Evaluation of chemiluminescent estradiol conjugates by using a surface plasmon resonance detector

Article abstract of DOI:10.1016/S0039-128X(00)00091-X

A series of chemiluminescent 17β-estradiol probes were synthesized. Relative equilibrium dissociation constants (K(D)) for the interaction of an anti-E₂ Fab fragment for the probes in solution were evaluated using a single E₂-analog biosensor surface on a BIAcore surface plasmon resonance instrument. The results show the antibody fragment binds all chemiluminescent conjugates tested with high affinity showing only minor preferences for site of substitution (C6 versus C7), stereochemistry (α versus β), or linker moiety. (C) 2000 Elsevier Science Inc.

Full text of DOI:10.1016/S0039-128X(00)00091-X

Steroids 65 (2000) 295–303
Evaluation of chemiluminescent estradiol conjugates by using a surface
plasmon resonance detector
Maciej Adamczyk*, Yon-Yih Chen, John C. Gebler, Donald D. Johnson,
Phillip G. Mattingly, Jeffrey A. Moore, Rajarathnam E. Reddy, Jiang Wu, Zhiguang Yu
Department of Chemistry, Diagnostics Division, Abbott Laboratories, 100 Abbott Park Road, Abbott Park, Illinois, USA
Received 18 October 1999; received in revised form 22 November 1999; accepted 7 December 1999
Abstract
A series of chemiluminescent 17␤-estradiol probes were synthesized. Relative equilibrium dissociation constants (KD) for the interaction
of an anti-E2 Fab fragment for the probes in solution were evaluated using a single E2-analog biosensor surface on a BIAcore surface
plasmon resonance instrument. The results show the antibody fragment binds all chemiluminescent conjugates tested with high affinity
showing only minor preferences for site of substitution (C6 versus C7), stereochemistry (␣ versus ␤), or linker moiety. © 2000 Elsevier
Science Inc. All rights reserved.
Keywords: Steroid; 17␤-Estradiol; Chemiluminescent probes; Anti-estradiol mAb; BIAcore; Affinity
1. Introduction
phobic steroidal backbone playing critical roles in assay
development.
The concentration of 17␤-estradiol (E2, 1) in serum is
often measured while addressing fertility problems and
monitoring follicle maturity during in vitro fertilization pro-
cedures [1]. Historically, the concentration of E2 in serum
has been measured by extracting the serum with an organic
solvent and subsequent quantification of the hormone using
radioimmunoassay. Depending on the specificity of the an-
tibody, chromatographic separation of cross-reacting estro-
gens may have been necessary. Such labor-intensive proce-
dures have now been largely supplanted by highly
automated nonisotopic immunoassays that use very specific
monoclonal antibodies (mAb) and suitably nonradiolabeled
estradiol conjugates (tracers) [2,3]. Often, specific anti-E2
antibodies are induced by immunogenic proteins conjugated
to commercially available 17␤-estradiol 6-carboxymethyl-
oxime [2,4]. The selection of appropriate E2 tracers to be
paired with a chosen antibody is more challenging with
factors such as the site of substitution, the stereochemistry at
the site of substitution, linker length and structure, and the
characteristics of the label being incorporated on the hydro-
Recently, we have prepared several E2 derivatives mod-
ified at the C6 [5–7], C7 [8], and C17 [9] positions that can
be conjugated to a label and used to evaluate such critical
factors in an antibody/tracer binding interaction. We now
describe the synthesis of novel chemiluminescent tracers
derivatized at the C6 and C7 position of E2 and studies
designed to evaluate the critical factors involved in an E2
tracer/anti-E2 monoclonal Fab fragment binding interaction.
2. Experimental
2.1. Materials and methods
* Corresponding author. Tel.: ϩ1-847-937-0225; fax: ϩ1-847-938-
8927.
All reagents for synthesis were purchased from Aldrich
Chemical Co. (Milwaukee, WI, USA), Sigma Chemical Co.
E-mail address: maciej.adamczyk@abbott.com (M. Adamczyk).
0039-128X/00/$ – see front matter © 2000 Elsevier Science Inc. All rights reserved.
PII: S0039-128X(00)00091-X

296
M. Adamczyk et al. / Steroids 65 (2000) 295–303
(St. Louis, MO, USA), or Steroloids, Inc. (Newport, RI,
USA) and used without purification, except where noted.
All solvents used were of HPLC grade from EM Science
(Gibbstown, NJ, USA) and used as received except CH2Cl2,
which was freshly distilled from CaH2 under nitrogen
before use. 9-[[(3-Carboxypropyl)[(4-ethylphenyl)sulfonyl]
amino]carbonyl]-10-(3-sulfopropyl)-acridinium inner salt
(2) [10] and the N10-(3-sulfopropyl)-N-sulfonylacridinium-
9-carboxamide conjugates 14 [6] and 15 [7] were prepared
as described previously. Preparative reversed-phase HPLC
was performed using a Waters (Millford, MA, USA)
␮Bondapak C18 10 ␮m column (40 ϫ 100 mm) eluting
under the conditions reported. Analytical RP-HPLC was
performed using a Waters ␮Bondapak C18 10-␮m column
(8 ϫ 100 mm) unless otherwise stated. Proton NMR (300
MHz), 13C NMR (75 MHz), and 19F NMR (282 MHz)
spectra were recorded on a Varian Gemini 2300 spectrom-
eter (Palo Alto, CA). Chemical shifts (␦) are reported in
parts per million (ppm) relative to tetramethylsilane (TMS,
1H), residual solvent (13C), or ␣,␣,␣-trifluorotoluene (19F).
Coupling constants (J) are reported in Hz. Electrospray
ionization mass spectra (ESI-MS) were obtained on a Per-
kin-Elmer/Sciex (Foster City, CA, USA) API 100 Benchtop
system using the Turbo IonSpray ion source. HRMS and
FAB mass spectra were obtained on a Nermang 3010
MS-50 or a JEOL SX102-A mass spectrometer. IUPAC
names of all new compounds were obtained using the ACD/
ILab Web service version 3.5 at http://www.acdlabs.com/
ilab. Surface plasmon resonance (SPR) measurements were
carried out on a BIAcore 2000 (BIAcore, Inc., Piscataway,
NJ, USA) automated system at 25°C using CM-5 four-
channel sensor chips. Reagents for the BIAcore instrument
consisted of HBS buffer [10 mM HEPES (pH 7.4), 150 mM
NaCl, 3.4 mM EDTA, and 0.05% surfactant P-20] and a
coupling kit containing N-hydroxysuccinimide (NHS),
N-ethyl-N-(3-dimethylaminopropyl)-carbodiimide (EDC),
and ethanolamine hydrochloride all purchased from BIA-
core, Inc.
2.3. 3-[9-({{4-[(2,5-dioxo-1-pyrrolidinyl)oxy]-4-
oxobutyl}[(4-methylphenyl)sulfonyl]amino}carbonyl)-10-
acridiniumyl]-1-propanesulfonate (4)
Acridinium acid 2 [10] (8.0 g, 13.68 mmol) was sus-
pended in DMF (80 ml) and pyridine (11 ml). A solution of
TFA-NHS (3, 29.0 g, 137.38 mmol) in DMF (80 ml) was
added. All the solids gradually dissolved after which a
yellow precipitate formed. After stirring the mixture for
24 h in the dark, the precipitate was collected by filtration
and washed sequentially with ethyl acetate (200 ml), cold
ethanol (60 ml) and again with ethyl acetate (2 ϫ 200 ml).
Drying in vacuo afforded the desired active ester 4 (9.5 g,
99%) as a yellow solid. Analytical RP-HPLC [MeCN/0.1%
aq formic acid (30:70), 1 ml/min, 254 nm, ␮Bondapak C18
1
10 ␮m column (4.6 ϫ 300 mm)]: Rt 4.34 min, 97%. H
NMR (CDCl3): ␦ 9.05–9.00 (m, 2 H), 8.52–8.40 (m, 2 H),
8.19 (d, 1 H, J ϭ 8.4), 8.50–7.80 (m, 4 H), 7.67 (d, 1 H, J ϭ
8.5), 7.22–7.10 (m, 2 H), 5.63 (m, 2 H), 4.42–4.26 (m, 2 H),
3.01 (t, 2 H, J ϭ 7.0), 2.87 (s, 4 H), 2.32 (s, 3 H), 1.80–1.60
(m, 2 H). ESI-MS: m/z 683 (M ϩ H)ϩ. Anal. calcd. for
C32H31N3O10S2: Theory: C, 56.38; H, 4.58; N, 6.16; S,
9.41. Found: C, 56.15; H, 4.70; N, 6.24; S, 9.31.
2.4. N-(2-aminoethyl)-2-({[(8R,9R,13S,14R,17S)-3,17-
dihydroxy-13-methyl-7,8,9,11,12,13,14,15,16,17-
decahydro-6H}-cyclopenta[a]phenanthren-6-
ylidene]amino}oxy)acetamide (7)
NHS (86 mg, 0.75 mmol) and EDC (72 mg, 0.38 mmol)
were added sequentially to a solution of 17␤-estradiol 6-car-
boxymethyloxime (5, 135 mg, 0.38 mmol) dissolved in dry
DMF (1.9 ml) under nitrogen. The mixture was stirred for
16 h, and the resulting crude succinimidyl active ester so-
lution was treated with triethylamine (52 ␮l, 0.38 mmol)
and tert-butyl N-(2-aminoethyl)carbamate (60 mg, 0.38
mmol) under nitrogen. The reaction mixture was stirred an
additional 16 h and then filtered through a pipette fitted with
a cotton plug. The crude product was purified by preparative
RP-HPLC [MeOH/0.5% aq acetic acid (65:35), 45 ml/min,
254 nm] to afford the Boc-protected derivative 6 (125 mg,
65%). Analytical RP-HPLC [MeOH/0.1% aq acetic acid
2.2. N-trifluoroacetoxysuccinimide (3) (TFA-NHS) [11]
1
(65:35), 1.0 ml/min, 225 nm]: Rt 7.46 min, 98%. H NMR
Trifluoroacetic anhydride (150 ml, 1.06 mol) was cooled
to 0°C in an ice bath under nitrogen. NHS (81.5 g, 0.71 mol)
was added in one portion. The ice bath was removed and
stirring was continued for 18 h. Excess solvent was removed
in vacuo. The residue was dissolved in toluene (100 ml) and
remaining amounts of trifluoroacetic anhydride/trifluoro-
acetic acid were coevaporated on a rotary evaporator. The
process was repeated twice with toluene (2 ϫ 100 ml) and
three times with CH2Cl2 (3 ϫ 100 ml) providing 3 as a
(CD3OD ϩ 3 drops of CDCl3): ␦ 7.58 (br s, 1 H), 7.36 (d,
1 H, J ϭ 2.7), 7.18 (d, 1 H, J ϭ 8.6), 6.84 (dd, 1 H, J ϭ 8.6,
2.7), 6.24 (br s, 1 H), 3.50–2.96 (m, 6 H), 2.39–1.10 (m, 20
H), 1.40 (s, 9 H), 0.76 (s, 3 H). MS (FAB): m/z 524 (M ϩ
Na)ϩ.
The Boc-protected compound 6 (120 mg, 0.23 mmol)
was dissolved in a mixture of CH2Cl2 and trifluoroacetic
acid (1:1 ratio, 4.0 ml) and stirred for 30 min. The solvent
was removed in vacuo to afford the amine 7 (115 mg, 95%)
as the TFA salt. Analytical RP-HPLC [MeCN/0.1% aq TFA
1
white solid (148 g, 99%). mp 73–75°C (lit. oil) [11]. H
NMR (CDCl3): ␦ 2.94 (s, 4 H). 13C NMR (CDCl3): ␦ 167.3,
153.6 (q, JCF ϭ 45.9), 113.8 (q, JCF ϭ 286.0), 25.4. 19F
NMR (CDCl3): ␦ –9.47.
1
(20:80), 2.0 ml/min, 225 nm]: Rt 10.68 min, 99.6%. H
NMR (CD3OD): ␦ 7.30 (d, 1 H, J ϭ 2.7), 7.20 (d, 1 H, J ϭ
8.7), 6.80 (dd, 1 H, J ϭ 8.7, 2.7), 4.63 (s, 2 H), 3.67 (t, 1 H,

M. Adamczyk et al. / Steroids 65 (2000) 295–303
297
J ϭ 8.4), 3.54 (t, 2 H, J ϭ 5.7), 3.07 (t, 2 H, J ϭ 6.0),
2.38–1.22 (m, 13), 0.76 (s, 3 H). ESI-MS: m/z 402 (M ϩ
H)ϩ, 424 (M ϩ Na)ϩ.
2.7. 3-[9-({(4-{[(6S,8R,9R,13S,14R,17S]-3-hydroxy-13-
methyl-17-acetoxy-8,9,11,12,13,14,15,16,17-decahydro-
6H-cyclopenta[a]phenanthren-6-yl]amino}-4-oxobutyl)[(4-
methylphenyl)sulfonyl]amino}carbonyl)-10-acridiniumyl]-
1-propanesulfonate (12)
2.5. 3-[9-({{4-[(2-{[2-({[(8R,9R,13S,14R,17S]-3,17-
dihydroxy-13-methyl-7,8,9,11,12,13,14,15,16,17-
decahydro-6H-cyclopenta[a]phenanthren-6-
ylidene]amino}oxy)acetyl]amino}ethyl)amino]-4-
oxobutyl}[(4-methylphenyl)sulfonyl]amino}carbonyl)-10-
acridiniumyl]-1-propanesulfonate (8)
To a solution of 6-amino-17␤-estradiol acetate (11,
6␣/␤, 88:12) [12] (33 mg, 0.1 mmol) in DMF (0.5 ml) under
nitrogen were sequentially added triethylamine (84 ␮l, 0.6
mmol) and acridinium active ester 4 (68 mg, 0.1 mmol). The
reaction mixture was stirred for 20 h, and the solvent was
then removed in vacuo. The resulting residue was purified
by preparative RP-HPLC [MeCN/0.1% aq TFA (50:50),
45.0 ml/min, 225 nm] to afford compound 12 (41 mg, 48%)
as a yellow powder. Analytical RP-HPLC [MeCN/0.1% aq
TFA (50:50), 2.0 ml/min, 225 nm]: Rt 5.03 min, 98.4%. 1H
NMR (CD3OD): ␦ 8.96–8.82 (m, 2 H), 8.46–7.62 (m, 8 H),
7.22–7.02 (m, 3 H), 6.71–6.56 (m, 2 H), 5.76–5.62 (m, 2
H), 5.24–5.18 (m, 1 H), 4.65 (t, 1 H, J ϭ 7.0), 4.31 (t, 2 H,
J ϭ 7.5), 3.29–3.10 (m, 2 H), 2.70–1.20 (m, 19 H), 2.37 (s,
3 H), 2.03 (s, 3 H), 0.89 and 0.87 (2 s, 3 H). ESI-MS: m/z
897 (M ϩ H)ϩ, 918 (M ϩ Na)ϩ.
To a solution of amine-TFA salt 7 (19.8 mg, 36 ␮mol) in
MeCN (1.0 ml) under nitrogen were added acridinium ac-
tive ester 4 (25 mg, 36 ␮mol) and N,N-diisopropylethyl-
amine (30 ␮l, 200 ␮mol). The flask was covered with
aluminum foil, and the mixture was stirred for 18 h. The
reaction mixture was subsequently diluted with MeCN/
0.05% aq TFA (35:65, 1.0 ml) and purified by preparative
RP-HPLC [MeCN/0.05% aq TFA (35:65), 45 ml/min, 240
nm] to afford compound 8 (10 mg, 29%) as a yellow
powder. Analytical RP-HPLC [MeCN/0.1% aq TFA (35:
65), 2.0 ml/min, 240 nm]: Rt 3.9 min, 98%. 1H NMR
(CD3CN ϩ CF3CO2D): ␦ 8.72 (d, 2 H, J ϭ 9.5), 8.36–8.30
(m, 2 H), 8.12 (d, 50/100 H, J ϭ 8.3), 7.91–7.98 (m, 2 H),
7.77–7.71 (m, 2 H), 7.58 (d, 50/100 H, J ϭ 8.3), 7.27 (s, 1
H), 7.17 (d, 1 H, J ϭ 8.2), 7.04 (s, 3 H), 6.86–6.79 (m, 1
H), 5.64–5.54 (m, 2 H), 4.64 (s, 2 H), 4.20–4.15 (m, 2 H),
3.80–2.80 (m, 9 H), 2.65–1.14 (m, 17 H), 2.54 and 2.30 (2
s, 3H), 0.67 (s, 3 H). ESI-MS: m/z 968 (M ϩ H)ϩ.
2.8. 3-[9-({(4-{[(6S,8R,9R,13S,14R,17S]-3,17-dihydroxy-
13-methyl-8,9,11,12,13,14,15,16,17-decahydro-6H-
cyclopenta[a]phenanthren-6-yl]amino}-4-oxobutyl)[(4-
methylphenyl)sulfonyl]amino}carbonyl)-10-acridiniumyl]-
1-propanesulfonate (13)
Compound 12 (38 mg, 42 ␮mol) was dissolved in meth-
anol (20 ml). 1 N HCl (20 ml) was added, and the reaction
mixture was stirred at 50°C for 15 h. The reaction mixture
was concentrated to dryness in vacuo, and the crude product
was purified by preparative RP-HPLC [MeCN/0.1% aq
TFA (35:65), 45.0 ml/min, 225 nm] to afford compound 13
(26 mg, 74%) as a yellow powder. Analytical HPLC
[MeCN/0.1% aq TFA (35:65), 2.0 ml/min, 225 nm]: Rt 8.59
2.6. 3-[9-({{4-[(6-{[(6S,8R,9R,13S,14R,17S]-3,17-
dihydroxy-13-methyl-7,8,9,11,12,13,14,15,16,17-
decahydro-6H-cyclopenta[a]phenanthren-6-
yl]amino}-6-oxohexyl)amino]-4-oxobutyl}[(4-
methylphenyl)sulfonyl]amino}carbonyl)-10-acridiniumyl]-
1-propanesulfonate (10)
1
min, Ͼ 99%. H NMR (CD3OD): ␦ 8.96–8.84 (m, 2 H),
8.48–7.62 (m, 8 H), 7.22–7.02 (m, 3 H), 6.71–6.56 (m, 2
H), 5.76–5.62 (m, 2 H), 5.22–5.18 (m, 1 H), 4.31 (t, 2 H,
J ϭ 7.5), 3.64 (t, 1 H, J ϭ 8.7), 3.29–3.10 (m, 2 H),
2.70–1.20 (m, 19 H), 2.37 (s, 3 H), 0.81 and 0.79 (2 s, 3 H).
ESI-MS: m/z 855 (M ϩ H)ϩ, 876 (M ϩ Na)ϩ.
To a solution of aminoestradiol derivative 9 [5] (12 mg,
30 ␮mol) in DMF (300 ␮l) under nitrogen were sequentially
added triethylamine (42 ␮l, 300 ␮mol) and acridinium ac-
tive ester 4 (20 mg, 30 ␮mol). After stirring the reaction
mixture for 22 h, the solvent was removed in vacuo. The
resulting residue was purified by preparative RP-HPLC
[MeCN/0.1% aq TFA (32:68), 45.0 ml/min, 225 nm] to
afford compound 10 (5.4 mg, 20%) as a pale yellow pow-
der. Analytical RP-HPLC [MeCN/0.1% aq TFA (35:65),
2.0 ml/min, 225 nm]: Rt 8.21 min, Ͼ 99%. 1H NMR
(CD3OD): ␦ 8.98–8.91 (m, 2 H), 8.47–7.62 (m, 8 H),
7.22–7.08 (m, 3 H), 6.68–6.56 (m, 2 H), 5.78–5.68 (m, 2
H), 5.14–5.08 (m, 1 H), 4.30–4.22 (m, 2 H), 3.60 (t, 1 H,
J ϭ 8.1), 3.29–3.10 (m, 4 H), 2.70–1.20 (m, 27 H), 2.37 (s,
3 H), 0.78 and 0.67 (two s, 3 H). ESI-MS: m/z 968 (M ϩ
H)ϩ.
2.9. 3-[9-({{[4-N-(2-aminoethyl)aza]-4-oxobutyl}[(4-
methylphenyl)sulfonyl]amino}carbonyl)-10-acridiniumyl]-
1-propanesulfonate (16)
Ethylenediamine (1 ml, 15 mmol) was added to a solu-
tion of acridinium active ester 4 (1.02 g, 1.5 mmol) in DMF
(10 ml) under nitrogen, and the reaction mixture was stirred
in the dark for 1 h. The mixture was then filtered, and the
filtrate was evaporated in vacuo. The residue was purified
by preparative RP-HPLC [step gradient using MeCN:0.1%
aqueous formic acid (initial 5:95, then at 8 min 15:85, at 10
min 20:80), 45 ml/min, 225 nm] to afford compound 16

298
M. Adamczyk et al. / Steroids 65 (2000) 295–303
(530 mg, 52%) as a yellow solid. Analytical RP-HPLC
[MeCN/0.1% aq formic acid (25:75), 2 ml/min, 225 nm]: Rt
4.18 min, 98%. 1H NMR (DMSO-d6, CF3CO2D): ␦ 9.00 (m,
2 H), 8.47–8.37 (m, 2 H), 8.15 (d, 1 H, J ϭ 8.4), 8.07–7.96
(m, 2 H), 7.85–7.72 (m, 1 H), 7.68 (d, 1 H, J ϭ 8.2), 7.61
(d, 1 H, J ϭ 8.7), 7.16–7.06 (m, 2 H), 5.67 (m, 2 H),
4.23–3.84 (m, 2 H), 3.40–1.13 (m, 17 H). ESI-MS: m/z 627
(M ϩ H)ϩ.
H, J ϭ 8.4, 2.7), 6.46 (d, 1 H, J ϭ 2.4), 6.66 (t, 1 H, J ϭ
8.1), 3.39 (t, 2 H, J ϭ 6.0), 2.98 (t, 2 H, J ϭ 6.0),
2.90–2.68 (m, 2 H), 2.36–0.98 (m, 18 H), 0.78 (s, 3 H).
13C NMR (CD3OD): ␦ 177.7, 156.2, 137.6, 131.9, 128.1,
117.1, 114.1, 82.7, 47.8, 44.6, 43.6, 40.8, 39.7, 38.2,
38.1, 37.2, 35.6, 34.6, 30.7, 28.6, 26.4, 25.3, 23.5, 11.6.
ESI-MS: m/z 401 (M ϩ H)ϩ, 801 (2 M ϩ H)ϩ.
2.12. Preparation of a 17␤-estradiol analog biosensor
2.10. 3-[9-({(4-{ [2-({4-[(7R,8R,9R,13S,14R,17S]-3,17-
dihydroxy-13-methyl-7,8,9,11,12,13,14,15,16,17-
decahydro-6H}-cyclopenta[a]phenanthren-7-
yl]butanoyl}amino)ethyl]amino}-4-oxobutyl)[(4-
methylphenyl)sulfonyl]amino}carbonyl)-10-acridiniumyl]-
1-propanesulfonate (18)
surface
A biosensor surface for determination of free anti-estra-
diol Fab fragment concentration in solution was prepared by
amine coupling of estradiol analog 20 to an activated CM-5
sensor chip. Specifically, a continuous flow of HBS buffer
at 10 ␮l/min was initiated over the surface of a CM-5 sensor
chip. The carboxymethylated dextran matrix on the sensor
surface was activated by a 5-min injection of a solution of
0.2 M EDC and 0.05 M NHS. A solution of 20 (50 ␮M) and
ethanolamine (950 ␮M; 1 mM total amine in 100 mM
borate buffer, pH 8.3) was then injected (6 min), followed
by a 7-min injection of 1.0 M ethanolamine hydrochloride
(pH 8.5) to consume the remaining unreacted active ester
groups. A blank surface was generated under identical con-
ditions without 20 in the coupling solution.
To a solution of 7␣-(3Ј-carboxypropyl)-17␤-estradiol
(17 [8], 10 mg, 28 ␮mol) in DMF (1.0 ml) under nitrogen
were sequentially added acridinium amine 16 (22 mg, 33.5
␮mol), 1,3-dicyclohexylcarbodiimide (8.6 mg, 41.9 ␮mol),
1-hydroxybenzotriazole (57 mg, 41.9 ␮mol), and triethyl-
amine (20 ␮l, 140 ␮mol). The resulting mixture was stirred
for 17 h, and the solvent was removed in vacuo. The residue
was purified by preparative RP-HPLC [MeCN/0.1% TFA
(30:70), 45 ml/min, 225 nm] to afford compound 18 (15.6
mg, 57%) as a pale yellow powder. Analytical RP-HPLC
[MeCN/0.1% aq TFA (30:70), 2.0 ml/min, 225 nm]: Rt
13.20 min, Ͼ 99%. 1H NMR (CD3OD): ␦ 8.93 (d, 2 H, J ϭ
9.6), 8.47–8.36 (m, 2 H), 8.20 (d, 50/100 H, J ϭ 8.7), 8.06
(d, 2 H, J ϭ 8.4), 7.95–7.76 (m, 2 H), 7.63 (d, 50/100 H, J ϭ
8.1), 7.22–7.02 (m, 3 H), 6.54–6.38 (m, 2 H), 5.78–5.64
(m, 2 H), 4.26 (t, 2 H, J ϭ 7.8), 3.68–3.18 (m, 9 H),
2.80–0.95 (m, 24 H), 2.37 (s, 3 H), 0.77 and 0.72 (2 s, 3 H).
ESI-MS: m/z 968 (M ϩ H)ϩ.
2.13. Preparation of anti-estradiol mAb Fab fragment
Anti-estradiol mAb produced from an Armenian hamster
cell line using estradiol-6-CMO-thyroglobulin as immuno-
gen was obtained from the Abbott Laboratories cell culture
facility and digested to obtain anti-E2 Fab fragment. Spe-
cifically, papain (1 mg/ml) was activated by incubation at
37°C for 10 min in digestion buffer (50 mM sodium phos-
phate, 10 mM cysteine, pH 7.0). Anti-estradiol mAb (1.7
mg) in digestion buffer was combined with the activated
papain (1.7 ␮g) and then incubated at 37°C for 75 min. The
resulting papain-digested anti-estradiol mAb fragments
were purified by gel permeation chromatography on a 7.8 ϫ
300 mm Bio-Select SEC 125–5 column (Bio-Rad Labora-
tories, Hercules, CA, USA) eluting with 100 mM sodium
phosphate, 100 mM NaCl, pH 7.0 at a flow rate of 0.8
ml/min. The concentration of the anti-E2 Fab fragment was
0.16 mg/ml as determined by the micro-BCA method using
mouse IgG (FabЈ)2 as the standard. Purified anti-E2 Fab
fragment was characterized by SDS-PAGE (Ͼ95% purity)
and LC/ESI mass spectrometry (MW ϭ 47 977). All studies
described were conducted with the Fab fragment of anti-E2
mAb to eliminate the complexity associated with bivalency
of the mAb.
2.11. N-(2-aminoethyl)-4-[(7R,8R,9R,13S,14R,17S)-3,17-
dihydroxy-13-methyl-7,8,9,11,12,13,14,15,16,17-
decahydro-6H-cyclopenta[a]phenanthren-7-yl]butanamide
(20)
Ethylenediamine (0.5 ml, 8.3 mmol) was added to a
solution of 3,17␤-bis(2-trimethylsilyl)ethoxymethyl-7␣-
(3Ј-carbomethoxypropyl)estradiol (19 [8], 132 mg, 0.20
mmol) in MeOH (0.75 ml) under nitrogen. The mixture
was stirred for 24 h, and additional ethylenediamine (0.5
ml, 8.3 mmol) was added. After stirring the mixture for
29 h it was concentrated in vacuo. The residue was
redissolved in MeCN (7 ml) and 48% HF (0.13 ml) was
added. The mixture was stirred for 4 h and concentrated
in vacuo. The residue was redissolved in MeCN/0.1%
aqueous TFA (40:60, 10 ml), filtered, and purified by
preparative RP-HPLC [MeCN/0.1% aq TFA (20:80), 45
ml/min, 225 nm] to afford amine 20 (29 mg, 36%) as a
colorless powder. Analytical RP-HPLC [MeCN/0.1% aq
TFA (20:80), 2.0 ml/min, 225 nm]: Rt 10.74 min, 99%.
1H NMR (CD3OD): ␦ 7.08 (d, 1 H, J ϭ 8.7), 6.54 (dd, 1
2.14. Determination of relative KD
2.14.1. Generation of a calibration curve
Six concentrations [0, 31.25, 62.5, 125, 250, and 500
pM] of anti-E2 Fab fragment were serially diluted into HBS

M. Adamczyk et al. / Steroids 65 (2000) 295–303
299
Scheme 1.
buffer and injected individually (250 ␮l) over the 17␤-
estradiol analog biosensor surface at 30 ␮l/min. Buffer
alone was then injected for an additional 3 min. The bio-
sensor surface was regenerated after each run with a 6 M
guanidine hydrochloride pulse (1 min). The resulting bind-
ing response obtained for each anti-E2 Fab fragment con-
centration examined was measured over a 30-s window
beginning immediately after the sample plug had been re-
placed with HBS buffer flow (General fit-average; BIA-
evaluation 3.0). A calibration curve was obtained from the
4-parameter logistic fit of a plot of measured binding re-
sponses versus concentration of added anti-E2 Fab frag-
ment.
2.14.2. Determination of solution affinities of the
acridinium-17␤-estradiol conjugates
Ten to twelve concentrations of each estradiol conjugate
were diluted serially into HBS buffer and mixed with 500
pM anti-E2 Fab fragment. The solutions were allowed to
equilibrate (Ն4 h) and then injected over the calibrated
E2-analog biosensor surface at 30 ␮l/min as above. The
binding response was measured, and the concentration of

300
M. Adamczyk et al. / Steroids 65 (2000) 295–303
Scheme 2.
free anti-E2 Fab fragment in solution was calculated from
the calibration curve. A plot of free anti-E2 Fab fragment in
solution versus concentration of added estradiol conjugate
provided competition binding curves. Relative equilibrium
dissociation constants (KD) were obtained by nonlinear re-
gression analysis of the data using the solution affinity
model built into BIAevaluation 3.0 software (BIAcore, Inc.).
amino-17␤-estradiol acetate (11) [12] were coupled with the
active ester 4 in a similar fashion to give conjugates 10 and
12, respectively. Preparative RP-HPLC purification of com-
pound 12 and subsequent methanolysis of the acetate group
in 12 provided the desired stereochemically pure chemilu-
minescent conjugate 13. Chemiluminescent 17␤-estradiol
conjugates 14 and 15 derivatized from the 6␤-position were
synthesized as described previously [6,7].
Preparation of the chemiluminescent C7-substituted
17␤-estradiol conjugate 18 was achieved as shown in
Scheme 2. Briefly, acridinium active ester 4 was reacted
with excess ethylenediamine in DMF to afford N10-(3-sul-
fopropyl)acridinium-9-carboxamide amine 16 in 52% yield.
Subsequent coupling of 7␣-(3Ј-carboxypropyl)-17␤-estra-
diol (17) [8] with amine 16 in the presence of DCC and
HOBt provided chemiluminescent conjugate 18 in 57% yield.
3. Results and discussion
3.1. Synthesis of chemiluminescent 17␤-estradiol
conjugates
Chemiluminescent C6-substituted 17␤-estradiol conju-
gates were synthesized as shown in Scheme 1. Briefly,
N10-(3-sulfopropyl)acridinium-9-carboxamide free acid (2)
was activated using N-trifluoroacetoxysuccinimide (3). The
resulting active ester 4 precipitated from the reaction mix-
ture in nearly quantitative yield without need for further
purification and could be used directly to label estradiol
haptens containing free amines. Thus, commercially avail-
able 17␤-estradiol 6-carboxymethyloxime (5) was activated
with EDC and NHS to give the corresponding succinimidyl
ester that was reacted directly with tert-butyl N-(2-amino-
ethyl)carbamate in DMF to give 6 in 65% yield. Hydrolysis
of the Boc protecting group in 6 with trifluoroacetic acid in
CH2Cl2 provided amine hapten 7 as its TFA salt in 95%
yield. Coupling of N10-(3-sulfopropyl)acridinium-9-carbox-
amide active ester 4 with amine hapten 7 in DMF containing
a tertiary amine provided chemiluminescent conjugate 8.
6␣-[N-(6-aminohexamido)]-17␤-estradiol (9) [5] and 6␣-
3.2. Preparation of an immobilized 17␤-estradiol analog
biosensor surface
17␤-Estradiol analog 20 containing a terminal amino
functionality from the C7 position was synthesized by cou-
pling ethylenediamine with the bis-SEM-protected methyl
ester of 7␣-(3Ј-carboxypropyl)-17␤-estradiol (19) (Scheme
3) [8]. Subsequent amine coupling of compound 20 via
NHS and EDC to the carboxymethyl dextran surface of a
CM-5 sensor chip provided the immobilized estradiol ana-
log biosensor surface used in all studies described here
(Scheme 3). The surface generated had a maximum binding
capacity for the anti-E2 Fab fragment of ϳ200 RUs and
could be regenerated after each binding event with a single
pulse of 6 M guanidine hydrochloride.

M. Adamczyk et al. / Steroids 65 (2000) 295–303
301
Scheme 3.
3.3. Evaluation of anti-E2 Fab fragment/17␤-estradiol
conjugate binding affinities by BIAcore
subtraction of the sensorgram data from the blank surface,
and a binding response for each concentration of anti-E2
Fab fragment was subsequently determined (Fig. 1). Six
identical calibration studies were conducted during the
course of this work with no significant changes in binding
response obtained for a given concentration of anti-E2 Fab
fragment. The binding responses obtained for each concen-
tration of anti-E2 Fab fragment were averaged and plotted
against the Fab fragment concentration. A 4-parameter lo-
gistic fit (general model in BIAevaluation 3.0) of the data
provided the average calibration curve depicted in Fig. 2.
A fixed concentration of anti-E2 Fab fragment (500 pM)
was then mixed with 10–12 concentrations of each chemi-
luminescent 17␤-estradiol conjugate, and the solutions were
Binding interactions between the anti-E2 Fab fragment
and chemiluminescent 17␤-estradiol conjugates were eval-
uated in a solution competition format using the estradiol
analog biosensor surface to determine the concentration of
free anti-E2 Fab fragment in equilibrium solutions with a
chemiluminescent 17␤-estradiol conjugate [13,14]. Specif-
ically, the biosensor surface was calibrated by injecting
known concentrations of anti-E2 Fab fragment (0Ϫ500 pM)
over the blank and immobilized estradiol analog surfaces.
Data from the estradiol analog surface was corrected to
reflect only true anti-E2 Fab fragment binding events by
Fig. 1. Overlay plot of subtracted sensorgrams generated with 0, 31.2, 62.5, 125, 250, and 500 pM anti-E2 Fab fragment binding to the immobilized E2 analog
biosensor surface. Binding responses were obtained from the shaded portion of each sensorgram.

302
M. Adamczyk et al. / Steroids 65 (2000) 295–303
Table 1
Relative equilibrium dissociation constants between an anti-E2 Fab
fragment and estradiol or chemiluminescent conjugatesa
Conjugate
Linker (L)
Equilibrium dissociation
constant (pM)
Fig. 2. Average standard curve obtained for binding of anti-E2 Fab frag-
ment to the immobilized E2 analog biosensor surface.
Estradiol (1)
8
None
179 Ϯ 19.8
394 Ϯ 16.3
332 Ϯ 68.6
220 Ϯ 2.10
435 Ϯ 202
1610 Ϯ 270
256 Ϯ 72.1
ϭNOCH2CONH(CH2)2NH-
6␣-HCO(CH2)5NH-
6␣-NH-
6␤-NH-
6␤-CH2CONH(CH2)2NH-
7␣-(CH2)3CONH(CH2)2NH-
allowed to equilibrate. The concentration of free anti-E2 Fab
fragment remaining in each equilibrium mixture was sub-
sequently determined from the resulting corrected binding
response obtained and the calibration curve. A plot of free
anti-E2 Fab fragment versus total concentration of added
chemiluminescent 17␤-estradiol conjugate provides a com-
petition curve. Fig. 3 shows representative competition
curves for two of the analogs tested here. The data points
represent the experimentally determined concentrations of
free anti-E2 Fab fragment in equilibrium solutions at a given
concentration of chemiluminescent 17␤-estradiol conjugate.
The curves represent the best nonlinear fit of the data using
Eq. 1 (solution affinity model in BIAevaluation 3.0 soft-
ware)
10
13
14
15
18
a Values are average relative equilibrium dissociation constants (KD)
obtained from duplicate measurements.
where Fabf is the concentration of free anti-E2 Fab frag-
ment in equilibrium solutions; FabT is the total concen-
tration of anti-E2 Fab fragment in solution (500 pM); A is
the total concentration of added chemiluminescent con-
jugate, and KD is the relative equilibrium dissociation
constant for the binding of the conjugate to anti-E2 Fab
fragment in solution. The results of all binding studies are
summarized in Table 1.
2
FabT Ϫ A Ϫ KD
͑FabT ϩ A ϩ KD͒
Fabf ϭ
ϩ
Ϫ A*FabT
(1)
ͱ
2
4
3.4. Interpretation of anti-E2 Fab fragment/17␤-estradiol
conjugate binding results
The results of the binding studies show anti-E2 Fab
fragment binds all chemiluminescent conjugates tested with
relatively high affinity. Anti-E2 Fab fragment binds native
17␤-estradiol the tightest with KD ϭ 179 Ϯ 19.8 pM.
Homologous chemiluminescent conjugate 8 was prepared
from the same amino hapten used for preparation of the
immunogen used to raise the antibody in this study. Not
surprisingly, the antibody binds chemiluminescent conju-
gate 8 tightly with KD ϭ 394 Ϯ 16.3 pM. The remaining C6
modified conjugates (heterologous tracers) were all de-
signed to retain the sp3 nature of the C6 position of E2 and
thereby more closely mimic the structure and conformation
of native 17␤-estradiol. However, generation of conjugates
at the C6 position of E2 introduces concerns of preferred
stereochemistry and potential electronic effects on the ad-
jacent A-ring phenol. It was anticipated that ␣-stereo-
chemistry at the C6 position would be preferred because of
the possible spatial interactions between a 6␤-substituent
Fig. 3. Representative solution competition curves for anti-E2 Fab fragment
binding interactions with chemiluminescent estradiol conjugates 14 (F)
and 15 (Œ). The fixed concentration of anti-E2 Fab fragment was 500 pM.

M. Adamczyk et al. / Steroids 65 (2000) 295–303
303
and the 13␤-methyl and 17␤-hydroxyl functionalities that
References
are potential critical antibody binding determinants. A di-
rect comparison of equivalent chemiluminescent conjugates
13 and 14 derived from 6␣- and 6␤-amino-E2, respectively,
shows only a minor preference for the 6␣-derived conjugate
13 over the 6␤-derived conjugate 14. Incorporation of an
aminocaproate linker into the 6␣-amino-E2 hapten and in-
troduction of the chemiluminescent label to produce conju-
gate 10 also had little effect on the resulting binding inter-
action with the anti-E2 Fab fragment (KD ϭ 332 Ϯ 68.6
pM). Conjugates derived from 6-amino-E2 haptens may
result in an altered electronic A-ring environment attribut-
able to the electron-withdrawing nature of the resulting
amide functionality at the C6 position of any conjugates.
Incorporation of a linker through the C6-position of E2
originating in an alkyl functionality such as in conjugate 15
was anticipated to minimize electronic effects on the A-ring
environment. Interestingly, chemiluminescent conjugate 15
derived from a 6␤-carboxymethyl modified E2 hapten
showed a significantly reduced binding interaction with the
anti-E2 Fab fragment (KD ϭ 1610 Ϯ 270 pM). Whether this
reduction in binding recognition by anti-E2 Fab fragment
for conjugate 15 is attributable to an altered electronic
effect, an unfavorable steric interaction resulting from the
alternative linkage and 6␤-orientation, or both is uncertain
at this time. In contrast, incorporation of the chemilumines-
cent label from the 7␣-position such as in conjugate 18 also
minimizes effects on the A-ring while maintaining the pre-
ferred ␣-stereochemistry. Anti-E2 Fab fragment binds con-
jugate 18 with KD ϭ 256 Ϯ 72.1 pM.
[1] Wheeler MJ. In: Wild D, editor. The immunoassay handbook. New
York: Stockton Press, 1994. pp. 373–4.
[2] de Lauzon S, Desfosses B, Moreau MF, Le Trang N, Rajkowski K,
Cittanova N. Comparison of monoclonal antibodies to estradiol ob-
tained from structurally different immunogens. Hybridoma 1990;9:
481–91.
[3] el Jabri J. Enzyme immunoassay for plasma estradiol using a mono-
clonal antibody. J Steroid Biochem Mol Biol 1991;38:339–43.
[4] Tamate K, Charleton M, Gosling JP, et al. Direct colorimetric mono-
clonal antibody enzyme immunoassay for estradiol-17 beta in saliva.
Clin Chem 1997;43:1159–64.
[5] Adamczyk M, Mattingly PG, Reddy RE. An efficient stereoselective
synthesis of 6-␣-aminoestradiol: preparation of estradiol fluorescent
probes. Steroids 1997;62:462–7.
[6] Adamczyk M, Mattingly PG, Reddy RE. Synthesis of 6␤-aminoestra-
diol and its biotin, acridinium, and fluorescein conjugates. Steroids
1998;63:130–4.
[7] Adamczyk M, Johnson DD, Reddy RE. Synthesis of 6␤-[(2Ј-amino-
ethyl)carboxamidomethyl]estradiol and preparation of estradiol
probes. Bioconjugate Chem 1998;9:403–8.
[8] Adamczyk M, Johnson DD, Reddy RE. A stereoselective synthesis of
7␣-(3Ј-carboxypropyl)estradiol from a noncontrolled substance. Ste-
roids 1997;62:771–5.
[9] Adamczyk M, Chen YY, Moore JA, Mattingly PG. Estradiol-mimetic
probes. Preparation of 17␣-(6-aminohexynyl)estradiol biotin, fluores-
cein, and acridinium conjugates. Bioorg Med Chem Lett 1998;8:
1281–4.
[10] Adamczyk M, Chen YY, Mattingly PG, Pan Y. Neopentyl 3-trifly-
oxypropanesulfonate. A reactive sulfopropylation reagent for the
preparation of chemiluminescent labels.
1998;63:5636–9.
J
Org Chem
[11] Sakaibara S, Inukai N. The trifluoroacetate method of peptide syn-
thesis. I. The synthesis and use of trifluoroacetate reagents. Bull
Chem Soc Jpn 1965;38:1979–84.
[12] Hamcher H, Christ E. Potential antineoplastics. Arzneim-Forsch/
Drug Res 1983;33:347–52.
[13] Friguet B, Chafotte AF, Djavadi–Ohaniance L, Goldberg ME. Mea-
surements of the true affinity constant in solution of antigen-antibody
complexes by enzyme-linked immunosorbent assay. J Immunol
Methods 1985;77:305–19.
[14] Karlsson R. Real-time competitive kinetic analysis of interactions
between low-molecular weight ligands in solution and surface-immo-
bilized receptors. Anal Biochem 1994;221:142–51.
In summary, a series of conjugates between 17␤-estra-
diol and a hydrophilic acridinium-9-carboxamide chemilu-
minescent label were synthesized, and their solution binding
affinities for an anti-E2 Fab fragment were measured using
a single E2 analog biosensor surface on a BIAcore surface
plasmon resonance instrument. All of the conjugates bound
to the antibody fragment tightly, though slightly less effi-
ciently than unmodified estradiol, and represent suitable
candidates for further immunoassay development.