces direct effects on CD40, but only modest up-regulation of MHC II and no NO production. Autocrine production of TNF-α is required for NO production, maximal CD40 up-regulation and MHC II up-regulation, and this is differently regulated by TNFRp55 and TNFRp75.

FIGURE 7.

Selective autocrine signaling from p55 and p75 induced by IFN-γ. Ligation of the IFN-γ receptor produces direct effects on CD40, but only modest up-regulation of MHC II and no NO production. Autocrine production of TNF-α is required for NO production, maximal CD40 up-regulation and MHC II up-regulation, and this is differently regulated by TNFRp55 and TNFRp75.

Close modal

We thank Prof. Geoff Hale at the Therapeutic Ab Centre (Oxford, U.K.). for the gift of the sTNFr-Ig fusion protein. We thank David Copland for technical assistance.

The authors have no financial conflict of interest.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1

This work was supported in part by a grant from the National Multiple Sclerosis Society (RG 3257) to L.B.N. A.D.D. is supported by the National Eye Research Centre.

3

Abbreviations used in this paper: EAU, experimental autoimmune uveoretinitis; EAE, experimental autoimmune encephalomyelitis; IRBP, interphotoreceptor retinoid-binding protein; BM-Mφ, bone marrow-derived macrophage; PPD, purified protein derivative; WT, wild type; LN, lymph node; LNC, LN cell; PAMP, pathogen-associated molecular pattern.

1
Caspi, R. R., C. C. Chan, Y. Fujino, F. Najafian, S. Grover, C. T. Hansen, R. L. Wilder.
1993
. Recruitment of antigen-nonspecific cells plays a pivotal role in the pathogenesis of a T cell-mediated organ-specific autoimmune disease, experimental autoimmune uveoretinitis.
J. Neuroimmunol.
47
:
177
.-188.
2
Forrester, J. V., I. Huitinga, L. Lumsden, C. D. Dijkstra.
1998
. Marrow-derived activated macrophages are required during the effector phase of experimental autoimmune uveoretinitis in rats.
Curr. Eye Res.
17
:
426
.-437.
3
Robertson, M. J., L. P. Erwig, J. Liversidge, J. V. Forrester, A. J. Rees, A. D. Dick.
2002
. Retinal microenvironment controls resident and infiltrating macrophage function during uveoretinitis.
Invest. Ophthalmol. Vis. Sci.
43
:
2250
.-2257.
4
Hoey, S., P. S. Grabowski, S. H. Ralston, J. V. Forrester, J. Liversidge.
1997
. Nitric oxide accelerates the onset and increases the severity of experimental autoimmune uveoretinitis through an IFN-γ-dependent mechanism.
J. Immunol.
159
:
5132
.-5142.
5
Liversidge, J., A. Dick, S. Gordon.
2002
. Nitric oxide mediates apoptosis through formation of peroxynitrite and Fas/Fas-ligand interactions in experimental autoimmune uveitis.
Am. J. Pathol.
160
:
905
.-916.
6
MacMicking, J., Q. W. Xie, C. Nathan.
1997
. Nitric oxide and macrophage function.
Annu. Rev. Immunol.
15
:
323
.-350.
7
Santos Lacomba, M., C. Marcos Martin, J. M. Gallardo Galera, M. A. Gomez Vidal, E. Collantes Estevez, R. Ramirez Chamond, M. Omar.
2001
. Aqueous humor and serum tumor necrosis factor-α in clinical uveitis.
Ophthalmic Res.
33
:
251
.-255.
8
Dick, A. D., P. G. McMenamin, H. Korner, B. J. Scallon, J. Ghrayeb, J. V. Forrester, J. D. Sedgwick.
1996
. Inhibition of tumor necrosis factor activity minimizes target organ damage in experimental autoimmune uveoretinitis despite quantitatively normal activated T cell traffic to the retina.
Eur. J. Immunol.
26
:
1018
.-1025.
9
Robertson, M., J. Liversidge, J. V. Forrester, A. D. Dick.
2003
. Neutralizing tumor necrosis factor-α activity suppresses activation of infiltrating macrophages in experimental autoimmune uveoretinitis.
Invest. Ophthalmol. Vis. Sci.
44
:
3034
.-3041.
10
Murphy, C. C., K. Greiner, J. Plskova, L. Duncan, A. Frost, J. D. Isaacs, P. Rebello, H. Waldmann, G. Hale, J. V. Forrester, A. D. Dick.
2004
. Neutralizing tumor necrosis factor activity leads to remission in patients with refractory noninfectious posterior uveitis.
Arch. Ophthalmol.
122
:
845
.-851.
11
Greiner, K., C. C. Murphy, F. Willermain, L. Duncan, J. Plskova, G. Hale, J. D. Isaacs, J. V. Forrester, A. D. Dick.
2004
. Anti-TNFα therapy modulates the phenotype of peripheral blood CD4+ T cells in patients with posterior segment intraocular inflammation.
Invest. Ophthalmol. Vis. Sci.
45
:
170
.-176.
12
Dick, A. D., L. Duncan, G. Hale, H. Waldmann, J. Isaacs.
1998
. Neutralizing TNF-α activity modulates T-cell phenotype and function in experimental autoimmune uveoretinitis.
J. Autoimmun.
11
:
255
.-264.
13
Loetscher, H., Y. C. Pan, H. W. Lahm, R. Gentz, M. Brockhaus, H. Tabuchi, W. Lesslauer.
1990
. Molecular cloning and expression of the human 55 kd tumor necrosis factor receptor.
Cell
61
:
351
.-359.
14
Schall, T. J., M. Lewis, K. J. Koller, A. Lee, G. C. Rice, G. H. Wong, T. Gatanaga, G. A. Granger, R. Lentz, H. Raab.
1990
. Molecular cloning and expression of a receptor for human tumor necrosis factor.
Cell
61
:
361
.-370.
15
Vandenabeele, P., W. Declercq, R. Beyaert, W. Fiers.
1995
. Two tumour necrosis factor receptors: structure and function.
Trends Cell Biol.
5
:
392
.-399.
16
Sheehan, K. C., J. K. Pinckard, C. D. Arthur, L. P. Dehner, D. V. Goeddel, R. D. Schreiber.
1995
. Monoclonal antibodies specific for murine p55 and p75 tumor necrosis factor receptors: identification of a novel in vivo role for p75.
J. Exp. Med.
181
:
607
.-617.
17
Barbara, J. A., W. B. Smith, J. R. Gamble, X. Van Ostade, P. Vandenabeele, J. Tavernier, W. Fiers, M. A. Vadas, A. F. Lopez.
1994
. Dissociation of TNF-α cytotoxic and proinflammatory activities by p55 receptor- and p75 receptor-selective TNF-α mutants.
EMBO J.
13
:
843
.-850.
18
Rothe, J., W. Lesslauer, H. Lotscher, Y. Lang, P. Koebel, F. Kontgen, A. Althage, R. Zinkernagel, M. Steinmetz, H. Bluethmann.
1993
. Mice lacking the tumour necrosis factor receptor 1 are resistant to TNF-mediated toxicity but highly susceptible to infection by Listeria monocytogenes.
Nature
364
:
798
.-802.
19
Peschon, J. J., D. S. Torrance, K. L. Stocking, M. B. Glaccum, C. Otten, C. R. Willis, K. Charrier, P. J. Morrissey, C. B. Ware, K. M. Mohler.
1998
. TNF receptor-deficient mice reveal divergent roles for p55 and p75 in several models of inflammation.
J. Immunol.
160
:
943
.-952.
20
Ruuls, S. R., R. M. Hoek, V. N. Ngo, T. McNeil, L. A. Lucian, M. J. Janatpour, H. Korner, H. Scheerens, E. M. Hessel, J. G. Cyster, et al
2001
. Membrane-bound TNF supports secondary lymphoid organ structure but is subservient to secreted TNF in driving autoimmune inflammation.
Immunity
15
:
533
.-543.
21
Grell, M., E. Douni, H. Wajant, M. Lohden, M. Clauss, B. Maxeiner, S. Georgopoulos, W. Lesslauer, G. Kollias, K. Pfizenmaier, et al
1995
. The transmembrane form of tumor necrosis factor is the prime activating ligand of the 80 kDa tumor necrosis factor receptor.
Cell
83
:
793
.-802.
22
Avichezer, D., P. B. Silver, C. C. Chan, B. Wiggert, R. R. Caspi.
2000
. Identification of a new epitope of human IRBP that induces autoimmune uveoretinitis in mice of the H-2b haplotype.
Invest. Ophthalmol. Vis. Sci.
41
:
127
.-131.
23
Dick, A. D., Y. F. Cheng, J. Liversidge, J. V. Forrester.
1994
. Immunomodulation of experimental autoimmune uveoretinitis: a model of tolerance induction with retinal antigens.
Eye
8
:(Pt. 1):
52
.-59.
24
Munder, M., K. Eichmann, M. Modolell.
1998
. Alternative metabolic states in murine macrophages reflected by the nitric oxide synthase/arginase balance: competitive regulation by CD4+ T cells correlates with Th1/Th2 phenotype.
J. Immunol.
160
:
5347
.-5354.
25
Erwig, L. P., D. C. Kluth, G. M. Walsh, A. J. Rees.
1998
. Initial cytokine exposure determines function of macrophages and renders them unresponsive to other cytokines.
J. Immunol.
161
:
1983
.-1988.
26
Eugster, H. P., K. Frei, R. Bachmann, H. Bluethmann, H. Lassmann, A. Fontana.
1999
. Severity of symptoms and demyelination in MOG-induced EAE depends on TNFR1.
Eur. J. Immunol.
29
:
626
.-632.
27
Suvannavejh, G. C., H. O. Lee, J. Padilla, M. C. Dal Canto, T. A. Barrett, S. D. Miller.
2000
. Divergent roles for p55 and p75 tumor necrosis factor receptors in the pathogenesis of MOG(35–55)-induced experimental autoimmune encephalomyelitis.
Cell. Immunol.
205
:
24
.-33.
28
Kagi, D., A. Ho, B. Odermatt, A. Zakarian, P. S. Ohashi, T. W. Mak.
1999
. TNF receptor 1-dependent β cell toxicity as an effector pathway in autoimmune diabetes.
J. Immunol.
162
:
4598
.-4605.
29
Espey, M. G., K. M. Miranda, R. M. Pluta, D. A. Wink.
2000
. Nitrosative capacity of macrophages is dependent on nitric-oxide synthase induction signals.
J. Biol. Chem.
275
:
11341
.-11347.
30
Hankey, D. J., S. L. Lightman, D. Baker.
2001
. Interphotoreceptor retinoid binding protein peptide-induced uveitis in B10.RIII mice: characterization of disease parameters and immunomodulation.
Exp. Eye Res.
72
:
341
.-350.
31
Munder, M., M. Mallo, K. Eichmann, M. Modolell.
1998
. Murine macrophages secrete interferon γ upon combined stimulation with interleukin (IL)-12 and IL-18: a novel pathway of autocrine macrophage activation.
J. Exp. Med.
187
:
2103
.-2108.
32
Wesemann, D. R., Y. Dong, G. M. O’Keefe, V. T. Nguyen, E. N. Benveniste.
2002
. Suppressor of cytokine signaling 1 inhibits cytokine induction of CD40 expression in macrophages.
J. Immunol.
169
:
2354
.-2360.
33
Nguyen, V. T., E. N. Benveniste.
2002
. Critical role of tumor necrosis factor-α and NF-κB in interferon-γ-induced CD40 expression in microglia/macrophages.
J. Biol. Chem.
277
:
13796
.-13803.
34
Krieg, A. M..
2002
. CpG motifs in bacterial DNA and their immune effects.
Annu. Rev. Immunol.
20
:
709
.-760.
35
Kassiotis, G., G. Kollias.
2001
. Uncoupling the proinflammatory from the immunosuppressive properties of tumor necrosis factor (TNF) at the p55 TNF receptor level: implications for pathogenesis and therapy of autoimmune demyelination.
J. Exp. Med.
193
:
427
.-434.
36
Dick, A. D., J. V. Forrester, J. Liversidge, A. P. Cope.
2004
. The role of tumour necrosis factor (TNF-α) in experimental autoimmune uveoretinitis (EAU).
Prog. Retin. Eye Res.
23
:
617
.-637.
37
Kelly, L. M., R. D. Wheeler, T. Owens.
2005
. Regulation of interferon γ expression by tumor necrosis factor receptor type 1 signalling.
Clin. Immunol.
1
:(Suppl.):
S208
.
38
Caspi, R. R..
2002
. Th1 and Th2 responses in pathogenesis and regulation of experimental autoimmune uveoretinitis.
Int. Rev. Immunol.
21
:
197
.-208.
39
Gordon, S..
2003
. Alternative activation of macrophages.
Nat. Rev. Immunol.
3
:
23
.-35.
40
Benveniste, E. N., V. T. Nguyen, D. R. Wesemann.
2004
. Molecular regulation of CD40 gene expression in macrophages and microglia.
Brain Behav. Immun.
18
:
7
.-12.
41
Nguyen, V. T., W. S. Walker, E. N. Benveniste.
1998
. Post-transcriptional inhibition of CD40 gene expression in microglia by transforming growth factor-β.
Eur. J. Immunol.
28
:
2537
.-2548.
42
Vieira, L. Q., M. Goldschmidt, M. Nashleanas, K. Pfeffer, T. Mak, P. Scott.
1996
. Mice lacking the TNF receptor p55 fail to resolve lesions caused by infection with Leishmania major, but control parasite replication.
J. Immunol.
157
:
827
.
43
Flynn, J. L., M. M. Goldstein, J. Chan, K. J. Triebold, K. Pfeffer, C. J. Lowenstein, R. Schreiber, T. W. Mak, B. R. Bloom.
1995
. Tumor necrosis factor-α is required in the protective immune response against Mycobacterium tuberculosis in mice.
Immunity
2
:
561
.
44
Gordon, S..
2002
. Pattern recognition receptors: doubling up for the innate immune response.
Cell
111
:
927
.-930.