Syntheses of novel hapten-protein conjugates for production of highly specific antibodies to formononetin, daidzein and genistein

Article abstract of DOI:10.1016/S0040-4020(99)00993-X

The syntheses of novel hapten-carrier protein conjugates are described. Isoflavones functionalised in C2 were coupled to bovine serum albumin or swine thyroglobulin. The antibodies obtained were used for the development of ELISAs (enzyme-linked immunosorb

Full text of DOI:10.1016/S0040-4020(99)00993-X

TETRAHEDRON
Pergamon
Tetrahedron 56 (2000) 295–301
Syntheses of Novel Hapten–Protein Conjugates for Production of
Highly Specific Antibodies to Formononetin, Daidzein
and Genistein
a
b,c,
c
b
*
´
´
Cyril Le Houerou, Catherine Bennetau-Pelissero,
Valerie Lamothe, Franc¸oise Le Menn,
Pierre Babind and Bernard Bennetaua,
*
a
´
´
´
Laboratoire de Chimie Organique et Organometallique UMR 5802 CNRS, Universite Bordeaux I, 351 cours de la Liberation,
33405 Talence, France
b
´
´
Laboratoire de Biologie de la Reproduction des Poissons, Universite Bordeaux I, Avenue des Facultes, 33405 Talence, France
c
´ ´
ENITA de Bordeaux, 1 cours du General de Gaulle, 33175 Gradignan, France
d
´
Laboratoire de Pharmacie Chimique, Universite Bordeaux II, Place de la Victoire, 33000 Bordeaux, France
Received 3 August 1999; accepted 3 November 1999
Abstract—The syntheses of novel hapten–carrier protein conjugates are described. Isoflavones functionalised in C2 were coupled to bovine
serum albumin or swine thyroglobulin. The antibodies obtained were used for the development of ELISAs (enzyme-linked immunosorbent
assays) to measure the three isoflavones with IC50 ranging from 15.6 to 100 ng/mL. ᭧ 1999 Elsevier Science Ltd. All rights reserved.
Introduction
to higher phyto-oestrogen concentrations than Western
consumers. They also present lower incidence of
oestrogen-dependent diseases such as breast, uterus or
colon cancer.8
Isoflavones such as formononetin 1, daidzein 2 and genis-
tein 3 are natural oestrogenic compounds present in large
amounts (up to 1–4 mg/g)1 in seeds, fruits and vegetables.2,3
Phyto-oestrogens have oestrogen-like activity and may
interfere with the reproductive cycle in animals.4–7
However, during the last decade, an interest in the study
of these compounds has increased due to their potential
effects as health protecting dietary factors in human
populations.
More recently, soy extracts were shown to decrease the
incidence of hot flushes9 and to reduce osteoporosis in
menopausal women.10 Today, soy extracts are com-
mercialised for the prevention of female menopausal
troubles. Lately, concern has grown about the potential
adverse effects of these compounds in man considering
the high amounts of both genistein and daidzein found in
infant formula milk substitutes, based on soy.11,12 Due to the
potential effects of phyto-oestrogens as health dietary
factors, the demand for analytical methods which are
more cost-effective and less time-consuming than GC or
HPLC is increasing. This demand could be filled by use of
enzyme-linked immunosorbent assays (ELISAs). As a
matter of fact, analytical methods based on immuno-
chemistry are widely used in biochemistry, endocrinology
or medical chemistry13 and have been extended to
environmental and toxicological areas.14
Epidemiological studies dealt with comparisons between
Western and Asian consumers. These showed that
vegetarian and Asian consumers were currently exposed
The initial step in the development of an immunoassay is the
production of specific antibodies, but small molecules such
as isoflavones are non-immunogenic. Hence, it is necessary
to design appropriate chemical structures15 that can be
covalently coupled to a carrier protein. Various methods
are available for coupling haptens and proteins with
carboxylic acid groups in target molecules and amino
Keywords: hapten–protein conjugates; antibodies; isoflavones.
* Corresponding author. Fax: ϩ556-846-646; e-mail: b.bennetau@lcoo.u-
bordeaux.fr
0040–4020/00/$ - see front matter ᭧ 1999 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020(99)00993-X

´
C. Le Houerou et al. / Tetrahedron 56 (2000) 295–301
296
Scheme 1. Reagents: (i) BF3.Et2O then H2O, 70%; (ii) LiHMDS, ClCO(CH2)2CO2Me, AcOH/H2SO4, 32%; (iii) BBr3, CH2Cl2, 87%; (iv) coupling with
proteins, see ratio in Table 1.
groups in protein.16 Using these strategies, several authors
have described the synthesis of isoflavone–protein conju-
gates to develop radioimmunoassays (RIAs).17–20 However
a direct conjugation with a functional group in the target
molecule and amino groups in protein should be avoided
because an important antigenic determinant of the molecule
disappears making the resulting antibodies less specific.
Moreover, in the RIA technique, radioisotopes are used
which are not very convenient for control assays in the
diet industry.
with six equivalents of boron tribromide at low temperature
gave the acid 12 in 87% yield.
A similar strategy has been described for the synthesis of
hydroxylated flavones22 with acetophenones and aroyl
chlorides.
The synthesis of the hapten 17 of genistein is given in
Scheme 2. Initially, it was anticipated that the deoxybenzoin
15 could be made by variation of the known procedure with
phloroglucinol but the very poor yields observed prompted
us to undertake another process. With the trimethoxyether
13, the hydroxylated deoxybenzoin 15 was obtained in 33%
yield with BF3.Et2O as catalyst.
Results
Based on these considerations, in order to develop an
ELISA test, we designed the synthesis of isoflavone–protein
conjugates by introducing an ethylene carboxyl group at C2
carbon for conjugating with BSA or Thyr while all func-
tional groups of isoflavones were preserved (Scheme 1).
As previously observed, a selective mono-demethylation
occurs in the synthesis of chalcones.23 The acid 17 was finally
obtained, after cyclisation and demethylation, following the
same procedure as described above. A few examples of
isoflavones, with a side chain containing four methylene
groups and a carboxylic group, were synthesised by a differ-
ent method but specific antibodies were not obtained since
the acids have not been conjugated with proteins.24
The synthesis of 4 and 5 was achieved by using resorcinol as
the starting material. Deoxybenzoin 9 was prepared accord-
ing to a known procedure21 in 70% yield with BF3.Et2O as
catalyst. From the enolate of 9, generated with lithium
hexamethyldisilazane (LiHMDS), addition of methoxy-
carbonylpropionyl chloride produced the cyclised product
11 in 32% yield. All attempts to isolate the intermediate
diketone 10 were unsuccessful. However, treatment of 11
The three haptens (11, 12 and 17) were coupled with BSA
(4–6a) and Thyr (4–6b). The efficiency of the conjugation
was tested spectrophotometrically25 and is summarised in
Table 1.
Scheme 2. Reagents: (i) BF3.Et2O then H2O, 33%; (ii) LiHMDS, ClCO(CH2)2CO2Me, AcOH/H2SO4, 30%; (iii) BBr3, CH2Cl2, 27%; (iv) coupling with
proteins, see ratio in Table 1.

´
C. Le Houerou et al. / Tetrahedron 56 (2000) 295–301
297
Table 1. Conjugation efficiency expressed as the ratio of haptens coupled to
(BSA) or (Thyr)
The data show clearly that the antibodies obtained are valu-
able for assays of phyto-oestrogens. Because of some high
cross-reactions, separation of the compounds prior to assay
might be worthwhile. For soy, however, sample treatments
are not needed since soy only contains daidzein and genis-
tein and cross-reactions between the two are very low. A
simple extraction is therefore sufficient for assay in food and
animal tissues. Comparing our sensitivity results with
ELISA of two herbicides, Norflurazon᭨ and Bromacil᭨,
described by other authors,26,27 IC50 between 1 and
10 ng/mL, i.e. 1–10 mg/L or 1–10 ppb were reported. In
that respect, our ELISAs are of the same order of sensitivity.
When we compare our assays to RIAs of phyto-oestrogens
we observe that the latter are generally 10–100 times more
sensitive than ELISAs are. However, a very high sensitivity
is not required. Indeed, previous data1,28 indicate that phyto-
oestrogens are present in high concentrations in plant as
well as in animal and human fluids. Moreover, phyto-oestro-
gens are usually considered to be 1000–10 000 fold less
oestrogenic than oestradiol.28 We thus made the choice of
a very easy and convenient method which does not need
radioisotopes. This method is suitable to be used by diet
manufacturers to control their materials before and after
processing, and for detection in animal fluids with correla-
tion to biological activities.
Haptens
Ratio haptens/BSA
Ratio haptens/Thyr
11
12
17
48.7 (4a)
44.5 (5a)
47.2 (6a)
191.1 (4b)
231 (5b)
200 (6b)
The BSA conjugates were repeatedly injected individually
to New Zealand rabbits affording three polyclonal anti-
bodies. Their titre, specificity and sensitivity were examined
by ELISA. The principle of the assay lies in the competition
between free isoflavones and coated hapten–Thyr for
specific antibodies. All the antibodies exhibited extremely
high titres (optimal final dilution 1:60 000; 1:160 000;
1:240 000 for antisera 4a, 5a and 6a, respectively). The
optimal assay conditions for each of the antibodies are
presented in Table 2.
Experimental
Melting points were determined on Kofler apparatus or a
Mettler FP 62 capillary melting point apparatus and were
uncorrected. Infrared spectra were recorded on a Paragon
1000 Perkin–Elmer infrared spectrophotometer with poly-
styrene as standard; wavenumber (cmϪ1) values are given.
Proton magnetic resonance spectra (1H NMR) were
recorded on a Bruker AC 200 at 200 MHz or on a Bruker
AC 250 spectrometer at 250 MHz. Carbon magnetic reso-
nance spectra (13C NMR) were recorded on a Bruker AC
250 spectrometer at 62.9 MHz or on a Bruker AC 200 at
50 MHz. Chemical shifts (d) are given in parts per million
(ppm) using tetramethylsilane as the internal standard at
0.00 ppm. Electron impact mass spectra were determined
on a VG AUTO SPEC-Q apparatus at 70 eV; data are
reported as m/z (relative intensity).
The sigmoid competition curves obtained from the anti-
bodies produced are shown in Fig. 1.
Once the optimal conditions were determined, the specifi-
cities of the antibodies were determined (see Table 3) with
the parent compounds (structures given in Fig. 2).
Whereas the titres of our antibodies were extremely high for
antisera 4a and antisera 5a, the displacement of the standard
curve was lower than that of anti 6a. This suggests that
the antibodies raised against BSA conjugates recognise
unspecifically the Thyr conjugates coated onto the plate.
The specificities of our antibodies are good, however, in each
case a strong cross-reaction with one of the isoflavones tested
was detected. These strong cross-reactions were observed
between daidzein and formononetin on the one hand and
between genistein and biochanin A on the other hand.
The ELISA technique was carried out on microtitration
plates (NUNC maxisorp) with 96 wells. The optical
densities were read on a Dynex MRX II microtitration
plate reader at 490 nm. The standard curves were semi-
log expression of OD sample/OD positive controlˆ
f(log(concentrations)).
The cross-reactivity values were calculated for each antisera
according to the following equation:
IC50 of the isoflavone to which is directed the antibody
×100:
1-(2,4-Dihydroxyphenyl)-2-(4-methoxyphenyl)ethanone
9. BF3.Et2O (120 mL, 0.91 mol) was added dropwise to a
mixture of resorcinol 7 (5 g, 45.41 mmol) and benzylic acid
8 (7.5 g, 45.13 mmol). The mixture was heated for 12 h at
Ϫ70ЊC. After cooling, the mixture was poured onto icy
water (800 mL). The product was filtered off and
recrystallized from ethanol–water (1:1) to give compound
9 as white crystals (8.2 g, 70%); mp 163ЊC (Kofler); IR
(KBr) 3361, 1623, 1613, 1511, 1435, 1352, 1242, 1175,
1129, 1024; dH (250 MHz, acetone-d6/CDCl3) 3.29 (s,
2H), 3.69 (s, 2H), 5.88 (d, 1H, Jˆ2.4 Hz), 5.96 (dd, 1H,
IC50 of the free phyto-oestrogens
Table 2. Optimal conditions for each of the assay procedures (IC50: concen-
tration of free antigen including 50% displacement of the antibody binding
to the coating)
Optimal conditions
Antiserum 4a Antiserum 5a Antiserum 6a
Coating conc. (mg/mL) 0.1
0.4
0.1
Std curve limit (ng/mL) 2000–0.97
1000–0.48
40
250–0.12
15.6
IC50 (ng/mL)
100
Slope
Ϫ0:75 ^ 0:19 Ϫ0:66 ^ 0:11 Ϫ0:60 ^ 0:08

´
C. Le Houerou et al. / Tetrahedron 56 (2000) 295–301
298
Figure 1. Standard curves obtained with the different antibodies.
Jˆ2.4 Hz, Jˆ8.8 Hz), 6.38 (AA0BB0, 2H), 6.74 (AA0BB0,
2H), 7.36 (d, 1H, Jˆ8.8 Hz), 8.92 (s, 1H), 12.08 (s, 1H); dC
(50 MHz, acetone-d6/CDCl3 43.2, 54.4, 102.6, 107.7, 112.2,
113.5, 126.4, 129.9, 132.6, 158.2, 164.2, 165.4, 201.9.
(m, 2H), 3.78 (s, 3H), 6.82 (d, 1H, Jˆ2.0 Hz), 6.89 (dd, 1H,
Jˆ8.7 Hz, Jˆ2.0 Hz), 6.97 (AA0BB0, 2H), 7.17 (AA0BB0,
2H), 7.86 (d, 1H, Jˆ8.7 Hz); 10.75 (s, 1H), 12.3 (s, 1H); dC
(62.9 MHz, DMSO-d6) 27.4, 30.5, 54.9, 101.8, 113.4, 114.8,
115.5, 121.7, 124.8, 127.0, 131.5, 156.9, 158.5, 162.3,
163.5, 173.0, 175.1.
2-(2-Carboxyethyl)-7-hydroxy-3-(4-methoxyphenyl)-
4H-chromen-4-one 11. A solution of LiHMDS [prepared
by addition of BuLi (2.5 M in hexane, 18.6 mL, 46.47 mmol)
to hexamethyldisilazane (9.8 mL, 46.47 mmol) at Ϫ70ЊC]
in THF (40 mL) was added dropwise to a solution of the
deoxybenzoin 9 (3 g, 11.62 mmol) in THF (40 mL) at 70ЊC.
After return to room temperature, the mixture was cooled to
Ϫ70ЊC and a solution of methoxycarbonylpropionyl
chloride (1.5 mL, 12.2 mmol) in THF (20 mL) was added
dropwise. The reaction mixture was stirred for 4 h, then
poured onto water (300 mL) and acidified with HCl (1 M).
The solution was extracted with chloroform (3×100 mL).
Removal of the organic phase under reduced pressure
gave a brown residue that was treated with acetic acid
(100 mL) and sulphuric acid (1 mL). The mixture was
heated at reflux for 2 h. After removal of the solvent
under reduced pressure, the crude product was diluted
with chloroform (100 mL). The organic phase was washed
with water (3×80 mL) and dried over magnesium sulphate.
Removal of the solvent under reduced pressure gave a
brown residue which was purified by column chromato-
graphy eluting with 20% diethyl ether in dichloromethane
and then 20% ethanol in dichloromethane to give a brown
powder which was purified again by recrystallisation from
ethanol–water (1:1) to give the product 11 as a white
powder (1.27 g, 32%): mp 255ЊC (Mettler); IR (KBr)
3144, 1695, 1626, 1607, 1584, 1510, 1468, 1392, 1294,
1233, 1172; dH (200 MHz, DMSO-d6) 2.61 (m, 2H), 2.74
2-(2-Carboxyethyl)-7-hydroxy-3-(4-hydroxyphenyl)-4H-
chromen-4-one 12. BBr3 1 M in dichloromethane (22 mL,
21.5 mmol) was added dropwise with a syringe to a stirred
solution of acid 11 (1.2 g, 3.52 mmol) in dichloromethane
(60 mL). After complete addition, the mixture was stirred
for 24 h at room temperature, then poured onto icy water
(400 mL). Filtration of the crude product gave a yellow
residue which was recrystallized from ethanol–water (1:1)
to give the product 12 as a white powder (1 g, 87%): mp
250–300ЊC (Mettler); IR (KBr) 3450, 3215, 1712, 1621,
1610, 1562, 1553, 1510, 1401, 1211, 1168; dH (250 MHz,
DMSO-d6) 2.57–2.62 (m, 2H), 2.73–2.78 (m, 2H), 6.79–
6.82 (m, 3H), 6.88 (d, 1H, Jˆ8.6 Hz), 7.05 (AA0BB0, 2H),
7.86 (d, 1H, Jˆ8.6 Hz); 9.52 (s, 1H), 10.75 (s, 1H), 12.3 (s,
1H); dC (62.9 MHz, DMSO-d6) 27.4, 30.6, 101.8, 114.7,
114.8, 127.0, 131.4, 115.5, 122.0, 123.1, 156.6, 156.9,
162.3, 163.4, 173.1, 175.2; MS m/z (relative intensity) 326
(56.4 Mϩ), 281 (100MϩϪCO2H), 137 (19.7); C18H14O6
required C 66.26%, H 4.32%, O 29.42%, found C
65.77%, H 4.74%, O 29.29%.
1-(2-Hydroxy-4,6-dimethoxyphenyl)-2-(4-methoxy-
phenyl) ethanone 15. According to the procedure described
above for 9, trimethoxyether 13 (5 g, 29.73 mmol) was
treated with BF3.Et2O (80 mL, 0.65 mol) and benzylic
acid 14 (4.9 g, 29.5 mmol). The solution was extracted
Table 3. Cross-reactivity between the antibodies and the free compounds (values are % cross-reaction at 50% binding ^SD calculated on three determinations)
Antisera
Formononetin 1
Daidzein 2
65^1
Equol 19
Biochanin A 18
Genistein 3
4a
5a
6a
0.36^0.1
0.19^0.05
0.062^0.02
7.9^0.03
1.78^0.5
53^12
7.9^0.03
5.22^1
51.5^10
2.08^1.1
2.05^1.2

´
C. Le Houerou et al. / Tetrahedron 56 (2000) 295–301
299
Figure 2. Structures of phyto-oestrogens used for cross-reactivity.
with diethyl ether (2×80 mL). The organic phase was
washed with water (2×50 mL) and dried over magnesium
sulphate. Removal of the solvent gave a yellow solid residue
which was purified by column chromatography eluting with
dichloromethane to give a yellow powder product which
was recrystallized from ethanol–water (10:1) to give
compound 15 as yellow crystals (3 g, 33%): mp 88ЊC
(Kofler); IR (KBr) 3448, 1625, 1590, 1509, 1418, 1279,
1226, 1166, 1145, 1117, 1030; dH (250 MHz, CDCl3)
3.76 (s, 6H), 3.87 (s, 3H), 4.27 (s, 2H), 5.93 (d, 1H,
4Jˆ2.1), 6.07 (d, 1H, 4Jˆ2.1), 6.89 (AA0BB0, 2H), 7.16
(AA0BB0, 2H), 14.99 (s, 1H); dC (62.9 MHz, CDCl3)
49.4, 55.2, 55.5, 90.8, 93.8, 105.6, 118.8, 127.5, 130.7,
158.4, 162.7, 166.3, 167.9, 203.2.
1H, Jˆ2.0 Hz), 6.93 (AA0BB0, 2H), 7.21 (AA0BB0, 2H),
8.46 (s, 1H), 9.57 (s, 1H), 13.0 (s, 2H); dC (50.3 MHz,
acetone-d6) 28.5, 31.1, 94.3, 99.5, 105.1, 116.0, 122.4,
123.7, 132.7, 158.0, 158.6, 163.6, 164.8, 166.1, 173.4,
182.1; MS m/z (relative intensity) 342 (90.6Mϩ), 297
(100MϩϪCO2H), 153 (25.3); C18H14O7 required
C
63.16%, H 4.12%, O 32.72%, found C 62.87%, H 4.31%,
O 31.62%.
Preparation of Hapten–Protein conjugates
Haptens were covalently attached to BSA using a variation
of two procedures previously reported.26,27 Peptidic bonds
were formed between the carboxylic function of the haptens
and the free amino groups of lysine residues present on
BSA. A protein solution of BSA (6.1 mg, 0.072 mmol)
was prepared in 5 mL borate buffer (0.2M borate–boric
buffer, pH 8.7). Tributylamine (17.3 mL, 0.072 mmol),
isobutyl chloroformate (4.7 mL, 0.036 mmol) and hapten
(0.036 mmol) were mixed in 1 ml DMF for 30 min on an
ice bed. The hapten solution in DMF was then added drop-
wise to the borate buffer solution of protein and continu-
ously mixed for 6 h at room temperature. This solution was
then dialysed, first against PBS (0.01 M phosphate buffer
with 0.9% NaCl, pH 7.4), and then against distilled water.
The conjugates were lyophilised and stored in aliquots at
Ϫ20ЊC. The stoichiometric ratio was 1:50.
2-(2-Carboxyethyl)-5,7-dimethoxy-3-(4-methoxyphe-
nyl)-4H-chromen-4-one 16. According to the procedure
described above for 11, compound 15 (3 g, 9.92 mmol)
in THF (40 mL) was treated with
a solution of
LiHMDS [prepared by addition of BuLi (2.5 M in hexane,
12 mL, 29.77 mmol) to hexamethyldisilazane (6.3 mL,
29.77 mmol)] then with a solution of methoxycarbonyl-
propionyl chloride (2.5 mL, 19.85 mmol) in THF (20 mL).
Removal of the solvent under reduced pressure gave a
brown residue which was purified by column chromato-
graphy eluting with 20% diethyl ether in dichloromethane
and then 20% ethanol in dichloromethane to give a brown
powder which was purified again by recrystallisation from
ethanol–water (1:1) to give the cyclisation product 16 as a
white powder (1.1 g, 30%): mp 236ЊC (Kofler); IR (KBr)
3421, 1712, 1644, 1615, 1572 1514, 1460, 1422, 1246,
1216, 1196, 1176; 1164, 1115; dH (250 MHz, DMSO-d6)
2.59–2.62 (m, 2H), 2.66–2.69 (m, 2H), 3.87 (s, 6H), 3.87 (s,
3H), 6.47 (d, 1H, Jˆ2.1 Hz), 6.64 (d, 1H, Jˆ2.1 Hz) 6.96
(AA0BB0, 2H), 7.14 (AA0BB0, 2H), 12.3 (s, 1H); dC
(62.9 MHz, DMSO-d6) 27.0, 30.4, 54.8, 55.7, 55.8, 92.2,
95.8, 107.7, 113.3, 123.0, 125.0, 131.5, 158.4, 158.7,
160.4, 161.0, 163.4, 173.0, 174.1.
The coupling of haptens with Thyr (48.2 mg, 0.072 mmol)
was realised according to the above experimental procedure.
The coupling efficiencies are summarised in Table 1.
Coating antigens and immunisation
Haptens were covalently attached to swine thyroglobulin
(Thyr), through their carboxylic function to the lysine resi-
dues of Thyr, forming a peptidic bond. The procedure
follows the method previously described. In this case, the
stoichiometric ratio was 1:500. The coupling efficiency is
summarised in Table 1.
2-(2-Carboxyethyl)-5,7-dihydroxy-3-(4-hydroxyphenyl)-
4H-chromen-4-one 17. According to the procedure
described above for 12, acid 16 (1.7 g, 4.42 mmol) in
dichloromethane (60 mL) was treated with BBr3 1 M in
dichloromethane (31 mL, 31 mmol). Filtration of the
crude product gave a yellow residue which was recrystal-
lized from ethanol–water (1:1) to give the product 17 as a
brown powder (320 mg, 27%): mp 246ЊC (Kofler); IR (KBr)
3355, 3280, 1710, 1654, 1609, 1562, 1513, 1367, 1215,
1197, 1162; dH (200 MHz, acetone-d6) 2.71–2.79 (m,
2H), 2.89–2.97 (m, 2H), 6.26 (d, 1H, Jˆ2.0 Hz), 6.42 (d,
Polyclonal antibodies were chosen for their better specifi-
city. Immunisation of the rabbits was done on New Zealand
rabbits from CEGAV (France). Rabbits were first ear-
sampled for pre-immunoserum test. For each injection
500 mg of the BSA conjugates were injected. On the first
injection antigens were dissolved in 1 mL PBS-complete
Freund’s adjuvant (v–v). For the subsequent injections anti-

´
C. Le Houerou et al. / Tetrahedron 56 (2000) 295–301
300
gens were dissolved in PBS-incomplete Freund’s adjuvant
(v–v). Injections were performed at multiple points accord-
ing to the following injection schedule. The first three injec-
tions were performed at one-week intervals. Two additional
injections followed thereafter at three-week intervals. After
the fifth injection a test was done on a 5 mL sample of serum
to check the specificity and titre of the antibodies. Two last
injections were then performed at one-month intervals. Each
blood sampling was carried out one week after the previous
injection; 50 mL of blood were finally collected. Serum was
obtained after clotting for 24 h at 4ЊC and centrifuged at
3000 g for 10 min at 4ЊC. Aliquots of sera were stored
frozen at Ϫ20ЊC. When in use, the aliquots were diluted
v–v with glycerol and stored at Ϫ20ЊC.
prepared in PBS–T–PS–DMSO and each IC50 was deter-
mined in the assay experiment. The cross-reactivity values
were calculated for each antisera according to the following
equation: (IC50 of the isoflavone to which is directed
the antibody/IC50 of the phyto-oestrogens listed in
Table 3)×100.
Acknowledgements
This work was supported by a grant from the French
Ministry of Agriculture, by the French Ministry of
´
´
‘Recherche et Enseignement Superieur’, by the ‘Region
Aquitaine’ and by the ‘ENITA de Bordeaux’.
ELISA procedure
References
Microtitration plates were coated with 200 mL/well of a
Thyr conjugates solution at the concentrations indicated in
Table 2. This operation was performed in coating buffer
(0.1 M carbonate bicarbonate buffer, pH 9.6), overnight, at
4ЊC covered with aluminium sheet. The following day the
plates were saturated with PBS–T–PS–DMSO (Phosphate
Buffer Saline solution containing Tween, Pork serum and
DMSO; 0.01 M phosphate buffer, 0.9% NaCl, 1.6% crude
pork serum, pH 7.4, 0.05% Tween 20, 1% DMSO), for
30 min at 37ЊC and then washed three times in PBS-T-
DMSO) (PBS, 0.05% Tween 20%, 1% DMSO). Serial dilu-
tions of the analyte were prepared in PBS–T–PS–DMSO as
standard curves and added in a 100 mL/well volume to the
coated plates. This operation was followed by the addition
of 100 mL/well of antibody solution to PBS–T–PS–DMSO.
The final optimal dilutions of the antibodies in the wells are
presented in Table 2. After 2 h incubation at 37ЊC, the plates
were washed three times with PBS–T–DMSO. Incubation
with 200 mL/well of a second antibody solution (1/1000 in
PBS–T–PS–DMSO) was performed. Second antibody is
swine immunoglobulin anti-rabbit immunoglobulin coupled
to horseradish peroxidase (HRP) (Dako). Incubation lasted
30 min at 37ЊC. Finally, the plates were washed three times
with PBS–T–DMSO, and the revelation of the peroxidase
activity was effected by adding 200 mL of ortho-phenylene-
diamine (OPD) 0.5 M, in citrate–phosphate buffer (0.15 M,
pH 5) containing 0.025% H2O2 30% at room temperature.
Revelation lasted for 30 min and the reaction was stopped
by adding 50 mL to H2SO4 4 M to each well.
1. Anderson, R. L.; Wolf, W. J. J. Nutr. 1995, 125, 581S.
2. Farsworth, N. R.; Bingel, A. S.; Cordell, G. A.; Crane, F. A.;
Fong, H. H. S. J. Pharm. Sci. 1975, 64, 535.
3. Farsworth, N. R.; Bingel, A. S.; Cordell, G. A.; Crane, F. A.;
Fong, H. H. S. J. Pharm. Sci. 1975, 64, 717.
4. Hugues Jr., C. L.; Kaldas, R. S.; Weisinger, A. S.; McCants,
C. E.; Basham, K. B. Reprod. Toxicol. 1991, 5, 127.
5. Hugues Jr., C. L.; Chakinala, M. M.; Reece, S. G.; Miller, R. N.;
Schomberg Jr., D. W.; Basham, K. B. Reprod. Toxicol. 1991, 5,
133.
6. Medlock, K. L.; Branham, W. S.; Sheehan, D. M. Proc. Soc.
Exp. Biol. Med. 1995, 208, 67.
7. Levy, J. R.; Faber, K. A.; Ayash, L.; Hughes Jr., C. L. Proc. Soc.
Exp. Biol. Med. 1995, 208, 60.
8. Tham, D. M.; Gardner, C. D.; Haskell, W. L. J. Clin. Endo-
crinol. Metab. 1998, 83, 2223.
9. Murkies, A. L.; Lombard, C.; Strauss, B. J.; Wilcox, G.; Burger,
H. G.; Morton, M. S. Maturitas 1995, 21, 189.
10. Anderson, J. J.; Garner, S. C. Baillieres Clin. Endocrinol.
Metab. 1998, 12, 543.
11. Setchell, K. D. R.; Zimmer-Nechemias, L.; Cai, J.; Heubi, J. E.
Am. J. Clin. Nutr. 1998, 68, 1453S.
12. Knight, D. C.; Eden, J. A.; Huang, J. L.; Waring, M. A. J.
Paediatr. Child Health 1998, 34, 135.
13. Marco, M.-P.; Hammock, B. D.; Kurth, M. J. J. Org. Chem.
1993, 58, 7548.
14. Szurdoki, F.; Bekheit, H. K. M.; Marco, M.-P.; Goodrow,
M. H.; Hammock, B. D. New Frontiers in Agrochemical Immuno-
assay; Kurtz, D. A.; Skerritt, J. H.; Stanker, L. Eds.; AOAC
International: 1995, p 39 and cited references.
15. Sherry, J. P. Crit. Rev. Anal. Chem. 1992, 23, 217.
16. Erlanger, B. F. Methods in Enzymology. 1980, 70, 85.
17. Wang, W.; Tanaka, Y.; Han, Z.; Cheng, J. J. Agric. Food.
Chem. 1994, 42, 1584.
18. Hampl, R.; Lapcik, O.; Wahala, K.; Starka, L.; Adlercreutz, H.
Chem. Listy 1998, 92, 44.
Analysis of the antibody titre
The titre of the antibodies in the serum of animals was
determined following the ELISA procedure by measuring
the binding of a 1/4000 antibody dilution to a 0.25 mg/mL
concentration of Thyr conjugates. Coating of Thyr
conjugates was performed in coating buffer. The antibody
incubation was performed in PBS–T–PS–DMSO.
ˇ´
¨ ¨ ¨
19. Lapcık, O.; Hampl, R.; Al-Maharik, N.; Salkla, A.; Wahala,
K.; Adlercreutz, H. Steroids 1997, 62, 315.
ˇ´
¨ ¨ ¨
20. Lapcık, O.; Hampl, R.; Hill, M.; Wahala, K.; Al Maharik, N.;
Cross-reactivity determinations
Adlercreutz, H. J. Steroid Biochem. Mol. Biol. 1998, 64, 261.
¨ ¨ ¨
21. Wahala, K.; Hase, T. J. Chem. Soc., Perkin Trans. 1991, 1,
Stock solutions (1 mg/mL) of the phyto-oestrogens listed in
Table 3 were prepared in ethanol. Seria dilutions, from
250 mg/mL to 0.0076 mg/mL (dilution stepˆ4), were
3005.
22. Cushman, M.; Nagarathnam, D. Tetrahedron Lett. 1990, 31,
6497.

´
C. Le Houerou et al. / Tetrahedron 56 (2000) 295–301
301
23. Schiemenz, G. P.; Schmidt, U. Liebigs Ann. Chem. 1982,
1509.
24. Pelter, A.; Ward, R. S.; Whalley, J. L. Synthesis 1998, 1793.
25. Erlanger, B. F.; Borek, F.; Beiser, S. M.; Lieberman, S. J. Biol.
Chem. 1957, 228, 713.
26. Riggles, B.; Dunbar, B. J. Agric. Food. Chem. 1990, 38, 1922.
27. Szurdoki, F.; Bekheit, H. K. M.; Marco, M.-P.; Goodrow,
M. H.; Hammock, B. D. J. Agric. Food. Chem. 1992, 40, 1459.
28. Setchell, K. D. R. Estrogens in the Environment II; McLachlan
Ed.; Elsevier: Amsterdam, 1985, p 69.