The activation of Myc induces apoptosis of human ovarian adenocarcinoma N.1 cells when serum factors are limited. However, the downstream mechanism that is triggered by Myc is unknown. Myc-activation and treatment with the proapoptotic ligands TNFalpha, FasL, and TRAIL induced H-ferritin expression under serum-deprived conditions. H-ferritin chelates intracellular iron and also intracellular iron sequestration by deferoxamine-induced apoptosis of N.1 cells. Supplementation of serum-free medium with holo-transferrin blocked apoptosis of N.1 cells that was induced by Myc-activation or by treatment with TNFalpha, FasL, and TRAIL, whereas apotransferrin did not prevent apoptosis. This suggests that intracellular iron depletion was a trigger for apoptosis and that transferrin-bound iron rescued N.1 cells. Furthermore, apoptosis of primary human ovarian carcinoma cells, which was induced by TNFalpha, FasL, and TRAIL, was also inhibited by holo-transferrin. The data suggest that Myc-activation, FasL, TNFalpha, and TRAIL disturbed cellular iron homeostasis, which triggered apoptosis of ovarian carcinoma cells and that transferrin iron ensured survival by re-establishing this homeostasis.

No abstract

A mucin-like carcinoma-associated antigen (MCA) was recently identified on the surface of established breast cancer cell lines by several monoclonal antibodies. The antibody b-12 was used in a sandwich enzyme immunoassay to measure MCA concentrations in serum samples and other biological fluids. The upper limit for noncancerous women and men was 14 U/ml. MCA levels were independent of estrogen or prolaction secretion. 63% of patients with metastatic breast cancer had elevated serum concentrations of MCA. Elevated MCA levels were also associated with cervical, ovarian or endometrial cancer and carcinoma of the prostate. In metastatic breast cancer, MCA and CA 15.3 showed similar sensitivity. Carcinoembryonic antigen levels did not correlate with MCA. Serum concentrations of MCA increased during pregnancy and remained elevated in nursing mothers. Amniotic fluid was found to be rich in MCA. In contrast, CA 15.3 showed only small changes during pregnancy and was low in amniotic fluid. From binding tests with antibodies used in the MCA and CA 15.3 assays, we conclude that the monoclonal antibodies b-12 as well as 115-D8 and DF3 (CA 15.3 assay) recognize coexisting epitopes on mucin-like antigens, which belong to a polymorphic family of glycoproteins suitable for tumor monitoring. Differences in the behavior of MCA and CA 15.3 may emerge from the complexity and heterogeneity of these mucin-like antigens.

No abstract