Role of Antigen‐Presenting Cells in the Cytotoxic T‐Cell Response to Minor Histocompatibility Antigens (MIHA)

Although antigen‐presenting cells (APC) appear to be able to process minor histocompatibility antigens (MIHA) expressed on allogeneic cells and present them in association with intrinsic H‐2 of the APC in vivo, this does not occur in vitro. This could be due to fundamentally different mechanisms of antigen handling by APC in vitro, or it could represent compromised APC function. In order to distinguish these possibilities, we designed a system in which BALB/c spleen cells are transferred into an MIHA‐disparate irradiated host and allowed to reside for 2‐3 days; the spleen cells of the repopuiated host are removed and used as stimulator cells for the CTL response of primed BALB/c responder cells to DBA/2 MIHA. These cells are referred to as in vivo‐pulsed APC (IVP‐APC). Donor BALB/c cells are able to pick up DBA/2 MIHA after a passage in DBA/2 hosts and efficiently present MIHA to primed CTL precursors to generate DBA/2‐specific CTL. The donor cell type able lo pick up and present MIHA is present in spleen bul not thymus or bone marrow, is Thy‐1.2 negative, Ia+, and nylon wooladherent. Its stimulatory capacily is as efficient on a per cell basis as that of DBA/2 spleen cells. The generation of IVP‐APC requires repopulation of the irradiated host, which must express MIHA foreign to the BALB/c donor cells. When we attempled to generate IVP‐APC in H‐2 incompatible hosts, we found that, although the IVP‐APC could present H‐2 antigens, they were unable to present MIHA in association with intrinsic APC H‐2 antigens. Use of intra‐H‐2 recombinant strains as host mice indicated that compatibility of donor and host at the KI region of H‐2 was essential for the generation of IVP‐APC able to present apparently unprocessed MIHA to primed BALB/c responder cells. Thus, we were unable to reproduce the antigenprocessing function of APC encountering antigen in situ using an adoptive transfer method of pulsing APC with MIHA in vivo. In addition, we suggest that our results may impose constraints on the formulation of models to account for the association of MIHA with H‐2 antigens.