Intravenous injection of human sex steroid hormone-binding globulin in mouse decreases blood clearance rate and testicular accumulation of orally administered [2-125I]iodobisphenol A

Article abstract of DOI:10.1016/S0039-128X(02)00014-4

Bisphenol A, an environmental compound with estrogenic activity, has been shown to bind human sex steroid hormone-binding globulin (hSHBG), the main plasma transport protein which regulates the metabolism of androgens and estrogens and limits their access

Full text of DOI:10.1016/S0039-128X(02)00014-4

Steroids 67 (2002) 637–645
Intravenous injection of human sex steroid hormone-binding globulin in
mouse decreases blood clearance rate and testicular accumulation of
orally administered [2-125I]iodobisphenol A
Nadia Bendridia,b, Elisabeth Mappusb, Catherine Grenotb, Herve´ Lejeunea,c,
Claude Yves Cuilleronb, Michel Pugeata,b,*
aHospices Civils de Lyon, Laboratoire de la Fe´de´ration d’Endocrinologie, Hoˆpital de l’Antiquaille, Lyon, France
bINSERM U 329, Lyon, France
cINSERM U 418, Hoˆpital Debrousse, Lyon, France
Received 31 July 2001; received in revised form 14 December 2001; accepted 19 December 2001
Abstract
Bisphenol A, an environmental compound with estrogenic activity, has been shown to bind human sex steroid hormone-binding globulin
(hSHBG), the main plasma transport protein which regulates the metabolism of androgens and estrogens and limits their access to target
organs. The present study was conducted to determine whether physiologically relevant concentrations of hSHBG can influence the blood
clearance rate of bisphenol A and its accumulation in the testes. A radioactive [2-125I]iodobisphenol tracer was synthesized with an
association constant (Ka) for binding to hSHBG of 0.14 Ϯ 0.01 ϫ 106 MϪ1 at 37°C, a value much lower than for [2-125I]iodoestradiol,
which was also synthesized. We used i.v. injection of immunopurified hSHBG in adult male mice to maintain hSHBG levels within the
physiologically possible range for humans (27–267 nM) before gavage administration of [2-125I]iodobisphenol or [2-125I]iodoestradiol, for
measuring the blood clearance rate of radioactive signal in blood samples taken during the following 120 min. Testicular accumulation of
radioactivity was measured 24 h and 48 h after gavage of [2-125]iodobisphenol A. In mice receiving immunopurified hSHBG or vehicle,
the time-dependent blood clearance of radioactivity exhibited a bi-exponential decrease which indicated ␣-diffusion and ␤-elimination
phases for both radioactive ligands. The presence of circulating hSHBG significantly and dose-dependently lowered the clearance rate of
radioactivity. However, much higher circulating levels of hSHBG were required to retard the blood clearance of [2-125I]iodobisphenol A
as compared to those required for [2-125I]iodoestradiol, in keeping with the important difference in their respective Ka value for binding
to SHBG. In addition, mice treated with hSHBG exhibited significantly (P ϭ 0.036) reduced testicular accumulation of radioactivity 24 h
and 48 h after ingestion of [2-125I]iodobisphenol A. Provided that the binding properties of bisphenol A for hSHBG are not substantially
different from those measured for [2-125I]iodobisphenol A, these findings suggest that, although hSHBG binds 2-mono-iodobisphenol A
with a relatively low binding affinity, high enough concentrations of circulating hSHBG (range concentrations between 85 and 267 nM) are
potentially able to exert a protective effect against exposure to bisphenol A. © 2002 Elsevier Science Inc. All rights reserved.
Keywords: Endocrine disruptor; Bisphenol A; Iodinated bisphenol A derivative; Human SHBG; Steroid; Clearance
1. Introduction
to xenoestrogens has been postulated as playing a role in the
decline of testicular function and increase in genital tract
malformations, but experimental support is lacking.
Although the question is still debated, several studies
have claimed that sperm counts in human males have de-
creased by an average of 50% since 1940 [1,2]. In addition,
increased frequencies of cryptorchidism and hypospadia, as
well as of cancer of testis [3] have been reported. Exposure
Bisphenol A (4,4Ј-isopropylidenediphenol) is a mono-
mer widely employed in the manufacture of polycarbonate,
epoxy, and polyester–styrene resins. Trace levels of bisphe-
nol A leach from the linings of food cans and dental resins
[4,5]. High doses were reported to cause reproductive tox-
icity and affect cellular development in rats and mice [6,7].
Recently, low levels of bisphenol A administered orally (2
ng/g) to mice (CF-1) during pregnancy have been found to
* Corresponding author. Tel.: ϩ33-4-72-38-50-70; fax: ϩ33-4-72-25-
65-86.
E-mail address: michel.pugeat@chu-lyon.fr (M. Pugeat).
0039-128X/02/$ – see front matter © 2002 Elsevier Science Inc. All rights reserved.
PII: S0039-128X(02)00014-4

638
N. Bendridi et al. / Steroids 67 (2002) 637–645
reduce daily sperm production in male offspring and to
reduce the size of the accessory reproductive organs [8].
Estrogenic activity of bisphenol A has been demonstrated in
water-sterilized polycarbonate flasks [9]. In addition, bis-
phenol A has been shown to bind estrogen receptors [10], to
stimulate cell proliferation, and also to induce the synthesis
of progesterone receptor in estrogen-dependent MCF-7
breast cancer cell lines [9,11,12]. Since these initial find-
ings, estrogenic activity has been reported for many natural
environmental or synthetic nonsteroidal substances. Some
of these xenoestrogens are structurally similar to diethylstil-
bestrol and can be active as estrogen agonists or antagonists
[13]. More specifically, the binding of some chemical es-
trogens to estrogen receptor-␣ has been shown to be in-
volved in the activation of the uterine insulin-like growth
factor 1 signaling pathway [14]. In vivo experiments have
shown that bisphenol A-treated Fischer 344 rats have stim-
ulated proliferation of pituitary lactotroph cells with in-
creased expression and secretion of prolactin, these effects
being species-specific and not observed in Sprague Dawley
rats [15].
Danzo [16] reported that bisphenol A binds human sex
hormone-binding globulin (hSHBG) with a much lower
binding affinity than its natural ligands, testosterone and
17␤-estradiol. These findings were confirmed by our labo-
ratory [17]. It has been postulated that SHBG-binding might
be a limiting step for intracellular access and biological
activity of several xenoestrogens that specifically bind
SHBG [16,18]. However, most in vivo studies reporting the
effects of xenoestrogen exposure were performed in rodents
which have low, if any, circulating levels of SHBG during
adult life, although they do have intra-testicular androgen-
binding protein (ABP) [19,20]. Therefore, the question is
whether the main function of hSHBG, which is to regulate
the bioavailability of steroids to their target tissues [21,22],
might be relevant for low binding affinity ligands such as
bisphenol A in a natural endogenous steroid environment.
To elucidate this question, radioinert and radioactive iodin-
ated derivatives of bisphenol A (2-mono-iodobisphenol A
and 2,2Ј-di-iodobisphenol A) and estradiol (2-iodoestradiol
and 2,4-di-iodoestradiol) were prepared and employed to
determine the influence of iodination on the binding affinity
for SHBG. In a second step, [2-125]iodobisphenol A and
[2-125I]iodoestradiol were orally administered to adult mice
which were i.v. injected with immunopurified hSHBG or
vehicle, in order to study the effects of hSHBG-binding on
the time-dependent clearance rate of total blood radioactiv-
ity and testicular accumulation.
4-androsten-3-one) and estradiol (1,3,5(10)-estratriene-
3,17␤-diol) (Sigma Chemical Co, Saint-Louis, MO, USA)
were dissolved in ethanol prior to appropriate dilutions.
Bisphenol A (4,4Ј-isopropylidenediphenol) (purity Ͼ 99%)
was supplied by Fluka (France), Concanavalin A-Sepharose
(ConA-Sepharose) from Pharmacia Fine Chemicals AB
(Uppsala, Sweden), Cytoscintꢀ TM scintillation fluid from
ICN Pharmaceuticals (France), NaI from Riedel-De Ha¨en
(Hanover, Germany) and Chloramine T and sodium meta-
bisulfite from Aldrich-Chemie (Steinheim, Germany).
125NaI was purchased from ICN pharmaceutical (100 mCi/
ml). Radioactivity was counted in a ␥ counter (1900 model
or Cobra II auto-␥, Packard Instrument Co., Downers
Grove, IL, USA). NMR spectra were recorded at the Centre
Commun de RMN (ESCPE-Universite´ Claude Bernard,
1
Villeurbanne, France). H NMR spectra of iodinated estra-
diol and bisphenol A derivatives were recorded at 300 MHz
on a Bru¨ker (Wissenbourg, France) spectrometer equipped
with an Aspect 300 computer. Chemical shifts were mea-
sured relative to tetramethylsilane Si(CH). Samples were
prepared by dissolving 10 mg of steroid in 0.75 ml of
CD3OD (99.8%), purchased from CEA (Saclay, France).
1
The H chemical shifts were estimated to be accurate to Ϯ
0.01 ppm, and coupling constants to Ϯ 0.5 Hz.
2.2. Preparation of iodinated estradiol and bisphenol A
derivatives, 2-iodoestradiol and 2,4-di-iodoestradiol
A solution of estradiol (544 mg, 2.13 mmol) in 1,4-
dioxane (10 ml) was mixed with a solution of NaI (150 mg,
2 mmol), in 2 ml of 50 mM sodium phosphate buffer pH 7.6
(PBS), before adding a solution of Chloramine T (540 mg,
2.37 mmol) in the same buffer (5 ml). The reaction was
allowed to proceed at 20°C for 5 min with constant mixing,
before the addition of a solution of sodium metabisulfite
(1.9 g, 9.99 mmol) in the same buffer (5 ml). Distilled water
(20 ml) was added and the mixture was extracted twice with
ethyl acetate (50 ml). The pooled extracts were evaporated
to dryness and the residue dissolved in a small volume of
methanol/water (7:3 v/v) mixture. The reaction products
were separated by preparative high-performance liquid
chromatography (HPLC) on a 250 ϫ 10 mm ␮Bondapak
C18 reverse phase column (Waters RCM - 10 ␮M - 125 Å)
using methanol/water (7:3 v/v) mixture as eluent (elution
rate 8 ml/min with continuously monitored effluent at 280
nm). Separate peaks were collected and evaporated to dry-
ness. The reaction products were also separated by thin-
layer chromatography (TLC) on silica plates, using toluene/
ethyl acetate (3:1 v/v) as the developing mixture (Rf of
2-iodoestradiol: 0.44; Rf of 2,4-di-iodoestradiol: 0.60). The
corresponding [2-125I]iodoestradiol was prepared with
2. Experimental
[
125I]Na using the same procedure and separated by TLC
2.1. Reagents
using radioinert 2-iodoestradiol as reference compound. 1H
1
[1,2,6,7-3H]Testosterone (3.44 TBq/mmol) (Amersham
NMR 2-iodoestradiol: H NMR (CD3OD) ␦0.77 (3H, s,
France SA, Les Ulis, France), testosterone (17␤-hydroxy-
CH3-18), 3.65 (H, m, H-17), 6.53 (H, s, H-4), 7.50 (H, s,

N. Bendridi et al. / Steroids 67 (2002) 637–645
639
H-1). 1H NMR 2,4-di-iodoestradiol: 1H NMR (CD3OD)
␦0.87 (3H, s, CH-8), 3.65 (H, m, H-17), 7.60 (1H, s, H-1).
2-Mono-iodobisphenol A and 2,2Ј-di-iodobisphenol A
were prepared according to the above method. The crude
reaction products were dissolved in a minimum volume of
methanol and separated using TLC on silica gel with a
chloroform/ether (4:1 v/v) mixture (Rf of 2-mono-iodobis-
phenol A: 0.54; Rf of 2,2Ј-di-iodobisphenol A: 0.65). The
corresponding 125I-labeled derivatives were prepared with
a mixture of [3H]testosterone (50 000 cpm in 100 ␮l of
Tris-HCl buffer), and various final concentrations of the
radioinert ligand: 0–500 ␮M for bisphenol A, 2-mono-
iodobisphenol A, 2,2-di-iodobisphenol A, and 0–0.03 ␮M
for 17␤-estradiol, 2-iodoestradiol, 2,4-di-iodoestradiol or
testosterone, and incubated in a total volume of 2.3 ml of
Tris-HCl buffer, for 90 min at 37°C under shaking. After
centrifugation for 5 min at 3000 ϫ g, the pellets were
counted for radioactivity in CytoScintꢀ scintillation fluid in
a beta counter (Packard Instrument Co.). Specific binding,
in the presence of competitor, was expressed as a percentage
of the maximum binding that was calculated by subtracting
non-specific binding (estimated in the presence of 100-fold
excess of testosterone) from total binding. Relative binding
affinities (RBA) of ligands to hSHBG were calculated as
follows:
[
125I]Na and analyzed by TLC using radioinert 2-mono-
iodobisphenol A and 2,2Ј-di-iodobisphenol A as reference
compounds ([2-125I]iodobisphenol A (a.s.: 0.013 ϫ 1013
1
1
dpm/mmol)). H NMR bisphenol A: H NMR (CD3OD)
␦1.57 (6H, s, CH), 7.01 (4H, d: J ϭ 8.8 Hz, arom. H-meta),
6.66 (4H, d: J ϭ 8.8 Hz, arom. H-ortho). 1H NMR 2-mono-
1
iodobisphenol A: H NMR (CD3OD) ␦1.56 (6H, s, CH),
6.68 (2H, d: J ϭ 8.7 Hz, H-2Ј-6Ј), 6.70 (1H, d: J ϭ 8.4 Hz,
H-6), 7.00 (1H, dd: J ϭ 8.6 and 2.3 Hz, H-5), 7.01 (2H, d:
RBA ϭ [IC50, Testosterone/IC50, Competitor] ϫ 100.
1
J ϭ 8.8 Hz, H-3Ј-5Ј), 7.47 (1H, d: J ϭ 2.3 Hz, H-3). H
The IC50 value is the concentration of radioinert competitor
required to reduce [3H]testosterone specific binding by
50%. The RBA value for testosterone was arbitrarily set at
100%.
NMR 2,2Ј-di-iodobisphenol A: 1H NMR (CD3OD) ␦.55
(6H, s, CH), 6.72 (2H, d: J ϭ 8.5 Hz, H-6), 6.99 (2H, dd:
J ϭ 8.6 and 2.3 Hz, H-5), 7.49 (2H, d: J ϭ 2.4 Hz, H-3).
2.3. Purification and measurement of hSHBG
2.6. Influence of iodinated bisphenol A derivatives on
testosterone binding to immunopurified hSHBG
Purification of hSHBG was performed by immunoaffin-
ity chromatography of a pool of human serum using an
immobilized anti-hSHBG antibody, as previously described
[23]. The serum concentration of hSHBG was measured
using an immunoradiometric assay for hSHBG (125I-labeled
SHBG-Coatria, BioMe´rieux, Marcy l’Etoile, France).
Immunopurified hSHBG (100 nM in 100 ␮l of Tris-HCl
pH 7.6 buffer) was incubated with ConA-Sepharose (500 ␮l
of 50% diluted gel) for 30 min at 20°C, and the solid phase
was washed three times with 2 ml of Tris-HCl and centri-
fuged for 5 min at 3000 ϫ g. The pellets were incubated
with a mixture of [3H]testosterone (50 000 cpm in 100 ␮l of
Tris-HCl buffer), increasing radioinert testosterone for final
concentrations ranging from 0 to 30 nM, and fixed concen-
trations of bisphenol A (20 ␮M), 2-mono-iodobisphenol A
(20 and 25 ␮M) or 2,2Ј-di-iodobisphenol A (10 and 25 ␮M)
in a total volume of 2.5 ml of Tris-HCl buffer. After incu-
bation for 90 min at 37°C and centrifugation, the radioac-
tivity of pellets was counted. Changes in association con-
stant and/or in number of hSHBG binding sites for
testosterone were studied by comparing Scatchard plots
established in the presence or in the absence of bisphenol A
or iodinated bisphenol A derivatives.
2.4. Binding assays
2.4.1. Scatchard plot analysis
The respective association constants (Ka) of 2-mono-
iodobisphenol A and 2,2Ј-di-iodobisphenol A for binding to
immunopurified hSHBG were calculated by Scatchard plot
analysis using the corresponding radioiodinated tracers [24]
and a solid phase binding assay [25].
2.5. Relative binding assays of iodinated bisphenol A and
estradiol derivatives
The competition curves were obtained from binding as-
says, performed on immunopurified hSHBG incubated with
[3H]testosterone, in the presence of increasing concentra-
tions of each studied ligand (bisphenol A, iodinated bisphe-
nol A derivatives, 17␤-estradiol, iodinated estradiol deriv-
atives and testosterone), as previously described [25,26].
Briefly, 500 ␮l of 50% Con A-Sepharose in Tris-HCl buffer
were added to an immunopurified hSHBG solution (100 nM
in 100 ␮l of Tris-HCl buffer). After 30 min of incubation at
20°C, 2 ml of Tris-HCl was added. The pellets obtained
after centrifugation for 5 min at 3000 ϫ g were washed
three times with Tris-HCl buffer. Pellets were then added to
2.7. Animals, treatments, and sample collection
Adult male mice (Balb/C), 60–70 days old, with weights
ranging from 17 to 24 g (purchased from Elevage Scienti-
fique des Dombes, Chaˆtillon sur Chalaronne, France), were
housed individually at 20°C with alternating darkness and
artificial light cycles, with food and water available ad
libitum. In a first set of experiments, one group of male
adult Balb/C mice (n ϭ 6) was given a single oral bolus of
500 000 cpm of [2-125I]iodobisphenol A and a second group
(n ϭ 6) was given a single oral bolus of 500 000 cpm of

640
N. Bendridi et al. / Steroids 67 (2002) 637–645
[2-125I]iodoestradiol, in 0.1 ml of 5% glucose vehicle.
Blood samples were taken (200–300 ␮l) by perocular sinus
puncture at 5, 10, 15, 20, 25, 30, 40, 60, and 80 min after
gavage, placed in preweighted tubes containing heparin
(Sanofi-Winthrop, l5 I.U.), and counted for total radioactiv-
ity without any prior purification. Thus, the radioactivity
measured in the samples may correspond either to metabo-
lites or unmetabolised iodinated bisphenol A derivatives. In
a second set of experiments, two groups of adult male
Balb/C mice were treated by tail-vein injection of purified
hSHBG (9 ␮M) in 0.1 ml of PBS. One hour later, a blood
sample was taken to measure serum hSHBG concentration,
before oral gavage administration of [2-125I]iodobisphenol
A (n ϭ 7) or [2-125I]iodoestradiol (n ϭ 2). In a further set
of experiments, one group (n ϭ 18, n ϭ 3 per time tested,
6 times tested) of male adult Balb/C mice was injected with
purified hSHBG or with vehicle (n ϭ 18, n ϭ 3 per time
tested, 6 times tested) before receiving an oral bolus of
500 000 cpm of [2-125I]iodobisphenol A. The animals were
sacrificed by cardiac puncture for blood sampling at 1.5, 3,
5, 15, 24 and 48 h after gavage, and the tissue distribution
and accumulation patterns of radioactivity were determined
by immediate sampling of the main organs for weighing and
counting.
Fig. 1. Competition curves (B/B0 ϫ 100) of tritiated [3H]testosterone from
human SHBG by increasing concentrations of radioinert testosterone (●),
estradiol (Œ), 2-iodoestradiol (E), 2,4-di-iodoestradiol (ϫ), bisphenol A
(⌬), 2-mono-iodobisphenol A (▫), 2,2Ј-di-iodobisphenol A (᭛). Purified
human SHBG (100 nM in Tris-HCl buffer) was incubated with [3H]tes-
tosterone (50 000 cpm) in the presence of estradiol, iodinated estradiol,
bisphenol A and iodinated bisphenol A derivatives or unlabeled testoster-
one for 1.5 h at 37°C. Bound/free separation was achieved with Con-A
Sepharose. The data are the average Ϯ S.D. of 2–4 independent
experiments.
2.8. Statistics
0.01 ϫ 106 MϪ1 or 0.62 ϫ 106 Ϯ 0.07 M–1 at 37°C,
respectively. The differences between these two affinity
constants was significant (P Ͻ 0.0005, n ϭ 4).
Data are expressed as the mean Ϯ S.D. Multifactorial
analysis of variance (ANOVA) was performed for: (i) anal-
ysis of hSHBG binding affinity of iodinated bisphenol A
and estradiol derivatives; (ii) analysis of the relative binding
affinities (RBA) of ligands to hSHBG; (iii) assessment of
the effect of fixed concentrations of bisphenol A and iodin-
ated bisphenol A derivatives on testosterone-binding to hS-
HBG. Multifactorial analysis of variance (ANOVA) was per-
formed to assess the effects of hSHBG on the time-dependent
disappearance of [2-125I]iodoestradiol. The correlation be-
tween the time-dependent disappearance of [2-125I]iodobis-
phenol A and SHBG blood concentrations was established
using the least square method, which could be applied in
this case, owing to the higher number of experiments con-
ducted for [2-125I]iodobisphenol A as compared with
[2-125I]iodoestradiol. These two relationships were studied
at 40 min, 60 min and 80 min after oral administration. In
both cases, comparisons yielding more than 95% statistical
confidence (P Ͻ 0.05) were considered significant.
3.2. Competitive binding studies to hSHBG
The relative binding affinities (RBAs) of bisphenol A,
2-mono-iodobisphenol A, and 2,2Ј-di-iodobisphenol A for
human SHBG were determined from their ability to com-
pete with [3H]testosterone (Fig. 1). RBA was also measured
for estradiol, 2-iodoestradiol and 2,4-di-iodoestradiol. All of
the ligands tested were able to compete significantly (P Ͻ
0.001) with tritiated testosterone for binding to human
SHBG. Bisphenol A and the two iodinated bisphenol A
derivatives showed low binding affinities. The RBA values
(Table 1) were 0.20% Ϯ 0.01 for bisphenol A, 0.32% Ϯ
0.10 for 2-mono-iodobisphenol, 0.47% Ϯ 0.12 for 2,2Ј-di-
iodobisphenol A, 175.5% Ϯ 27.6 for 2-iodoestradiol, 23.8%
Ϯ 3.1 for 2,4-di-iodoestradiol and 58.9% Ϯ 1.6 for estra-
Table 1
Relative binding affinities for hSHBG
3. Results
Ligands
RBA (%)
Testosterone
Estradiol
2-Iodoestradiol
2,4-Di-iodoestradiol
Bisphenol A
100
3.1. Scatchard plot analysis of iodinated bisphenol A
derivatives
58.9 Ϯ 1.6*
175.5 Ϯ 27.6*
23.90 Ϯ 3.06*
0.20 Ϯ 0.01*
0.325 Ϯ 0.100*
0.48 Ϯ 0.12*
Scatchard plot analysis, using either [2-125I]iodobisphe-
nol A or [2,2Ј-125I]iodobisphenol A, showed a single pop-
ulation of hSHBG binding sites with Ka values of 0.14 Ϯ
2-Mono-iodobisphenol A
2,2Ј-Di-iodobisphenol A

N. Bendridi et al. / Steroids 67 (2002) 637–645
641
Table 2
Relative constant affinity for testosterone binding to hSHBG
Ligands
Ka of testosterone (37°C)
(X109) (MϪ1
)
Testosterone
Testosterone ϩ 20 ␮M of
bisphenol A
0.395 Ϯ 0.048
0.185 Ϯ 0.005*
Testosterone ϩ 20 ␮M of
2-mono-iodobisphenol A
Testosterone ϩ 25 ␮M of
2-mono-iodobisphenol A
Testosterone ϩ 10 ␮M of
2,2Ј-di-iodobisphenol A
Testosterone ϩ 25 ␮M of
2,2Ј-di-iodobisphenol A
0.107 Ϯ 0.016*
0.093 Ϯ 0.023*
0.073 Ϯ 0.026*
0.056 Ϯ 0.016*
Conversely, the differences in the inhibitory effect observed
with the three ligands, bisphenol A, 2-mono-iodobisphenol
A, and 2,2Ј-di-iodobisphenol A, were statistically signifi-
cant, as established for the comparison of the effects of a 20
␮M concentration of 2-mono-iodobisphenol A or bisphenol
A (P Ͻ 0.01) or of a 25 ␮M concentration of 2,2Ј-di-
iodobisphenol A and 2-mono-iodobisphenol A (P Ͻ 0.01).
Taken together, these results indicate that iodinated bisphe-
nol A derivatives and testosterone compete for the same
hSHBG steroid binding site.
Fig. 2. Scatchard plot analysis of testosterone binding in the presence or in
the absence of constant concentrations of radioinert bisphenol A or iodin-
ated bisphenol A derivatives. Purified hSHBG (100 nM in Tris-HCl buffer
pH 7.6) bound to Con-A was incubated for 1.5 h at 37°C, with [3H]tes-
tosterone in 100 ␮l of Tris-HCl buffer pH 7.6, with increasing amounts (0–
30 nM) of unlabeled testosterone (▫) in the absence or in the presence of
bisphenol A 20 ␮M (ꢁ), 2-mono-iodobisphenol A 20 ␮M (⌬) or 25 ␮M
(ϫ), and 2,2Ј-di-iodobisphenol A 10 ␮M (␱) or 25 ␮M (●) in 100 ␮l of the
same buffer. Separation of bound and free was achieved with Con-A
Sepharose. The data are on average Ϯ S.D. of 3 experiments.
3.4. Effects of hSHBG on blood clearance of [2-
125I]iodinated bisphenol A and estradiol derivatives
diol. The affinity ranking for the various tested ligands on
the basis of statistical differences in RBA value was 2-io-
doestradiol Ͼ testosterone Ͼ estradiol Ͼ 2,4-di-iodoestra-
diol Ͼ 2,2Ј-di-iodobisphenol A Ͼ 2-mono-iodobisphenol
A Ͼ bisphenol A.
Typical time-dependent clearance curves for radioactiv-
ity in the blood after oral administration of [2-125I]iodinated
bisphenol A or [2-125I]iodinated estradiol derivatives are
shown in Fig. 3 and Fig. 4. Radioactive material was de-
tectable in the blood shortly after oral administration. The
3.3. Influence of iodinated bisphenol A derivatives on
testosterone binding to purified hSHBG
Scatchard plots for testosterone binding to purified
hSHBG in the presence or the absence of iodinated bisphe-
nol A derivatives are shown in Fig. 2. The presence of
bisphenol A (20 ␮M), or 2-mono-iodobisphenol A (20 and
25 ␮M) or 2,2Ј-di-iodobisphenol A (10 and 25 ␮M) signif-
icantly decreased (P Ͻ 0.0001) the apparent Ka for testos-
terone binding to hSHBG, but with an essentially un-
changed number of binding sites (Table 2). This inhibitory
effect was more pronounced for 2,2Ј-di-iodobisphenol A
than for 2-mono-iodobisphenol A, a result in agreement
with the differences between their respective binding affin-
ities for hSHBG. Even the lowest concentration of 2,2Ј-di-
iodobisphenol A (10 ␮M) exerted a stronger inhibition than
the highest concentration of 2-mono-iodobisphenol A (25
␮M). However, the differences in inhibiting effects between
the two tested concentrations of each of these two com-
pounds did not reach statistical significance (P Ͼ 0.1).
Fig. 3. Time-dependent blood clearance of orally ingested [2-125I]io-
doestradiol in the absence of blood hSHBG (Ϫ Ϫ Ϫ) (the broken lines
represent the highest and the lowest values of radioactivity measured in the
absence of circulating hSHBG) or in the presence of 31 nM (⌬) and 38 nM
(ꢁ) of blood hSHBG concentrations. The maximum blood concentration of
[2-125I]iodoestradiol at t ϭ 20 min is taken as 100%.

642
N. Bendridi et al. / Steroids 67 (2002) 637–645
active material (at 60 min or 80 min after gavage) due to the
presence of hSHBG was significant (P ϭ 0.002) for both
iodinated derivative ligands. Regression coefficients for to-
tal radioactivity in blood sampling at 40, 60 and 80 min after
the oral administration of [2-125I]iodobisphenol, in the pres-
ence of various concentrations of hSHBG, were obtained
using the linear least square method: (i) r ϭ 0.972, P ϭ
0.0002, n ϭ 7 at 40 min; (ii) r ϭ 0.946, P ϭ 0.0177, n ϭ
5 at 60 min; (iii) r ϭ 0.919, P ϭ 0.0095, n ϭ 6 at 80 min.
These results indicate that the highest hSHBG blood con-
centrations corresponded to the lowest [2-125I]iodobisphe-
nol A blood clearance rates.
3.5. Effects of hSHBG on the distribution and organ
accumulation of orally ingested [2-125I]iodobisphenol A
in mice
Fig. 4. Time-dependent blood clearance of orally ingested [2-125I]iodobis-
phenol A in the absence of blood hSHBG (Ϫ Ϫ Ϫ) (the broken lines
represent the highest and the lowest values of radioactivity measured in the
absence of circulating hSHBG), or in the presence of 27 nM (␱), 85 nM
(⌬), 132 nM (᭛), 190 nM (ϫ) and 267 nM (▫) of blood hSHBG concen-
trations. The maximal blood concentration of [2-125I]iodobisphenol A at
t ϭ 20 min is taken as 100%.
[2-125I]Iodobisphenol A orally ingested by mice was
readily adsorbed from the gastrointestinal tract and distrib-
uted to organs. About 6% of the radioactivity was accumu-
lated in the liver during the first 3 h, but accumulation
reached 41% within 48 h. After 15 h, 31% of the radioac-
tivity was eliminated in the feces, which suggested entero-
hepatic circulation of the radioactive material. In addition,
only 3% of the radioactivity was detected in the urine after
5 h. The kinetics of accumulation of [2-125I]iodobisphenol
A in the testes and epididymes in the presence or the
absence of hSHBG were studied in three different experi-
ments. In the absence of injected hSHBG, a time-dependent
accumulation of [2-125I]iodobisphenol A in the testes and
epididymes was observed up to 24 h. At this point in time,
the percentages of the total orally administered radioactivity
recovered in the testes and in epididymes were 0.25% Ϯ
0.028 and 0.29% Ϯ 0.35, respectively (Table 3). In the
presence of i.v. hSHBG, although hSHBG concentration
decreased time-dependently (158 nM Ϯ 32 at 3 h, 109
nM Ϯ 2 at 15 h, 65 nM Ϯ 6 at 24 h and 55 nM Ϯ 21 at 48 h),
the accumulation of [2-125I]iodobisphenol A in testes and
epididymes, up to 24 h, was less than 0.06% in both cases.
A multifactorial analysis of variance taking into account the
differences in hSHBG concentration confirmed that the
presence of hSHBG significantly decreased (P ϭ 0.036) the
testicular accumulation of radioactive material after admin-
time-dependent decrease exhibited a three-phase profile cor-
responding to: (i) resorption (data not shown), with a radio-
active peak appearing in less than 20 min, (ii) ␣-diffusion
from the vascular compartment during the next 15 min, (iii)
␤-elimination (reflecting metabolic degradation, mostly in
liver, kidney and intestinal compartments) during the fol-
lowing 35 min, respectively. The presence of hSHBG at
concentrations ranging from 31 to 27 nM significantly re-
duced the time interval for ␣-diffusion and ␤-elimination
phases of total radioactivity in blood sampling after oral
administration of [2-125I]iodoestradiol, as compared to con-
trol animals receiving no hSHBG (Fig. 3). However, the
presence of 20 or 27 nM concentrations of hSHBG did not
cause any significant decrease in the ␣-diffusion or ␤-elim-
ination phases of [2-125I]iodobisphenol A derivatives. In the
latter case, much higher hSHBG concentrations, ranging
from 85 to 267 nM, were required to achieve a significant
increase in [2-125I]iodobisphenol A blood concentration
(Fig. 4). A multifactorial analysis of variance taking into
account the concentrations of hSHBG for the two radiola-
beled derivatives showed that the increase in blood radio-
Table 3
Percentage of ingested radioactivity measured in testes and epididymes
Time after ingestion of
Testes
Epididymes
[2-125I]iodobisphenol A (h)
Absence of
hSHBG
Presence of
hSHBG
Absence of
hSHBG
Presence of
hSHBG*
1.5
3
5
15
24
48
0.020 Ϯ 0.010
0.017 Ϯ 0.012
0.010 Ϯ 0.014
0.280 Ϯ 0.014
0.250 Ϯ 0.028
0.370 Ϯ 0.335
0.020 Ϯ 0.010
0.007 Ϯ 0.006*
0.010 Ϯ 0.000*
0.017 Ϯ 0.015*
0.053 Ϯ 0.040*
0.190 Ϯ 0.130*
0.050 Ϯ 0.014
0.030 Ϯ 0.000
0.015 Ϯ 0.021
0.330 Ϯ 0.000
0.290 Ϯ 0.350
0.285 Ϯ 0.370
0.063 Ϯ 0.042*
0.040 Ϯ 0.030*
0.043 Ϯ 0.029*
0.040 Ϯ 0.010*
0.057 Ϯ 0.060*
0.237 Ϯ 0.159*

N. Bendridi et al. / Steroids 67 (2002) 637–645
643
istration of an oral dose of radiolabeled [2-125I]iodobisphe-
nol A as compared to control mice, untreated with hSHBG.
cipitated 22 or 36% of total radioactivity. Although not
extensive, these results suggested that a significant propor-
tion of radioactive [2-125I]iodobisphenol A (and/or the me-
tabolites) may circulate in the blood in a partly hSHBG-
bound form. They also suggested that, even for low hSHBG
concentrations, hSHBG binding sites were not saturated by
mouse-endogenous sex steroid hormones [33], a finding in
agreement with the model of physiological plasma steroid-
transport developed by Dunn et al. [25] in humans.
4. Discussion
Human SHBG binds estradiol with twice as low a bind-
ing affinity as testosterone [27]. By this property, SHBG
influences the androgen-to-estrogen balance, at least in vitro
[28]. In screening the biochemical activities of xenoestro-
gens, several groups have reported that bisphenol A, as well
as some phytoestrogens, can interact specifically with hu-
man SHBG but with low binding affinity [16,17,29]. This
raises the general question of the effects of hSHBG on
metabolism, blood clearance and organ accumulation of low
affinity SHBG-ligands. The aim of the present study was to
use radioactive iodinated bisphenol A derivative and a
mouse model with various concentrations of hSHBG ob-
tained by i.v. tail injection of purified hSHBG with full
binding activity.
Because 2-iodoestradiol exhibited a much higher affinity
than estradiol, as expected from a previous study by Fern-
lund and Gershagen [30], it was important to investigate
whether iodinated derivatives of bisphenol A had increased
hSHBG binding affinity. In fact, [2,2Ј-125I]iodobisphenol A,
a derivative of bisphenol A symmetrically iodinated on
both phenolic rings, exhibited a slightly higher affinity for
hSHBG than [2-125I]iodobisphenol A. This suggested a lim-
ited effect of iodination, when a phenolic ring remains
unsubstituted as in [2-125I]iodobisphenol A. Moreover,
measured RBAs for bisphenol A, 2-mono-iodobisphenol A,
and 2,2Ј-di-iodobisphenol A, by using a tritiated testoster-
one tracer, indicated stepwise but only slight increased af-
finity for hSHBG. Therefore, the small increase in binding
affinity, resulting from 2-iodination of bisphenol A, is un-
likely to lead to overestimating the hSHBG binding effect
on clearance and testicular accumulation of the 2-mono-
iodobisphenol A tracer. Since mice do not secrete hepatic
SHBG, i.v. tail injection of immunopurified hSHBG that
conserved full binding activity [31] was performed to
achieve serum hSHBG concentrations that were between 27
and 267 nM - i.e. within the physiological range for normal
humans [25]. Although measured in a single experiment, the
half-life of hSHBG was found to be close to 20 h in mouse,
in agreement with the half-life value of hSHBG that we had
previously reported in rabbit [32]. This result indicates that
hSHBG concentration decreased by at least 50% during our
24–48 h experimental protocols, minimizing its potential
activity as compared to the normal human situation of
constant liver production of hSHBG that maintains a rela-
tively stable hSHBG concentration [27].
The measurement of blood-clearance after a bolus of
[2-125I]iodobisphenol A in mice suggests that radioactive
material follows a classic pathway of resorption, with ap-
pearance in the blood in the 20 min following oral admin-
istration, diffusion from the vascular compartment, and
␤-elimination, found mostly in liver, renal and intestinal
compartments. The rapid first-pass elimination by the liver
and the evidence for enterohepatic recirculation of radioac-
tive material are in agreement with the toxicokinetic prop-
erties recently reported for bisphenol A in DA/Han rats [34].
The presence of circulating levels of hSHBG in mice was
found to induce a significant increase in the blood retention
time of [2-125I]iodobisphenol A (Fig. 4) whereas a much
stronger effect was observed for [2-125I]iodoestradiol, as
expected from the higher hSHBG binding affinity for 2-io-
doestradiol than for 2-mono-iodobisphenol A. However, the
24 h testicular accumulation of 2-mono-iodobisphenol A,
that was 0.25% in mice with no circulating hSHBG, was no
longer detectable in the presence of high circulating con-
centrations of hSHBG. These results suggested that the low
hSHBG binding affinity for 2-mono-iodobisphenol A could
be overcome by high hSHBG concentrations so as to sig-
nificantly enhance blood retention time and prevent testic-
ular accumulation. One can speculate that, during prenatal
and prepubertal exposure to some environmental xenoestro-
gens [35], SHBG levels, which are much higher than dur-
ing adult life [36,37], may have the capacity to limit the
biologic effects of xenoestrogens that specifically bind
hSHBG. This protective effect of hSHBG was not antici-
pated from the low hSHBG binding affinity for bisphenol A.
It appears to be strong enough to counterbalance a potential
‘shuttle’ mechanism of testicular accumulation of bisphenol
A. Indeed, high testicular concentration of testosterone, one
of the best natural ligands for hSHBG, could have displaced
low-affinity hSHBG-bound molecules, facilitating their tes-
ticular accumulation. This mechanism may even be further
amplified by the presence of intra-testicular ABP/SHBG
[20]. The use of a transgenic mouse model that expresses
human hSHBG transgenes [38] and maintains SHBG con-
centrations in the physiological range for humans would
certainly be a powerful tool for assessing the influence of
hSHBG on bisphenol A metabolism, cellular disposal, and
testicular toxicity. It might also highlight the importance of
the first liver by-pass for drugs or xenobiotics interaction
with hSHBG during oral exposure.
Because of the small sample size, only total radioactivity
could be measured in blood samples. We found that, in mice
serum samples with hSHBG concentrations of 12 or 25 nM,
treatment with Con A-Sepharose, a lectin that has been used
to adsorb serum glycoproteins including hSHBG [25], pre-
In conclusion, the present data suggest that the presence
of circulating hSHBG in mice facilitates sequestration of

644
N. Bendridi et al. / Steroids 67 (2002) 637–645
chemicals and phytoestrogens with estrogen receptor. Endocrinology
1998;39:4252–63.
[11] Soto AM, Sonnenschein C, Chung KL, Fernandez MF, Olea N,
Serrano FO. The E-SCREEN assay as a tool to identify estrogens: an
update on estrogenic environmental pollutants. Environ Health Per-
spect 1995;3:113–22.
[12] Villalobos M, Olea N, Brotons JA, Olea-Serrano F, Ruiz de Almo-
dovar JM, Pedraza V. The E-screen assay: a comparison of different
MCF7 cell stocks. Environ Health Perspect 1995;3:844–50.
[13] Colborn T, vom Saal FS, Soto AM. Developmental effects of endo-
crine-disrupting chemicals in wildlife and humans. Environ Health
Perspect 1993;1:378–84.
[14] Klotz DM, Hewitt SC, Korach KS, Diaugustine RP. Activation of a
uterine insulin-like growth factor I signalling pathway by clinical, and
environmental estrogens: requirement of estrogen receptor-␣. Endo-
crinology 2000;41:3430–9.
bisphenol A (and/or metabolites) in blood, despite a low
binding affinity for hSHBG. The data suggest that SHBG-
binding could exert a preventive effect on testicular bisphe-
nol A accumulation. However, potential extrapolation from
animal models to humans may be hazardous because of the
great variability of endogenous sex steroid hormones [39]
and of liver metabolism [40]. Even so, however controver-
sially, it has been claimed that the deleterious effects of
exposure to bisphenol A could occur at extremely low
doses, at least in some mice strains [41]. Therefore, in a
human context, a potential protective effect of hSHBG
could become particularly relevant in low bisphenol A ex-
posure conditions - especially when hSHBG levels are very
high, as during the prepubertal period.
[15] Steinmetz, R, Brown NG, Allen DL, Bigsby RM, Ben-Jonathan N.
The environmental estrogen bisphenol A stimulates prolactin in vitro
and in vivo. Endocrinology 1997;38:1780–6.
[16] Danzo BJ. Environmental xenobiotics may disrupt endocrine function
by interfering with the binding of physiological ligands to steroid
receptors and binding protein. Environ Health Perspect 1997;5:294–
301.
[17] De´chaud H, Ravard C, Claustrat F, Brac de la Perrie`re A, Pugeat M.
Xenoestrogen interaction with human sex hormone-binding globulin
(hSHBG). Steroids 1999;64:328–34.
[18] Hodgert Jury HH, Zacharewski TR, Hammond GL. Interactions be-
tween human plasma sex hormone-binding globulin, and xenobiotic
ligands. J Steroid Biochem Mol Biol 2000;75:167–76.
[19] Gunsalus GL, Musto NA, Bardin CW. Immunoassay of androgen
binding protein in blood: a new approach for the study of the semi-
niferous tubule. Science 1978;65–6.
Acknowledgments
The authors are much indebted to Geoffrey L. Hammond
(London, Ontario, Canada) for numerous suggestions about
all aspects of this work, Henri De´chaud for his kind help in
binding assays, and Christine Baret for excellent technical
assistance.
This work was supported by grant N°EN98D08 from the
French ‘Ministe`re de l’Environnement et de l’Ame´nagement
du Territoire.’
[20] Joseph D. Structure, function and regulation of androgen-binding
protein/sex hormone-binding globulin. Vitamin Horm 1994;9:180–
97.
References
[21] Hammond GL. Molecular properties of corticosteroid-binding glob-
ulin and the sex-steroid binding proteins. Endocrine Rev 1990;1:65–
79.
[22] Rosner W. The functions of corticosteroid-binding globulin and sex
hormone-binding globulin: recent advances. Endocrine Rev 1990;1:
80–91.
[23] Grenot C, de Montard A, Blachere T, Rolland de Ravel M, Mappus
E, Cuilleron CY. Characterization of Met-139 as the photolabeled
amino acid residue in the steroid-binding site of sex hormone-binding
globulin using ⌬-6 derivatives of either testosterone or estradiol as
unsubstituted photoaffinity labeling reagents. Biochemistry 1992;1:
7609–21.
[24] Scatchard G. The attractions of proteins for small molecules and ions.
Ann NY Acad Sci 1949;1:660–72.
[25] Dunn JF, Nisula B, Rodbard D. Transport of steroid hormones:
binding of 21 endogenous steroids to both testosterone-binding glob-
ulin and corticosteroid-binding globulin in human plasma. J Clin
Endocrinol Metab 1981;3:58–68.
[26] Pugeat M, Dunn JF, Nisula BC. Transport of steroid hormones:
interaction of 70 drugs with testosterone-binding globulin and corti-
costeroid-binding globulin in human plasma. J Clin Endocrinol
Metab 1981;3:69–75.
[27] Rosner W. Plasma steroid-binding proteins. Endocrinol Metab Clin
North Am 1991;20:697–720.
[1] Carlsen A, Giwercman A, Keiding N, Skakkebaek NE. Evidence for
decreasing quality of semen during the past 50 years. Br Med J
1992;5:609–13.
[2] Sharpe RM. Declining sperm counts in men: is there an endocrine
cause? J Endocrinol 1993;36:357–60.
[3] Forman D, Moeller H. Testicular cancer. Cancer Surv 1994;9/20:
323–41.
[4] Brotons JA, Olea-Serrano MF, Villalobos M, Pedraza V, Olea N.
Xenoestrogens released from lacquer coatings in food cans. Environ
Health Perspect 1995;3:608–12.
[5] Olea N, Pulgar R, Perez P, Olea-Serrano F, Rivas A, Novillo-Fertrel
A, Pedraza V, Soto AM, Sonnenschein C. Estrogenicity of resin-
based composites and sealants used in dentistry. Environ Health
Perspect 1996;4(3):298–305.
[6] Reel JR, George JD, Lawton AD, Myers CB, Lamb JC. Bisphenol A:
reproduction and fertility assessment in CD-1 mice when adminis-
tered in feed. NTP/NIEHS Report 1985.
[7] Morrisey RE, George JD, Price CJ, Tyl RW, Marr MC, Kimmel CA.
The developmental toxicity of bisphenol A in rats and mice. Funda-
ment Appl Toxicol 1987:571–82.
[8] Vom Saal FS, Cooke PS, Buchanan DL, Palanza P, Thayer KA, Nag
SC, Parmigiani S, Welshons WV. A physiologically based approach
to the study of bisphenol A and other estrogenic chemicals on the size
of reproductive organs, daily sperm production and behavior. Toxicol
Ind Health 1998;14:239–60.
[28] Burke CW, Anderson DC. Sex hormone-binding-globulin is an oes-
trogen amplifier. Nature 1972;240:38–40.
[29] Kloas W, Schrag B, Ehnes C, Segner H. Binding of xenobiotics to
hepatic estrogen receptor, and plasma sex steroid-binding protein in
the teleost fish, the common carp (Cyprinus carpis). General and
Comparative Endocrinol 2000;119:287–99.
[9] Krishnan AV, Stathis P, Permuth SF, Tokes L, Feldman D. Bisphenol
A, an estrogenic substance, is released from polycarbonate flasks
during autoclaving. Endocrinology 1993;32:2279–86.
[10] Kuiper GG, Lemmen JG, Carlsson B, Corton JC, Safe SH, van der
Saag PT, van der Burg B, Gustafsson J. Interaction of estrogenic

N. Bendridi et al. / Steroids 67 (2002) 637–645
645
[30] Fernlund P, Gershagen S. 2-Iodoestradiol binds with high affinity to
human sex hormone-binding globulin SHBG. J Steroid Biochem
1990;36:75–81.
[31] Grenot C, de Montard A, Blachere T, Mappus E, Cuilleron CY.
Identification of a photolabeled site of the plasma binding protein for
testosterone and estradiol (SBP) using tritiated 17 ␤-hydroxy-4,6-
androstadien-3-one. C R Acad Sci III 1988;307:391–6.
[32] Cousin P, De´chaud H, Grenot C, Lejeune H, Pugeat M. Human
variant sex hormone-binding globulin (SHBG) with an additional
carbohydrate chain has a reduced clearance rate in rabbit. J Clin
Endocrinol Metab 1998;83:235–40.
[33] Jean-Faucher C, Berger M, de Turckheim M, Veyssiere G, Jean C.
Permanent changes in the functional development of accessory sex
organs and in fertility in male mice after neonatal exposure to cypro-
terone acetate. J Endocrinol 1985;104:113–20.
[34] Upmeier A, Degen GH, Diel P, Michna H, Bolt HM. Toxicokinetics
of bisphenol A in female DA/Han rats after a single i.v. and oral
administration. Arch Toxicol 2000;4(8):431–6.
[35] Atanassova N, McKinnell C, Turner KJ, Walker M, Fisher JS, Morley
M, Millar MR, Groome NP, Sharpe RM. Comparative effects of
neonatal exposure of male rats to potent, and weak (environmental)
estrogens on spermatogenesis at puberty, and the relationship to adult
testis size, and fertility: evidence for stimulatory effects of low es-
trogen levels. Endocrinology 2000;141:3898–907.
[36] Apter D, Bolton NJ, Hammond GL, Vihko R. Serum sex hormone-
binding globulin during puberty in girls and in different types of
adolescent menstrual cycles. Acta Endocrinol 1984;107:413–9.
[37] Lee IR, Lawder LE, Townend DC, Wetherall JD, Hahnel R. Plasma
sex hormone-binding globulin concentration and binding capacity in
children before and during puberty. Acta Endocrinol 1985;109:276–
80.
[38] Ja¨nne M, Deol HK, Power SG, Yee S, Hammond GL. Human sex
hormone-binding globulin gene expression in transgenic mice. Mol
Endocrinol 1998;12:23–136.
[39] Howdeshell KL, Hotchkiss AK, Thayer KA, Vandenberg JG, vom
Saal FS. Exposure to bisphenol A advances puberty. Science 1999;
1:763–4.
[40] Long X, Steinmetz R, Ben-Jonathan N, Caperell-Grant A, Young PC,
Nephew KP, Bigsby RM. Strain differences in vaginal responses to
the xenoestrogen bisphenol A. Environ Health Perspect
2000;8:243–7.
[41] Kaiser J. Panel cautiously confirms low-dose effects. Science 2000;
90:695–6.