Acute and chronic allograft rejection can occur in HLA-identical sibling transplants implicating the importance of immune response against non-HLA targets. Non-HLA, complement and non-complement-fixing antibodies, might be responsible for a variety of allograft injuries. It reflects the complexity of their acute and chronic actions. Non-HLA antibodies might occur as alloantibodies or autoantibodies. Their antigenic targets include various minor histocompatibility antigens, vascular receptors, adhesion molecules, and intermediate filaments[13]. Blockade of the co-stimulatory pathways CD28-B7 by either CTLA-4Ig or AdCTLA-4Ig alone has been shown to improve allograft survival. However, single pathway blockade did not prevent the eventual development of chronic rejection, as assessed by the presence of vasculopathy. Histologically, the most striking finding concerning on the rejected cardiac allografts from the AdCTLA-4Ig-treated group was the marked ICOS-positive mononuclear cell infiltration. Therefore, we examined the effect on the development of chronic rejection of simultaneously blocking both B7-CD28 and ICOS-ICOSL pathways employing AdCTLA-4Ig and anti-ICOS antibodies. Compared with allografts from animals treated with single agent CTLA-4Ig, those treated with both CTLA-4Ig and anti-ICOS displayed a significant decrease in ICOS-positive mononuclear cell infiltration and vascular wall lesions as well as reduction in the number of occluded vessels. Nukada et al[14] investigated the signaling mechanisms underlying the morphological polarization of activated T cells, initiated by ICOS signaling. ICOS signaling induces the activation of phosphoinositide-3 (PI3)-kinase[14]. This may be the mechanism of the inhibition of allograft vasculopathy caused by anti-ICOS mAb. Our data suggested that direct inhibition of intragraft ICOS-positive mononuclear cell infiltration would play an important role in preventing chronic rejection. CD28-mediated co-stimulation plays a critical role in naive and primary T-cell activation. However, there is evidence that not all T cell activation depends on the B7-CD28 pathway[15]. ANG II, NADPH oxidase, and reactive oxygen species also modulate the activation of T cell[16]. In allogeneic T cell responses, some T cell activation is likely to occur by escaping from the inhibition of B7-CD28 blockade. Hence, those T cells might develop into effector T cells. It is also likely that ICOS-ICOSL co-stimulation of T cells in the absence of B7/CD28 interactions is sufficient to induce migration of such effector T cells to the graft site, and subsequently to contribute to the chronic intragraft infiltration, and even to allograft failure. The present data extend our previous study, which the development of CD4+CD25+ regulatory T cells under the combined AdCTLA-4Ig and anti-ICOS antibody therapy appeared to play an important role in tolerance induction and maintenance. However, some other possible mechanisms should be considered for the development of chronic rejection. Guillonneau et al[17] recently reported that chronic rejection lesions were prevented by blocking the ICOS pathway with mouse anti-ICOS antibodies and CD40-Ig. This approach resulted in the inhibition of anti-donor CTL responses as well as reduction alloantibody levels.
It remains controversial as to how the naive T cells and effector T cells differ in their co-stimulatory requirements for activation and expansion. CD28 is generally thought to be the most important molecule in the initiation of T cell responses. Studer et al[18] investigated that CD28 down-regulation on CD4+cells was associated with poor outcomes in lung transplantation, whereas ICOS was considered to act during the effector phase. Effector cells require lower numbers of engaged T-cell antigen receptor (TCR), shorter duration of TCR ligation, and have a lesser requirement for lesser CD28 co-stimulation[19]. Studies in vitro performed in murine suggested that the ICOS-ICOSL interaction regulated the differentiation of naive cells into effector cells as well as Th2 cells, but did noy support Th1 function. However, more recent studies suggested that the situation may be more complex in vivo, such as inhibition of ICOS also inhibits Th1-associated EAE and allograft rejection[10,13, 20-22]. van Berkel et al[23] showed that ICOS was not only important for T cell effector function but also contributed to the primary expansion of T cell responses depending on CD28 signaling. Taken together, these studies suggested a critical role of ICOS-ICOSL interactions in the effector phase of T cell responses.
Together with our previous reports, we conclude that infiltration of ICOS-positive effector T cells into the graft plays an important role in the development of chronic rejection and cardiac allograft failure. Simultaneous blockade of ICOS/ICOSL and B7-CD28 pathways with anti-ICOS antibody and AdCTLA-4Ig during allograft transplantation induced stable immune tolerance without chronic rejection, which was associated with the development of CD4+CD25+ regulatory T cells and modulation of effector T cell infiltration. Thus, either inhibition of ICOS-positive effector T cell development at an early phase of the allograft response, or/ and prevention of ICOS-positive effector T cell infiltration from progressive chronic rejection would be effective in protecting cardiac allografts from chronic rejection.
