Assessment of mixed lymphocyte reactivity in human bone marrow cell cultures.

The use of monoclonal anti-T-cell antibodies has been proposed as a means of eliminating T cells from bone marrow inocula, and thereby avoiding graft-versus-host reactions following transplantation. Since the mixed lymphocyte reaction (MLR) provides a measurement of alloreactive mature T cells, we have attempted to apply this assay to bone marrow populations before and after treatment with anti-T-cell antibodies and complement. However, initial studies showed MLR to be very difficult to measure using bone marrow as a responder cell population, and a systematic analysis of the reasons for this difficulty was therefore carried out. The first major problem in performance of standard one-way MLRs using bone marrow responder cells was found to be due to the presence of numerous non-T marrow cells that maintained high levels of background proliferation. Proliferation of these populations was found to be variable during MLR culture, leading to aberrant results. This problem was overcome by removing the rapidly proliferating population of marrow cells either by density centrifugation or by susceptibility to cryopreservation and thawing. A second problem causing variability even after removal of these proliferating cells was found to be due to additional non-T cells in the marrow that responded to soluble mediators produced by peripheral blood lymphocyte stimulator cells during an MLR. Such non-T-cell stimulation was not eliminated by removal of bone marrow T cells, obscuring the results of T cell depletion of the marrow. This problem was overcome by the use of HLA-defined B cell lines as stimulators. A mixture of such lines provided a reliable stimulator source that did not produce soluble mediators capable of stimulating additional marrow cells. These refinements of MLR conditions permit a reproducible and reliable assay of bone marrow MLR, and provide a means for assessment of elimination of such alloreactive cells by monoclonal antibodies and complement.