Despite several attempts to isolate a mAb specific for human dendritic cells, none currently exists. Recent attempts have utilized an improved dendritic cell purification method to prepare immunogens and a rapid two-color flow cytometric screening procedure that allows large numbers of hybridoma supernatants to be examined in each fusion. Yet these improvements have also failed, yielding only hybridomas that bind "shared" antigens expressed by both dendritic cells and other leukocytes. Dendritic cells express many shared antigens, including CD45 [leukocyte common antigen], CD40, leukocyte [beta 2] integrins CD11a and CD11c, CD54 [ICAM-1], CD44 [Pgp-1], CD58 [LFA-3], and the B7/BB1 antigen. Therefore, we are attempting to bias the immune response toward rarer, dendritic cell-specific clones by tolerizing or immunosuppressing our animals to shared antigens. In one approach, adult mice held in barrier cages are injected with "nondendritic" cells and cyclophosphamide [CP], in order to ablate responding "nonspecific" B cell clones. Fifteen days after the last dose of CP, they are challenged with nondendritic cells. A week later they are bled, and serum antibody titers against nondendritic cells are determined by FACS, in order to demonstrate tolerance compared to controls injected with CP alone. In the second approach, neonatal mice are injected with human T lymphoblasts at birth, followed by boosting at 1 week. In adulthood, they are challenged sequentially with sheep erythrocytes [sRBC], then with T blasts, to demonstrate that they can respond to unrelated cells but not to tolerogenic cells. One week after each kind of challenge, mice are bled and serum antibody levels are determined for treated and sham-injected mice. When these two approaches were compared, CP led only to nonspecific immunosuppression, while neonatal injections produced selective, antigen-specific nonresponsiveness to the tolerizing T blasts.