Rapid Method for the Estimation of Total Free Estrogens in Plasma Eugene Cerceo and Cipriano Elloso Department of Pathology, Thomas Jefferson University Hospital, Philadelphia, Pa. 19 107
In recent years, much work has been devoted to the development of suitable radioimmunoassays for estrogens (1-9). Many of these investigations have focused on the determination of estradiol-17& the physiological most potent of the three estrogens. These assays have been specific, sensitive, and accurate-and also time-consuming. In this investigation, a radioimmunoassay is described for the rapid estimation of the total free (or unconjugated) estrogen content of blood (estrone, estradiol, and estriol) using an antibody with a high cross-reactivity to all three forms. With the procedure described here, no elaborate extraction of the plasma or chromatographic separation prior to the assay is necessary, with the exception of precipitation of plasma proteins through the addition of methanol to a small plasma aliquot. Aside from the interference caused by the presence of plasma proteins, other factors influence the antigen-antibody binding. As an example, coupling of the antigen to the antibody is optimum a t a p H between 7 and 8. Recent studies (10, 11) have shown that there is increased antigen-antibody binding if the buffer contains neutral salts. Thus, binding is a function of ionic strength, although too high an ionic strength can lead to nonspecific binding. Even though in a radioimmunoassay a tracer with the maximum specific activity is desirable, too high a specific activity in the antigenic part of the molecule can cause a decrease in binding to the antibody. This is a steric effect 04 OH
LSIQA3IOL
-
1 1 '
1SlROUt
tSIRlO,
due to the larger size of the radioisotope bound to the tracer molecule. Specifically, in this case the use of 2,4,6,7,-3H estradiol-17P as a tracer yielded a lesser percent binding to antibody than 6,7-3H estradiol-17P. The small amount of binding inhibition is the result of two extra tritium atoms on the antigenic part of the steroid nucleus. This suggests a high tightness-of-fit between the antibody's active site and the aromatic ring of the steroid (12). (1) G. Abraham and D. Tulchinsky, J. Clin. Endocrinol.. 33, 775 (1971). (2) G. Abraham, W. Odell, R. Edwards, and J. Purdy. Acta Endocrinol. (Copenhagen) Suppb, 143 (7), 332 (1970). (3) N. S.Jiang and R. Ryan, Mayo Clin. Proc., 44, 461 (1961). (4) G. Abraham, J. Clin. Endocrinol., 29, 866 (1969). (5) E. Gurpride eta/., Amer. J. Obstet. Gynecol., 109, 897 (1971). (6) G. Mikhail, M. Ferin. and R. Vande Wiele, Acta Endocrinol. (Copenhagen) Suppl., 143 (7), 347 (1970). (7) A. R. Midgley and G. D. Niswender, Acta Endocrinol. (Copenhagen) Suppl., 146, 320 (1971). (8) D. A . Shutt, Steroids, 13, 69 (1969). (9) B. England, G. Niswender, and A. Midgley, J. Clin. Endocrinol. 38, 42 (1974). (10) M. Ganguly and U. Westphal, J. Bid. Chem., 243, 6130 (1968). (1 1) M. Ganguly, R. Carneghan, and U. Westphal. Biochemistry, 6, 2803 (1967). (12) F. G. Peron and B. V. Burton, Ed., "Immunologic Methods in Steroid Determination," Meredith Corporation, New York, N.Y., 1970, p 99.
1578
Cholesterol, despite its very low immunoreactivity with the antibody (0.01%) and its slight solubility in methanol, can become a factor due to its very large concentration (approximately 250 mg/100 ml). Suitable aliquots of methanol-plasma solution are taken that will yield optimum reTsults. Aliquots suggested in this assay have shown no interference due to the presence of cholesterol. Other factors affecting radioimmunoassays of steroids have been previously reported (13). Repeat determinations have shown that the assay yields reasonably accurate data on the level of total estrogens in plasma and also amniotic fluid. The assay can be modified to give the concentrations of the individual estrogens utilizing suitable chromatographic techniques (14, 15).
EXPERIMENTAL Reagents. T h e estrogen standard for the assay was made from estradiol-170 and pig skin gelatin, type 11, used as t h e carrier protein. These were both obtained from Sigma Chemical Company, St. Louis, Mo. 63118. T h e stock concentration for the estrogen standard was 10.0 ng/ml. Methanol, spectroanalyzed, was purchased from Fisher Scientific Company, King of Prussia, Pa. 19406. [6,7-3H]Estradiol-17S(55 Ci/mmol) was obtained from New England Nuclear, Cambridge, Mass. 02118, while (2,4,6,7-3H]estradiol- 17p was purchased from Amersham/Searle Corporation, Arlington Heights, Ill. 60005. Antiserum S310 No. 5 was acquired from Harbor General Hospital, Torrance, Calif. 90509. This antibody was produced in rabbits through repeated injections using estriol 3,16 e,17~-trihemisuccinate-horseserum albumin. Its crossreactivity with estrogens and other steroids was obtained from Harbor General Hospital and is shown in Table I. Charcoal was coated with dextran T-70 by mixing equal volumes of 5.0 g/l. dextran solution (powdered dextran T - 7 0 was obtained from Pharmacia Fine Chemicals, Piscataway, N.J. 08854) with 50 g/1. suspension of Norit-A charcoal (Fisher Scientific Company). Scintillation "cocktail" was made by dissolving 4.0 g of 2,5-diphenyloxazole (PPO) and 0.2 g of p-bis[2-(5-phenyloxalyl)]benzene (POPOP), both from New England Nuclear, into 1.0 liter of scintillation grade toluene. To this solution was added 500 ml of Triton X-100 (New England Nuclear). T h e resulting solution was stored in dark bottles. E q u i p m e n t . Radioactive counting of the samples from t h e assays was conducted on the Mark I1 liquid scintillation counter supplied by Searle Analytic (formerly Nuclear Chicago Corporation), Des Plaines, Ill. 60018. P r o c e d u r e . To 200 /11 of plasma is added 1.8 ml of methanol for the purpose of protein precipitation. T h e mixture is carefully agitated and then centrifuged. Then 400.~1aliquots of the clear methanol-plasma solution are pipetted into separate test tubes and taken to dryness under a stream of nitrogen. During pregnancy, smaller aliquots or titering of the specimen may be necessary. A working standard of 5.0 ng/ml estradiol-17p' is made by mixing equal volumes of the 10 ng/ml standard with methanol. Appropriate volumes are pipetted into duplicate test tubes to generate the standard curve. These standards are taken to dryness under a stream of nitrogen. Buffer reagent is prepared by adding tritiated estradiol and a n tiserum S-310 No. 5 t o 0.10M phosphate buffer, p H 7.5 containing 0.15M sodium chloride and 0.05% pig skin gelatin. T h e gelatin acts (13) E . Cerceo and C. Elloso, Clin. Chem., 18, 537 (1972). (14) H. Van Baelen. W. Heyns, and P. De Moor, J. Chromatogr.. 30, 226 (1967). (15) J. Touchstone, T. Murawec, and 0. Brual, J. Chromatogr., 37, 359 (1968).
ANALYTICAL CHEMISTRY, VOL. 46, NO. 11, SEPTEMBER 1974
Table I. Cross Reaction of Various Steroids with Antiserum S-310 No. 5 a Cross reaction, R
Steroid
Co7steroids Cholesterol steroids Progesterone 17-Hydroxyprogesterone 5-Pregnenolone Aldosterone Cortisol C i 9steroids Androsterone Dehy droepiandrosterone (DHEA) Testosterone Cis steroids Estrone Estradiol-17 Estriol 16-Epiestriol 16,17-Epiestriol 17-Epiestriol a
Table 11. Protocol for t h e Radioimmunoassay of Total Estrogens
0.01
< o ,001
