Development and application of a two-site enzyme immunoassay for the determination of 'total' activin-A concentrations in serum and follicular fluid.

The performance of existing immunoassays and bioassays for activins is compromised by the presence of activin-binding proteins such as follistatin and alpha 2 macroglobulin (alpha 2M) in biological fluids. To overcome this problem we have developed a novel two-site enzyme immunoassay procedure for activin-A which incorporates an analyte denaturation and oxidation step. The optimized assay is sensitive (detection limit approximately 10 pg/well), precise (mean within- and between-plate coefficients of variation 4.9 and 9.1% respectively) and accurate (activin-A recovery values of 102 +/- 3 and 96 +/- 5% for bovine follicular fluid (FF) and human serum respectively). In specificity tests, high concentrations of follistatin (500 ng/ml) and alpha 2M (100 microgram/ml) did not interfere with the response signal to activin-A. In addition, no significant cross-reactivity was observed with a range of related molecules including inhibin-A, inhibin-B, activin-B (all < 0.5%), bovine pro-alpha C and follistatin (both < 0.1%). Response curves parallel to the activin-A standard curve were obtained for a variety of test samples including bovine, human, ovine and porcine FF, human sera and conditioned medium from cultured bovine and human granulosa cells. Fractionation of bovine FF by SDS-PAGE confirmed assay specificity since only one peak of activin-A immunoreactivity was detected (M(r) approximately 25 k) in eluted gel slices. However, gel-permeation chromatography showed that under physiological conditions all of the detectable activin-A in bovine FF eluted with apparent M(r) values of > 700 and 60-200 k reflecting its association with binding protein(s). Analysis of bovine FF samples (n = 76) from morphologically dominant follicles during the luteal phase showed that activin-A levels were positively correlated with inhibin-A (r = +0.54; P < 0.001) and total beta subunit immunoreactivity (r = +0.32; P < 0.005) but not with total alpha subunit immunoreactivity (r = -0.09). Classification of these follicles according to oestrogenic status showed that activin-A, inhibin-A and total beta subunit levels were highest in oestrogen-inactive follicles (P < 0.01) whereas total alpha subunit levels were lowest in these follicles (P < 0.001). Activin-A levels were measurable in all human serum samples analysed, ranging from 128 pg/ml during the normal menstrual cycle, 210 pg/ml in women undergoing ovarian hyperstimulation and approximately 500 pg/ml in postmenopausal women to over 4000 pg/ml during pregnancy. In conclusion, the present assay provides a reliable method for quantitating total (i.e. bound+free) activin-A concentrations in a variety of biological samples and should prove useful for further in vivo and in vitro studies in a range of species including man.