HME—a novel derivative for direct iodination in steroid radioimmunoassays

J. steroid Biochem.

Vol. 28, No. 6, pp. 663467, Printed in Great Britain. All rights reserved

1987

0022-4731/87

53.00 + 0.00

Copyright 0 1987Pergamon JournalsLtd

HME-A NOVEL DERIVATIVE FOR DIRECT IODINATION IN STEROID RADIOIMMUNOASSAYS R.

STUPNICKI$,

T. KOWALCZYK-F’RZEEWLOKA,

E. KULA*

and L. Y.

KAS~NG~~

Institute of Sports, 01-809 Warsaw and *Institute of Animal Physiology and Nutrition, OS-110 Jablonna, Poland (Received

23

March

1987)

Summary-Steroid derivatives containing histidine methyl ester (HME), instead of histamine, were prepared by mixed anhydride coupling. The derivatives were crystalline, and when labelled in microgram quantities by using Iodo-gen (exposure time 1 h) the yield of the immunoreactive fraction was 4&50%. The products were similar in immunoreactivity and stability to the known histamine derivatives. Assay parameters obtained with HME-derivatives were compared with those obtained with tritiated steroids and with analogous TME-derivatives. A heterologous assay of progesterone (3 antisera against 12or-succinyl-BSA, and methylsuccinyl derivatives for labelling substituted at 1la-position), and a homologous assay of cortisol (4 antisera against 21-succinyl-BSA, and 21-carbonyl-derivatives for labelling) were studied. The HME-derivatives produced logit-log curves with slopes comparable to those in tritium-based assays. The sensitivity, as expresssed by ED, values, was by 66% higher than in tritium assays, and by 50% higher than for the tyramine derivative in a similar heterologous assay. The HME-based assay for cortisol was by 26% less sensitive compared to tritium, but several times more sensitive than in case of analogous TME-based assay.

INTRODUcTION

Histamine-substituted steroid derivatives are known to be most suitable for iodine labelling due to their excellent assay characteristics [l]. However, the twostage preparation of labelled material, i.e. first labelling histamine which is being subsequently reacted with milligram quantities of the appropriate steroid derivative [2, 3,4], makes the assay rather complicated. An attempt was made at using histidine methyl ester (HME) in place of histamine, in the hope that

the product would bring an improvement in the radioiodine-based steroid radioimmunoassay. NO mention was found in the available literature concerning the use of HME as iodine acceptor in any kind of steroid or non-steroid derivative. The properties of HME-derivatives of two steroids-progesterone and cortisol-were studied. The progesterone assay was based on positionand/or bridge-heterologous, or on homologous reactions, employing positions 11~ and 12a, and three bridge structures: succinyl, methylsuccinyl and carbonyl. The cortisol assay was bridge-heterologous at carbon 21. EXPERIMENTAL

Antisera

Antisera (Rl, R2,

against R314),

progesterone-l la:-succinyl-BSA progesterone-12a-succinyl-BSA

TIAEA-fellow. Present address: Nuclear Energy Regional Center, Kinshasa, Zaire. $Reprint requests to be sent to: Romuald Stupnicki, Department of Endocrinology and Doping Control, Institute of Sports, 01-809 Warsaw, Poland.

(R41, R78, R79, R80), and against cortisol-21succinyl-BSA (R63, R74, R76) were obtained in rabbits as described earlier [5, 61. Derivatives for labelling

The derivatives used throughout the study are shown in Fig. 1 and listed in the list of trivial names. They were prepared by coupling the respective succinic or methylsuccinic-hydroxysteroid monoesters with tyramine (TA) or tyrosine methyl ester (TME) by the conventional mixed anhydride method [7] or with histidine methyl ester (HME) by a modified method of Fridkin and Goren [8] as follows: steroid hemiester, HME-hydrochloride, and p-(N-dimethylamino)-pyridine, 0.1 mmol each, were dissolved in 3 ml of dry dimethyl-formamide (DMF). To this was added N-ethylmorpholine (26~1), the mixture was cooled down to O”C, and 22mg of dicyclohexylcarbodiimide in 1 ml of DMF was added. The mixture was stirred for 1 h at 0°C and then for 24 h at room temperature. Cortisol-21-chloroformate, prepared by phosgenation of cortisol, was dissolved in dry dioxane and added dropwise to a solution of IOOmg HMEhydrochloride in 6ml of 50% dioxane adjusted to pH = 8.0 by NaOH (1 mol/l) and left for 2 h at room temperature with constant stirring. The reactions were monitored by silica gel TLC in chloroform-ethanol, 9: 1. DMF or dioxane was partially evaporated in a vacuum rotary evaporator, the remaining liquid was transferred to 50 ml of water and extracted 3 times with 25 ml of chloroform. The combined organic phases were extracted with water once, dried with anhydrous MgSO,, evaporated, transferred to a preparative silica gel column and eluted with chloroform+thanol, 19 : 1. The overall 663

R. STUPNICKIet al.

664

i, --e:‘i 5

i

0

HN

.OH

/

0

III

.OH

V Fig. I. Steroid derivatives used for lx Iodine-labelling (see list of trivial names). efficiency after recrystallization was about 70%. The identity of the products was confirmed by NMR spectra in CDCl, and by mass fragmentography (see Appendix). Both free amines and their hydrochlorides were crystalline and yellowish in colour. LabeIling

Tyramine and TME-derivatives were labelled by a conventional chloramine-T technique as described earlier [9], which rendered low yields of iodinated

products when applied to HME-derivatives. The yield was markedly improved by using tetrachlorodiphenylglycoluril (Iodo-gen, Pierce) as follows. To a small glass test-tube, coated with 1.5-2 fig of Iodogen, were added: 25 ~1 phosphate buffer (0.05 mol/l pH 7.3, 4pg of HME-derivative in dioxane and 0.1 mCi of Na’251 in a volume of 10 ~1. The mixture was incubated at room temperature for about 1 h with occasional stirring. Water (0.5 ml) and a drop of NaOH were added, the contents were transferred to

Steroid-HME derivatives for direct iodination

J.L

@

I

665

curves were prepared from serial dilutions ranging from 800 to 12.5 (progesterone) or 1600 to 12.5 pg/tube (cortisol). Appropriate dilution of the standard (100 PI), labelled steroid derivative (100 ~1) and antiserum (100 ~1) were pipetted into test tubes. The contents were thoroughly mixed, incubated for 90 min, then 0.5 ml of charcoal suspension (Norit A, O.l5-0.3%) was added, mixed, and incubated in an ice bath for 15 min. After centrifugation 0.5 ml of the supernatant was taken for radioactivity measurements in an automatic gamma counter (Riagamma, Beckman). The results were presented in the logit-log form using decimal logits [5].

_B

RESULTSAND

All the logit-log curves were linear, and therefore for a given system the essential average characteristics (slope and ED,,) were calculated and presented in Table 1, and the mean logit-log curvesin Fig. 3. Results obtained with tritiated tracers served as a basis of comparisons. All iodine-based assays but one (No. 8 in Table 1) involved position and/or bridge heterology. Position-homologous ligands containing succinyl or methylsuccinyl bridges produced, in general, flat dose-response curves, with slopes ranging from -0.7 to -0.53, and were thus unsuitable for practical purposes. Best results were obtained with position-heterologous assays. The effects of HME-substitution can be directly evaluated by comparing parameters of assays Nos 6 and 7: the tracers are identical with respect to both position of substitution and bridge structure, but the iodine acceptors are different. The phenolic substituent (tyramine) produced results only slightly better in terms of ED, values relative to the 3H-assay (1 lo%), while the imidazolic one (HME) shifted the logit-log curves to the left giving an EDso value as high as 165% of the tritium assay (cf. Fig. 4). It is interesting to note that the reverse system, i.e. antibodies against progesterone substituted at 1lu -position, tracer-at 12a-position (Table 1, assay No. 8) produced results identical with those of the tritium assay. A better

t t F u.v Fig. 2. Thin-layer chromatography of ‘251-iodinatedsteroid derivatives. Upper graph: progesterone-l la-methylsuccinyltyramine-[‘251]. Acetone-chloroform, I :4. Lower graph: progesterone-l la-methylsuceinyI-HME-[‘251]. Methanolchloroform 1:9. S: start. F: front. u.v.: spot of unreacted substrate visible under a u.v.-lamp. Arrows indicate imt

S

munoreactive fractions.

a larger test tube and extracted as in the chloramineT technique. After evaporation of the solvent the residue was purified by silica gel TLC in chloroform-methanol, 9: 1. A single peak of radioactivity was observed (Fig. 2) which after elution contained 4060% of the total radioactivity used. The iodinated compound was eluted from the plate with methanol and stored in refrigerator. Conditions of the assay

All solutions (labelled steroids, antisera, standards) were made in phosphate buffer (0.05 mol/l, pH 7.5) containing 0.2% gelatin. Antiserum dilution was selected so as to bind 30-35% of the tracer. Standard Table 1. Basic characteristics Antibody

to

Assay no.

P-lla-S-BSA

I

P-12~~S-BSA

2 3 4 5 6

F-21-S-BSA

Steroid derivative

l’ I-W'

of iodine-based n

d

[‘W

40 I 5 5 40 13 5 11 5 20

2.5 60 3

V IV

12

20

I II III

[‘HIP I

7

II

8 9 10 II

III

DISCUSSION

3 20 20 4 3 3

steroid

radioimmunoassays

b

ED,.

-0.90 f 0.06 -0.6 -0.70 +0.10 -0.90 + 0.05 -0.98 i 0.07 -1.00~0.10 - I .05 + 0.09 -0.86+_O.l0 -0.53 + 0.09 -0.98 * 0.05 no reliable -0.84 + 0.05

100 700 425 100 89 81

n: number of assays. d: working dilution of antiserum (in thousands). b: slope of the logit-log curve (+ SD). ED,-displacement (pg/tube) at 50% binding vs zero-dose.

54 360 174 235

Antisera

nos

RI, R2, R314 RI RI, R2 RI R41, R79, R80 R41 R41, R78 R79 R41, R78, R79, R80 R63, R74 obtained R63. R74. R76

666

R. STUPNKX et al. -I2o-S-HME-“’

I

-Ilo-MS-HMEmi5i -II.

-MS P-lla

-TA-

”

I

-MS-TAG”? - MS - HME-‘“I

100 200

50

500

IO00

pg/tube

.-

E

F-21-C-HME+5

1

I

I

IO

I

I

I

I

I

I

20

50

100

200

500

1000

I

pg /tube

Fig. 3. Mean logit-log curves obtained with different tracers of progesterone (upper graph) and cortisol (lower graph). Numbers in brackets refer to assay Nos as in Table 1. Results of assays Nos 1 and 4, being almost identical with those of assay No. 5, were omitted from the graph. HOMO

ED50

cm 8

3

0

HETERO

II -

50

0

m

._

-

F ‘H

7 -

4 -

Fig. 4. Effects of traceis on relative displacement (ED,,) abilitv.

Relative Upper logous

displacement

=

ED,, (‘H-steroid) ED,, (iodinated

APPENDIX

characteristics of the steroid derivatives used for labelling. The NMR-spectra were recorded on JEOL 100 MHz spectrometer. Basic

I50

100

performance of the 12a(antibody)/l lcr(tracer) system could be due partly to the difference in bridge structures (methylsuccinyl in case of the 12cr/l lo! system, succinyl in case of the reverse one), and partly to the better overall performance of antibodies directed against progesterone substituted at 12cc-position, as compared with 1 ICC.This latter remark coincides with an earlier observation made on tritium-based progesterone assays [6]. Two cortisol-2 1-carbonyl derivatives were studied: TME- and HME-substituted. Of the 3 labellings of the TME-derivative only in one instance was a reasonable standard curve obtained, while in all instances of using the HME-derivative very consistent results were observed. Although the ED,, value was lower than in case of the ‘H-based assay (74’S, cf. Fig. 4) the slope was acceptable and the sensitivity was sufficient for a direct assay of cortisol in 1 PI/tube of human plasma (Stupnicki, unpublished results). It should be mentioned that the results presented were in most cases obtained with up to four different antisera directed against the same immunogen, and only in one case (assay No. 7 in Table 1) was a difference in the performance of different antisera observed. White et al. [lo] reported a direct labelling of histamine-substituted steroid derivatives in microgram quantities. However, these derivatives had to be stored deep-frozen in alcoholic solution for they were not crystalline compounds. In contrast to that histidine methyl ester, which is not a hazardous compound, can be easily coupled to carboxylic derivatives forming stable, crystalline compounds. A slightly yellowish colour which develops with time does not affect the iodine-acceptor properties. The steroid-HME derivatives can be directly iodinated in microgram quantities with good yield. They combine al the advantageous characteristics of imidazolic and phenolic iodine acceptors, being free from the disadvantageous ones.

derivative)

graph: homologous systems. Lower graph: heterosystem. Numbers under the bars refer to assay Nos as in Table 1.

I m.p. 105-108°C NMR: G(DMS0); ppm: 0,65(s,3H,C18); l.l2(s,3H, C19); 1.98(s,3H,C21) 5.60(s,lH,C4); 6.65(d,2H,arom.); 6,97(d,2H,arom.) II m.p. 95°C NMR: S(CDC1,); ppm: 0,75(s,3H,C18); I .30(s,3H, C19); 2.20@,3H,C21) 3.75(s,3H,-COOCH,); 5.85(s,lH, C4); 6.90(s,lH,imidazole-ring); 7,70(s,lH,imidazolering) IIJ m.p: 1355138°C NMR: 6(CDCl,); ppm: 0,75(s,3H,C18); 1.32(s,3H, C19); 2,20(s,3H,C21) 3.78(s,3H,-COOCH,); 5.90(s,lH, C4); 6.90(s,lH,imidazole-ring); 7.70(s,lH,imidazolering) IV m.p. 178%180°C NMR: 6(CDCl,); ppm: 0,76(s,3H,C18); 1,35(s,3H, C19); 3.75(s,3H,-COOCH,); 4.95(s,2H,C21); 5,85(s,lH, C4); 6.95(s,lH,imidazole-ring); 7,72(s,lH,imidazolering)

Steroid-HME derivatives for direct iodination V m.p. 13&135”C NMR: 6(CD,),CO; ppm: 0.76(s,3H,C18); 1.35(s,3H, C19); 4.80(d,2H,C21); 5.55(s,lH,C4); 6,75(d,2H,arom.); 7,lO(d,2H,arom.) Trivial names: I (P-l la-MS-TA): 1la-hydroxy-pregn-C en-3,20-dione-1 l[(N-tyramine)-2smethyl succinamide]hemisuccinate. II (P-l la-MS-HME): 1lc~-hydroxy-pregn-4-en-3,20-dione1l[N-(2’-histidine-methyl ester)-2smethyl succinamide]hemisuccinate. III (P-12a-S-HME): 12a-hydroxy-pregn-4-en-3,20-dione12[N-(2’-histidine methyl ester)-succinamide]-hemisuccinate. IV (F-21-C-HME): ll/J,l7a,21-trihydroxy-pregn-4-en3,20-dione-21[N-(2’-histidine methyl ester)]-carbamate. V (F-21-C-TME): 1lb,17a,21-trihydroxy-pregn-4-en3,20-dione-21(N-tyrosine methyl ester)-carbamate. P-lla-S-BSA: lla-hydroxy-pregn-4-en-3,20-dione-llhemisuccinate-BSA. F-21-S-BSA: 11/3,17a,21-trihydroxy-pregn-4-en-3,20-dione21-hemisuccinate-BSA. BSA: bovine serum albumin. REFERENCES

1. Jeffcoate S. L.: Use of ‘rSI-iodine tracers in steroid radioimmunoassays. In Radioimmunoassay of Steroid Hormones (Edited by D. Gupta). Verlag Chemie, Base1 (1980) pp. 209-219. 2. Cameron E. H. D., Scarisbrick J. I., Morris S. E. and Read G.: iz51-iodohistamine derivatives as tracers for the radioimmunoassay of progestagens. In Steroid Immunoassav (Edited bv E. H. D. Cameron, S. G. Hillier and K. ‘Grifliths).~Alpha Omega, Cardiff (1975) pp. 153-164.

667

3. Hunter W. M., Nars P. W. and Rutherford F. J.: Preparation and behaviour of ‘251-labelled radioligands for phenolic and neutral steroids. In Steroid Immunoassay (Edited by E. H. D. Cameron, S. G. Hillier and K. Griffiths). Alpha Omega, Cardiff (1975) pp. 141-152. 4. Thorell J. I., Ekman R and Malmquist M.: Technical aspects of the production and application of iodinated steroids for radioimmunoassay. In Radioimmunoassays and Related Procedures in Medicine (Proceedings Series). I.A.E.A. Vienna, Austria (1982) pp. 147-160. 5. Stupnicki R.: Direct radioimmunoassay of progesterone in plasma of farm animals, Endokrinologie 66 (1975) pp. 145-151. 6. Stupnicki R. and Kula E.: Direct radioimmunoassay of progesterone in human plasma. Endokrinologie 80 (1980) pp. l-7. 7. Erlanger B. T., Beiser S. M., Borek F., Edel F. and Lieberman S.: The preparation of steroid-protein conjugates to elicit hormonal antibodies. In Methods in Immunology and Immunochemistry (Edited by C. A. Williams and M. W. Chase). Academic Press, N.Y. Vol. 1 (1967) pp. 144150. 8. Fridkin M. and Goren H. J.: Use of N-t-butyloxycarbonyl in the synthesis of poly-1-histidine. Can. J. Chem. 49 (1971) 1578. 9. Stupnicki R. and Kula E.: Direct radioimmunoassay of progesterone in milk using a radioiodinated progesterone derivative. Anim. Reprod. Sci. (1980) pp. 113-l 18. 10. White A., Smith G. N., Crosby S. R. and Ractliffe W. A.: A study of 19-O-carboxymethyl ether and 19-hemisuccinate derivatives of testosterone, their immunogenicity and use as iodinated radioligands for radioimmunoassay of testosterone. J. steroid Biochem. 23 (1985) pp. 981-989.