A radioimmunoassay of plasma unconjugated and conjugated estetrol

521

2132

A RADIOIMMUNOASSAY

OF PLASMA UNCONJUGATED AND CONJUGATED ESTETROL

Konbai Den, Hiroshi Matsumoto, Kiyoshi Tom Fujii, Kiyohide Furuya, Takao Yoshida and Shigeo Takagi Department of Obstetrics and Gynecology, Nihon University, School of Medicine, Akira Kanbegawa and Tomoko Kokubu Department of Chemistry, Research Laboratories, Teikoku Hormone Mfg. Co., Ltd., Tokyo, Japan ABSTRACT -

Rec'd.: 5-16-77

A radioimmunoassay for the measurement of both unconjugated and conjugated estetrol in plasma has been developed. The antiserum obtained after 6 months of immunization with 6-oxoestetrol-6-(O-carboxymethyl)oxime-BSA was used at a final dilution of 1:90,000 and showed almost no cross reaction with other steroids except for estriol at 1.24%. Estetrol-glucosiduronate was synthesized by incubating with adrenalectomized rat liver homogenate and uridine diphosphoglucuronic acid. Then, plasma estetrol-glucosiduronate was measured in the same manner for unconjugated estetrol after hydrolysis with 8-glucuronidase. Sephadex LH-20 column chromatography (7x110 mm, benzene:methanol, 85: 15) was employed for accurate assessment. The sensitivity was 10 pg and the smallest amount measurable was 40 pglsample. The method blank was consistently negligible. The intra and inter assay precision was 11.8% and 14.2% for unconjugated estetrol and that for estetrol-glucosiduronate was 13.5% and 17.1%. - INTRODUCTION Estetrol

[1,3,5(10)-estratriene-3,15o.,16o,178-tetro1, E&l is a

highly polar CIB-metabolite mono-glucuronoside

(1,2,3), excreted preponderantly

as the

of ring D (4,5) and suggested that the level of E4

in pregnancy may serve as a possible superior index of fetal viability (6,7,8). A number of publications have reported on the radioimmunoassay (RIA) for measuring plasma or serum unconjugated El, (Eb-U) using the antiserum obtained by estriol (E3)-16,17-dihemisuccinate-BSA Eb-3-0-carboxymethyl

ether derivative

(9,10,11).

(7,8) or

Recently, Kundu and

Grant (12) reported a simple RIA without chromatography using a more specific antiserum obtained by 6-oxo-Ek-6-(0-carboxymethyl)oxime-BSA.

Vo'owne 30, Nwnber 4

S

‘P=ROXD=

October,

2977

S

522

TDROIDS

However, the values reported by these various authors are not in agreement and we know of no reports concerning conjugated EI+. We have developed a RIA for plasma Et+-U and Es-glucosiduronate (EQ-G) using an antiserum from rabbits immunized with 6-0x0-Es-6-(0carboxymethyl)oxime-BSA

(Eb-6CMO-BSA) and we wish to report our

findings. - MATERIALS AND METHODS EI, was obtained from Steraloids Inc. E3 and other steroids were provided by Teikoku Hormone Pharmaceutical Co. E3-6,7-3H (53.1 Ci/m (30 mole,.New England Nuclear, Ea-3H), Ea-6,7-3H-16a-8-D-glucuronide Ci/m mole, New England Nuclear, Ea-3H-16G) and its acetate-methyl ester (Teikoku Hormone Pharmaceutical Co.) was obtained. E3-6,9-3H (40.0 Ci/m mole, The Radiochemical Center Ltd., EQ-~H) was purified every second month. E4-3H-glucosiduronate (E4-3H-G) was made by a modification of the method of Nambara -et al. (13). Adrenalectomized rat liver was homogenized with 10 ml of 0.05M phosphate buffer (pH 7.4) per gm liver and 200 pg centrifuged at 800xG for 60 min. One ml of the supematant, of uridine diphosphoglucuronic acid and 200 nCi of EI,-~H were incubated at 37°C for 60 min in a Shaker-Incubator with air as the gas phase. Then, ethanol was added to deproteinize the mixture, which was centrifuged and the supematant was dried in vacua. The residue was transfered to a sephadex LH-20 column chromatography (7x55 mm), eluted with 8 ml of benzene:mathanol (85:15) and 1 ml benzene;methanol (50: 50) to isolated the free fraction. Further elution with 8 ml of benzene:methanol (50:50) isolated the conjugated fraction and provided a 9% yield (Fig.1). Es-G

I1

0

0

2

6

8

3

10

12

e-(50 : 50) (85 :15) benzene : methanol Fig.1 Separation of Et+-% and its conjugate by sephadex LH-20 column chromatography. +

1

-I~

4

(ml)

0.2

0%

0.6

-d-

~~

,

(Rf)

Fig.2 Thin layer chromatography using solvent system, CHCl,:iso-PrOH:HCOOH (15:5:4). 1) E3-3H 2) E3-3H-16G 3) E4-3H 4) E4-3H-conjugate

Identification of this conjugate as E4-3H-G was carried out by First, this conjugate and standard grade the following procedures. Ex-~H-~~G were developed on a Silica gel G thin layer chromatography

S

TmROIDS

523

(E. Merck A.G., Darmstadt, TLC), using a solvent system of chloroform: isopropano1:formi.c acid (15:5:4), and the Rf values obtained were 0.12 and 0.14 respectively (Fig.2). Then, we transformed this conjugate into an acetate-methyl ester with diazomethane and acetic anhydridepyridine. The product was developed on TLC using methano1:chlorofor-m (1:9), yielding a single spot with an Rf of 0.52, whereas the Rf obtained from the standard grade Es-~H-~~G acetate-methyl ester was 0.60. Next, we submitted an aliquot of the conjugate in O.lM acetate buffer (pH 4.7) to the hydrolytic activity of beef-liver 8-glucuronidase (6,000 Fishman unit). The mixture was incubated at 5O'C for 3 hrs. An auxiliary reaction was performed on the above mixture in the presence of D-glucaro-1,4-lactone, a hydrolytic inhibitor of B-glucuronidase. The rates of hydrolysis was ca.83% and with D-glucaro-1,4lactone was 27%. Preparation of antigens Ek-6CM0 (III) was synthesized by a modification of the method of Longwell and Wintersteiner (14). Es-tetraacetate (150 mg, Ib), obtained from E4 by acetic anhydride and pyridine, was oxidized with chromium trioxide in glacial acetic acid. The ketone substance obtained (II,) was allowed to stand for 4 hrs with 10% methanolic sodium hydroxide at It was dissolved room temperature in a NP atmosphere for hydrolysis. in water, neutralized with CO2 and extracted with ethyl acetate. The crude residue obtained (17 mg) was chromatographed on Silica gel (2gm) in hexane:chloroform (1:20). A fraction corresponding to 6-0x0-E4 (12 mg) with a m.p. of 163'C-167°C was obtained (IIb). A mixture of 6-

0

I, : R=H Ib : R = AC Fig.3

II, : R = AC IIb : R=H

N-0-CH~COOH III

N-0-CHz-C=O BSA-NH IV

The sequence for synthesis of the 6-oxo-derivatives and the position for coupling to BSA that was used in the synthesis of the antigen.

oxo-El, (10 mg) in methanol (0.8 ml), carboxymethoxylamine hemihydrochloride (12 mg) and 2N-NaOH (0.06 ml) was let stand overnight at room After evaporating the methanol in vacua, the residue was temperature. dissolved in water, buffered to pH 8.5 with sodium bicarbonate and then extracted with ethyl acetate. The aqueous solution was acidified with O.lN-HCl to pH 1.0-2.0, the precipitate was then extracted with ethyl acetate. Following evaporation, the residue was purified on Silica gel TLC (methanol:chloroform, 1:5, Rf=0.42) and a product (12 mg) with a m.p. of 215"C-22O'C was obtained (III). Coupling of EI+-6CM0 to BSA was done as follows. To a mixture of Eb-6CM0 in dioxane and tri-n-butylamine (0.01 ml), isobutyl-chloroformate (0.01 ml) was added and stirred in an ice water bath for 20 min. After which, the mixture was added gradually to a solution of BSA (30 mg) in water (0.8 ml), followed by dioxane (0.6 ml) and lN-NaOH

S

TIIEOXDm

(0.03 ml) and stirred continuously for 4 hrs, then dialized against water for 24 hrs and titrated to pH 4.5 with lN-HCl. After centrifugation of the precipitate, the residue was dissolved in water, adjusted to pH 7.5 with sodium bicarbonate and lyphophilyzed yielding 40 mg of Eb-6CMO-BSA (IV). The product consisted of on the average 22 molecules of EI, coupled to each BSA molecule. Preparation of antisera The antiserum to Eb was produced in 4 month old male New Zealand white rabbits. One mg of Ek-6CMO-BSA was emulsified with 0.5 ml of physiological saline and 0.5 ml complete Freund's adjuvant. A portion of this solution was injected into the front and hind foot pads, and intradermally into the back divided into more than twenty sites. Further booster injections were given twice a month for a further two months and then subsequently once a month. The rabbits were bled at 10 days intervals after each injection from marginal veins and the serum collected was stored at -80°C. - ASSAY PROCEDURES Extraction and Purification 1,000 dpm of EQ-~H was added to the test plasma sample (usually 0.25-0.50 ml of plasma) which was then extracted with 5 vol of ethyl ether using a vortex mixer. The extracts were transferred to a conical tube and dried under a N2 stream at 45°C and then submitted to sephadex LH-20 column chromatography (7x11O mm, benzene:methanol, 85:15). One half of the eluate was used for recovery and the other half for RIA. Eb-G was measured as total Et, after hydrolysis with b'-glucuronidase. The condition between the concentration of enzyme, incubation time and temperature are summarized in Tab.1. We used 6,000 Fishman units of 8-glucuronidase at 50°C for 3 hrs. After ether extraction, the same procedure as for Ek-U was performed. Tab.1

Percent hydrolysis of Ek-3H-G with 8-glucuronidase at 50°C.

enzyme concentration (Fishman unit) 1 1,500 3,000 6,000 12,000

84.1 86.5 89.4 93.9

incubation time (hrs) 3 6 12 24 81.4 82.0 83.8 84.3

86.9 87.1 91.6 92.5

88.8 90.8 92.3 97.7

88.1 91.2 92.5 97.8

48 89.1 91.1 92.0 97.5

Standard curve In order to obtain a standard curve, 5,500 dpm of EI+-~H and Et+ over the range of O-500 pg were pipetted into assay tubes and evaporated under a NZ stream. A 0.25 ml aliquot of antiserum diluted to 90,000 was added to each tube. After incubation at 24°C for 45 min, 0.25 ml of 0.025% cold dextran coated charcoal suspension was added. After frequent shaking, the contents were centrifuged at 2,500 rpm for 5 min. The supematant was decanted into counting vials containing 10 ml of scintillation fluid (250 mg of dimethyl POPOP, 10 gm of DPO and 100 mg of naphthalene in 1 liter of dioxane) and the radioactivity was counted for 10 min in Aloka LSC(The efficiency was 45% and back ground

S

TmEOID=

525

was 20 cpm. The external standard method was employed for quenching correction). The affinity constant estimated from Scatchard plots was 2.88 ~10' liter/mole (Fig.4).

3.0

Y

K = 2.88 xlOgl/mole (3~-E4, 13.9 pg)

A

,

1.0

0.5

I 1.5

Fig.4

10

20

50

100

200

500

E4 added (pg)

mole

Calibrated standard curve Eor EI, using an antiserum to Et,6CMO-BSA at a dilution of 1:90,000 (v/v). Each point represents the mean (tstandard deviation) of bound percentage in quadruplicate. - RESULTS -

Time course of immunization The antibody titer determinations were based on a standard dilution of 0.25 ml of antiserum corresponding to 70% binding of the EI+3H at 5,500 dpm. The antibody titer increased after the 2nd month and it was 5,000 at 3 months, 15,000 at 5 months, 90,000 at 6 months, continued to increase to 8 months and then tapered off. It was noted that the antibody titer and its specificity declined if the blood was collected beyond 10 days after each booster injection. Sensitivity At the 95% confidence limits, 10 pg significantly differed from zero. The coefficients of variation at each point ranged from 10 pg to 500 pg on the standard curve assayed in quadruplicate was always less than 5%. Hence, a values of 40 pg/sample may be taken as the limit of sensitivity of the assay if it is assumed that a 50% recovery was obtained. Specificity The cross reaction was determined at 50% displacement of Eb-3H with various steroids according to the method of Abraham (15). The main cross reactant of this antiserum was E3 at 1.24%. The other steroids tested showed little or no detectable cross reaction. Interference and Separation The cross reaction of E3 with this antiserum lead to possible overestimations of valid values for Et+ because of the relative

abundance of E3 in pregnancy plasma (16). In order to investigate the interference of Es, a study was carried out by measuring the recovery of known amounts of EQ added to 0.25 ml of male plasma each containing a 10 fold amount of E3. The recovery following hexane wash and ether extraction was found to have poor reproducibility and plasma blank values were usually over 20 pg, even though water blanks were always negligible, indicating a high degree of interference by E3 and plasma. Sephadex LH-20 column chromatography (7x110 mm, benzene:methanol, 85:15) was used as the procedure for removal of plasma blanks and isolating the Es. In the serial elution of E,+, 15 ml-19 ml of the eluted fraction were chosen as the optiomal range. Tab.2

The percent cross reaction of selected compound with antiserum produced with Es-6CMO-BSA. compounds

3-Hydroxy-1,3,5(10)-estratrien-17-one 3,15a-Dihydroxy-1,3,5(10)-estratrien-17-one 2-Methoxy-3-hydroxy-l,3,5(10)-estratrien-l7-one 1,3,5(10)-Estratriene-3,17&diol 1,3,5(10)-Estratriene-3,17a-diol 2-Methoxy-1,3,5(10)-estratriene-3,17S-diol 1,3,5(10)-Estratriene-3,15a,l7@-trio1 1,3,5(10)-Estratriene-3,16a,l7@-trio1 1,3,5(10)-Estratriene-3,16fi,l7R-trio1 17b-Hydroxy-4-androsten-3-one 178-Hydroxy-5a-androstan-3-one 4-Androstene-3,17-dione 38-Hydroxy-5a-androstan-17-one 5u-Androstene-3P,17@diol 3@-Hydroxy-5-androsten-17-one 3@,16a-Hydroxy-5-androsten-l7-one

% 0.04 0.05 CO.005 co.005 co.005 0.03 0.1 1.24 0.01 0.01 co.005 0.03 co.005 co.005 co.005 co.005

Recovery and Accuracy The mean recovery after the over all procedures was 77.4% for EI+-U and 63.0% for E,+-G. The blank arising from plasma and from assay procedure were found to be the least detectable values. Recovery experiments were performed by measuring the levels of both E J,in quadruplicate adding known amount of E+ to 0.25 ml of male plasma. The recovery correction coefficient was 0.951, the regression line was y=1.109f0.72 for E4-U and 0.982, y=1.320*4.84 for E+-G. Precision The intra and inter assay variance was tested by measurement of the 25 to 40 week pregnancy plasmas in duplicate simultaneously, then separately according to the procedure proposed by Snedecor (17). The intra and intra assay variance was 11.8%, 14.2% for EI+-U and 13.5%, 17.1% for E4-G. Maternal peripheral plasma levels of E4 in normal pregnancy The mean EI, levels in maternal peripheral veins showed a rapid increase during the 3rd trimester and at term, E4-U was 0.67iO.33 ng/ml and Ek-G was 4.57k2.84 ng/ml which is about 4 to 5 times higher than at 25 to 28 weeks of pregnancy of 0.13?0.05 and 1.02kO.68 ng/ml re-

S

527

TDICOID=

spectively. The diurnal variations of plasma E4 levels in the last trimester were both statistically insignificant. Tab.3

Maternal peripheral plasma levels of Et, during the 3rd trimester in normal pregnancy

weeks of gestation

number of sample

25- 28

21

29- 32

23

33- 36

24

37-40

21

type

mean + S.D (ng/ml)

E4-IJ EI,-G Es-U E4-G E4-U Es-G E4-U Et,-G

0.13+0.05 1.02 +0.68 0.19 kO.12 1.02kO.44 0.34iO.20 1.8Ok1.31 0.6JkO.33 4.57+ 2.84

range 0.07

- 0.26 0.25-2.88 0.07-0.58 0.48-1.91 0.12-0.93 0.60-4.16 0.42-1.84 1.84-9.09

- DISCUSSION The specificity and titer of an antibody are important requirements for RIA, and are dependent on the site of conjugation. shown that conjugation at C-6 is superior to C-3 and C-17

It has been for the pro-

duction of antisera to estrogens as it leaves the functional group intact.

For example, coupling BSA to EJ at the C-6 position has been

successfully

utilized to prepare highly specific antisera (16).

In

this experiment, we inserted a ketone at C-6 of EL,with chromic acid oxidation and conjugated the carbonyl group to BSA via its O-(carboxymethyl)oxime

and obtained the antiserum by immunizing rabbits.

The reactivity titer of this antiserum showed a significance elevation and the sensitivity of the standard curve was very high (10 pg), however, this antiserum showed cross reactivity with EJ at 1.24% which is higher than that reported by Kundu and Grant.

This cross reactivity

was thought to make a direct assay without prior purification

and sepa-

ration impossible, since E3 is present in high concentrations

(8-16

ng/ml> during the 3rd trimester of pregnancy

(16) and the 0.25-0.50ml

of control plasma which was routinely used for assay showed relatively higher blank values.

Therefore, we attempted separation of E3 and re-

S

528

TRROXDb

moved the plasma blank by sephadex LH-20 column chromatography.

This

process is relatively simple, not too time consuming, requires only 20 ml of solvent, and eliminated EJ and plasma blanks completely.

The

accuracy and precision suggested the reliability of our method, and the levels of EL,-U are in near agreement with or are slightly lower than those of Kundu and Grant's report. However, there are several points which must be taken into consideration when attempting an accurate measurement of plasma EI,. One is the stability and the specific activity of 3H-labelled Es used in the RIA.

Eb-6,9-3H dissolved in ethanol remains stable for at least 3

months and is not affected by assay procedures including g-glucuronidase hydrolysis.

Another is the storage of sample plasma :EI+-U values

are overestimated when sample plasmas are stored for long periods,i.e, longer than 3 months, or if a given sample is processed It was hypothesized glucosiduronate (4,5).

repeatedly.

that circulating E4 is predominantly

a mono-

of ring D and is higher in concentration than is E4-U

Thus, we measured plasma Es-G as total EL, after hydrolysis with

!3-glucuronidase.

EI+-~H-G was made by incubating EI+-~H with an adrena-

lectomized rat liver homogenate in the presence of uridine diphosphoglucuronic acid.

And tentative identification of the product was made

by observing the chromatographic behavior of both EI+-G and its acetatemethyl ester.

Further confirmation was acquired by observing its

hydrolytic cleavage with b-glucuronidase.

The results were compared

with those obtained from identical procedures performed on a standard grade Es-~H-~~G and its acetate-methyl ester.

The above results lead

to the suggestion that it is reasonable that our product is an Es-3H glucosiduronate.

However,

the site of conjugation could not be pre-

S

TD&OXDlB

529

cisely identified. Results show that Eh-G levels increase more rapidly than Es-U during the 3rd trimester in normal pregnancy and is about 7 fold higher than E+-U levels at term.

These results suggest that E+-G and Eb-U

may serve as monitors of fetal well-being, however, as Es-G requires lesser amounts of plasma for assay, clinically, Elt-G may perhaps be more acceptable.

- ACKNOWLEDGEMENT

-

This work was supported by a research grant from the Ministry of Health and Welfare of Japan for Research on Handicaped Children.

- REFERENCE 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17.

Hagen, A.A., Barr, M. and Diczfalusy, E., ACTA ENDOCR. 49, 207, (1967) Zucconi, G., Lisbon, B.P., Simonitsch, E., Roth, L., Hagen, A.A. and Diczfalusy, E., ACTA ENDOCR. 5&, 413, (1967) Hagen, A.A., J.CLIN.ENDOCRINOL.METAB. 30, 763, (1970). Nagatomi, K., Osawa, Y., Kirdani, R.Y. and Sandberg, A.A., J.CLIN. ENDOCRINOL.METAB. 37, 887, (1973). Jirku, H., Kadner, S. and Levitz, M., J.CLIN.ENDOCRINOL.METAB. 35, 785, (1972). Heikkila, J. and Adlerkreutz, H., J.STEROID BIOCHEM. 1, 243, (1970). Giebenhain, M.E., Tagatz, G.E. and Gurpide, E., J.STEROID BIOCHEM. 3, 707, (1972). Sciarra, J.J., Tagatz, G.E., Notation, A.D. and Depp, R., AMER.J. OBSTET.GYNECOL. 118, 626, (1974). Tulchinsky, D., Frigoletto, F.D.Jr., Ryan, K.J. and Fishman, J., J.CLIN.ENDOCRINOL.METAB. 3, 560, (1975). Fishman, J. and Guzik, H., J.CLIN.ENDOCRINOL.METAB. 2, 892, (1972). Korda, A.R., Challis, J.J., Anderson, A.B.M. and Tumbull, A.C., BRITISH J.OBST.GYN. 82, 882, (1975). Kundu, N. and Grant, M., STEROIDS 27, 785, (1976). Nambara, T., Matsuki, Y. and Kurata, M., CHEM.PHARM.BULL. 0, 2607, (1972). Longwell, B. and Wintersteiner, O., J.BIOL.CHEM. 183, 219, (1940). Abraham, G.E., J.CLIN.ENDOCRINOL.METAB. 7, 866, (1969). Den, K., Fujii, K.T., Yoshida, T. and Takagi, S., ENDOCRINOL_ JAPON 20, 315, (1973). Snedecor, G.W., BIOMETRICS 8, 85, (1952).