A sensitive and specific sandwich enzyme immunoassay for human thyroid-stimulating hormone

A sensitive and specific sandwich enzyme immunoassay (EIA) for human thyroid-stimulating hormone (hTSH) has been developed. hTSH is incubated with anti-hTSH IgG-coated polystyrene balls, and after washing they are further incubated with anti-hTSH Fab'-β-d-galactosidase conjugate. The β-d-galactosidase activity bound to the polystyrene balls is proportional to the amount of hTSH to be assayed. Polystyrene balls are coated with rabbit anti-hTSH IgG which had been affinity-purified and treated with human chorionic gonadotropin-Sepharose 4B to remove antibodies cross-reacting with structurally related hormones. Rabbit anti-hTSH Fab', which had been affinity-purified was conjugated with β-d-galactosidase from Escherichia coli using N,N′-o-phenylenedimaleimide.In the specific sandwich enzyme immunoassay developed, 1 nU (1 × 10^?9 U) of hTSH per tube can be measured and the sensitivity for serum hTSH level is 0.1 μU/ml when 10 μl of serum is used. No significant interference was observed in the presence of 1.3 mU hLH/tube, 0.5 mU hFSH/tube or 0.5 U hCG/tube. Recoveries of hTSH added to human sera were 95.3–104% with a standard deviation of 12.0–14.9%. The coefficients of within-assay and between-assay variations were 6.0–7.5% and 4.9–8.7%, respectively. The regression equation and coefficient for correlation to radioimmunoassay (RIA) were y (RIA) = 0.95x (EIA) + 3.2 and 0.97, respectively.Serum levels of hTSH in normal male and female adults were 2.4 ± 1.0 (SD) (n = 41) and 2.9 ± 1.3 (n = 46) μU/ml, respectively; those in hyperthyroidism and hypothyroidism were 0.28 ± 0.06 (n = 20) and 49.6 ± 24.7 (n = 22) μU/ml, respectively; and those in pregnant and postmenopausal women were 2.5 ± 1.2 (n = 7) and 2.7 ± 1.0 (n = 35) μU/ml, respectively, indicating that high serum levels of hCG or hLH and hFSH under these conditions did not significantly interfere with the present assay of hTSH at normal levels.