Cytomegalovirus: Another Source of Antibody-mediated Graft Injury?

Abstract

Despite improvements in immunosuppression and organ preservation, rejection remains the bane of solid organ transplantation. There has been a longstanding interest in antibody-mediated rejection, with more recent interest focusing on the role of antibodies in chronic rejection. Given the allogeneic nature of transplantation, HLA antigens and alloantibody have naturally been the primary focus of this work,1 and little attention has been paid to potential alternate antigens and antibodies. Concomitantly, there is increasing recognition of the influence of cytomegalovirus (CMV) on organ transplantation outcomes and rejection.2 CMV reactivation has long been associated with allograft rejection, and the debate continues about cause or effect. Because studying this association is not ethically possible in clinical constructs, investigators have turned to murine models that allow isolation of single variables after allogeneic transplantation. Using such models, it has been recently suggested that CMV reactivation accelerates renal allograft injury at least in part by enhancing T cell–mediated rejection.3 Renewed interest in antibody-mediated mechanisms next begged the question: Can antibodies to chronic viruses like CMV contribute to antibody-mediated rejection when the allograft has infection/reactivation? In the current issue, Saunders4 shares results suggesting that CMV infection/reactivation induces antibodies that deposit in CMV+ allografts.4 This represents the continuation of their previous work, leveraging a murine model of renal transplantation to study the impact of virus-induced antibody. Put simply, they show that renal allografts from mice previously infected with CMV have antibody and complement deposition that is absent in grafts from noninfected donors. This deposition is accentuated in recipients that are CMV+. To confirm that this is both virus and antibody related, they adoptively transferred CMV immune sera into µ-chain deficient mice recipients that lack B-cell function and cannot make an antibody. When CMV+ allografts are transplanted into this model, only µ-chain deficient mice receiving CMV-reactive antibody show antibody deposition. Thus, CMV-induced antibodies and also complement accumulate in previously infected grafts after allotransplantation. The question then becomes whether CMV-induced sera contain virus-specific or allospecific antibodies induced by CMV infection. Although this question is not completely answered, their results with isografts following D+R− transplantation confirm that either viral antigen or some non-major histocompatibility complex molecule expressed during viral reactivation are sufficient target(s) for antibody/complement deposition. So even if CMV infection does trigger alloantibody in small amounts (as suggested by results in Figure 4A), alloantigen is not requisite for antibody/complement deposition.4 So what is the commonality to these allograft/isograft findings? It does not seem to be a latent virus because the concomitant evaluation of native kidneys showed that despite comparable viral DNA loads, antibody/complement did not accumulate in CMV+ native kidneys in graft recipients. It is clear from seminal work by Hummel and Abecassis that allogeneic transplantation can trigger intragraft CMV reactivation after renal transplantation.5 More recent work from the same group shows that ischemia and reperfusion in isografts can also trigger CMV reactivation.6 Taken together, these results strongly suggest that virus reactivation sets the stage for antigen presentation and an antiviral immune response that leads to antibody/complement deposition in these grafts. What remains to answer is whether the observed antibody accumulation is a contributor to worsened outcomes or simply an epiphenomenon of viral reactivation. Work from the current authors has suggested that CMV accelerates renal allograft injury and that this is in part mediated by T cells.3 However, just as we have learned that not all T cell–infiltrating grafts are antigraft and thus not “harmful,” it is possible that the antibody deposition described by the authors is not relevant to graft rejection. In fact, if viral reactivation is a priori harmful to renal allografts, then control of the virus through the accumulation of antibody/complement may actually be helpful to graft survival. Conversely, antiviral responses might be additive to allogeneic responses, and the ex vivo cytotoxicity studies in tubular epithelial cells suggest a mechanism by which that damage may occur. Thus, while it is tempting to assume that antibody deposition in these grafts is detrimental, this remains to be proven. Fortunately, the authors have the tools at their disposal to address these questions in future studies.