Estradiol-Testosterone Binding Interactions and Free Plasma Estradiol under Physiological Conditions*

We have employed a flow dialysis technique (FDT) to describe the interactions of estradiol (E2), testosterone (T), testosterone-estradiol-binding globulin (TEBG), and albumin in undiluted, 37 C plasma with negligible disturbance of the equilibrium. The percentages of free E2 and free T were measured simultaneously, both with and without the addition of exogenous E2 and T, and were found to correlate closely in a rectangular hyperbolic relationship which closely approximated linearity in the normal range. A physical chemical model was developed to fit these data. Under these physiological conditions, we found that E2 and T binding are related in a fashion consistent with a model in which they compete for TEBG-binding sites and in which albumin is a source of unsaturable low affinity binding sites. The experimentally derived value of the TEBG association constant for E2 (KtE = 5.0 × 108 M-1) is 50% that for T (KtT = 9.8 × 108 M-1). The albumin apparent association constants for E2 and T are comparable (3.5–3.6 × 104 M-1). T binding to TEBG was also evaluated by a modification of our previously described charcoal adsorption assay (CA). The relationship between T binding assessed by the modified CA assay and that assessed by FDT was then determined. Subsequently, E2 binding as well as T binding under physiological conditions could be calculated from assay of T binding by the CA method, since E2 and T compete for TEBG-binding sites. Using this approach, the plasma free E2 and T levels of normal women, hirsute women, and normal men were computed after determination of total plasma E2 and T concentrations. The free E2 concentrations of normal follicular phase females, hirsute follicular phase or anovulatory females, and normal males averaged 1.9 ± 1.3, 1.4 ± 0.6, and 0.76 ± 0.24 (SD) pg⁄ml, respectively; the total E2 levels in these groups were 80 ± 57, 45 ± 21, and 23 ± 5 pg⁄ml. The differences between these values for normal and hirsute women were not significant. The free T levels of these groups averaged 7.4 ± 2.6, 22 ± 9, and 220 ± 82 pg⁄ml. Our data show that the imbalance between plasma free T and E2 in hirsutism is due solely to the increase in free T. However, hirsute women who have anovulatory cycles tend to have an E2 deficit which is the consequence of failure of the late follicular and luteal E2 surges to occur.