Korngold and Jonathan Sprent

Unprimed T cells transferred to heavily irradiated H-2-compatible mice cause a high incidence of lethal graft-vs.-host disease (GVHD)a in certain strain combinations (1-3). GVHD in this setting is a consequence of mature post-thymic donor T cells responding to the multiple minor histocompatibility antigen (minor HA) differences of the host. With the combination of CBA and B10.BR (both H-2~), doses of as few as 106 CBA T cells regularly kill close to 100% of irradiated B10.BR mice. Recent studies with this strain combination demonstrated that T cells eliciting lethal GVHD to minor HA are subject to H-2-restriction (4). To examine this question, CBA T cells were reeirculated through irradiated mice of the B10 H-2 congenic lines and then tested for their capacity to kill B10.BR mice. When CBA T cells were filtered from blood to lymph for 1 d through irradiated B10.BR or H-2-semisyngeneic (CBA × B10)F1 mice, negative selection occurred, i.e., the filtered T cells failed to kill B10.BR mice on further transfer. Selection was not apparent, however, when CBA T cells were filtered through totally H-2-different irradiated mice, e.g., B10 (H-2°), B10.D2 (H-2a), or B 10.S (H-if). Selection thus depended upon a sharing of H-2 determinants between the donor and host. These findings raised a number of questions, including: (a) What is the relationship between T cells causing GVHD to minor HA and minor HA-specific cytotoxic lymphocytes (CTL)? (b) Do GVHD-indueing T cells, like CTL, comprise discrete subgroups of H-2Kand H-2D-restricted cells? (c) Are H-2I-restricted cells involved in GVHD to minor HA? (d) What cells present minor HA to T cells during negative selection? (e) Is antigen processing involved during negative selection? This paper attempts to provide answers to these questions.