Expression of tissue factor and factor VIIa/tissue factor inhibitor activity in endotoxin or phorbol ester stimulated U937 monocyte-like cells

Previously, unstimulated cells of the human monocytic tumor cell line U937 have been shown to possess a negligible cell-surface tissue factor (TF) activity, and to secrete a small amount of factor VIIa/tissue factor (VIIa/TF) inhibitor activity. On stimulation with endotoxin or with phorbol myristate acetate (PMA), TF of these cells is known to be increased approximately fourfold. In this report, we demonstrate that VIIa/TF inhibitor is also increased on stimulation of U937 cells with endotoxin (approximately equal to threefold) or with PMA (approximately equal to 20-fold). Notably, the secretion of the inhibitor persisted after the cell surface TF had started to decline. Further, when serum- free media from PMA stimulated cells was electrophoresed on a sodium dodecyl sulfate (SDS) gel, we eluted two inhibitor activity peaks corresponding to Mr approximately equal to 47,000 and Mr approximately equal to 36,000. The molecular weights of these peaks are similar to those obtained earlier from human plasma for this inhibitor(s).     

cis-Retinoic acid stimulates the clonal growth of some myeloid leukemia cells in vitro

We studied the effects of cis-retinoic acid (cisRA) on the clonogenic growth of samples of leukemic cells from 35 patients with acute nonlymphocytic leukemia (ANLL). We observed significant inhibition of leukemic colony growth in 17 samples by 10(-7) to 10(-6)M cisRA. However, we found that retinoid exposure resulted in striking stimulation of clonal growth in ten samples at the same drug concentrations. With the exception of cases with promyelocytic features, there was no morphologic or functional evidence that cisRA induced the leukemic blasts to differentiate. Both inhibition and stimulation were dose-dependent and observable at pharmacologically achievable levels of cisRA. Leukemic cells with monocytic features more frequently demonstrated a stimulatory response than did those without monocytic features. Depletion of T lymphocytes and monocytes did not alter the type of growth response. Assays for cellular retinoic acid- binding protein (CRABP) were performed on five samples (two with inhibitory growth responses, two with stimulatory responses, and one with no growth) and failed to reveal detectable levels of CRABP in any case. The addition of cisRA to liquid suspensions of leukemic cells produced no significant change in the number of viable cells. We conclude that the effects of cisRA on leukemic colony growth are not cytotoxic and not mediated by T lymphocytes, monocytes, or CRABP. More importantly, cisRA appears to enhance the growth of certain human leukemia cells in vitro. Taking into account the increasing use of retinoids in clinical trials for patients with leukemia, the latter findings may represent a significant cautionary note.