Effect of estrogens and antiestrogens (AEs) on estrogen receptor (ER) half-life was analyzed in MCF-7 cells by assessing its progressive disappearance after covalent labeling in situ with [3H]tamoxifen aziridine ([3H]TAZ). Cells were incubated for 1 h with 20 nM [3H]TAZ either in the absence or presence of a 500-fold excess of unlabeled estradiol (E2) (non-specific binding). The entire ER population was labeled by this method as established by subsequent incubation of the cells with [125I]E2. [3H]TAZ labeled cells were maintained in culture for additional 5 h in the absence (control) or presence of increasing amounts (0.1 nM - 1 microM) of either a given estrogen (E2, estrone, diethylstilbestrol, bisphenol), a pure AE (RU 58 668, ICI 164 384) or an AE with residual estrogenic activity (RU 39 411, 4-hydroxytamoxifen, keoxifene). The progressive disappearance of nuclear and cytosolic [3H]TAZ-ER complex during 5 h incubation were assessed by their immunoprecipitation with anti-ER monoclonal antibody (H 222) followed by scintillation counting or SDS-PAGE and fluorography. Fading of labeled receptors was extremely slow (approximately 10% loss after 6 h) in absence of any hormone/antihormone indicating a long half-life of the [3H]TAZ-ER complex. Addition of estrogens as well as pure AEs led to a dramatic reduction of the half-life while AEs with residual estrogenic activity were extremely less efficient in this regard providing an explanation for the ability of latter compounds to up-regulate the receptor since they do not affect ER mRNA synthesis and stability. Receptor disappearance induced by estrogens was closely related to their binding affinity for ER. Newly synthesized ER emerged during the treatment with hormones or antihormones seems to be implicated in the phenomenon since [3H]TAZ was covalently bound and could, therefore, not be displaced by these compounds. Induction of synthesis of a short half-life peptide(s) with degradative activity was demonstrated by addition of cycloheximide or puromycine (both at 50 microM) which completely blocked ER disappearance. The fact that no cleavage products of ER were detected by SDS-PAGE suggested a lysosomial hydrolysis. Hence, hormonal modulation of only a part of ERs may down-regulate their total population until it reaches the steady-state level.