The CD134-CD134 ligand (CD134L) costimulatory pathway has been shown to be critical for both T and B cell activation; however, its role in regulating the alloimmune response remains unexplored. Furthermore, its interactions with other costimulatory pathways and immunosuppressive agents are unclear. We investigated the effect of CD134-CD134L pathway blockade on allograft rejection in fully MHC-mismatched rat cardiac and skin transplantation models. CD134L blockade alone did not prolong graft survival compared with that of untreated recipients, and in combination with donor-specific transfusion, cyclosporine, or rapamycin, was less effective than B7 blockade in prolonging allograft survival. However, in combination with B7 blockade, long-term allograft survival was achieved in all recipients (>200 days). Moreover, this was synergistic in reducing the frequency of IFN-γ-producing alloreactive lymphocytes and inhibiting the generation of activated/effector lymphocytes. Most impressively, this combination prevented rejection in a presensitized model using adoptive transfer of primed lymphocytes into athymic heart transplant recipients. In comparison to untreated recipients (mean survival time (MST): 5.3 ± 0.5 days), anti-CD134L mAb alone modestly prolonged allograft survival (MST: 14 ± 2.8 days) as did CTLA4Ig (MST: 21.5 ± 1.7 days), but all grafts were rejected within 24 days. Importantly, combined blockade further and significantly prolonged allograft survival (MST: 75.3 ± 12.7 days) and prevented the expansion and/or persistence of primed/effector alloreactive T cells. Our data suggest that CD134-CD134L is a critical pathway in alloimmune responses, especially recall/primed responses, and is synergistic with CD28-B7 in mediating T cell effector responses during allograft rejection. Understanding the mechanisms of collaboration between these different pathways is important for the development of novel strategies to promote long-term allograft survival.
It is now firmly established that costimulatory signals, primarily through the ligation of B7-CD28 and CD40-CD154, are necessary for full T cell activation and play a crucial role in mediating allograft rejection. In several models, CD28-B7 blockade alone or in combination with donor-specific transfusion (DST),3 has been shown to prevent acute rejection, induce long-term allograft survival (1, 2, 3, 4), and prevent the development and progression of chronic rejection (5, 6). However, the same degree of efficacy is not seen when more stringent transplantation models have been used (7, 8, 9, 10). Moreover, CD28- or CD154-deficient animals are still capable of rejecting allografts acutely, albeit at a slower tempo, as well as developing chronic rejection (9, 11, 12). Therefore, it appears that CD28-B7- and CD40-CD154-independent mechanisms can mediate full T cell activation. These include effector cells that are less reliant on CD28 and CD154 signaling for activation, such as CD8+ cells, NK cells, and memory cells, or activation using alternative costimulatory pathways, such as inducible costimulator-B7h, CD134-CD134 ligand (CD134L), and CD70-CD27, which may play a greater role in the maintenance phase of the immune response (13). Of course, these mechanisms may not be mutually exclusive.
Recent data have demonstrated that numerous members of the TNF superfamily are capable of providing costimulatory signals and regulating T cell immune responses (13, 14). The most studied of these pathways is CD40-CD154 (15), blockade of which is synergistic with B7 blockade in preventing allograft rejection (7, 16, 17). Newer members of the superfamily are CD134 and CD134L (also termed OX40 and OX40 ligand), which are expressed, respectively, on activated T cells (18, 19, 20, 21) and APC (22, 23) (including activated B cells and dendritic cells) as well as vascular endothelial cells (24). In vitro and in vivo evidence has demonstrated that the CD134-CD134L interaction can provide a costimulatory signal to T cells, increasing T cell proliferation and cytokine production (25, 26) as well as influencing B cell proliferation and Ig production (27, 28). CD134+ T cells have been implicated in various immune-mediated diseases, such as experimental autoimmune encephalomyelitis (29, 30), rheumatoid arthritis (31, 32), inflammatory skin disease (33), as well as graft-vs-host disease (34). In experimental autoimmune encephalomyelitis, administration of a neutralizing anti-CD134L mAb or OX40-Fc fusion protein ameliorates disease (35, 36), while in animals lacking CD28 costimulation, CD134-CD134L blockade can completely abrogate disease (37). These data suggest that CD134-CD134L costimulation may be in part responsible for the CD28-B7-independent activation of T cells, at least in an autoimmune model. In the present study, we used a newly developed anti-rat CD134L mAb, ATM-2 (38), in allogenic cardiac and skin transplantation models, as well as in a more stringent sensitized model, to explore the role of the CD134-CD134L pathway in mediating allograft rejection and its interaction with the CD28-B7 pathway.
Materials and Methods
Eight- to 10-wk-old inbred male Wistar-Furth (WF, RT1u) or ACI (RT1a) rats were used as heart or skin allograft donors and Lewis (LEW, RT1l) animals as recipients (all from Harlan Sprague-Dawley, Indianapolis, IN). Athymic LEW/Mol-rnu rats were purchased from M&B (Ry, Denmark) and bred in our facility. All animals were housed and cared for under National Institutes of Health guidelines.
Heterotropic vascularized cardiac transplants were performed by standard microvascular techniques (39, 40). Donor hearts were engrafted into the recipient’s abdomen using an end-to-side anastomosis of the donor ascending aorta and pulmonary artery with the recipient’s abdominal aorta and vena cava. Transplantation tolerance was confirmed by engrafting long-term transplant survivors with a second original donor or third-party donor heart. Graft function was assessed by daily palpation, and rejection was confirmed at autopsy.
Full thickness skin grafts were prepared from the lateral thoracic skin of the donor, cut into 2.5-cm circular pieces, and then maintained at 4°C until transplantation. Donor skin was engrafted onto the recipient’s lumbar region and checked daily for signs of rejection (39, 40).
Immunospot plates (Cellular Technology, Cleveland, OH) were coated with purified anti-rat IFN-γ Ab (BD PharMingen, San Diego, CA) and stored at 4°C overnight. The next day, they were washed three times with PBS and blocked for at least 2 h with PBS containing 1% BSA. After further washing with PBS, irradiated WF spleen cells (5 × 105/well) and splenocytes (5 × 105/well) from naive LEW rats, or those of LEW recipients of skin transplants (7, 14, or 30 days posttransplantation), which received treatment with or without anti-CD134L Ab, CTLA4Ig, or a combination of both, were added to some wells, each in 100 μl of complete RPMI 1640 medium containing 10% FCS (Sigma-Aldrich, St. Louis, MO), 2 mM l-glutamine, 100 U/ml penicillin/streptomycin (BioWhittaker, Walkersville, MD), and 50 mM 2-ME (Sigma-Aldrich). Control wells containing stimulator cells alone or responder cells alone were also plated out. The plate was incubated at 37°C for 36 h, and then washed three times with PBS and then PBS containing 0.05% Tween. Biotin-labeled secondary anti-IFN-γ mAb, diluted in PBS-1% BSA-Tween, was added to each well, and the plate was incubated overnight at 4°C. After further washing, HRP conjugate (DAKO, Carpinteria, CA) was added for 2 h at room temperature. Development was performed with 3-amino-9-ethylcarbazole (Sigma-Aldrich). The resulting spots were counted on a computer-assisted ELISASpot Image Analyser (Cellular Technology). The results were then calculated as cytokine-producing cells per half million splenocytes.
Splenocytes were isolated 14 or 30 days following skin transplantation from animals treated with human CTLA4Ig (hCTLA4Ig) alone, anti-CD134L alone, or combined therapy, or from untreated allograft controls (three animals in each group), and were stained with anti-CD3-PE and anti-CD44-FITC mAb (BD PharMingen). Cells were analyzed on a FACSCalibur (BD Biosciences, San Jose, CA) using CellQuest software (BD Biosciences). The percentages of splenocytes expressing a CD3+CD44high phenotype were then calculated. The expansion and survival of T cells in the peripheral blood and spleen of the athymic heart recipients, adoptively transferred with sensitized splenocytes and treated with the different regimens, was examined by staining for T cells using an anti-CD3-PE mAb 18 days after transplantation and adoptive transfer.
Sensitization and adoptive transfer of sensitized spleen cells
Fourteen days following skin transplantation of a WF graft onto a LEW animal, the recipients were sacrificed. Spleens were removed, and single-cell suspensions were obtained. RBC were removed by incubating with ACK lysing buffer (BioWhittaker) for 10 min at room temperature. After several washes, cells were counted and resuspended in PBS. Immediately following transplantation of a WF heart into an athymic nude rat of LEW background, 4 × 107 of the sensitized splenocytes were adoptively transferred by i.v. injection into the nude rats.
Abs and immunosuppressive agents
Hybridoma cell lines producing mAbs directed against rat CD134L (ATM-2) (38), rat B7-1 (3H5), and B7-2 (24F) (41) were generous gifts from Dr. H. Yagita (Juntendo University, Tokyo, Japan) and were produced by Bioexpress (West Lebanon, NH). hCTLA4Ig was a kind gift from Dr. R. Peach (Bristol-Myers Squibb, Princeton, NJ). Anti-CD134L Ab (1 mg) was given by injection (i.p.) on the day of transplantation and on days 2, 4, 6, 8, 10, and 12 following transplantation. Anti-rat B7-1 or B7-2 mAb (0.5 mg), or hCTLA4Ig (0.5 mg) were administrated as a single i.p dose on day 2, in cardiac transplant recipients, or as multiple doses (0.5 mg each) on days 0, 2, 4, and 6 in the skin transplant recipients and in athymic nude rats. Cyclosporine (CsA; 10 mg/kg/day; Bedford Laboratories, Bedford, OH) and rapamycin (0.3 mg/kg/day; Wyeth-Ayerst, Madison, NJ) were given for 4 consecutive days following transplantation.
The Kaplan-Meier test was used to calculate the graft survival and the log-rank (Mantel-Cox) test was applied to compare differences in survival between groups. A value of p < 0.05 was considered significant.
The effect of CD134-CD134L and CD28-B7 blockade on cardiac allograft rejection
First, we compared the effects of B7 vs CD134L blockade on vascularized cardiac allograft rejection. In the WF into LEW rat transplant model, untreated control cardiac allografts were all rejected promptly, with a mean survival time (MST) of 8.6 ± 1.2 days (n = 8). Following treatment with a single injection of hCTLA4Ig on posttransplant day 2, allograft survival was significantly prolonged (MST: 35.7 ± 13.1 days; n = 6, p < 0.001) (Fig. 1⇓), while the administration of anti-CD134L mAb alone had no significant effect on allograft survival (MST: 9.3 ± 3.5 days; n = 6, p = NS compared with allograft controls) (Fig. 1⇓).
As we have previously described (39), administration of DST on the day of transplant followed by a single injection of hCTLA4Ig on day 2 resulted in all of the allografts achieving long-term survival (MST > 200, n = 6, p < 0.001 compared with untreated controls and hCTLA4Ig alone). However, DST showed no significant additional effect on graft survival when combined with anti-CD134L Ab (MST: 15.5 ± 1.732 days; n = 4, p = NS) (Fig. 1⇑).
A short course of CsA (10 mg/kg i.p. on days 0, 1, 2, and 3) resulted in some prolongation of allograft survival but all grafts were ultimately rejected (MST: 20.5 ± 6.5 days; n = 6, p = 0.0004 compared with control allografts). When CsA was used in combination with hCTLA4Ig, all of the grafts achieved long-term survival (>200 days, n = 6, p < 0.001 compared with either CsA or hCTLA4Ig treatment alone). However, the combination of CsA with anti-CD134L mAb resulted in no further survival benefit (MST: 16.5 ± 0.5 days; n = 6, p = NS compared with CsA alone) (Fig. 2⇓A).
Rapamycin (0.3 mg/kg i.p. on days 0, 1, 2, and 3) treatment alone significantly prolonged graft survival (37.4 ± 11.7 days, n = 5, p = 0.001 compared with allograft controls). When recipients were treated with both rapamycin and hCTLA4Ig, allograft rejection was further delayed (MST: 87.5 ± 10.5 days; n = 6, p = 0.0005 and 0.0007, compared with hCTLA4Ig or rapamycin alone) (Fig. 2⇑B). Moreover, rapamycin in combination with anti-CD134L mAb also prolonged allograft survival more than rapamycin alone (MST: 56.5 ± 6.3; n = 4, p = 0.0049 compared with rapamycin alone) (Fig. 2⇑B).
Synergy between CD134-CD134L and CD28-B7 blockade
Although CD134-CD134L blockade alone had no effect on allograft survival, the combination of anti-CD134L mAb and hCTLA4Ig demonstrated significant synergy (Fig. 3⇓A) with indefinite survival being achieved in all animals (MST > 200 days, n = 8, p < 0.001 when compared with hCTLA4Ig alone or anti-CD134L alone). Moreover, donor-specific tolerance was demonstrated by the acceptance of a second WF heart graft (>100 days) and rejection of a third-party ACI strain heart within 10 days (Table I⇓). To further elucidate the mechanism of this synergy between hCTLA4Ig and anti-CD134L, blocking anti-B7-1 and -B7-2 mAbs were used. Anti-B7-1 and anti-B7-2 mAb alone or in combination (0.5 mg for each, i.p on posttransplantation day 2) modestly prolonged allograft survival (MST: 13.5 ± 1.3 days, n = 4; 15.4 ± 1.8, n = 5; and 18 ± 1.8, n = 4, respectively). However, while the addition of anti-CD134L to either anti-B7-1 (MST: 13.3 ± 3.6 days, n = 7) or anti-B7-2 (23.1 ± 13.6 days, n = 7) produced no further prolongation of allograft survival (p = NS when compared with isolated anti-B7-1 or -B7-2 mAb treatment, respectively), coadministration of anti-CD134L with both anti-B7-1 and -B7-2 mAbs resulted in long-term allograft survival in all recipients (MST > 200 days, n = 6, p < 0.001 compared with the combination of CD134L mAb with either anti-B7-1 or anti-B7-2 alone or with anti-B7-1 and anti-B7-2 combined) (Fig. 3⇓B). Therefore, blockade of both B7-1 and B7-2 is necessary to achieve long-term allograft survival with anti-CD134L mAb.
The effect of CD134-CD134L and CD28-B7 blockade on skin allograft rejection
Mirroring the results from the cardiac transplant model, prolongation of skin transplantation was best achieved with a combination of anti-CD134L and hCTLA4Ig. All of the untreated allograft control animals rejected their grafts within 10 days (MST: 8.2 ± 1.0; n = 6), while anti-CD134L mAb alone had no effect on graft survival (MST: 8.7 ± 1.2; n = 6, p = NS) and hCTLA4Ig marginally delayed graft rejection (MST: 13.5 ± 1.4 days; n = 6, p = 0.001 compared with allograft controls). However, the combination of anti-CD134L mAb and hCTLA4Ig dramatically prolonged graft survival in this stringent transplant model, with 33% of allografts surviving over 100 days (MST: 59.1 ± 32.7 days; n = 6, p < 0.0007 compared with control and hCTLA4Ig groups) (Fig. 3⇑C).
In vivo blockade of CD134-CD134L and CD28-B7 diminishes the frequency of alloreactive IFN-γ-producing lymphocytes and prevents the development of activated cells
Using ELISPOT analysis, the frequency of alloreactive IFN-γ-producing lymphocytes from skin transplant recipients, treated with or without anti-CD134L mAb, hCTLA4Ig, or their combination, was measured 7, 14, and 30 days after transplantation. On day 7, the allograft control group had an elevated frequency of alloreactive IFN-γ-producing lymphocytes compared with that of naive control animals (Fig. 4⇓). Treatment with anti-CD134L mAb alone diminished the IFN-γ frequency by 21%, but a greater reduction was achieved with hCTLA4Ig and combined hCTLA4Ig and anti-CD134L mAb (74.3 and 64.2% reduction, respectively). However, by day 14, when animals treated with hCTLA4Ig were rejecting their grafts, the frequency of alloreactive IFN-γ-producing cells in this group increased significantly, while the combined treatment group of hCTLA4Ig and anti-CD134L mAb maintained low alloreactive frequencies consistent with the frequencies on day 7. By day 30, a time when most grafts treated with combined blockade are in the process of rejection, the alloreactive T cell frequency in untreated recipients was unchanged, while the frequency in the combined treatment group had increased, compared with day 14 (Fig. 4⇓). Furthermore, 14 days posttransplantation, the combined treatment group demonstrated the greatest reduction of activated/effector T cells, as judged by a CD3+CD44high phenotype (43% reduction) (Fig. 5⇓), while single therapy with hCTLA4Ig (22% reduction) or anti-CD134L (26% reduction) resulted in more modest decreases in the percentage of effector CD44high cells. Moreover, this effect persisted up to 30 days posttransplantation, when there was still a significant (27%) reduction in the CD3+CD44high T cells in the combined treatment group compared with the allograft controls (p = 0.038, compared with control, by two-tailed Mann-Whitney U test).
Effect of CD134-CD134L and CD28-B7 blockade on primed alloreactive lymphocytes
LEW background athymic nude rats, lacking T cells, failed to reject MHC-mismatched WF cardiac grafts (MST > 100 days, n = 5). Upon adoptive transfer of 4 × 107 lymphocytes from naive LEW rats, allografts were rejected in 12–23 days after transplantation (MST: 17 ± 4.6 days; n = 4). If the adoptively transferred cells were obtained from LEW rats presensitized with WF skin grafts, rejection was significantly accelerated (MST: 5.3 ± 0.5 days; n = 4, p < 0.0058 compared with naive lymphocyte transfused recipients) (Fig. 6⇓). We used this model to examine the role of B7 and CD134L on transplant rejection mediated by primed/effector T lymphocytes. Both anti-CD134L mAb and hCTLA4Ig when administered alone delayed the allograft rejection (MST: 14 ± 2.8 days, n = 4, p = 0.009; and 21.5 ± 1.7 days, n = 4, p < 0.001, for anti-CD134L mAb and hCTLA4Ig, respectively, compared with the untreated control group) mediated by sensitized cells. When anti-CD134L and hCTLA4Ig were combined, allograft survival was further and significantly prolonged (MST: 75.3 ± 12.7 days; n = 6, p < 0.001 compared with anti-CD134L mAb- or hCTLA4Ig-treated group alone) (Fig. 6⇓). Consistent with the survival data, the percentage of recovered CD3+ T cells in the peripheral blood of reconstituted nude recipients 18 days following allogenic cardiac transplantation, was greater in the untreated allograft controls (3.27%) and in the anti-CD134L mAb (4.54%)- and CTLA4Ig (7.51%)-treated nude recipients than in the combination group (0.96%). These data demonstrate that combined B7 and CD134L blockade prevented the expansion and/or persistence of the effector/primed alloreactive T cells in vivo.
CD134 and CD134L are members of the TNF and TNFR superfamilies, respectively, and have been recently recognized as efficient T cell costimulatory molecules (25, 42, 43). CD134 is expressed on activated T cells, primarily on CD4+ cells, although CD8+ T cell expression is also found but to a lesser degree (18, 44). CD134L has also been found on activated B cells (22, 23), dendritic cells (22), and vascular endothelial cells (24), and has been shown to be up-regulated following the engagement of the TCRs with MHC-peptide complexes. The expression peaks 24–48 h after T cell engagement with Ag and is down-regulated by 72–96 h (25). CD134L-transfected APC have little ability to induce either IL-2 secretion or the proliferation of naive T cells; however, enhanced proliferation is observed when B7 is coexpressed on the APC, and thereafter proliferation persists for several days (25). Thus, signaling through this pathway promotes T cell survival and helps determine the development of CD4+ T cell memory by regulating primary clonal expansion (26). Furthermore, signaling through the CD134-CD134L pathway has been shown to reverse peripheral tolerance (45). CD134+ T cells have also been associated with various autoimmune diseases (30, 32, 33, 34), and targeting of the CD134-CD134L pathway is believed to hold therapeutic potential in the treatment of autoimmunity and possibly in enhancing vaccine efficacy (46).
Although numerous in vitro studies have shown that CD134-CD134L interaction plays an important role in T cell activation to alloantigens (25, 47, 48), there is no clear evidence of its effect on in vivo alloimmune responses. Using our allogenic transplantation models, we found that, in both heart and skin transplants, CD134-CD134L blockade alone had no significant effect on allograft rejection, and its combination with DST, CsA, or rapamycin was not as efficacious as CD28-B7 blockade in prolonging allograft survival. However, the combination of anti-CD134L mAb and hCTLA4Ig (or both anti-B7-1 and anti-B7-2) induced donor-specific tolerance in all cardiac transplant recipients, as evidenced by long-term graft survival, acceptance of second donor WF hearts, and rejection of third-party ACI allografts. Moreover, it significantly prolonged survival in a more stringent model using fully allogenic skin grafts. Following in vivo blockade of both CD134-CD134L and CD28-B7 pathways, the frequency of alloreactive T cells was markedly reduced and was maintained at low levels. Furthermore, the percentage of T cells expressing an activated/effector phenotype was also maximally inhibited by combined CD134-CD134L and CD28-B7 blockade. Interestingly, the decrease in activated phenotype persisted, at a time when the frequency of IFN-γ-producing alloreactive lymphocytes began to recover to control values. Taken together, these data demonstrate that the effect of combined blockade is on the development and persistence of effector/memory T cells.
Furthermore, unlike other therapies, this combination was able to suppress transplant rejection mediated by primed/effector T cells from presensitized recipients. The extent of graft survival prolongation achieved has never been reported using other sensitized models. Onodera et al. (49), using a LBNF1 (LEW × BN F1 hybrid) donor heart transplanted into a LEW recipients presensitized with Brown-Norway skin grafts a week before, demonstrated that grafts were rejected within 36 h. Treatment with CTLA4Ig in the sensitization phase resulted in a cardiac graft survival of 4.4 ± 1.3 days, while treatment in both sensitization and effector phases prolonged survival to only 10.5 ± 8.5 days. Even the highly efficacious combination of anti-CD40L Ab and DST, which can induce long-term survival of allografts in fully MHC-mismatched mouse heart transplant models, failed to prevent cardiac graft rejection in a presensitized model (50). The demonstration of very low percentages of recovered CD3+ T cells from reconstituted nude recipients treated with the combination of anti-CD134 mAb and hCTLA4Ig, as compared with untreated controls and animals treated anti-CD134L mAb or hCTLA4Ig alone, suggests that only the combined therapy effectively inhibits T cell expansion and/or decreases cell survival. These findings are consistent with the known functions of CD28-B7 (51) and CD134-CD134L (25, 26) in promoting expansion and survival of Ag-specific T cells in vivo. Specifically, the CD134-CD134L pathway appears to be crucial for regulating the extent of CD4+ T cell expansion in the primary immune response, and thus the ability of T cells to persist as a population over time (26). CD134-deficient T cells secrete IL-2 and proliferate normally during the initial period of activation, but cannot sustain this during the latter phases of the primary response and exhibit decreased survival over time. Mice lacking CD134 generated lower frequencies of Ag-specific CD4+ T cells late in the primary response in vivo and generated lower frequencies of surviving memory cells as compared with WT mice (26).
This study represents the first demonstrating the role of the CD134-CD134L pathway in transplant rejection, and defines the interactions between CD134-CD134L and CD28-B7 pathways in naive and primed/effector alloimmune responses in vivo. Moreover, it indicates that combined blockade of CD134L and B7 is synergistic in promoting allograft survival even in presensitized recipients, and may thus represent one clinically relevant approach for overcoming the CD28-B7-independent allograft rejection mechanisms, responsible for hampering the development of tolerance induction protocols (12).
↵1 This work was supported by National Institutes of Health Grant RO1 AI34965.
↵2 Address correspondence and reprint requests to Dr. Mohamed H. Sayegh, Laboratory of Immunogenetics and Transplantation, Brigham and Women’s Hospital, 75 Francis Street, Boston, MA 02115. E-mail address:
↵3 Abbreviations used in this paper: DST, donor-specific transfusion; CD134L, CD134 ligand; MST, mean survival time; hCTLA4Ig, human CTLA4Ig; CsA, cyclosporine.
- Received September 9, 2002.
- Accepted January 3, 2003.
- Copyright © 2003 by The American Association of Immunologists