BACKGROUND Allograft survival can be prolonged by the administration of alloantigen(s) before transplantation. Blood transfusion is the commonest form of alloantigen pretreatment currently used in clinical practice. However, for recipients of organs from cadaver donors, it is not possible to predict the identity of the organ donor in advance. Therefore, it is highly unlikely that all the alloantigens expressed by a cadaver organ donor will be represented in the alloantigen pretreatment inoculum. We have previously shown that it is not necessary to expose the recipient to the full complement of donor alloantigens to induce long-term survival of a subsequent cardiac allograft. Here, we investigated the in vivo mechanism responsible for this phenomena. METHODS Unresponsiveness to the mouse MHC class I molecule Kb was induced in CBA.Ca (H2k) recipients by administration of bone marrow cells from transgenic CBA mice, CBK (H2k + Kb) before transplantation of fully allogeneic and F1 vascularized cardiac allografts. RESULTS Pretreatment with CBK bone marrow cells resulted in the long-term survival of all cardiac allografts expressing H2-Kb. For example, C57BL/10 (H2b) and (CBKxBALB/c) F1 (H2k,d + Kb) cardiac allografts were accepted by recipients treated with CBK bone marrow cells before transplantation. In contrast, allografts that did not express Kb, such as BALB/c (H2d) or (CBAxBALB/c) F1 (H2k,d), were rejected acutely, with a median survival time (MST) of 7 and 6 days, respectively, in recipients treated with CBK bone marrow cells. Furthermore, when recipients pretreated with CBK bone marrow cells were grafted with a BALB/c heart and a CBK heart simultaneously, the BALB/c hearts were rejected (MST=10 days), whereas the CBK hearts were accepted. By contrast, in the maintenance phase (i.e., after transplantation), recipients with long-term surviving (CBKxBALB/c) F1 hearts (> 100 days) were found to accept BALB/c hearts indefinitely, whereas fourth-party B10.S (H2s) grafts were rejected (MST=7.5 days). This indicated that the allografts bearing Kb could tolerize recipients to other alloantigens expressed by the transplanted heart. CONCLUSIONS These data provide clear evidence for linked epitope suppression in the induction of operational tolerance in vivo.