0021-972X/78/4705-1092$02.00/0 Journal of Clinical Endocrinology and Metabolism Copyright © 1978 by The Endocrine Society
Vol. 47, No. 5 Printed in U.S.A.
A Specific Radioimmunoassay for Estrone Sulfate in Plasma and Urine without Hydrolysis* K. WRIGHT.f D. C. COLLINS, P. I. MUSEY, AND J. R. K. PREEDY Departments of Medicine and Biochemistry, Emory University School of Medicine, Atlanta, Georgia 30322 ABSTRACT. Estrone sulfate, quantitatively the most important estrogen in plasma, has previously been determined only after hydrolysis and chromatography. An antiserum raised against estrone glucosiduronate-bovine thyroglobulin was found to be suitable for the specific RIA of estrone sulfate both in plasma and urine. Plasma levels were measured after solvent extraction without hydrolysis or chromatography. The mean (±SE) was 972 ± 79 pg/ml (range, 537-1581) in 15 women in the follicular phase, 1806 ± 160 pg/ml (range, 814-3358) in 15 women in the luteal phase, and 922 ± 62 pg/ml (range, 461-1238) in 13 men. The urinary excretion of estrone sulfate, measured after simple chromatographic sepa-
E
STRONE sulfate has been shown to be a major metabolite of estrone and 17/?estradiol in man (1-3) and is quantitatively the most important estrogen in peripheral plasma (4-6). In addition, estrone sulfate itself is further metabolized (2, 3). Such metabolism includes hydrolysis to estrone (2, 3). A variety of techniques have been used to measure this conjugate in plasma. Brown and Smyth (4) used solvent partition to separate the estrone sulfate, followed by fluorimetric estimation. Loriaux et al. (5) isolated the steroid sulfates by chromatography and estimated the estrone released on solvolysis by RIA. Hawkins and Oakey (6) used solvolysis of the sulfates, followed by competitive binding analysis after reduction of the purified estrone to 17/?-estradiol. These procedures are all relatively laborious and time consuming. The method of Received April 11, 1978. Address all correspondence and requests for reprints to: Dr. D. C. Collins, Emory University School of Medicine, 69 Butler Street, S.E., Atlanta, Georgia 30303. * These data were presented in part at the 60th Annual Meeting of The Endocrine Society, June, 1978. This work was supported in part by Contract N01-CB-74101 and Grant R01-HL 16394-02 from the NIH. t Present address: Department of Obstetrics and Gynecology, Yale University School of Medicine, New Haven, Connecticut 06510.
ration, ranged from 0.8-7.9 /ig/24 h in men and 5.1-18.7 fig/24 h in nonpregnant women. This was generally oneseventh to one-half the simultaneous estrone glucosiduronate excretion rate. An approximate mean renal clearance of estrone sulfate calculated from the above values was 2.7 ml/min. The low clearance rate is taken to reflect extensive binding of estrone sulfate by plasma proteins. The splanchnic extraction of estrone sulfate measured in 6 patients undergoing hepatic vein catheterization for diagnostic purposes was 29.8 ± 11.1%, indicating net uptake of this compound by the splanchnic area. (J Clin Endocrinol Metab 47: 1092, 1978)
Brown and Smyth (4) is also relatively insensitive. All attempts to raise a specific antiserum for estrone sulfate have thus far been unsuccessful. We examined several antisera raised against estrone glucosiduronate1-bovine thy1
The following trivial names were used: androsterone3-glucosiduronate, 17-oxo-5a-androstan-3a-yl-/?-D-glucopyranosiduronate; dehydroisoandrosterone-3-sulfate, 17oxoandrost-5-en-3/?-yl-sulfate; 16-epiestriol, estra-1,3,5(10)-triene-3,16/?,17/8-triol; 16,17-epiestriol, estral,3,5(10)-triene-3,16/?,17a-triol; 17-epiestriol, estral,3,5(10)-triene-3,16a,17a-triol; estetrol, estra-l,3,5(10)triene-3,15a,16a,17/?-tetrol; 17/?-estradiol-3-glucosiduronate, 17/?-hydroxyestra-l,3,5(10)-trien-3-yl-/?-D-glucopyranosiduronate; 17/?-estradiol-17-glucosiduronate, 3hydroxyestra-1,3,5(10) -trien-17/?-yl-/?-D-glucopyranosiduronate; estriol-3-glucosiduronate, 16a,17/?-dihydroxyestra-l,3,5(10)-trien-3-yl-/?-D-glucopyranosiduronate; estriol- 16a-glucosiduronate, 3,17/?-dihydroxyestra-1,3,5( 10) trien- 16a-yl-/?-D-glucopyranosiduronate; estriol- 17-glucosiduronate, 3,16a-dihydroxyestra-l,3,5(10)-trien-17/?-yl/8-D-glucopyranosiduronate; estrone glucosiduronate, 17oxoestra-l,3,5(10)-trien-3-yl-/?-D-glucopyranosiduronate; etiocholanolone-3-glucosiduronate, 17-oxo-5/?-androstan3a-yl-/?-D-glucopyranosiduronate; 2-hydroxyestradiol, estra-l,3,5(10)-triene-2,3,17/?-triol; 2-hydroxyestrone, 2,3dihydroxyestra-1,3,5(10)-trien-17-one; 16a-hydroxy estrone, 3,16a-dihydroxyestra-l,3,5(10)-trien-17-one; 16-ketoestradiol, 3,17/?-dihydroxyestra-l,3,5(10)-trien-16-one; 16-ketoestrone, 3-hydroxyestra-l,3,5(10)-triene-16,17dione; mestranol, 17a-ethinylestra-l,3,5(10)-trien-17/?-ol3-methyl ether; estrone-2-methyl ether, 3-hydroxyestral,3,5(10)-trien-17-one-2-methyl ether; estrone-3-methyl
1092
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 November 2015. at 16:12 For personal use only. No other uses without permission. . All rights reserved.
RIA OF ESTRONE SULFATE roglobulin while developing a specific RIA procedure for estrone glucosiduronate (7). As we characterized these antisera, we noted one antiserum that showed a very high cross-reaction with estrone sulfate but which retained specificity in relation to other estrogens and estrogen conjugates. In this paper, we describe the use of this antiserum to develop a RIA procedure for estrone sulfate which does not require prior hydrolysis or (in the case of plasma) chromatography. This procedure was used to measure peripheral levels of estrone sulfate in normal men and nonpregnant women and to determine the percentage of extraction of estrone sulfate by the splanchnic area. Urinary levels of estrone sulfate were measured after simple chromatography on DEAE-Sephadex in 10 urine samples from men and nonpregnant women and compared to both the plasma levels of estrone sulfate and to the urinary levels of estrone glucosiduronate. Materials and Methods Nonradioactive steroids and steroid conjugates were obtained from Sigma Chemical Co. (St. Louis, MO) or Steraloids (Wilton, NH) and purified by recrystallization. Purity was confirmed by melting point determination and thin layer chromatography. [6,7-3H]Estrone sulfate (SA, 54 Ci/mmol), obtained from New England Nuclear Co. (Boston, MA), was purified by chromatography on DEAESephadex (8). Ethyl acetate and ethanol were fractionally distilled. Anesthesia grade diethyl ether was used without purification. Steroids and steroid conjugates were stored at 4 C in ethanol. Estrone sulfate was stable for at least 3 months in ammoniacal ethanol and was prepared and stored at 4 C in that solvent. Phosphate-buffered saline [PBS; 0.1 M sodium phosphate, pH 6.8, containing (wt/vol) 0.9% NaCl, 0.1% ethylmercurithiosalicylate, and 0.1% bovine serum albumin (Sigma; RIA grade)] was used as diluent for the antiserum and in the assay. The radioactive solution of [3H]estrone sulfate (400,000 dpm/ml PBS) was prepared fresh daily because of ether, 17-oxoestra-l,3,5(10)-triene-3-methyl ether; pregnanediol-3-glucosiduronate, 20a-hydroxy-5/?-pregnan-3ayl-/?-D-glucopyranosiduronate; testosterone- 17/?-glucosiduronate, 3-oxoandrost-4-en-17/?-yl-/3-D-glucopyranosiduronate.
1093
the increased rate of breakdown with prolonged storage in PBS. Radioactivity was measured in Packard liquid scintillation spectrometers (models 3320 or 3255) using Scintiverse universal scintillation cocktail (Emory formula, Fisher Scientific).
Antiserum Antisera against estrone glucosiduronate-bovine thyroglobulin were produced in six male rabbits, as previously described (7). These antisera were tested for their ability to bind [3H]estrone sulfate. One antiserum was found to have sufficient specificity to be used in the RIA of estrone sulfate. The crossreactivities with other steroids and steroid conjugates for this antiserum were determined as previously described (9) and expressed as the percent equivalents of estrone sulfate. Plasma RIA of estrone sulfate Samples of plasma (0.05, 0.1, or 0.2 ml) were added to assay tubes containing 0.1. ml PBS or 0.1 ml [3H]estrone sulfate in PBS. After the addition of 0.5 ml 3 M NaCl, the samples were mixed thoroughly and allowed to equilibrate for 30 min at 22 C. The samples were extracted once with 4 ml ether and the ether extract was discarded. The samples were then extracted twice with 4 ml ethyl acetate and the ethyl acetate extract was transferred to tubes for assay or scintillation vials to assess recovery, dried under N2, and redissolved in 0.5 ml PBS. Antiserum (0.1 ml 1:10,000 dilution in PBS) and [3H]estrone sulfate (40,000 dpm or 125 pg in 0.1 ml PBS) were added to samples and standards, mixed thoroughly, and incubated at 4 C for 60 min or 18 h. Bound and free steroids were separated by centrifugation (1500 X g) after incubation with 1 ml dextran-coated charcoal (0.5% Norit and 0.1% dextran in PBS) for 15 min at 0 C. The supernatant was decanted into scintillation vials and counted. The accuracy of the method was determined by assay of a male plasma sample to which known amounts of estrone sulfate had been added. Intraassay and interassay variations were assessed at three levels of estrone sulfate within an assay and in a series of assays, respectively. Solvolysis Samples were hydrolyzed and the released estrone was measured using a method similar to that described by Hawkins and Oakey (5), as follows. One milliliter of plasma was mixed with 4 ml H2O together with [3H]estrone sulfate and allowed to
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 November 2015. at 16:12 For personal use only. No other uses without permission. . All rights reserved.
1094
WRIGHT ET AL.
equilibrate 30 min at room temperature. Ten microliters of 3 N NaOH were added and the samples were extracted twice with 20 ml diethyl ether. Ammonium sulfate (1.1 g), 25 /xl 66% H2SO4, and 5 ml ethyl acetate were added to the aqueous residue, mixed thoroughly, and incubated for 18 h at 37 C. Hydrolyzed estrone was extracted twice with 10 ml ether. The ether extracts were combined and washed sequentially with 2 ml H2O and 2 ml 8% NaHCO3. The washed extracts were dried, redissolved in 1 ml H2O, and extracted twice with 5 ml diethyl ether. The ether extract was dried under N2 and redissolved in 1 ml ethanol. Aliquots of 50 and 100 /xl were assayed for estrone using a specific RIA (9). Aliquots of 500 fi\ were counted to assess recovery. The values obtained for estrone were corrected for the molecular weight difference between estrone and estrone sulfate by multiplying by 1.38. Urinary RIA of estrone sulfate Collections of 24-h urine samples were carried out by a group of normal men and women and stored at —20 C until analyzed. A 15-ml aliquot to which [3H]estrone sulfate had been added was extracted with 30 ml ethanol by mixing vigorously and standing overnight at —20 C. The sample was centrifuged and the supernatant was removed and taken to dryness in a flash evaporator. The residue was applied to a 1 x 60-cm DEAE-Sephadex column and eluted sequentially with 500 ml deionized distilled H2O, 1000 ml 0.1 M NaCl, and 500 ml 0.2 M NaCl (8). Aliquots of each fraction from the 0.2 M NaCl elution step were counted to locate the estrone sulfate peak. The tubes comprising the [3H]estrone sulfate peak were pooled and an aliquot of the pool was counted to assess recovery. The purified estrone sulfate fraction was quantitated by direct RIA of 0.5-ml aliquots after dilutions in PBS of 1:5; 1:10, and 1:20. The amount of estrone sulfate present in each urine sample was compared to that of estrone glucosiduronate, measured as previously described (7). Splanchnic extraction Plasma samples were collected simultaneously from the hepatic vein and from a peripheral artery in a group of patients undergoing cardiac catheterization for diagnostic purposes. The samples were stored at -20 C until analyzed. The level of estrone sulfate present in the simultaneously collected samples was determined as described above. The percentage of extraction of estrone sulfate by the splanchnic area was calculated from the standard
JCE & M • 1978 Vol 47 • No 5
formula as follows (10, 11): % extraction = 100 [1 — (HV/A)], where A and HV are the endogenous levels of estrone sulfate in a peripheral artery and the hepatic vein, respectively.
Results Antiserum The cross-reactions of the antiserum used for the RIA of estrone sulfate are shown in Table 1. The antiserum shows relatively little cross-reactivity for the unconjugated estrogens, except for estrone and estrone-3-methyl ether which showed cross-reactions of greater TABLE 1. Percentage of cross-reaction of various steroids with an antiserum against estrone glucosiduronate-bovine thyroglobulin used for the RIA of estrone sulfate Compound Estrone 17/?-Estradiol Estriol 17a-Estradiol 2-Hydroxyestrone 2-Hydroxyestradiol 2-Hydroxyestriol Estrone-2-methyl ether Estrone-3-methyl ether 16-Ketoestrone 16-Ketoestradiol 16a-Hydroxyestrone 16-Epiestriol 17-Epiestriol 16,17-Epiestriol Estetrol
% Cross-reaction 108.3 2.8 0.0 3.1 3.8 3.5 0.3 14.2 >200.0 0.6 4.5 1.4 3.2 0.4 0.4 0.2
Progesterone Testosterone Cortisol
0.0 0.4 0.0
Ethinylestradiol Mestranol Diethylstilbestrol
1.4 0.2 0.0
Estrone glucosiduronate Estrone sulfate 17/?-Estradiol-17-glucosiduronate Estriol- 16a-glucosiduronate 17/?-Estradiol-3-glucosiduronate Estriol-3-glucosiduronate Estriol-3-sulfate Estriol-17-glucosiduronate Estriol-16-sulfate Androsterone-3-glucosiduronate Etiocholanolone-3-glucosiduronate Testosterone-17/?-glucosiduronate Pregnanediol-3-glucosiduronate Dehydroisoandrosterone-3-sulfate
>200.0 100.0 0.0 0.0 9.7 1.0 0.4 0.0 5.7 0.6 0.7 0.6 0.0 0.8
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 November 2015. at 16:12 For personal use only. No other uses without permission. . All rights reserved.
RIA OF ESTRONE SULFATE
1095
than 100%. None of the nonestrogenic steroids TABLE 3. The mean values (±SE) measured in a sample of male plasma after addition of known amounts of estested showed any significant cross-reaction. trone sulfate The antiserum showed low cross-reactivity pg Added pg Recovered (mean ± SE) with the conjugated estrogens, except for es29.0 31.8 ± 5.8 (6) trone glucosiduronate where the cross-reac67.0 66.4 ± 4.9 (15) tion was greater than 200%. 140.0 133.5 ± 8.2 (18) Number of samples analyzed is shown in parentheses.
RIA of estrone sulfate A logit-log transform of a typical standard curve is shown in Fig. 1. The standard curve was useful over the range 15-350 pg, with 50% displacement at 125 pg. The sensitivity was found to be 15 pg in a series of six assays. The intraassay and interassay variations are shown in Table 2. The intraassay variation was calculated from samples at three different points on the standard curve. The coefficients of variation (CV = SD X 100 -*- mean) were less than 15% at each point. The interassay variation was calculated from multiple analyses of three different plasma samples. The CVs were found to be less than 20% for all samples.
~ 2400r Id 2000 1600 I2OO 800
= 0 88x + 151 r=O95
400
1600 2400 800 SOLVOLYSIS AND RIA OF ESTRONE (pg/ml)
FIG. 2. Comparison of plasma values of estrone sulfate obtained by specific RIA with values obtained by a different method requiring solvolysis and assay of estrone. The line shown is the calculated regression line.
Recovery experiments were carried out on a sample of male plasma to which three difZ. 60ferent amounts of estrone sulfate had been o 50added. The plasma was analyzed and the reZ 40O 30sults are shown in Table 3. In each case, the * ~ amount recovered was not significantly differ10ent from the amount added. l The correlation in 12 samples of values for estrone sulfate obtained by specific RIA and I 10 50 100 250 300 500 700 after solvolysis and assay of estrone by RIA pg ESTRONE SULFATE (log) FIG. 1. A typical standard curve for the specific RIA of (9) is shown in Fig. 2. The correlation coefficient (0.95) was significantly different from estrone sulfate. zero (P < 0.01). The regression coefficient, TABLE 2. The intraassay and interassay variation in the [0.88 ± 0.09 (±SE)], was not significantly difRIA of plasma estrone sulfate ferent from 1 (P > 0.05), and the intercept Mean ± SD No. of (±SE), 151 ± 129 pg/ml, was not significantly CV (pg/ml) assays different from zero (P > 0.05), indicating no Intraassay systematic difference between the two meth12.20 Sample 1 344 ± 42 Sample 2 810 ± 79 9.75 ods. 90-
Sample 3 Interassay Sample 1 Sample 2 Sample 3
3 4 3
2473 ± 318
12.81
628 ± 118 1072 ± 92 1390 ± 220
18.79 8.58 15.83
Peripheral plasma values The level of estrone sulfate was measured in peripheral plasma samples from 13 men, 15
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 November 2015. at 16:12 For personal use only. No other uses without permission. . All rights reserved.
WRIGHT ET AL.
1096
JCE&M • 1978 Vol 47 • No 5
women in the follicular phase, and 15 women in the luteal phase of the menstrual cycle. The results are summarized in Table 4. Although there was considerable individual variation, estrone sulfate was significantly (P < 0.01) elevated in the luteal phase (mean ± SE; 1806 ± 160 pg/ml) when compared to the follicular phase (mean ± SE; 972 ± 79 pg/ml). No significant difference was seen between men (mean ± SE; 922 ± 62 pg/ml) and women in the follicular phase.
women at various times during the menstrual cycle are shown in Table 5. The amount of estrone sulfate excreted by men ranged from 0.8-7.9 jiig/24 h; that excreted by women ranged from 5.1-18.7 /i,g/24 h. The amount of estrone glucosiduronate (7) excreted was in general 2 to 7 times the level of estrone sulfate excreted by both men and women at various stages of the menstrual cycle. In one man, however, the estrone glucosiduronate was 15.5 times the estrone sulfate level.
Urinary levels
Splanchnic extraction
The urinary levels of estrone sulfate measured after chromatography in men and in
The concentrations of estrone sulfate in a peripheral artery and in the hepatic vein, together with the splanchnic extraction, in six patients undergoing cardiac catheterization for diagnostic reasons are shown in Table 6. In each case, the concentration in the artery was significantly higher than that in the hepatic vein. The mean splanchnic extraction of 29.8 ± 11.1 was significantly different from zero (P < 0.01), indicating net uptake by the splanchnic area.
TABLE 4. Mean concentration (±SE) and ranges of estrone sulfate in the plasma of men and of nonpregnant women in the follicular and luteal phases of the menstrual cycle No. of sub- Mean ± SE jects (pg/ml) 972 ± 79 Follicular female 15 1806 ± 160 Luteal female 15 922 ± 62 Men 13 Source
Range (pg/ml) 537-1581 814-3358 461-1238
TABLE 5. Urinary excretion (micrograms per 24 h) of estrone sulfate (ES) and estrone glucosiduronate (EG) in normal men and women Subject Male Male Male Male Male Female Female Female Female Female
(day 16) (day 13) (day 9) (day 24) (day 22)
ES (Mg/24 h)
EG (Mg/24 h)
Ratio EG:ES
7.9 2.3 4.8 0.8 2.0 5.1
18.9 12.0 15.7 12.4 11.8 34.2 89.7 23.3 46.9 29.4
2.39 5.22 3.27 15.5 5.90 6.71 4.80 3.53 2.71 3.27
18.7 6.6
17.3 9.0
Discussion The antiserum against estrone sulfate was prepared by linking the protein to the carboxylic acid group of glucosiduronic acid rather than introducing a carboxylic acid group into the estrone molecule at another point. This procedure has been used in several laboratories to prepare antisera against steroid glucosiduronates (7,12-14). However, the antiserum described here is the first to show specificity characteristics suitable for the analysis of estrone sulfate. Based on our previous experience with the
TABLE 6. Concentrations in the peripheral artery (A) and hepatic vein (HV) and splanchnic extraction of estrone sulfate in men and women undergoing cardiac catherization for diagnostic purposes Plasma concentration (pg/ml) Patient no. 11 12 13 14 15 16 Mean ± SE
Sex Female Female Female Female Male Male
A
HV
352 433 403 564 566
262 186 352 443
534 492 + 47.7
460 460 361 ± 47.2
Splanchnic extraction (%) 25 58 13 21 31 14 29.8 ± 11.1
Splanchnic extraction = 100 [1 - (HV/A)].
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 November 2015. at 16:12 For personal use only. No other uses without permission. . All rights reserved.
RIA OF ESTRONE SULFATE
1097
development of antisera against estrogen glu- laboratory. Their luteal phase values were cosiduronates, we expected the greatest cross- 1247 pg/ml (range, 392-2800), whereas our reactivity to occur in the region where the values were 1806 pg/ml (range, 814-3358). In glucosiduronate molecule was attached to the a preliminary study of plasma levels throughestrogen. A review of the cross-reactivities in out the menstrual cycle in two subjects (5), Table 1 indicates that this is generally true. the expected cyclic variation has been shown The significant cross-reactions of the anti- with midcycle peaks of 3080 and 3200 pg/ml, serum with estrone and estrone glucosiduron- respectively. Cyclical variation had previously ate do not present a serious problem in our been shown by Brown and Smyth (6). When the values obtained by our specific specific assay for estrone sulfate in plasma. Estrone and other unconjugated estrogens are RIA procedure for estrone sulfate were comalmost completely removed (>95%) by preex- pared to the values obtained using the RIA of traction with diethyl ether. Estrone glucosi- estrone after solvolysis, good correlation was duronate, in common with other glucosiduron- observed (see Fig. 2). This suggests that the ates, is not extracted (
Vol. 47, No. 5 Printed in U.S.A.
A Specific Radioimmunoassay for Estrone Sulfate in Plasma and Urine without Hydrolysis* K. WRIGHT.f D. C. COLLINS, P. I. MUSEY, AND J. R. K. PREEDY Departments of Medicine and Biochemistry, Emory University School of Medicine, Atlanta, Georgia 30322 ABSTRACT. Estrone sulfate, quantitatively the most important estrogen in plasma, has previously been determined only after hydrolysis and chromatography. An antiserum raised against estrone glucosiduronate-bovine thyroglobulin was found to be suitable for the specific RIA of estrone sulfate both in plasma and urine. Plasma levels were measured after solvent extraction without hydrolysis or chromatography. The mean (±SE) was 972 ± 79 pg/ml (range, 537-1581) in 15 women in the follicular phase, 1806 ± 160 pg/ml (range, 814-3358) in 15 women in the luteal phase, and 922 ± 62 pg/ml (range, 461-1238) in 13 men. The urinary excretion of estrone sulfate, measured after simple chromatographic sepa-
E
STRONE sulfate has been shown to be a major metabolite of estrone and 17/?estradiol in man (1-3) and is quantitatively the most important estrogen in peripheral plasma (4-6). In addition, estrone sulfate itself is further metabolized (2, 3). Such metabolism includes hydrolysis to estrone (2, 3). A variety of techniques have been used to measure this conjugate in plasma. Brown and Smyth (4) used solvent partition to separate the estrone sulfate, followed by fluorimetric estimation. Loriaux et al. (5) isolated the steroid sulfates by chromatography and estimated the estrone released on solvolysis by RIA. Hawkins and Oakey (6) used solvolysis of the sulfates, followed by competitive binding analysis after reduction of the purified estrone to 17/?-estradiol. These procedures are all relatively laborious and time consuming. The method of Received April 11, 1978. Address all correspondence and requests for reprints to: Dr. D. C. Collins, Emory University School of Medicine, 69 Butler Street, S.E., Atlanta, Georgia 30303. * These data were presented in part at the 60th Annual Meeting of The Endocrine Society, June, 1978. This work was supported in part by Contract N01-CB-74101 and Grant R01-HL 16394-02 from the NIH. t Present address: Department of Obstetrics and Gynecology, Yale University School of Medicine, New Haven, Connecticut 06510.
ration, ranged from 0.8-7.9 /ig/24 h in men and 5.1-18.7 fig/24 h in nonpregnant women. This was generally oneseventh to one-half the simultaneous estrone glucosiduronate excretion rate. An approximate mean renal clearance of estrone sulfate calculated from the above values was 2.7 ml/min. The low clearance rate is taken to reflect extensive binding of estrone sulfate by plasma proteins. The splanchnic extraction of estrone sulfate measured in 6 patients undergoing hepatic vein catheterization for diagnostic purposes was 29.8 ± 11.1%, indicating net uptake of this compound by the splanchnic area. (J Clin Endocrinol Metab 47: 1092, 1978)
Brown and Smyth (4) is also relatively insensitive. All attempts to raise a specific antiserum for estrone sulfate have thus far been unsuccessful. We examined several antisera raised against estrone glucosiduronate1-bovine thy1
The following trivial names were used: androsterone3-glucosiduronate, 17-oxo-5a-androstan-3a-yl-/?-D-glucopyranosiduronate; dehydroisoandrosterone-3-sulfate, 17oxoandrost-5-en-3/?-yl-sulfate; 16-epiestriol, estra-1,3,5(10)-triene-3,16/?,17/8-triol; 16,17-epiestriol, estral,3,5(10)-triene-3,16/?,17a-triol; 17-epiestriol, estral,3,5(10)-triene-3,16a,17a-triol; estetrol, estra-l,3,5(10)triene-3,15a,16a,17/?-tetrol; 17/?-estradiol-3-glucosiduronate, 17/?-hydroxyestra-l,3,5(10)-trien-3-yl-/?-D-glucopyranosiduronate; 17/?-estradiol-17-glucosiduronate, 3hydroxyestra-1,3,5(10) -trien-17/?-yl-/?-D-glucopyranosiduronate; estriol-3-glucosiduronate, 16a,17/?-dihydroxyestra-l,3,5(10)-trien-3-yl-/?-D-glucopyranosiduronate; estriol- 16a-glucosiduronate, 3,17/?-dihydroxyestra-1,3,5( 10) trien- 16a-yl-/?-D-glucopyranosiduronate; estriol- 17-glucosiduronate, 3,16a-dihydroxyestra-l,3,5(10)-trien-17/?-yl/8-D-glucopyranosiduronate; estrone glucosiduronate, 17oxoestra-l,3,5(10)-trien-3-yl-/?-D-glucopyranosiduronate; etiocholanolone-3-glucosiduronate, 17-oxo-5/?-androstan3a-yl-/?-D-glucopyranosiduronate; 2-hydroxyestradiol, estra-l,3,5(10)-triene-2,3,17/?-triol; 2-hydroxyestrone, 2,3dihydroxyestra-1,3,5(10)-trien-17-one; 16a-hydroxy estrone, 3,16a-dihydroxyestra-l,3,5(10)-trien-17-one; 16-ketoestradiol, 3,17/?-dihydroxyestra-l,3,5(10)-trien-16-one; 16-ketoestrone, 3-hydroxyestra-l,3,5(10)-triene-16,17dione; mestranol, 17a-ethinylestra-l,3,5(10)-trien-17/?-ol3-methyl ether; estrone-2-methyl ether, 3-hydroxyestral,3,5(10)-trien-17-one-2-methyl ether; estrone-3-methyl
1092
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 November 2015. at 16:12 For personal use only. No other uses without permission. . All rights reserved.
RIA OF ESTRONE SULFATE roglobulin while developing a specific RIA procedure for estrone glucosiduronate (7). As we characterized these antisera, we noted one antiserum that showed a very high cross-reaction with estrone sulfate but which retained specificity in relation to other estrogens and estrogen conjugates. In this paper, we describe the use of this antiserum to develop a RIA procedure for estrone sulfate which does not require prior hydrolysis or (in the case of plasma) chromatography. This procedure was used to measure peripheral levels of estrone sulfate in normal men and nonpregnant women and to determine the percentage of extraction of estrone sulfate by the splanchnic area. Urinary levels of estrone sulfate were measured after simple chromatography on DEAE-Sephadex in 10 urine samples from men and nonpregnant women and compared to both the plasma levels of estrone sulfate and to the urinary levels of estrone glucosiduronate. Materials and Methods Nonradioactive steroids and steroid conjugates were obtained from Sigma Chemical Co. (St. Louis, MO) or Steraloids (Wilton, NH) and purified by recrystallization. Purity was confirmed by melting point determination and thin layer chromatography. [6,7-3H]Estrone sulfate (SA, 54 Ci/mmol), obtained from New England Nuclear Co. (Boston, MA), was purified by chromatography on DEAESephadex (8). Ethyl acetate and ethanol were fractionally distilled. Anesthesia grade diethyl ether was used without purification. Steroids and steroid conjugates were stored at 4 C in ethanol. Estrone sulfate was stable for at least 3 months in ammoniacal ethanol and was prepared and stored at 4 C in that solvent. Phosphate-buffered saline [PBS; 0.1 M sodium phosphate, pH 6.8, containing (wt/vol) 0.9% NaCl, 0.1% ethylmercurithiosalicylate, and 0.1% bovine serum albumin (Sigma; RIA grade)] was used as diluent for the antiserum and in the assay. The radioactive solution of [3H]estrone sulfate (400,000 dpm/ml PBS) was prepared fresh daily because of ether, 17-oxoestra-l,3,5(10)-triene-3-methyl ether; pregnanediol-3-glucosiduronate, 20a-hydroxy-5/?-pregnan-3ayl-/?-D-glucopyranosiduronate; testosterone- 17/?-glucosiduronate, 3-oxoandrost-4-en-17/?-yl-/3-D-glucopyranosiduronate.
1093
the increased rate of breakdown with prolonged storage in PBS. Radioactivity was measured in Packard liquid scintillation spectrometers (models 3320 or 3255) using Scintiverse universal scintillation cocktail (Emory formula, Fisher Scientific).
Antiserum Antisera against estrone glucosiduronate-bovine thyroglobulin were produced in six male rabbits, as previously described (7). These antisera were tested for their ability to bind [3H]estrone sulfate. One antiserum was found to have sufficient specificity to be used in the RIA of estrone sulfate. The crossreactivities with other steroids and steroid conjugates for this antiserum were determined as previously described (9) and expressed as the percent equivalents of estrone sulfate. Plasma RIA of estrone sulfate Samples of plasma (0.05, 0.1, or 0.2 ml) were added to assay tubes containing 0.1. ml PBS or 0.1 ml [3H]estrone sulfate in PBS. After the addition of 0.5 ml 3 M NaCl, the samples were mixed thoroughly and allowed to equilibrate for 30 min at 22 C. The samples were extracted once with 4 ml ether and the ether extract was discarded. The samples were then extracted twice with 4 ml ethyl acetate and the ethyl acetate extract was transferred to tubes for assay or scintillation vials to assess recovery, dried under N2, and redissolved in 0.5 ml PBS. Antiserum (0.1 ml 1:10,000 dilution in PBS) and [3H]estrone sulfate (40,000 dpm or 125 pg in 0.1 ml PBS) were added to samples and standards, mixed thoroughly, and incubated at 4 C for 60 min or 18 h. Bound and free steroids were separated by centrifugation (1500 X g) after incubation with 1 ml dextran-coated charcoal (0.5% Norit and 0.1% dextran in PBS) for 15 min at 0 C. The supernatant was decanted into scintillation vials and counted. The accuracy of the method was determined by assay of a male plasma sample to which known amounts of estrone sulfate had been added. Intraassay and interassay variations were assessed at three levels of estrone sulfate within an assay and in a series of assays, respectively. Solvolysis Samples were hydrolyzed and the released estrone was measured using a method similar to that described by Hawkins and Oakey (5), as follows. One milliliter of plasma was mixed with 4 ml H2O together with [3H]estrone sulfate and allowed to
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 November 2015. at 16:12 For personal use only. No other uses without permission. . All rights reserved.
1094
WRIGHT ET AL.
equilibrate 30 min at room temperature. Ten microliters of 3 N NaOH were added and the samples were extracted twice with 20 ml diethyl ether. Ammonium sulfate (1.1 g), 25 /xl 66% H2SO4, and 5 ml ethyl acetate were added to the aqueous residue, mixed thoroughly, and incubated for 18 h at 37 C. Hydrolyzed estrone was extracted twice with 10 ml ether. The ether extracts were combined and washed sequentially with 2 ml H2O and 2 ml 8% NaHCO3. The washed extracts were dried, redissolved in 1 ml H2O, and extracted twice with 5 ml diethyl ether. The ether extract was dried under N2 and redissolved in 1 ml ethanol. Aliquots of 50 and 100 /xl were assayed for estrone using a specific RIA (9). Aliquots of 500 fi\ were counted to assess recovery. The values obtained for estrone were corrected for the molecular weight difference between estrone and estrone sulfate by multiplying by 1.38. Urinary RIA of estrone sulfate Collections of 24-h urine samples were carried out by a group of normal men and women and stored at —20 C until analyzed. A 15-ml aliquot to which [3H]estrone sulfate had been added was extracted with 30 ml ethanol by mixing vigorously and standing overnight at —20 C. The sample was centrifuged and the supernatant was removed and taken to dryness in a flash evaporator. The residue was applied to a 1 x 60-cm DEAE-Sephadex column and eluted sequentially with 500 ml deionized distilled H2O, 1000 ml 0.1 M NaCl, and 500 ml 0.2 M NaCl (8). Aliquots of each fraction from the 0.2 M NaCl elution step were counted to locate the estrone sulfate peak. The tubes comprising the [3H]estrone sulfate peak were pooled and an aliquot of the pool was counted to assess recovery. The purified estrone sulfate fraction was quantitated by direct RIA of 0.5-ml aliquots after dilutions in PBS of 1:5; 1:10, and 1:20. The amount of estrone sulfate present in each urine sample was compared to that of estrone glucosiduronate, measured as previously described (7). Splanchnic extraction Plasma samples were collected simultaneously from the hepatic vein and from a peripheral artery in a group of patients undergoing cardiac catheterization for diagnostic purposes. The samples were stored at -20 C until analyzed. The level of estrone sulfate present in the simultaneously collected samples was determined as described above. The percentage of extraction of estrone sulfate by the splanchnic area was calculated from the standard
JCE & M • 1978 Vol 47 • No 5
formula as follows (10, 11): % extraction = 100 [1 — (HV/A)], where A and HV are the endogenous levels of estrone sulfate in a peripheral artery and the hepatic vein, respectively.
Results Antiserum The cross-reactions of the antiserum used for the RIA of estrone sulfate are shown in Table 1. The antiserum shows relatively little cross-reactivity for the unconjugated estrogens, except for estrone and estrone-3-methyl ether which showed cross-reactions of greater TABLE 1. Percentage of cross-reaction of various steroids with an antiserum against estrone glucosiduronate-bovine thyroglobulin used for the RIA of estrone sulfate Compound Estrone 17/?-Estradiol Estriol 17a-Estradiol 2-Hydroxyestrone 2-Hydroxyestradiol 2-Hydroxyestriol Estrone-2-methyl ether Estrone-3-methyl ether 16-Ketoestrone 16-Ketoestradiol 16a-Hydroxyestrone 16-Epiestriol 17-Epiestriol 16,17-Epiestriol Estetrol
% Cross-reaction 108.3 2.8 0.0 3.1 3.8 3.5 0.3 14.2 >200.0 0.6 4.5 1.4 3.2 0.4 0.4 0.2
Progesterone Testosterone Cortisol
0.0 0.4 0.0
Ethinylestradiol Mestranol Diethylstilbestrol
1.4 0.2 0.0
Estrone glucosiduronate Estrone sulfate 17/?-Estradiol-17-glucosiduronate Estriol- 16a-glucosiduronate 17/?-Estradiol-3-glucosiduronate Estriol-3-glucosiduronate Estriol-3-sulfate Estriol-17-glucosiduronate Estriol-16-sulfate Androsterone-3-glucosiduronate Etiocholanolone-3-glucosiduronate Testosterone-17/?-glucosiduronate Pregnanediol-3-glucosiduronate Dehydroisoandrosterone-3-sulfate
>200.0 100.0 0.0 0.0 9.7 1.0 0.4 0.0 5.7 0.6 0.7 0.6 0.0 0.8
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 November 2015. at 16:12 For personal use only. No other uses without permission. . All rights reserved.
RIA OF ESTRONE SULFATE
1095
than 100%. None of the nonestrogenic steroids TABLE 3. The mean values (±SE) measured in a sample of male plasma after addition of known amounts of estested showed any significant cross-reaction. trone sulfate The antiserum showed low cross-reactivity pg Added pg Recovered (mean ± SE) with the conjugated estrogens, except for es29.0 31.8 ± 5.8 (6) trone glucosiduronate where the cross-reac67.0 66.4 ± 4.9 (15) tion was greater than 200%. 140.0 133.5 ± 8.2 (18) Number of samples analyzed is shown in parentheses.
RIA of estrone sulfate A logit-log transform of a typical standard curve is shown in Fig. 1. The standard curve was useful over the range 15-350 pg, with 50% displacement at 125 pg. The sensitivity was found to be 15 pg in a series of six assays. The intraassay and interassay variations are shown in Table 2. The intraassay variation was calculated from samples at three different points on the standard curve. The coefficients of variation (CV = SD X 100 -*- mean) were less than 15% at each point. The interassay variation was calculated from multiple analyses of three different plasma samples. The CVs were found to be less than 20% for all samples.
~ 2400r Id 2000 1600 I2OO 800
= 0 88x + 151 r=O95
400
1600 2400 800 SOLVOLYSIS AND RIA OF ESTRONE (pg/ml)
FIG. 2. Comparison of plasma values of estrone sulfate obtained by specific RIA with values obtained by a different method requiring solvolysis and assay of estrone. The line shown is the calculated regression line.
Recovery experiments were carried out on a sample of male plasma to which three difZ. 60ferent amounts of estrone sulfate had been o 50added. The plasma was analyzed and the reZ 40O 30sults are shown in Table 3. In each case, the * ~ amount recovered was not significantly differ10ent from the amount added. l The correlation in 12 samples of values for estrone sulfate obtained by specific RIA and I 10 50 100 250 300 500 700 after solvolysis and assay of estrone by RIA pg ESTRONE SULFATE (log) FIG. 1. A typical standard curve for the specific RIA of (9) is shown in Fig. 2. The correlation coefficient (0.95) was significantly different from estrone sulfate. zero (P < 0.01). The regression coefficient, TABLE 2. The intraassay and interassay variation in the [0.88 ± 0.09 (±SE)], was not significantly difRIA of plasma estrone sulfate ferent from 1 (P > 0.05), and the intercept Mean ± SD No. of (±SE), 151 ± 129 pg/ml, was not significantly CV (pg/ml) assays different from zero (P > 0.05), indicating no Intraassay systematic difference between the two meth12.20 Sample 1 344 ± 42 Sample 2 810 ± 79 9.75 ods. 90-
Sample 3 Interassay Sample 1 Sample 2 Sample 3
3 4 3
2473 ± 318
12.81
628 ± 118 1072 ± 92 1390 ± 220
18.79 8.58 15.83
Peripheral plasma values The level of estrone sulfate was measured in peripheral plasma samples from 13 men, 15
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 November 2015. at 16:12 For personal use only. No other uses without permission. . All rights reserved.
WRIGHT ET AL.
1096
JCE&M • 1978 Vol 47 • No 5
women in the follicular phase, and 15 women in the luteal phase of the menstrual cycle. The results are summarized in Table 4. Although there was considerable individual variation, estrone sulfate was significantly (P < 0.01) elevated in the luteal phase (mean ± SE; 1806 ± 160 pg/ml) when compared to the follicular phase (mean ± SE; 972 ± 79 pg/ml). No significant difference was seen between men (mean ± SE; 922 ± 62 pg/ml) and women in the follicular phase.
women at various times during the menstrual cycle are shown in Table 5. The amount of estrone sulfate excreted by men ranged from 0.8-7.9 jiig/24 h; that excreted by women ranged from 5.1-18.7 /i,g/24 h. The amount of estrone glucosiduronate (7) excreted was in general 2 to 7 times the level of estrone sulfate excreted by both men and women at various stages of the menstrual cycle. In one man, however, the estrone glucosiduronate was 15.5 times the estrone sulfate level.
Urinary levels
Splanchnic extraction
The urinary levels of estrone sulfate measured after chromatography in men and in
The concentrations of estrone sulfate in a peripheral artery and in the hepatic vein, together with the splanchnic extraction, in six patients undergoing cardiac catheterization for diagnostic reasons are shown in Table 6. In each case, the concentration in the artery was significantly higher than that in the hepatic vein. The mean splanchnic extraction of 29.8 ± 11.1 was significantly different from zero (P < 0.01), indicating net uptake by the splanchnic area.
TABLE 4. Mean concentration (±SE) and ranges of estrone sulfate in the plasma of men and of nonpregnant women in the follicular and luteal phases of the menstrual cycle No. of sub- Mean ± SE jects (pg/ml) 972 ± 79 Follicular female 15 1806 ± 160 Luteal female 15 922 ± 62 Men 13 Source
Range (pg/ml) 537-1581 814-3358 461-1238
TABLE 5. Urinary excretion (micrograms per 24 h) of estrone sulfate (ES) and estrone glucosiduronate (EG) in normal men and women Subject Male Male Male Male Male Female Female Female Female Female
(day 16) (day 13) (day 9) (day 24) (day 22)
ES (Mg/24 h)
EG (Mg/24 h)
Ratio EG:ES
7.9 2.3 4.8 0.8 2.0 5.1
18.9 12.0 15.7 12.4 11.8 34.2 89.7 23.3 46.9 29.4
2.39 5.22 3.27 15.5 5.90 6.71 4.80 3.53 2.71 3.27
18.7 6.6
17.3 9.0
Discussion The antiserum against estrone sulfate was prepared by linking the protein to the carboxylic acid group of glucosiduronic acid rather than introducing a carboxylic acid group into the estrone molecule at another point. This procedure has been used in several laboratories to prepare antisera against steroid glucosiduronates (7,12-14). However, the antiserum described here is the first to show specificity characteristics suitable for the analysis of estrone sulfate. Based on our previous experience with the
TABLE 6. Concentrations in the peripheral artery (A) and hepatic vein (HV) and splanchnic extraction of estrone sulfate in men and women undergoing cardiac catherization for diagnostic purposes Plasma concentration (pg/ml) Patient no. 11 12 13 14 15 16 Mean ± SE
Sex Female Female Female Female Male Male
A
HV
352 433 403 564 566
262 186 352 443
534 492 + 47.7
460 460 361 ± 47.2
Splanchnic extraction (%) 25 58 13 21 31 14 29.8 ± 11.1
Splanchnic extraction = 100 [1 - (HV/A)].
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 21 November 2015. at 16:12 For personal use only. No other uses without permission. . All rights reserved.
RIA OF ESTRONE SULFATE
1097
development of antisera against estrogen glu- laboratory. Their luteal phase values were cosiduronates, we expected the greatest cross- 1247 pg/ml (range, 392-2800), whereas our reactivity to occur in the region where the values were 1806 pg/ml (range, 814-3358). In glucosiduronate molecule was attached to the a preliminary study of plasma levels throughestrogen. A review of the cross-reactivities in out the menstrual cycle in two subjects (5), Table 1 indicates that this is generally true. the expected cyclic variation has been shown The significant cross-reactions of the anti- with midcycle peaks of 3080 and 3200 pg/ml, serum with estrone and estrone glucosiduron- respectively. Cyclical variation had previously ate do not present a serious problem in our been shown by Brown and Smyth (6). When the values obtained by our specific specific assay for estrone sulfate in plasma. Estrone and other unconjugated estrogens are RIA procedure for estrone sulfate were comalmost completely removed (>95%) by preex- pared to the values obtained using the RIA of traction with diethyl ether. Estrone glucosi- estrone after solvolysis, good correlation was duronate, in common with other glucosiduron- observed (see Fig. 2). This suggests that the ates, is not extracted (
