Aggressive skin allograft rejection in CD28-/- mice independent of the CD40/CD40L costimulatory pathway.

CD28-/- mice have been utilized to study the role of B7/CD28 and B7-CTLA4 interactions. There is evidence that CTLA4 ligation may be critical for tolerance induction. The aim of the current study is to further investigate rejection responses of CD28-/- mice and to define the role of B7-CTLA4 interactions in the absence of the CD40 and CD28 pathways. Balb/c skin allografts were transplanted onto C57BL/6 (B6) wild type or CD28-/- mice treated with anti-CD40L, CTLA4-Ig, or combination blockade. To investigate the cellular mechanism of rejection in CD28-/- recipients, mice were treated with anti-CD4 or anti-CD8 antibodies prior to treatment with costimulation blockade. The fluoroscein dye CFSE was utilized to study T cell expansion in vivo. Surprisingly, treatment of B6 CD28-/- mice with CTLA4-Ig alone (MST 12d), anti-CD40L alone (MST 13d), or combined blockade (MST 13d) had no effect on allograft survival compared to untreated B6 CD28 mice (MST 11d). CD28-/- recipients depleted of CD4+ cells and treated with CTLA4-Ig, anti-CD40L, or combination blockade also did not have prolonged survival compared with untreated mice (MST 10d). In contrast, CD28-/- recipients depleted of CD8+ cells had markedly prolonged allograft survival when treated with either anti-CD40L alone (MST 49d) or with combination blockade (MST 57d). Studies utilizing CFSE demonstrated that CD28-/- CD8+ T cells are not defective in in vivo proliferation responses compared with wild type CD8 cells. Thus, CD28-/- CD8+ T cells are responsible for aggressive rejection responses of CD28-/- mice independent of the CD40 pathway. In addition, CD40L blockade does not result in CD4+ T cell tolerance in CD28 recipients, despite an intact B7-CTLA4 pathway.     

Normal Donor Bone Marrow is Superior to Flt3 Ligand-Mobilized Bone Marrow in Prolonging Heart Allograft Survival when Combined with Anti-CD40L (CD154)

Flt3 ligand (FL) administration markedly increases bone marrow (BM) stem cells and immature dendritic cells. We investigated the influence of CD40-CD40Ligand (CD154) pathway blockade on antidonor immunity, cytokine production, microchimerism and heart graft survival in BALB/c (H2d) recipients of fully allogeneic C57BL/10 (H2b) FL-mobilized BM (FL-BM) or normal BM. Anti-CD40L mAb strongly suppressed anti-donor T-cell proliferative responses in recipients of either normal or FL-BM, but was less efficient in inhibiting antidonor cytolytic T-cell (CTL) activity, especially in recipients of FL-BM. Interestingly, CD40L blockade was more effective in recipients of multiple compared with single donor BM infusions. Anti-donor cytokine responses revealed complete impairment of IFN-gamma, IL-4 and IL-10 production in recipients of normal BM and CD40L mAb. By contrast, and in agreement with the CTL data, mice given FL-BM retained ability to produce IFN-gamma CD40-CD40L blockade did not promote microchimerism, as evidenced by immunohistology and real time polymerase chain reaction. Nevertheless, anti-CD40L mAb enhanced heart allograft survival in recipients of FL-BM, but the effect was inferior to that achieved with normal BM. These data provide insight into the influence of growth factor-expanded donor BM and costimulation blockade on antidonor immune reactivity and transplant outcome. The comparatively poor outcome obtained using FL-BM plus anti-CD40L mAb in this model may be ascribed to the failure of effectively interdicting antidonor CTL activity.     

CD40 expression on graft infiltrates and parenchymal CD154 (CD40L) induction in human chronic renal allograft rejection

BACKGROUND: CD40-CD154 (CD40L) costimulatory signaling plays a pivotal role in the effector mechanisms of transplant graft rejection. In animal models, CD40-CD154 blockade induces long-term graft acceptance concurrent with an absence of chronic rejection (CR) lesions. Given the critical importance of CD40-CD154 interactions in the development of chronic transplant allograft rejection, the relevance of in situ CD40 and CD154 expression was assessed in human chronic renal allograft rejection. METHODS: The expression of CD40, CD154, CD68, and T-cell receptor (TCR)alpha/beta was analyzed by immunohistochemistry. Serial cryostat sections of snap-frozen core renal allograft biopsies were obtained from 30 renal transplant patients. Biopsy specimens received diagnoses of CR (N = 23) according to the Banff classification and were compared with controls (N = 7) consisting of stable allografts and normal kidney tissue. RESULTS: Striking CD40 staining of graft cellular infiltrates (P = 0.016) was observed in renal allografts with CR compared with controls. The CD40+ cellular infiltrates in CR were predominantly TCR alpha/beta + T cells and some CD68+ macrophages. These findings were contrasted by the low-level CD40 expression detected in glomeruli and tubules of CR and controls. However, glomerular induction of CD154 was observed in CR allografts (P = 0.028) as compared with controls. CD154 immunoreactivity was demonstrated on glomerular endothelial, epithelial, and mesangial cells. Moderate CD154 expression was detected on tubular epithelial cells, and only weak CD154 immunoreactivity was observed on the infiltrates in isolated CR cases. CONCLUSION: In human chronic renal allograft rejection, CD40 is expressed on graft-infiltrating cells of the T cell and macrophage compartments. CD154 expression is induced on glomerular and tubular epithelial cells during CR, demonstrating another novel source of CD154 expression. The data substantiate the potential contributory role of an interaction between CD40+ graft-destructive effector T cells and macrophages with CD154+ renal allograft parenchymal cells in the development of chronic renal allograft rejection.     

A novel bispecific antihuman CD40/CD86 fusion protein with t-cell tolerizing potential

Clinical trials designed to achieve tolerance in humans by selectively antagonizing one of the T-cell costimulatory pathways, CD40-CD40L or CD80/CD86-CD28, are pending. However, simultaneous blockade of both pathways synergistically prevented graft rejection and successfully induced donor-specific tolerance in animal models. Synergism is also supported in human T-cells in vitro following anti-CD86 mAb and anti-CD40 mAb blockade. Therefore, in our view the most promising clinical strategy would be to antagonize both CD40 and CD86. Fast clinical entrance of this anti-CD86 and anti-CD40 bidirectional concept is highly facilitated by a single molecule approach. In the present study, a single bispecific fusion protein was constructed that specifically binds human CD40 and CD86 and which combines the antagonistic activities of both anti-CD40 and anti-CD86 humanized mAb. The anti-CD40/86 fusion protein showed tolerance inducing potential as it prevented both allogeneic T-cell expansion and generation of cytotoxic effector T cells and induced anergic antigen specific regulatory T cells. These data provide proof of concept in successfully combining the antagonistic activity of two humanized mAb with great clinical potential in transplantation and autoimmunity, in one single molecule.     

The CD154-CD40 costimulation pathway in organ transplantation

The CD154-CD40 costimulation pathway is one of the most extensively studied in organ transplantation. However, despite impressive results reported in a number of allograft models, precise mechanisms behind its in vivo blockade remain unclear. Several issues related to its potential clinical application have been addressed, resulting in some intriguing findings, particularly on interaction with conventional immunosuppressive agents and differential effects on T cells under the influence of infections. This review focuses on recent data from rodent studies to provide new insights and builds a cohesive picture of the interlocked immune mechanisms in CD154-CD40 signaling at work in transplant recipients.     

Blockade of CD40-CD40 ligand protects against renal injury in chronic proteinuric renal disease

BACKGROUND: Interaction between CD40 and CD40 ligand (CD40L) is involved in both cognate and innate immune responses. Blockade of CD40-CD40L interactions reduces severity of renal injury in murine lupus nephritis and membranous nephropathy. We hypothesized that CD40-CD40L could contribute to renal injury in models that are not antibody-dependent, and that anti-CD40L could diminish inflammation and fibrosis in murine adriamycin nephropathy. METHODS: Male BALB/c mice were divided into three groups (N = 6 per group): (1). saline-treated, age-matched control; (2). adriamycin only; and (3). MR1 + adriamycin. In group 3, mice were treated with intraperitoneal injections of anti-CD40L antibody (clone MR1, 0.4 mg per mouse) after the onset of proteinuria at days 5, 7, 9, and 11 after adriamycin treatment. Animal subgroups were compared at 14 and 42 days after induction of adriamycin nephropathy. Functional and pathologic markers of disease severity, cellular components of interstitial inflammation, and the degree of CD40 expression were assessed. Relative cortical RNA expression of the chemokine monocyte-chemoattractant protein-1 (MCP-1) and regulated on activation normal T cell expressed and secreted (RANTES) was also compared between animal groups. RESULTS: CD40 was weakly expressed in tubules of normal mice but was expressed in tubules, interstitium, and glomeruli of mice with adriamycin nephropathy in a time-dependent manner. MR1 treatment resulted in a significant attenuation of the severity of adriamycin nephropathy at day 42 [e.g., glomerular sclerosis (%), group 3, 20.1 +/- 4.7 vs. group 2, 30.2 +/- 7.2, P < 0.001]. CD40L blockade significantly reduced tubulointerstitial injury as well [tubular diameter microm), group 3, 42.5 +/- 6.9 vs. group 2, 66.3 +/- 13.7, P < 0.001; and group 1, 37.3 +/- 5.7, P < 0.01; tubular cell height microm), group 3, 16.3 +/- 1.7 vs. group 2, 11 +/- 1.8, P < 0.01; and group 1, 18.2 +/- 1.9, P < 0.01; interstitial volume (%), group 3, 13.9 +/- 5.1 vs. group 2, 26.2 +/- 4.9, P < 0.001; and group 1, 1.3 +/- 0.7, P < 0.001; proteinuria (mg/24 hours), group 3, 1.8 +/- 0.6 vs. group 2, 4.3 +/- 0.8, P < 0.001; and group 1, 0.7 +/- 0.2, P < 0.05; and creatinine clearance microL/min), group 3, 75 +/- 4 vs. group 2, 35 +/- 2, P < 0.001; and group 1, 82 +/- 4, P < 0.01] were also improved by MR1. MR1 treatment also resulted in a significant reduction in the number of cortical macrophages at both 14 and 42 days after adriamycin (P < 0.01). Cortical expression of MCP-1 and RANTES was significantly reduced by MR1 treatment at 42 days after adriamycin (P < 0.01 and P < 0.05, respectively). CONCLUSION: Blockade of CD40-CD40L interaction protects against renal structural and functional injury in this murine model of chronic proteinuric renal disease.     

Immunological function of tubular epithelial cells: the functional implications of CD40 expression

Tubulo-interstitial inflammation in the kidney is characterized by the presence of activated T cells. Both by the local production of cytokines, as well as by direct cell-cell interactions, these activated T cells might affect the immune and inflammatory response. Recently it has been demonstrated that these kidney-infiltrating T cells express CD40 ligand and that tubular epithelial cells (TECs) express CD40. In the present review we will discuss the potential implications of CD40-CD40L interactions for the activation of TECs and its immunological function.     

CD154 Regulates Primate Humoral Immunity to Influenza

Current methods of immunosuppression for the purposes of allowing solid organ transplantation in humans are broadly inhibitory and thus are associated with an increased risk of opportunistic infections and neoplasia. We have shown previously that a selective blockade of CD40-CD154 interactions during heart transplantation in cynomolgus macaques significantly delays immune-mediated graft injury. Here, we determined the effect of anti-CD154 mAb therapy on primate serologic responses to immunization with influenza virus hemagglutinin (HA), a T-cell-dependent Ag. We found that CD154 blockade attenuated primary and secondary serum Ab responses of IgM and IgG isotypes to influenza, even when anti-CD154 treatment was discontinued prior to reimmunization. These findings show that in primates CD40-CD154 interactions are necessary for both primary and secondary Ab responses to viral Ags. Furthermore, the data suggest that viral Ag stimulation of primates in the absence of CD154 stimulation may have a tolerizing effect on that Ag.     

Impact of both donor and recipient strains on cardiac allograft survival after blockade of the CD40-CD154 costimulatory pathway

BACKGROUND: The effectiveness of anti-CD154 monoclonal antibodies in prolonging the survival of mouse allografts is dependent on the strain combination. In this report, we examined the impact of the donor and the recipient strains on the success of CD40-CD154 blockade. MATERIALS AND METHODS: Cardiac allograft survival was monitored in different donor/recipient strain combinations. Morphometric analyses on the allograft coronary arteries allowed quantification of vessel intimal thickening. RESULTS: Prolonged cardiac allograft survival after the administration of an anti-CD154 monoclonal antibody was found to be dependent on the donor and the recipient strains. The influence of the donor and the recipient strains lay in the ability of CD8 T cells to cause graft rejection despite CD40-CD154 blockade. Elimination of CD8 T cells before transplantation resulted in similar graft prolongation irrespective of the genotype of the donor or the recipient strain. CONCLUSION: These data show that both donor and recipient strains contribute to CD40-CD154-independent CD8 T-cell-mediated rejection.     

Mechanisms Involved in the Establishment of Tolerance Through Costimulatory Blockade and BMT: Lack of Requirement for CD40L-Mediated Signaling for Tolerance or Deletion of Donor-reactive CD4+ Cells

We have previously shown that high levels of multiline-age mixed hematopoietic chimerism and systemic T-cell tolerance can be achieved in mice without myeloablation through the use of anti-CD40L and costimulatory blockade alone (plus CTLA4Ig) or with recipient CD8 depletion and allogeneic bone marrow transplantation. Chimeric mice permanently accept donor skin grafts (> 100 days), and rapidly reject third-party grafts. The mechanisms by which costimulatory blockade facilitates the engraftment of allogeneic hematopoietic cells have not been defined. To further understand the in vivo mechanisms by which the administration of anti-CD40L mAb facilitates the engraftment of donor bone marrow and rapidly tolerizes CD4+ T cells, we analyzed the establishment of chimerism and tolerance in CD40L -/- mice. We demonstrate here that anti-CD40L mAb treatment is required only to prevent CD40L/CD40 interactions, and that no signal to the T cell through CD40L is necessary for the induction of CD4+ tolerance. Peripheral deletion of donor-reactive CD4+ T cells occurs rapidly in CD40L -/- mice receiving bone marrow transplantation (BMT), indicating that this deletion in the presence of anti-CD40L is not due to targeting of activated CD4+ cells by the antibody. Complete CD4+ cell tolerance is observed by both skin graft acceptance and in vitro assays before deletion is complete, indicating that additional mechanisms play a role in inducing CD4+ T-cell tolerance as the result of BMT in the presence of CD40/CD40L blockade.     

OX40 (CD134) blockade inhibits the co-stimulatory cascade and promotes heart allograft survival

BACKGROUND: CD134 (OX40) is a member of the tumor necrosis factor receptor superfamily that is expressed as a late event in T-cell activation and contributes to the generation of immunologic memory. CD134 blockade effectively ameliorates inflammation in models of autoimmune disease, but its role in transplantation has been less well studied. METHODS: The authors used an OX40-immunoglobulin (Ig) fusion protein to examine the contribution of this co-stimulatory molecule to the in vitro alloimmune response and studied the ability of CD134 blockade to prevent cardiac allograft rejection in mouse models of heart transplantation using strains representing both major histocompatibility complex (MHC) (BALB/c to CBA/Ca) and minor histocompatibility complex (mHC) (B10.BR to CBA/Ca) antigen mismatches. RESULTS: CD134 upregulation on in vitro alloactivated T cells was delayed compared with CD69 and CD25, and inhibition of T-cell proliferation was critically dependent on the timing of OX40-Ig administration. Heart allograft survival in a fully allogeneic, MHC-mismatched strain combination was not influenced by CD134 blockade alone, but OX40-Ig treatment in the mHC-mismatched model resulted in long-term graft survival (median survival time extended from 14 days to >100 days). Early mononuclear cell infiltration of the graft was similar in both rejecting and long-surviving heart grafts, but OX40-Ig treatment appeared to delay cellular infiltration. CONCLUSIONS: These results show that CD134-CD134L interaction plays an important role in the co-stimulatory cascade and that blockade of this molecular interaction may be of therapeutic value in helping to prevent allograft rejection.     

CD40-CD40L在儿童过敏性紫癜发病机制中的作用研究

目的探讨CD40-CD40L存儿童过敏性紫癜发病机制中的作用。方法研究对象为30例HSP患儿,正常对照组30例,应用流代细胞术检测其外周血单个核细胞中CD40－CD40L的表达情况,并观察实验组治疗前后CD40-CD40L的表达有无差异,结果与对照组比较,CD40L表达明妙增高,且治疗前表达高于治疗后（P〈0．05）,CD40表达增高,但与对照组及治疗前后比较差异均无显著性（P〉0．05）。结论HSP患儿外周血单个核细胞CD40L表达增高,且治疗前离于治疗后,表明其在HSP患儿免疫紊乱发病机制中可能起着重要作用。     

Analysis of the CD40/CD40L role in the sustenance of alloreactive antibody production

CD40 ligand (CD40L) is important for T/B lymphocyte interaction. To understand the cellular basis of humoral allosensitization we, therefore: (1) measured CD40L protein and gene expression in sensitized and non-sensitized uremic unactivated peripheral CD4+ T lymphocytes; (2) studied the impact of blocking the CD40/CD40L pathway on alloreactive antibody (allo-Ab) production by engrafted sensitized PBLs into severe combined immunodeficient (SCID) mice after in vitro preactivation with IL2/LPs/HLA class II allopeptides and adjuvants as a potent stimulus to produce allo-Ab (Shoker et al. Transplantation 1999;68;1188); and (3) studied the modifying effect of CD40/CD40L blockade on T helper type I and II cytokine gene expression in the respective mice spleen. The CD40L protein was measured by flow cytometry and the gene expression was measured by quantitative RT-PCR. Alloreactive antibodies (alo-Abs) produced by sensitized PBLs engrafted into SCID mice with and without blockade of the CD40 receptor were measured by the PRA-STAT ELISA method. The modifying effects of CD40 blocking on allo-Ab production and cytokine gene expression by the engrafted cells measured by RT-PCR were then compared. There was no detectable CD40L protein expression in either the uremic or the control groups. The CD40L gene expression of 0.04 +/- 0.02 attomoles (aM) in the sensitized group was significantly higher than in the non-sensitized patients (0.009 +/- 0.007 aM, P < 0.0001) or the control CD4+ T cells (0.016 +/- 0.004 aM, P < 0.001). Blockade of the CD40 receptor abrogated the production of allo-Ab antibodies by the engrafted sensitized cells in 60% of the tested mice (n = 10); decreased the mean +/- S.D. optic density of allo-Ab to 0.1 +/- 0.13 and the mean +/- S.D. PRA to 12 + 16). In the presence of the control Ab, allo-Ab production in SCID sera was present in 100% of the 10 SCID mice tested; the mean +/- S.D. PRA was 75 +/- 20, and the mean + S.D. OD activity was 0.412 +/- 0.17. All cytokine genes were, otherwise, expressed in the presence or absence of CD40 blockade. The results suggest a potential role of an enhanced CD40/CD40L interaction in the sustenance of alloreactive antibody production without significant deviation to T helper-like I or II responses. Blocking the CD40/CD40L pathway may have a potential therapeutic benefit to treat sensitized uremic patients.     

Long-term acceptance of rat cardiac allografts on the basis of adenovirus mediated CD40Ig plus CTLA4Ig gene therapies

BACKGROUND: We have previously demonstrated that blockade of either CD80/86-CD28 or CD40-CD154 costimulatory pathways by using adenovirus vector coding CTLA4Ig (AdCTLA4Ig) or CD40Ig (AdCD40Ig) genes induced donor-specific tolerance in rat liver transplantation. In this study, we asked whether these gene-therapy-based costimulation blockade would induce tolerance in cardiac transplantation. METHODS: Heterotopic heart transplantation was performed in a full major histocompatibility complex (MHC) barrier combination of ACI (RT1avl) to Lewis (LEW, RT1l) rats. Vector (1 x 10(9) plaque forming unit [PFU]), AdLacZ, AdCTLA4Ig, or AdCD40Ig, was administered intravenously to recipient animals immediately after grafting, and graft survival, serum CTLA4Ig/CD40Ig levels, and graft histology were assessed. Tolerance was determined by secondary skin-graft challenging. RESULTS: Allografts of both untreated and AdLacZ controls were promptly rejected within 7 days, whereas a single treatment with AdCTLA4Ig or AdCD40Ig significantly prolonged median graft survival to 55.5 and 28.5 days, respectively. In contrast, the combined AdCTLA4Ig and AdCD40Ig gene therapy maintained high CTLA4Ig and CD40Ig levels through the posttransplant period and allowed long-term cardiac allograft survival for more than 270 days. However, both donor and third-party skin grafts were rejected in the animals who harbored cardiac grafts over 150 days. Also, typical features of chronic rejection were evident in the long-term surviving grafts. CONCLUSION: Simultaneous blockade of CD28 and CD154 pathways by AdCTLA4Ig plus AdCD40Ig induces a strong immunosuppression that allows long-term acceptance of full MHC mismatched cardiac graft in rats. This strategy, however, was not enough to induce tolerance to skin grafts and to avoid chronic rejection, as shown in the liver-transplantation model.     

Liver Transplantation in a Patient With CD40 Ligand Deficiency and Hyper‐IgM Syndrome: Clinical and Immunological Assessments

Monoclonal antibodies that disrupt CD40–CD40 ligand (CD40L) interactions are likely to have use in human transplantation. However, the extent of the immunosuppressive effects of CD40–CD40L blockade in humans is unknown. Hyper‐IgM syndrome (HIGM) is a rare primary immunodeficiency syndrome characterized by defects in the CD40–CD40L pathway, severe immune deficiency (IgG), and high or normal IgM levels. However, the effects of CD40L deficiency on T‐ and natural killer (NK)‐cell function is not established. Here, we present a patient with HIGM syndrome who underwent liver transplantation for hepatitis C virus infection. Posttransplantation, NK‐cell antibody‐dependent cytokine release (γ‐interferon) to alloantigens and T cell responses to viral antigens and mitogens were assessed and showed normal CD4⁺, CD8⁺, and NK‐cell responses. We also examined antibody‐dependent cellular cytotoxicity against a CD40⁺ and HLA‐expressing cell line. These experiments confirmed that the patient's NK cells were equivalent to those of normal subjects in mediating antibody‐dependent cellular cytotoxicity despite the absence of CD40–CD40L interactions. Mitogenic stimulation of the patient's peripheral blood mononuclear cells showed no expression of CD40L on T and NK cells compared with increased expression in normal subjects. Taken together, these data suggest that absence of CD40L expression is responsible for aberrant B cell immunity but had little impact on NK‐ and T cell immune responses in vitro.     

The Immunobiology of Inductive Anti-CD40L Therapy in Transplantation: Allograft Acceptance is Not Dependent Upon the Deletion of Graft-Reactive T Cells

CD40–CD40L costimulatory interactions are crucial for allograft rejection, in that treatment with anti‐CD40L mAb markedly prolongs allograft survival in several systems. Recent reports indicate that costimulatory blockade results in deletion of graft‐reactive cells, which leads to allograft tolerance. To assess immunologic parameters that were influenced by inductive CD40–CD40L blockade, cardiac allograft recipients were treated with multiple doses of the anti‐CD40L mAb MR1, which was remarkably effective at prolonging allograft survival. Acute allograft rejection responses such as IL‐2 producing helper cell priming, Th1 priming, and alloantibody production were abrogated by anti‐CD40L treatment. Interestingly, the spleens of mice bearing long‐term cardiac allografts following inductive anti‐CD40L treatment retained precursor donor alloantigen‐reactive CTL, IL‐2 producing helper cells, and Th1 in numbers comparable to those observed in naïve mice. These mice retained the ability to reject donor‐strain skin allografts, but were incapable of rejecting the original cardiac allograft, or a second donor‐strain cardiac allograft. Further, differentiated effector cells were incapable of mediating rejection following adoptive transfer into mice bearing long‐term allografts, suggesting that regulatory cell function, rather than effector cell deletion was responsible for long‐term graft acceptance. Collectively, these data demonstrate that inductive CD40–CD40L blockade does not result in the deletion of graft‐reactive T cells, but induces the maintenance of these cells in a quiescent precursor state. They further point to a tissue specificity of this hyporesponsiveness, suggesting that not all donor alloantigen‐reactive cells are subject to this regulation.     

Inhibition of chronic rejection and development of tolerogenic T cells after ICOS-ICOSL and CD40-CD40L co-stimulation blockade

BACKGROUND.: Blockade of the CD40-CD40L pathway results in long-term allograft survival but does not prevent chronic rejection. ICOS-ICOSL are members of the CD28-B7 family that play an important role in T-cell activation. METHODS.: The authors analyzed the effect of single or combined treatment with an anti-ICOS monoclonal antibody and CD40 immunoglobulin (Ig) on acute and chronic rejection of heart allografts in rats. RESULTS.: Treatment with anti-ICOS resulted in a modest but significant prolongation of allograft survival. Treatment with CD40Ig resulted in long-term graft survival but the cardiac grafts developed chronic rejection lesions. Combined CD40Ig+anti-ICOS treatment led to indefinite graft survival in all recipients and a significant decrease of chronic rejection lesions compared with CD40Ig alone. Importantly, four of the seven CD40Ig+anti-ICOS-treated recipients showed a complete absence of chronic rejection lesions, whereas all of the CD40Ig-treated recipients showed chronic rejection. The CD40Ig+anti-ICOS group also showed significant decreased graft infiltration, decreased antidonor cytotoxic T-lymphocyte activity, and decreased alloantibodies compared with the CD40Ig-treated group. Adoptive transfer of splenocytes indefinitely prolonged allograft survival, whereas those depleted of T cells did not, suggesting the development of T-regulatory mechanisms. CONCLUSIONS.: These data indicate that the chronic rejection mechanisms that are CD40-CD40L independent are ICOS-ICOSL dependent. These results were obtained with conservation of cognate immune responses and development of tolerogenic T cells.