The measurement of transforming growth factor type beta (TGF beta) levels produced by peripheral blood mononuclear cells requires the efficient elimination of contaminating platelets.

In recent years there has been an increasing interest in measuring the levels of TGF beta produced by peripheral blood mononuclear cells (PBMC), since its abnormal regulation seems to be involved in several pathological states. Platelet-contamination, a common feature in PBMC populations isolated by the standard Ficoll-Paque method, would theoretically disturb the measurement of the levels of TGF beta produced by mononuclear cells, since platelets represent an important source of this cytokine. In this study, supernatants of PBMC cultures from healthy subjects, either platelet-contaminated or uncontaminated, were assayed for TGF beta activity in three different bioassays. We report that the presence of platelets led in most cases to an important overestimation of the TGF beta levels produced by MNC in the Swiss-3T3 bioassay and in a PBMC proliferation assay. In contrast, in the Mv1Lu bioassay these levels were significantly underestimated, an effect which we attribute to the presence of other platelet-derived growth factors. These results suggest that the elimination of platelets from PBMC cultures is essential if TGF beta production by mononuclear cells is to be studied.