Requirement of B7-Mediated Costimulation in the Induction of Experimental Autoimmune Anterior Uveitis

Abstract

purpose. To study the role of costimulatory signaling through the CD28-B7 interaction in experimental autoimmune anterior uveitis (EAAU).

methods. Naive Lewis rats were immunized with insoluble melanin-associated antigen (MAA) derived from bovine iris and ciliary body. CTLA4-Fc, a recombinant protein comprised of the extracellular domain of human CTLA4 bound to mouse IgG2a Fc, was used to block the CD28-B7 interaction. A mutant version (CTLA4-Fc-mutant) was used as a control. The effect of CTLA4-Fc on the in vivo induction of disease with MAA was studied. Subsequently, the mechanism by which CTLA4-Fc blocked the interaction of CD28 and B7 was investigated in vivo, using the adoptive transfer of T cells derived from CTLA4-Fc–treated rats, and in vitro, using the proliferative response and cytokine production of MAA-T cells in the presence of CTLA4-Fc.

results. CTLA4-Fc markedly reduced the incidence and severity of EAAU in Lewis rats after sensitization with MAA. The adoptive transfer of sensitized T cells from CTLA4-Fc–treated donors did not induce EAAU in naive recipients. CTLA4-Fc inhibited the expansion of antigen-specific MAA-T cells and the production of TNF-α.

conclusions. The costimulatory signal delivered through CD28-B7 is required for the induction and pathogenesis of EAAU. In the absence of this signal, antigen-specific expansion of MAA reactive T cells as well as production of TNF-α is inhibited. Abrogation of this costimulatory signal may be an important therapeutic option for EAAU.

Acute anterior uveitis (AAU), the most common form of uveitis, is an intraocular inflammatory disease of the anterior uvea of the eye. Although a single episode of the disease usually does not result in permanent visual loss, the recurrent nature of AAU is associated with significant clinical complications such as cystoid macular edema, cataract formation, and glaucoma. The cause of AAU is unknown. However, it is thought to be an autoimmune disease similar to the experimental model of autoimmune anterior uveitis (EAAU). 

Broekhuyse and colleagues, in 1991, first described an organ-specific autoimmune disease in the Lewis rat induced by a single injection of melanin-associated antigen (MAA). ¹  

Although they termed this disease EAAU, it subsequently became apparent that the experimental model that most closely resembled AAU was induced solely by a single injection of MAA derived from the bovine iris and ciliary body without melanin-associated antigen from the choroidal and retinal pigment epithelium. ^{2 3} Histopathologic examination of eyes with EAAU reveal a severely inflamed iris and ciliary body infiltrated by mostly mononuclear cells with minimal choroidal and no retinal involvement. Immunohistochemical study of inflamed eyes. ⁴ and the adoptive transfer of AAU by activated T cells ³ suggest that EAAU is a T-cell–mediated autoimmune disease. However, the precise mechanism by which autoimmunity to ocular antigens develops remains to be determined. 

Recently, the role of costimulatory signaling in the induction of autoreactive T cells in autoimmune diseases has been intensively studied. The activation and differentiation of T cells require both antigen/MHC recognition and costimulatory signals. ⁵ The signal delivered by the T-cell reactivity determines the antigen specificity of the response. The second signal, termed costimulation, is provided by accessory molecules on antigen-presenting cells (APCs) and appears to be necessary for functional T-cell activation. There are several receptor–ligand pairs that can provide costimulation. However, the signal provided by the B7 family of cell surface molecules, with its receptors on T cells, CD28, and CTLA-4, seems to be dominant for T-cell activation. ^{6 7 8}  

At least two members of the B7 family of CD28 ligands have been defined, B7-1 (CD80) and B7-2 (CD86). ^{9 10 11 12} These molecules, although homologous, are each capable of providing costimulation to T cells for proliferation and IL-2 production. A mouse genetically deficient for B7-1 is essentially immunocompetent. ⁹ Likewise, T cells from a mouse deficient for CD28 cannot produce IL-2 after stimulation with anti-CD3 antibody. ¹³ A mouse deficient for CTLA-4 has a fetal phenotype analogous a lymphoproliferative disorder. ¹⁴ Thus, the costimulatory signal derived through this receptor ligand is important for the development of an effective immunoresponse. 

CTLA4-Fc, a recombinant fusion protein consisting of the extracellular domain of human CTLA4 fused to mouse IgG2a Fc, binds to both B7-1 and B7-2 and can thereby block the interaction between B7 and CD28 or CTLA-4. ¹⁵ Administration of CTLA4-Ig prevented rat cardiac allograft rejection and pancreatic islet cell xenograft rejection in mice. ^{16 17} In both instances, it appeared the mechanism of suppression involved the induction of antigen-specific tolerance. The costimulation provided by B7 also appears to be important for the development of autoimmunity. CTLA4-Ig treatment, initiated the day before immunization of mice with myelin basic protein (MBP) and continued for 3 weeks, prevented development of the clinical and histologic manifestations of experimental allergic encephalomyelitis (EAE). ^{18 19} CTLA4-Ig treatment is also effective if initiated 10 days postimmunization with MBP, when clinical disease is event. ^{20 21 22 23} In the adoptive transfer model of EAE, moreover, treatment of donor cells with CTLA4-Ig during ex vivo exposure to MBP inhibited T-cell proliferation and IL-2 secretion in response to MBP and prevented transfer of EAE to recipients. ^{22 24} As is the case with EAE, treatment with CTLA4-Ig prevented the development of disease in other animal models of organ-specific autoimmunity—for example, collagen-induced arthritis in the BB/Wor rat, ²⁵ , autoimmune oophoritis induced by zona pellucida glycoprotein ZP23 peptide, ²⁶ and experimental antiglomerular basement membrane autoimmune glomerulonephritis. ^{27 28} A recent clinical study suggested a potential therapeutic use for this novel immunomodulatory approach in T-cell–mediated diseases. ²⁹ T-cell costimulation in patients with psoriasis vulgaris was blocked using CTLA4-Ig and was observed to have a beneficial effect. 

Using an animal model of AAU, we investigated the role of ongoing T-cell costimulation in the development and pathogenesis of ocular autoimmunity. In this article, we have shown that interference with the B7 costimulatory signal through CD28 by CTLA4-Fc reduced both the incidence and the severity of intraocular inflammation in EAAU. Furthermore, CTLA4-Fc resulted in a profound decrease in the proliferative response and TNF-α production by a rat MAA-specific T-cell line. Our results suggest a critical role for B7-mediated costimulation in the development of intraocular inflammation and identifies a promising therapeutic approach for the treatment of autoimmune uveitis. 

Methods

All animal studies conformed to the ARVO Statement on the Use of Animals in Ophthalmic and Vision Research. Institutional approval was obtained, and institutional guidelines regarding animal experimentation were followed. 

Rats

Pathogen-free male Lewis rats aged 5 to 6 six weeks (Harlan Sprague–Dawley, Inc., Indianapolis, IN) were housed and maintained at the animal facility of Washington University Medical School. 

CTLA4-Fc

Soluble recombinant proteins comprising the extracellular domain of wild-type human CTLA4 or a mutated form of CTLA4 lacking B7 binding activity (control) were obtained as fusion proteins to mouse IgG2a Fc. ¹⁵ In the experimental panel, 100 μg of CTLA4-Fc was injected intraperitoneally on days −1, 0, +1, +2, and every other day after MAA injection, for a total of eight doses (0.8 mg). The control panel received identical injections with 100 μg of the control CTLA4-Fc. 

Animal Model of EAAU

The animal model of EAAU in Lewis rats has been previously described. ^{2 3} Briefly, Lewis rats 5 to 6 weeks of age were immunized with 100 μg of insoluble MAA, isolated from bovine iris and ciliary body, in the hind footpad with a single dose. Animals were examined daily after 12 days postimmunization for clinical signs of acute anterior uveitis by slit-lamp biomicroscopy and graded using criteria previously published. ^{1 2} Animals typically develop acute inflammation localized to the anterior uvea by day 14 to 18 postimmunization, which peaks around day 20 to 22, and spontaneously resolves by day 30. 

Adoptive Transfer of EAAU

Draining popliteal lymph node cells derived from CTLA4-Fc–treated rats at 15 or 16 days postimmunization were cultured at 2 × 10⁶ cells/ml in DMEM (containing 10% FCS, 1% sodium pyruvate, 5 × 10⁻⁵ M 2-ME, 10 mM HEPES, 1% nonessential amino acid and antibiotics) with 5μ g/ml of soluble MAA. The soluble MAA was prepared by the methods previously described. ² After 3 days in culture the cells were harvested, washed, and purified from dead cells by Ficoll-density centrifugation. A minimum of 7.5 × 10⁶ viable cells was injected into naive Lewis rats through their tail veins. 

Establishment and Maintenance of MAA-Specific T-Cell Lines

The method described by Rozenszajn et al. ³⁰ was modified and used to generate MAA-specific T-cell line. Draining popliteal lymph nodes were collected at day 16 postimmunization with MAA. Single cell suspensions were placed into dishes and incubated for 4 to 6 hours to separate T cells from the adherent macrophages and B cells. The nonadherent cell population including lymphocytes were mixed at a concentration of 5 × 10⁶ cells/ml with 2.5 × 10⁶ cells/ml of naive syngeneic, irradiated (2500 rads) accessory spleen cells and 5 μg/ml of soluble MAA ³ in RPMI-1640 (supplemented with antibiotics, 50 μM 2-ME, and 3% syngeneic rat serum). The cells were cultured in 24-well tissue culture plates for 4 days at 37°C, 5% CO₂ and then for 72 hours in 24-well dishes with 5% conditioned medium (CM). The CM containing IL-2 was obtained from the supernatants of Lewis rat spleen cells stimulated with Con A (5μ g/ml) in a culture medium containing complete RPMI-1640 and 10% inactivated human pooled serum. ³⁰ Cells (2 × 10⁶ cells/ml) were restimulated with MAA (5μ g/ml) in the presence of 5 × 10⁶ syngeneic, irradiated antigen presenting spleen cells for 4 days and underwent continued expansion in medium containing IL-2. T-cell lines were maintained with repetitive stimulation and expansion cycles. After three cycles of expansion, cells were used to assess antigen-specific proliferative responses. 

Proliferative Response of Cell Lines to Antigen

T cells (2.5 × 10⁴ per well) were cultured with or without 5 μg/ml MAA in the presence of irradiated naive spleen cells (4 × 10⁵ per well) at 37°C for 72 hours. To test whether the inhibitory effect of CTLA4-Fc could be reversed by addition of IL-2, 5% or 10% of CM was added into the wells with CTLA4-Fc (10 μg/ml).[ ³H]Thymidine incorporation was assessed during the last 16 hours by a microplate scintillation counter (Packard, Meriden, CT). The proliferative response was expressed as mean cpm ± SD of the triplicate determination. 

ELISA

MAA T cells (1 × 10⁵ per well) in 96-well plate were cultured with irradiated syngeneic naive spleen cells (2 × 10⁶ cells/well) in the presence of 10 μg/ml of CTLA4-Fc or mutant CTLA4-Fc. Culture supernatants were collected after 48 hours. TNF-α released in the supernatants was assayed with the R&D (Minneapolis, MN) ELISA kit according to the manufacturer’s instruction. 

Statistical Analysis

Data were analyzed and compared using Student’s t-test, and differences were considered statistically significant with P < 0.05. 

Results

CTLA4-Fc on Incidence and Severity of EAAU

The role of B7 costimulation in EAAU in vivo was studied by observing the clinical effect of CTLA4-Fc treatment on the incidence and severity of the disease. Animals were injected with 100 μg CTLA4-Fc (experimental group) or CTLA4-Fc mutant (control group) intraperitoneally on days −1, 0, +1, +2, and every other day after MAA immunization. The animals were observed for clinical signs of inflammation by slit-lamp microscope starting day 12 postimmunization up to day 30 postimmunization. All animals (16 eyes) treated with the CTLA4-Fc mutant developed EAAU, with a mean onset of 17 ± 1 days postimmunization. The disease course is similar to the natural course of disease without treatment. In contrast, only 2 of 16 eyes in animals treated with CTLA4-Fc developed EAAU within 30 days postimmunization. Furthermore, the uveitis in these two eyes was very mild, in contrast to the severe inflammation observed in control eyes (Table 1) and disappeared over 1 week. 

Three rats from each of the two groups were killed on day 18. Their eyes were enucleated and studied histologically. Eyes obtained from animals treated with the CTLA4-Fc mutant (Fig. 1A) demonstrated marked infiltration of the anterior segment with inflammatory cells—primarily mononuclear cells. Cellular infiltration was noted in the anterior chamber, iris, and ciliary body and was associated with congestion and vascular engorgement of the iris and ciliary body. In contrast, eyes from recipients treated with CTLA4-Fc demonstrated no inflammation in the anterior segment (Fig. 1B) . Thus, our histopathologic studies confirmed our clinical observation—namely, that the anterior uveitis associated with EAAU was effectively prevented by treatment of the recipient with CTLA4-Fc during the induction of disease. This result suggested that the CD28-B7 interaction is required for the initiation of EAAU. 

Adoptive Transfer of MAA-T Cells from Rats Treated with CTLA4-Fc

We next explored whether MAA-responsive T cells from CTLA4-Fc–treated animals were able to transfer intraocular inflammation. First, we investigated whether T cells capable of inducing EAAU were still present and functional in CTLA4-Fc–treated rats. Draining popliteal lymph node cells from CTLA4-Fc–treated donors were cultured in vitro for 3 days with MAA before adoptive transfer into naive syngeneic Lewis rats. Animals receiving T cells were followed starting day 4 postadoptive transfer until 3 weeks after T-cell injection. As Table 2 demonstrates, all 9 recipients received 7.5 × 10⁶ activated lymphoid cells and none developed EAAU (0/18 eyes) within 20 days of posttransfer. In contrast, lymphoid cells from control (mutant CTLA4-Fc)-treated rats induced EAAU in all eight recipients (16/16 eyes) by day 7 ± 1 posttransfer. Five rats received equal numbers (7.5 × 10⁶) of both CTLA4-Fc–treated and control-treated lymphocytes. By day 5 ± 1 all 10 eyes developed EAAU. Histopathologic examination of the induced anterior uveitis revealed a pattern consistent with that typically observed in EAAU (Fig. 2) . We concluded that a long-term inhibitory effect of MAA-specific T cells by CTLA4-Fc was achieved. Inhibitory T cells were not formed because mixture T cells from control-treated and CTLA4-Fc–treated donors did not block development of uveitis in the naive recipients. 

CTLA4-Fc on MAA-Specific T-Cell Proliferation

MAA-responsive T-cell lines were cultured in vitro with MAA and CTLA4-Fc to address whether blockade of the CD28-B7 interaction inhibited expansion of MAA-responsive T cells was the mechanism of inhibition of disease. The proliferative response of rat MAA-specific T cells was evaluated in vitro in the presence of 10 μg/ml of CTLA4-Fc (experimental group) or mutant CTLA4-Fc (control group). Representative results of the proliferative response, as determined by[ ³H]thymidine incorporation, are shown in Figure 3 . The proliferative response of the MAA-specific T-cell line was significantly inhibited by CTLA4-Fc (P < 0.001) but not by mutant CTLA4-Fc. The inhibition of the proliferative response by CTLA4-Fc could be reversed by coculture with IL-2 in a dose-dependent fashion. These results suggest the effect of CTLA4-Fc in this model is due to inhibition of the expansion of MAA-T cells that cause EAAU. 

CTLA4-Fc on TNF-α Production of MAA-Specific T Cells

TNF-α is an important proinflammatory cytokine in EAAU. ³¹ To examine whether CD28-B7 blockade also inhibited the production of TNF-α, supernatants from T cells stimulated with MAA in the presence of CTLA4-Fc or its mutant were assessed for TNF-α by ELISA. As shown in Figure 4 , CTLA4-Fc significantly reduced TNF-α secretion. 

Discussion

T-cell costimulation caused by interaction between CD28 and B7 appears to be required for the development of many experimental autoimmune diseases. However, the requirement for the CD28-B7 interaction between MAA-responsive T cells and APCs in the development and pathogenesis of EAAU in the Lewis rat has not been studied. Previous work in other rat models of autoimmunity has demonstrated that CD28-B7 costimulatory blockade by CTLA4-Ig is effective in inhibiting disease. ^{27 28 32 33} The systemic administration of CTLA4-Ig was effective in preventing and treating EAE induced in the Lewis rat. ³² Administration of CTLA4-Ig from the time of immunization prevented the development of collagen type II–induced arthritis, whereas treatment during ongoing disease resulted in amelioration of arthritis. ³³ CTLA4-Ig treatment ameliorated experimental autoimmune glomerulonephritis in the WYK rat by reducing autoantibody production and cellular infiltration of glomeruli. ^{27 28} Lately, Silver and colleagues ³⁴ demonstrated that blockade of costimulation through B7-CD28 inhibited experimental autoimmune uveoretinitis (EAU) but did not reduce long-term tolerance. Similarly, in our experiments CTLA4-Fc administered during the priming of MAA-T cells prevented the induction of intraocular inflammation. Since CTLA4-Fc blocks the CD28 interaction with both B7-1 and B7-2 our observations suggest that a costimulatory signal delivered through the CD28-B7 interaction between T cells and APCs is necessary for the induction of EAAU. 

Although previous studies demonstrated that CTLA4-Fc treatment in vivo predominantly produced immunosuppression, the mechanism by which CTLA4-Fc inhibits autoimmunity in different animal models remains to be determined. We observed in our adoptive transfer experiments that T cells from CTLA4-Fc–treated rats, even after in vitro stimulation with MAA, were not able to adoptively transfer disease in naive recipients. These results suggest that CTLA4-Fc treatment results in the nonresponsiveness of MAA-specific T cells. Additionally, adoptive transfer experiments were performed in which lymphoid cells from both control (CTLA4-Fc mutant) and CTLA4-Fc–treated rats when combined before injection. These experiments did not prevent the induction of EAAU. The time course and severity of disease were similar to that observed after the adoptive transfer of T cells from control-treated rats alone. Thus, it appears that blockade of the CD28-B7 interaction during the induction of EAAU did not result in the generation of antigen-specific regulatory (i.e., suppressor) T cells. Regulatory cells are now known to express CTLA-4, ^{35 36} so their function may also be inhibited by CTLA4-Fc. The consequences of this for long-term antigen specific regulation within the patient remain to be further investigated. 

In vitro experiments have demonstrated that antigen-reactive T cells are rendered anergic if they are exposed to the peptide-MHC complex on APCs which lack B7 molecules. ⁵ The anergic T cells failed to expand and differentiate into functional T cells. In vivo experiments in autoimmune models of disease and solid organ transplantation also suggest that CTLA4-Fc can induce long-term unresponsiveness to autoantigens, ³⁷ as well as donor-specific tolerance, ¹⁶ presumably through T-cell anergy. In our in vitro experiment CTLA4-Fc inhibited the MAA-responsive T-cell proliferation and this inhibition was reversed by the addition of IL-2. This result suggests that CD28-B7 blockade inhibits MAA-specific T-cell expansion by inhibition of IL-2 production. In addition, the expansion and effector functions of autoreactive T cells can be sufficiently controlled by short-term CD28 blockade, which probably explains why T cells from rats treated with CTLA4-Fc were unable to induce disease in naive rats by adoptive transfer. These conclusions are consistent with the observations recently reported in EAU where it was found that B7-CD28 blockade prevented generation of effector T cells. ³⁴  

CTLA4-Fc treatment does not always induce T-cell anergy in autoimmune models of disease. In murine EAE, for example, CTLA4-Fc treatment prevented induction of disease, but T cells from immunized CTLA4-Fc–treated animals transferred EAE to naive recipients. ²⁰ Administration of CTLA4-Fc at the time of immunization reduced production of IFN-γ and IL-2 (secreted by Th1 cells) in the central nervous system but not IL-4 and IL-10 (secreted by Th2 cells). ²¹ These results indicated that B7 blockade by CTLA4-Fc may not just only inhibit a T-cell autoimmune response but also alter the dynamic balance of T-cell subsets and their secreted products. Previous studies in EAAU have identified the production of TNF-α during the induction of disease. ³¹ Administration of CTLA4-Fc significantly inhibited the in vitro production of TNF-α and may be a mechanism by which EAAU is inhibited. 

Differences in the kinetics and the level of expression of B7-1 and B7-2 on APCs have been reported during the generation of an immune response. ^{12 38} Consequently, the effect observed after blocking B7-1 or B7-2 in various models of autoimmune disease may differ. ^{39 40} The expression pattern and kinetics of B7-1 and B7-2 on APCs within the eye, an immunologically privileged site, may very well differ from that of a conventional site. Furthermore, the cell type expressing B7-1 or B7-2 may be important in determining the type of immune response generated. Because CTLA4-Fc binds both B7-1 and B7-2 on the APC, our current data suggest an essential role for B7-CD28–mediated costimulation in the induction of uveitis. However, the precise role of costimulatory signals on APCs in the draining lymph node and the eye is currently being studied. 

Blockade of CD28-B7 costimulation has the potential for inhibition of the T-cell–mediated autoimmune response in AAU in humans, regardless of the specific autoantigen(s) involved. Thus, blockade of B7-mediated costimulation by CTLA4-Fc may be a potential therapeutic approach in uveitis. 

 Supported in part by National Eye Institute Grants EY-51341 and EY-51461, as well as Research to Prevent Blindness, New York, New York.

 Submitted for publication December 21, 2000; revised April 11, 2001; accepted April 19, 2001.

 Commercial relationships policy: N.

 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.

 

Table 1.

 

Effects of CTLA4-Fc on the Clinical Course of EAAU

Treatment*	Eyes with EAAU					Day of Onset, †
		Incidence	Mild	Severe
CTLA4-Fc mutant	16/16	0	16	17 ± 1
CTLA4-Fc	2/16	2	0	18 ± 0

*  Treatment given on days −1, 0, 1, 2, 4, 6, 8, and 10 after induction with 100 μg insoluble MAA. The incidence of EAAU is given as the number of positive eyes/the number of total eyes, judged by slit-lamp biomicroscopy.

†  Mean ± SD.

Figure 1.

 

Histopathologic sections from CTLA4-Fc– and control (mutant CTLA4-Fc)-treated rats. (A) Lewis rat, 18 days postimmunization with bovine MAA, treated with mutant CTLA4-Fc. The iris and ciliary body (CB) were severely inflamed and infiltrated with inflammatory cells (arrows). (B) Lewis rat, 18 days postimmunization with bovine MAA, treated with CTLA4-Fc. No inflammation was noted in the anterior chamber, iris, or ciliary body. Magnification, ×100.

Table 2.

 

EAAU in Naive Rats Induced by Adoptive Transfer of Lymphocytes from Either Control (Mutant CTLA4-Fc) or CTLA4-Fc–treated Donors

Treatment*	Eyes with EAAU					Day of Onset, †
		Incidence	Mild	Severe
Control treated donor cells	16/16	0	16	7 ± 1
CTLA4-Fc–treated donor cells	0/18	0	0	0
Mixture of cells	10/10	0	10	5 ± 1

*  Naïve Lewis rats received lymphocytes derived from control (mutant CTLA4-Fc)-treated or CTLA4-Fc–treated rats as described in Material and Methods. The incidence of EAAU is given as the number of positive eyes/the number of total eyes judged by slit-lamp biomicroscopy.

†  Mean ± SD.

Figure 2.

 

Histopathologic sections of eyes after adoptive transfer of lymphocytes. Draining popliteal lymph node cells from control (mutant CTLA4-Fc)- and CTLA4-Fc–treated, MAA-immunized rats were cultured in vitro for 3 days with 5 μg/ml MAA before adoptive transfer into naive Lewis rats. Cells (7.5 × 10⁶ cells) were injected into the recipient’s tail vein, and clinical signs of EAAU followed starting day 4 posttransfer. On day 12 eyes were taken for histologic examination. (A) Rats receiving cells from control-treated rats. The anterior chamber, iris, and ciliary body (CB) were severely inflamed and infiltrated with inflammatory cells (arrows). (B) Rats receiving cells from CTLA4-Fc–treated rats. (C) Rats receiving a mixture of cells (7.5 × 10⁶ cells from each treatment) from both control and CTLA4-Fc–treated rats. Magnification, ×100.

Figure 3.

 

Effect of CTLA4-Fc on the proliferative response of a rat MAA-T cell line in vitro. T cells were plated in triplicate in 96-well plates with irradiated naive syngeneic spleen cells. They were cultured in the presence of medium, MAA, MAA + CTLA4-Fc, MAA + mutant CTLA4-Fc, or MAA+ CTLA4-Fc and either 5% or 10% CM (containing IL-2) for 3 days. During the final 18 hours, 0.5 μCi/well [³H]thymidine was added. The plates were harvested and counted by a microplate scintillation counter. Results are presented as the mean; error bars, SD.

Figure 4.

 

Effect of CTLA4-Fc on the TNF-α production of a rat MAA-specific T-cell line. T cells were plated in triplicate in 96-well plates with irradiated naive syngeneic spleen cells. They were cultured in the presence of medium, MAA, MAA + CTLA4-Fc, or CTLA4-Fc mutant for 2 days. TNF-α in the supernatants was determined by ELISA. Results are presented as the mean; error bars, SD.

The authors are grateful to Berlinda McMahan for her histology expertise and Russ Hornbeck for his photographs.