Chemiluminescence quenching of pteroic acid-N-sulfonyl-acridinium-9- carboxamide conjugates by folate binding protein

Article abstract of DOI:10.1016/j.bmcl.2004.01.105

N¹⁰-Trifluoroacetylpteroic acid was conjugated to chemiluminescent N-sulfonylacridinium-9-carboxamide labels at the N¹⁰ or 9-position carboxamide. Upon binding to folate binding protein the light output of the N¹⁰ derivative (9) was quenched up to 62% upon triggering with basic peroxide, while the 9-position carboxamide conjugate (7) was quenched only 12%. The utility of this effect was demonstrated in a model homogeneous chemiluminescent assay for folic acid.

Full text of DOI:10.1016/j.bmcl.2004.01.105

Bioorganic & Medicinal Chemistry Letters 14 (2004) 2313–2317
Chemiluminescence quenching of pteroic acid–N-sulfonyl-
acridinium-9-carboxamide conjugates by folate binding protein
Maciej Adamczyk,* James R. Fino, Phillip G. Mattingly, Jeffrey A. Moore and YouPan
Department of Chemistry (D09MD), Diagnostics Division, Abbott Laboratories, 100 Abbott Park Road, AP20,
Abbott Park, IL 60064-6016, USA
Received 8 January 2004; accepted 29 January 2004
Abstract—N10-Trifluoroacetylpteroic acid was conjugated to chemiluminescent N-sulfonylacridinium-9-carboxamide labels at the
N10 or 9-position carboxamide. Upon binding to folate binding protein the light output of the N10 derivative (9) was quenched up to
62% upon triggering with basic peroxide, while the 9-position carboxamide conjugate (7) was quenched only 12%. The utility of this
effect was demonstrated in a model homogeneous chemiluminescent assay for folic acid.
ꢀ 2004 Elsevier Ltd. All rights reserved.
Chemiluminescent
N-sulfonyl-acridinium-9-carbox-
system was chosen as the test case. Folic acid (pteroyl-
glutamic acid, PGA) (1) is the parent structure of a class
of enzyme cofactors important in one-carbon metabo-
lism and linked to serious disorders including congenital
malformations, cardiovascular disease, and carcinogen-
esis. A number of endogenous folate binding proteins
have been identified in a variety of tissues.4 The com-
mercially available folate binding protein (FBP) isolated
from cowÕs milk is a 222 amino acid glycoprotein5;6 with
amides are among the most efficacious labeling reagents
used worldwide in modern medical diagnostics. They
have achieved that status by virtue of the detection
sensitivity (equal or exceeding radioisotopes), specificity
(low sample matrix interference), dynamic range (106),
short read time (flash kinetics), and ease of use (variety
of reactive functional groups for conjugation, hydro-
philicity, simple equipment requirements).1
a reported association constant7 of 2 · 107–9 · 109 MÀ1
.
Recently, we communicated for the first time that the
chemiluminescence of acridinium salts can be quenched
by specific binding to a protein.2 In that example the
signal from an N10-biotinylated acridinium-9-carbox-
amide conjugate was quenched 92% in the presence of
avidin, laying the foundation for a homogeneous
chemiluminescent assay for biotin. In many respects
avidin–biotin receptor–ligand pairing is unique. Avidin
is a tetrameric glycoprotein that binds four biotin ligands
with extraordinarily high affinity (10À15 M). The ligand
Even though, the architecture of the binding site has not
been defined, FBP has been used successfully in many
assay formats for the detection of folic acid.7–10
H
H2N
N
N
N
N
CH2NH
CONH
O
HOOCCH2CH2
C
H
COOH
3
ꢀ
binding pocket is very deep, estimated to be 15 A.
1, Folic acid
In this report, we investigated whether a lower affinity
binding interaction could result in efficient chemilumi-
nescent quenching. The folate/folate binding protein
Two labeled ligands for FBP were prepared as shown in
Scheme 1. Pteroic acid (2) was first acylated with tri-
fluoroacetic anhydride to give N10-trifluoroacetyl pteroic
acid 3.11 Compound 3 was activated with N-hydroxy-
succinimide (NHS) and 1-[3-(dimethyl-aminopropyl)]-3-
ethylcarbodiimide hydrochloride (EDAC) and then
Keywords: Homogeneous chemiluminescent assays; Folate binding
protein.
* Corresponding author. Tel.: +1-847-937-0225; fax: +1-847-938-8927;
e-mail: maciej.adamczyk@abbott.com
0960-894X/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.01.105

2314
M. Adamczyk et al. / Bioorg. Med. Chem. Lett. 14 (2004) 2313–2317
H
H
N
O
CF3
H2N
N
N
H2N
N
N
N
H
N
N
N
N
a
OH
OH
O
O
O
O
2
3
H
N
O
CF3
H2N
N
N
N
N
b, c
NH(CH2)6NHR
O
O
4 R = tert-butoxycarbonyl
5 R = H
(CH2)3SO3
N
H
H2N
N
N
O
CF3
d
N
N
O
N
SO2
5
N
O
NH(CH2)6NHCO(CH2)3
O
7
H
N
H2N
N
N
O
CF3
N
N
O
NH(CH2)6NHCO(CH2)9
N
O
e
5
9
O
N
SO2
O3S(CH2)3
(CH2)3SO3
N
(CH2)9CO2C6F5
N
10
9
O
N SO2
O
O
N SO2
O CO(CH2)3
O3S(CH2)3
N
8
6
O
Scheme 1. Preparation of FBP ligands. (a) i. Trifluoroacetic anhydride, 60–65 ꢁC, sealed tube, 72 h; ii. aq ethanol, reflux, 1 h, 89%; (b) i. diiso-
propylethylamine, NHS, EDAC, DMF, 20 h; ii. BocNH(CH2)6NH2, 24 h, 94%; (c) trifluoroacetic acid, dichloromethane, 2 h, 76%; (d) 6, 50 mM
phosphate buffer (pH 8)/DMF, 16 h, 46%; (e) 8, DMF, triethylamine, 3 h.
condensed with 6-(N-tert-butoxycarbonylamino)hexyl-
amine to yield the N-protected pteroic acid derivative 4.
Acidolysis of the Boc group gave the amino function-
alized intermediate 5 as a trifluoroacetic acid salt.
Chemiluminescence of 7 and 9 were first evaluated in the
absence of FBP. As shown in Figure 1a, the time/
intensity profiles of the two compounds dissolved in
buffer (10 mM borate, 2 mM EDTA, 0.1% Tween-20,
pH 7.3) were similar upon reaction with basic hydrogen
peroxide, regardless of the regiochemistry of attachment
to the pteroic acid nucleus. Both tracers exhibited flash-
type luminescent profiles with all light being generated
within 5 s after addition of the triggering reagent. The
The acridinium-9-carboxamide active esters 612 and 82
bearing linkers extending from the 9- and 10-positions,
respectively, were conjugated to pteroylamine 5 to give
the chemiluminescent folate tracers 7 and 9.

M. Adamczyk et al. / Bioorg. Med. Chem. Lett. 14 (2004) 2313–2317
2315
40 nM
800000
600000
400000
200000
0
250
200
30 nM
20 nM
15 nM
10 nM
7.5 nM
5 nM
150
7
9
100
50
0
2.5 nM
1.25 nM
0.625 nM
0
2
4
6
8
10
0
100
200
300
400
(a)
(a)
Time (s)
Time (s)
1e+006
1e+005
10000
1000
100
20
16
12
8
9
7
7
9
3.2 × 1019 cts/mol
2.9 × 1019 cts/mol
4
10
0
1
1
10
100
1e-017
1e-016
1e-015
[moles]
1e-014
(b)
Folate Tracer (nM)
(b)
Figure 2. Binding of FBP to ligands 7 and 9. (a) BIACore sensorgram:
FBP (20 nm), ligand 9 (0.625–40 nM). (b) IC50 plot. (Data points
represent the average of duplicate values. The curve represents the best
nonlinear fit of the data using a four-parameter logistic.)
Figure 1. Chemiluminescence of ligands 7 and 9. (a) Time/relative light
units profile. The labeled ligands were triggered with 200 lL 0.18 N
NaOH, 0.7% H2O2, 1% Triton X-100, 0.05% diethylenetriaminepent-
acetic acid. (b) Dilution linearity. Ligands were diluted in 10 mM
borate, 2 mM EDTA, 0.1% Tween-20, pH 7.3, and triggered as above.
individually mixed with serially diluted folate binding
protein (0.08–40 nM) in assay buffer. After the solutions
were incubated for 45 min, the chemiluminescence
response from each mixture (25 lL) was recorded. Plots
of the normalized relative light units (RLU) obtained
versus folate binding protein concentration are shown in
Figure 3.
integrated signal for compound 7 was 3.2 · 1019 counts/
mol while compound 9 was essentially the same,
2.9 · 1019 counts/mol (Fig. 1b).
The ability of the acridinylated ligands to bind to FBP
was then tested independently of the chemiluminescent
signal from the acridinium moiety in a solution com-
petition surface plasmon resonance format on the
BIAcore instrument (Fig. 2).13 Thus, a high-density
biosensor was prepared by conjugating pteroylamine 5
to a CM-5 carboxylated chip according to the manu-
facturerÕs protocol. A solution of FBP (20 nM) was
incubated with varying concentrations of 7 or 9, and
then injected onto the biosensor. The resulting sensor-
grams (Fig. 2a) were used to construct a plot of the
concentration of free FBP versus ligand (Fig. 2b).
The IC50 for each ligand was determined from the plot.
The values for 7 (8.43 0.46 nM) and 9 (8.14 0.29 nM)
were essentially the same.
The effect of binding of the ligands to FBP on the
chemiluminescent signal from the acridinium-9-carbox-
amide label was next investigated. Specifically, the
acridinium-derivatized FBP ligands 7 and 9 (1 nM) were
Figure 3. Plot of normalized chemiluminescent response for (9) (ꢀ) and
(7) FBP ligands (ꢁ) versus folate binding protein concentration. (Data
points represent the average of duplicate values. The curve represents
the best nonlinear fit of the data using a four-parameter logistic.)

2316
M. Adamczyk et al. / Bioorg. Med. Chem. Lett. 14 (2004) 2313–2317
The data show that chemiluminescence from an acridi-
nium–FBP ligand can be quenched by folate binding
protein in a concentration-dependent manner. Signifi-
cantly, the chemiluminescent output derived from ligand
9 is quenched up to 62% upon titration with folate
binding protein. In contrast, ligand 7 was quenched only
12% upon titration with folate binding protein. Identical
studies utilizing the ligands 7 and 9 and bovine serum
albumin (BSA) were also conducted as controls.
Chemiluminescent signal produced from these controls
over the entire BSA concentration range tested (0.08–
40 nM) did not vary significantly. The results demon-
strate that chemiluminescence quenching is due to a
specific binding interaction.
Figure 4. Dose–response curve generated for folic acid. (Data points
represent the average of triplicate values. The curve represents the best
nonlinear fit of the data using a four-parameter logistic.)
The FBP specific modulation of acridinium chemilumi-
nescence was further demonstrated in a model homo-
geneous assay for folic acid. Thus, a fixed concentration
of ligand 9 (1 nM) was added to several concentrations
of folic acid (0.75–10 nM) in assay buffer. The folate
containing mixtures were subsequently added to a fixed
concentration of folate binding protein (20 nM) also in
assay buffer. The solutions were allowed to incubate for
45 min, and 25 lL of each sample was triggered as
above. The dose–response curve obtained is depicted in
Figure 4. The average % quenching value reported in
Figure 4 is relative to the chemiluminescent response
obtained for ligand 9 in buffer alone. As expected,
acridinium chemiluminescence quenching due to asso-
ciation of folate binding protein with the folate tracer
decreases with increasing concentrations of folic acid.
(9) results in more efficient chemiluminescence quench-
ing than that of the 9-position analog (7) is consistent
with the results previously reported in the biotin-acrid-
inium-9-carboxamide–avidin system2 albeit falling short
in magnitude. Based on the accepted mechanism of
acridinium salt chemiluminescence, we rationalize that
in both systems, the acridinium-ligand–protein complex,
when reacted with basic peroxide, gives rise to an excited
state acridone chemiluminophore, which is either free in
solution or left bound to the protein. This is illustrated
for compounds 7 and 9 in Figure 5.
In the bound state, the excited acridone from 9 must be
held in the proximity of a quenching element while the
acridone from 7 is free to dissociate from the FBP-
ligand complex.
The observation that linkage of pteroic acid through the
N10-position of the acridinium-9-carboxamide nucleus
Acridinium
Label
FBP
Ligand
7 or 9
H2O2 /HO
O
3S
O
NH
O2S
N
O
NH
Ligand
Ligand
FBP
FBP
O
Free Excited Acridone from 7
N
4
N
LIGHT
O
Bound Excited Acridone from 9
Figure 5. FBP-mediated quenching of ligands 7 and 9.

M. Adamczyk et al. / Bioorg. Med. Chem. Lett. 14 (2004) 2313–2317
2317
5. Svendsen, I.; Hansen, S. I.; Holm, J.; Lyngbye, J.
Carlsberg Res. Commun. 1984, 49, 123–131.
6. Adamczyk, M.; Gebler, J. C.; Shreder, K.; Wu, J. Rapid
Commun. Mass Spectrom. 2003, 17, 1504–1506.
7. Bachas, L. G.; Tsalta, C. D.; Meyerhoff, M. E. Biotech-
niques 1986, 4, 42–55.
In summary, the 62% chemiluminescence quenching
efficiency observed in the N10-acridinium-9-carbox-
amide-pteroic acid–FBP binding system was sufficient to
demonstrate a homogeneous chemiluminescent assay
format for the detection of folic acid.
8. Wilson, D.; Herrmann, R.; Hsu, S.; Biegalski, T.; Sohn,
L.; Forsythe, C.; Novotny, M.; Beggs, M.; Manderino, G.
Clin. Chem. 1995, 41, 1780–1781.
9. Huang, W.; Feltus, A.; Witkowski, A.; Daunert, S. Anal.
Chem. 1996, 68, 1646–1650.
We continue to explore other binding protein/ligand-
acridinium-9-carboxamide pairs to further define the
general scope of this technology.
10. Daunert, S.; Bachas, L. G.; Ashcom, G. S.; Meyerhoff,
M. E. Anal. Chem. 1990, 62, 314–318.
References and notes
11. Godwin, H. A.; Rosenberg, I. H.; Ferenz, C. R.; Jacobs,
P. M.; Meienhofer, J. J. Biol. Chem. 1972, 247, 2266–2271.
12. Adamczyk, M.; Chen, Y.-Y.; Fino, J. R.; Mattingly, P. G.
Microwave-assisted synthesis of chemiluminescent acridi-
nium salts. In Proceedings of the XIth international
symposium on bioluminescence andchemiluminescence ;
Case, J. F., Herring, P. J., Robison, B. H., Haddock,
S. H. D., Kricka, L. J., Stanley, P. E., Eds.; World
Scientific: Asilomar CA, 2001; pp 207–210.
1. Mattingly, P. G.; Adamczyk, M. Chemiluminescent
n-sulfonylacridinium-9-carboxamides and their applica-
tion in clinical assays. In Luminescence biotechnology:
Instruments andapplications ; Dyke, K. V., Ed.; CRC: Boca
Raton, 2002; pp 77–105.
2. Adamczyk, M.; Mattingly, P. G.; Moore, J. A.; Pan, Y.
Org. Lett. 2003, 5, 3779–3782.
3. Green, N. M. Methods in Enzymology 1990, 184, 51–67.
4. Lyngbye, J.; Hansen, S. I.; Holm, J. Methods Enzymol.
1980, 66, 694–709.
13. Adamczyk, M.; Moore, J. A.; Yu, Z. Methods 2000, 20,
319–328.